基于增量改进BP神经网络微波深度干燥模型及应用研究
|
摘要
BP神经网络作为人工智能研究领域的重要分支,具有良好的非线性映射能力和高度的并行信息处理能力,在多学科交叉技术领域得到广泛的应用。
微波干燥不同于传统干燥方式,其热传导方向与水分扩散方向相同。与传统干燥方式相比,具有干燥速率大、节能、生产效率高、干燥均匀、清洁生产、易实现自动化控制和提高产品质量等优点,但是在干燥过程中,影响微波干燥的因素包括微波输入功率、微波作用时间、物料初始含水量、物料质量、物料转速等,其在干燥过程中影响程度不同,致使微波预处理富硒渣的干燥过程试验周期长,试验量大且参数不易优化。选择具有非线性映射能力的BP神经网络,对微波干燥过程建立仿真模型,预测分析试验过程。
标准的BP算法是基于梯度下降法,通过计算目标函数对网络权值和阈值的梯度修正网络权值,在训练过程中存在收敛速度慢和局部最小的问题；且对于复杂的问题,在训练过程中会陷入局部最小点,以致无法收敛。因此,需对BP神经网络算法进行改进。
本论文以微波深度干燥富硒渣的工业化试验过程为研究对象,建立基于增量改进BP神经网络的预测模型和Smith补偿PID控制模型,及研究预测模型在微波煅烧领域的应用,主要研究内容是：
1)采用Levenberg-Marquardt (L-M)算法对BP神经网络加以改进,提高了神经网络的收敛速度,同时针对在训练神经网络的过程中,无法一次性提供所需的训练数据且当样本规模较大时,系统内存的限制使得对所有样本的训练不可行等问题,提出基于增量学习的BP神经网络。
结合在单个神经网络中设置权有效区域和添加隐含层单元数的方法实现增量学习,在新样本知识与原有样本知识接近时,可以在一定范围内修改权值和阈值,同时采用灵活的方式确定隐含层节点数目。在利用训练样本集对神经网络进行训练的过程中,不断增加隐含层节点数目,并计算输出误差,若误差达到要求就停止网络训练,此时网络隐含层节点数即为最优的隐含层节点数。此方法既可以使网络学习新样本知识,又可以使网络保持原有样本的知识。
基于增量学习和L-M优化算法的改进BP神经网络,弥补了传统BP神经网络的不足,它具有更快的收敛性,较好的预测精度和更好的拟合结果,并能够避免误差总和不再更新,网络不再训练以致使网络瘫痪等问题,在调整网络参数时不会陷入局部最小,使网络迅速收敛,能够有效地解决训练数据不能一次性提供,以较少的先验知识学习过程的特点等问题,选择有代表性的样本在占用较少内存的前提下训练神经网络,即能够保持原有知识,又能够学习新的知识。
2)在微波深度干燥富硒渣的工业化试验中,建立了增量改进BP神经网络的非线性系统预测模型,以微波输入功率、微波作用时间、物料初始含水量、物料质量和物料盘转速为输入条件,用以预测微波深度干燥富硒渣的工业化试验结果。
3)建立了增量改进BP神经网络的能耗预测模型,以微波干燥工业化试验中微波功率、微波作用时间、物料质量、物料初始含水率和所需物料最终含水率为输入条件,预测工业化试验过程的能耗。
4)在控制领域中,传统的PID控制器结构简单,对模型误差具有鲁棒性及易于操作等优点,被广泛应用于冶金、化工、电力、轻工和机械等工业过程控制领域中。随着工业的发展,被控对象的复杂程度不断加深,尤其对于大滞后、时变的、非线性的复杂系统,传统PID控制已经无法满足目标控制精确化的要求。
建立增量改进BP神经网络的Smith补偿PID控制模型,根据工业化试验过程测定的被控对象参数作为神经网络的输入,对控制系统进行离线系统辨识,利用简化了的微波深度干燥富硒渣工业化生产过程的控制模型对增量改进BP神经网络的Smith补偿PID控制器进行仿真研究,实现在线动态整定PID控制参数。
5)建立了增量改进BP神经网络反预测模型,以物料的最终质量、物料的最终温度和物料的相对脱水率为输入条件,预测微波干燥工业化试验所需的工作时间、物料初始含水率和试验能耗。
6)将工业化微波深度干燥富硒渣的神经网络预测模型应用于微波煅烧重铀酸铵(ammonium diuranate, ADU)和三碳酸铀酰铵(ammonium uranyl carbonate, AUC)试验中,用以预测微波煅烧ADU和AUC的试验过程结果。
As an important branch of artificial intelligence research area, Back-Propagation (BP) neural network has favourable nonlinear mapping and high parallel information processing capabilities, and it is widely used in technology areas of multi-disciplinary subjects.
Microwave drying technology, is different from the traditional drying methods, has characteristics of instantaneity, integrity, efficiency, safety, non-pollution, easy to realize the automatic control and improving the products quality, the direction of heat conduction is the same as the direction of moisture diffusion. But in the industrial process of microwave deep drying of selenium-rich slag, the influencing factors include the microwave input power, the acting time, the initial moisture content, the average material mass, the average material surface area and the rotational frequency. The affecting degrees of these factors are different, so that it can be guaranteed the testing cycle longer, the testing quantity larger and the parameters difficult to be optimized. The BP neural network, which has the ability of non-linear mapping, is chosen to build up the simulation model to predict the experimental process results.
However, the traditional BP algorithm based on gradient descent method revises the network weights through computing the network weights'and thresholds'values of the objective functions, needs more convergent time and sometimes the convergent results can not be obtained in local minimum areas. Therefore, the BP neural network algorithm should be improved.
In present thesis, the prediction model and the Smith compensation Proportion> Integration Differentiation (PID) control model based on incremental improved BP neural network are built up taking the industrial process of microwave deep drying of selenium-rich slag as the research target, and the universality of prediction model in the field of microwave calcination is researched. The main contributions are summarized as follows:
1) The improved BP neural network based on Levenberg-Marquardt (L-M) algorithm overcomes these limitations. In the process of training the network, many training data probably are offered by the way of increment batch and the limitation of the system memory can make the training data infeasible when the sample scale is large, the incremental learning is implemented by adjusting the weights of the BP neural network as demonstrated.
The incremental learning is realized by setting up the knowledge effective extent in single neural network and increasing the number of the nodes in hidden layers. When the new sample knowledge closes to the prior sample knowledge, the weights' and thresholds'value can be changed in the effective extent, and the number of the nodes in hidden layers is fixed flexibly. In the process of training neural network using the training sample, the number of the nodes in hidden layers is increased and the output error is calculated. If the indication value approaches to the target value, the training is ending and the number of the nodes in hidden layers is the optimization. By using the method, the network can not only learn the new sample knowledge but sustain the original knowledge.
The incremental improved BP neural network has the faster convergence, better prediction accuracy, better fitting results and can avoid the error sum squares no longer be updated, the phenomenon of network paralysis and the network not be trained, be out of the local minimum when adjusting the network parameters, make the network be converged rapidly, can effectively solve the problem such as the training data can not be provided for one-time, choose the representative samples to train the network in the case of occupying less memory source.
2) The incremental improved BP neural network non-linear prediction model is built up during the industrial process of microwave deep drying of selenium-rich slag. Taking the microwave input power, the acting time, the initial moisture content, the average material mass, the average material surface area and the rotational frequency as input variables, the model is used to predict the experimental results.
3) The incremental improved BP neural network prediction model for industrial electricity energy consumption is built up. Taking the microwave input power, the acting time, the initial moisture content, the final moisture content and the average material mass as input variables, the model is used to predict the experimental electricity energy consumption.
4) The conventional PID control model is widely applied in the fields of industry process control because of its simple structure, robustness and easily operation, such as metallurgy, chemical engineering and electric power and mechanics. With the development of industry, the complexity of the controlled plant is serious, especially for the time-delaying, time-varying and non-liner complex systems, the conventional PID control model can not be met the control accuracy of the target.
The industrial Smith compensation PID control model using incremental improved BP neural network is put forward. The parameters of controlled plant are evaluated in the experimental process and identified the neural network Smith compensation PID control off-line. Using the simplified control model of microwave deep drying selenium-rich slag to research the Smith compensation PID controller, the PID control parameters are tuned on-line.
5) The industrial anti-prediction model using incremental improved BP neural network is proposed in microwave deep drying of selenium-rich slag processing. Taking the final material mass, the final temperature and the final relative dehydration rate as input variables, anti-prediction model is used to predict the acting time, the initial dehydration rate and the electricity energy consumption.
6) The industrial prediction model of microwave deep drying of selenium-rich slag using incremental improved BP neural network is used for the microwave calcination of Ammonium Diuranate (ADU) and Ammonium Uranyl Carbonate (AUC) in order to investigate the universality of the prediction model. The model is used to predict the experimental processing results of microwave calcination.
微波干燥不同于传统干燥方式,其热传导方向与水分扩散方向相同。与传统干燥方式相比,具有干燥速率大、节能、生产效率高、干燥均匀、清洁生产、易实现自动化控制和提高产品质量等优点,但是在干燥过程中,影响微波干燥的因素包括微波输入功率、微波作用时间、物料初始含水量、物料质量、物料转速等,其在干燥过程中影响程度不同,致使微波预处理富硒渣的干燥过程试验周期长,试验量大且参数不易优化。选择具有非线性映射能力的BP神经网络,对微波干燥过程建立仿真模型,预测分析试验过程。
标准的BP算法是基于梯度下降法,通过计算目标函数对网络权值和阈值的梯度修正网络权值,在训练过程中存在收敛速度慢和局部最小的问题；且对于复杂的问题,在训练过程中会陷入局部最小点,以致无法收敛。因此,需对BP神经网络算法进行改进。
本论文以微波深度干燥富硒渣的工业化试验过程为研究对象,建立基于增量改进BP神经网络的预测模型和Smith补偿PID控制模型,及研究预测模型在微波煅烧领域的应用,主要研究内容是：
1)采用Levenberg-Marquardt (L-M)算法对BP神经网络加以改进,提高了神经网络的收敛速度,同时针对在训练神经网络的过程中,无法一次性提供所需的训练数据且当样本规模较大时,系统内存的限制使得对所有样本的训练不可行等问题,提出基于增量学习的BP神经网络。
结合在单个神经网络中设置权有效区域和添加隐含层单元数的方法实现增量学习,在新样本知识与原有样本知识接近时,可以在一定范围内修改权值和阈值,同时采用灵活的方式确定隐含层节点数目。在利用训练样本集对神经网络进行训练的过程中,不断增加隐含层节点数目,并计算输出误差,若误差达到要求就停止网络训练,此时网络隐含层节点数即为最优的隐含层节点数。此方法既可以使网络学习新样本知识,又可以使网络保持原有样本的知识。
基于增量学习和L-M优化算法的改进BP神经网络,弥补了传统BP神经网络的不足,它具有更快的收敛性,较好的预测精度和更好的拟合结果,并能够避免误差总和不再更新,网络不再训练以致使网络瘫痪等问题,在调整网络参数时不会陷入局部最小,使网络迅速收敛,能够有效地解决训练数据不能一次性提供,以较少的先验知识学习过程的特点等问题,选择有代表性的样本在占用较少内存的前提下训练神经网络,即能够保持原有知识,又能够学习新的知识。
2)在微波深度干燥富硒渣的工业化试验中,建立了增量改进BP神经网络的非线性系统预测模型,以微波输入功率、微波作用时间、物料初始含水量、物料质量和物料盘转速为输入条件,用以预测微波深度干燥富硒渣的工业化试验结果。
3)建立了增量改进BP神经网络的能耗预测模型,以微波干燥工业化试验中微波功率、微波作用时间、物料质量、物料初始含水率和所需物料最终含水率为输入条件,预测工业化试验过程的能耗。
4)在控制领域中,传统的PID控制器结构简单,对模型误差具有鲁棒性及易于操作等优点,被广泛应用于冶金、化工、电力、轻工和机械等工业过程控制领域中。随着工业的发展,被控对象的复杂程度不断加深,尤其对于大滞后、时变的、非线性的复杂系统,传统PID控制已经无法满足目标控制精确化的要求。
建立增量改进BP神经网络的Smith补偿PID控制模型,根据工业化试验过程测定的被控对象参数作为神经网络的输入,对控制系统进行离线系统辨识,利用简化了的微波深度干燥富硒渣工业化生产过程的控制模型对增量改进BP神经网络的Smith补偿PID控制器进行仿真研究,实现在线动态整定PID控制参数。
5)建立了增量改进BP神经网络反预测模型,以物料的最终质量、物料的最终温度和物料的相对脱水率为输入条件,预测微波干燥工业化试验所需的工作时间、物料初始含水率和试验能耗。
6)将工业化微波深度干燥富硒渣的神经网络预测模型应用于微波煅烧重铀酸铵(ammonium diuranate, ADU)和三碳酸铀酰铵(ammonium uranyl carbonate, AUC)试验中,用以预测微波煅烧ADU和AUC的试验过程结果。
As an important branch of artificial intelligence research area, Back-Propagation (BP) neural network has favourable nonlinear mapping and high parallel information processing capabilities, and it is widely used in technology areas of multi-disciplinary subjects.
Microwave drying technology, is different from the traditional drying methods, has characteristics of instantaneity, integrity, efficiency, safety, non-pollution, easy to realize the automatic control and improving the products quality, the direction of heat conduction is the same as the direction of moisture diffusion. But in the industrial process of microwave deep drying of selenium-rich slag, the influencing factors include the microwave input power, the acting time, the initial moisture content, the average material mass, the average material surface area and the rotational frequency. The affecting degrees of these factors are different, so that it can be guaranteed the testing cycle longer, the testing quantity larger and the parameters difficult to be optimized. The BP neural network, which has the ability of non-linear mapping, is chosen to build up the simulation model to predict the experimental process results.
However, the traditional BP algorithm based on gradient descent method revises the network weights through computing the network weights'and thresholds'values of the objective functions, needs more convergent time and sometimes the convergent results can not be obtained in local minimum areas. Therefore, the BP neural network algorithm should be improved.
In present thesis, the prediction model and the Smith compensation Proportion> Integration Differentiation (PID) control model based on incremental improved BP neural network are built up taking the industrial process of microwave deep drying of selenium-rich slag as the research target, and the universality of prediction model in the field of microwave calcination is researched. The main contributions are summarized as follows:
1) The improved BP neural network based on Levenberg-Marquardt (L-M) algorithm overcomes these limitations. In the process of training the network, many training data probably are offered by the way of increment batch and the limitation of the system memory can make the training data infeasible when the sample scale is large, the incremental learning is implemented by adjusting the weights of the BP neural network as demonstrated.
The incremental learning is realized by setting up the knowledge effective extent in single neural network and increasing the number of the nodes in hidden layers. When the new sample knowledge closes to the prior sample knowledge, the weights' and thresholds'value can be changed in the effective extent, and the number of the nodes in hidden layers is fixed flexibly. In the process of training neural network using the training sample, the number of the nodes in hidden layers is increased and the output error is calculated. If the indication value approaches to the target value, the training is ending and the number of the nodes in hidden layers is the optimization. By using the method, the network can not only learn the new sample knowledge but sustain the original knowledge.
The incremental improved BP neural network has the faster convergence, better prediction accuracy, better fitting results and can avoid the error sum squares no longer be updated, the phenomenon of network paralysis and the network not be trained, be out of the local minimum when adjusting the network parameters, make the network be converged rapidly, can effectively solve the problem such as the training data can not be provided for one-time, choose the representative samples to train the network in the case of occupying less memory source.
2) The incremental improved BP neural network non-linear prediction model is built up during the industrial process of microwave deep drying of selenium-rich slag. Taking the microwave input power, the acting time, the initial moisture content, the average material mass, the average material surface area and the rotational frequency as input variables, the model is used to predict the experimental results.
3) The incremental improved BP neural network prediction model for industrial electricity energy consumption is built up. Taking the microwave input power, the acting time, the initial moisture content, the final moisture content and the average material mass as input variables, the model is used to predict the experimental electricity energy consumption.
4) The conventional PID control model is widely applied in the fields of industry process control because of its simple structure, robustness and easily operation, such as metallurgy, chemical engineering and electric power and mechanics. With the development of industry, the complexity of the controlled plant is serious, especially for the time-delaying, time-varying and non-liner complex systems, the conventional PID control model can not be met the control accuracy of the target.
The industrial Smith compensation PID control model using incremental improved BP neural network is put forward. The parameters of controlled plant are evaluated in the experimental process and identified the neural network Smith compensation PID control off-line. Using the simplified control model of microwave deep drying selenium-rich slag to research the Smith compensation PID controller, the PID control parameters are tuned on-line.
5) The industrial anti-prediction model using incremental improved BP neural network is proposed in microwave deep drying of selenium-rich slag processing. Taking the final material mass, the final temperature and the final relative dehydration rate as input variables, anti-prediction model is used to predict the acting time, the initial dehydration rate and the electricity energy consumption.
6) The industrial prediction model of microwave deep drying of selenium-rich slag using incremental improved BP neural network is used for the microwave calcination of Ammonium Diuranate (ADU) and Ammonium Uranyl Carbonate (AUC) in order to investigate the universality of the prediction model. The model is used to predict the experimental processing results of microwave calcination.
引文
[1]W. S. McCulloch and W. Pitts. A logical calculus of ideas immanent in nervous activity [J]. Bulletin of Mathematical Biophysics,1990,52 (1-2):99-115.
[2]D. O. Hebb. The organization of behavior [M]. New York:Wiley,1949.
[3]F. Rosenblatt. The Perceptron:A probabilistic model for information storage and organization in the brain [J]. Psychological Review,1958,65 (6):386-408.
[4]B. Widrow and M. E. Hoff. Adaptive switching circuit [M]. New York:IRE,1960: 94-104.
[5]M. Minsky and S. Papert. Perceptrons [M]. MIT Press,1969.
[6]J. J. Hopfield. Neural networks and physical systems with emergent collective computational abilities [C]. Proceedings of the National Academy of Science, 1982,79:2554-2558.
[7]D. E. Rumelhart and J. L. McClelland. Parallel Distributed Processing: Explorations in the Microstructure of Cognition [M]. Cambridge MA:MIT Press, 1986.
[8]D. E. Rumelhart, G. E. Hinton and R. J. Williams. Learning representations by back-propagation errors [J]. Nature,1986,323:533-536.
[9]M. T. Leung. Fingerprint processing using back propagation neural networks [J]. International Joint Conference on Neural Networks,1990,90:15-20.
[10]K. Fukshima. Handwritten alphanumeric character recognition by the necognitron [J]. IEEE Transactions on Neural Networks,1991.
[11]S. C. Ahalt. Performance of synthetic neural network classification of noise radar signals [J]. Advances in Neural Information Processing Systems,1989,1: 181-288.
[12]E. Wacholder. A neural network algorithm for the multiple traveling salesman problem [J]. Biology Cybernetics,1989,61 (1):11-19.
[13]K. S. Narendra and K. Parthasarathy. Identification and control of dynamical systems using neural networks [J]. IEEE Transactions on Neural Networks,1990, 1(1):4-27.
[14]K. S. Narendra and K. Parthasarathy. Gradient methods for the optimization of dynamical systems containing neural networks [J]. IEEE Transactions on Neural Networks,1991,2 (2):252-262.
昆明理工大学博士学位论文 参考文献
[15]雷蕴奇.神经元自适应控制的概念[C].中国神经网络首届学术大会论文集,1990.
[16]A. Guez, J. Eilbert and M. Kam. Neuromorphic architecture for adaptive robot control:A preliminary analysis Proc [C]. IJCNN,1987,4:567-572.
[17]沈清,胡德文.神经网络应用技术[M].国防科技大学出版社,1993.
[18]N. V. Bhat. Modeling chemical process systems via neural computation [J]. IEEE Control Systems Magazine,1990,10 (3):24-30.
[19]A. Namphol. Higher order data compression with neural network [C]. International Joint Conference on Neural Networks,1991,1:55-59.
[20]J. D. Daugman. Complete discrete 2-D Gabor transforms by neural network for image analysis and compression [J]. IEEE Transactions on ASSP,1998,36 (7): 1169-1179.
[21]B. Kosko. Adaptive bi-directional associative memories [M]. Applied Optics, 1982.
[22]J. D. Provence. Neural network implementation for maximum-likelihood sequence estimation of binary signals in Gaussion noise [C]. Proceedings of the IEEE International Joint Conference on Neural Networks,1987,3.
[23]Y. T. Zhou. Image restoration using a neural network [J]. IEEE Transactions on ASSP.1988,36(7):1141-1151.
[24]S. U. H. Qureshi. Adaptive equalization [C]. Proceedings of IEEE,1985,73: 1349-1387.
[25]R. P. Lippmann and P. Bexkman. Adaptive neural net preprocessing for signal diction in non-Gaussion noise [M]. Advances in Neural Information Processing Systems,1989,1.
[26]B. Widrow and R. Winter Neural nets for adaptive filtering and adaptive pattern recognition [J]. Computer,1988,21 (3):25-29.
[27]Q. Z. Xue. Neural network based adaptive matched filtering for QRS detection [J]. IEEE Transactions on Biomedical Engineering,1992,39 (4):317-329.
[28]Q. Zhang. Adaptive equalization using the back-propagation algorithm [J]. IEEE Transactions on CAS,1990,37 (6):848-849.
[29]朱代根.微波干燥过程的神经网络模型[J].农机化研究,2010,6(6)：94-96.
[30]I. Alibas Ozkan, B. Akbudak, N. Akbudak. Microwave drying characteristics of spinach [J]. Journal of Food Engineering,2007,78 (2):577-583.
[31]许晓春,林朝朋.佛手瓜的微波真空干燥及其神经网络模拟的研究[J].韶关学院学报(自然科学),2007,28(12)：80-84.
[32]L. Momenzadeh, A. Zomorodian, D. Mowla. Experimental and theoretical investigation of shelled corn drying in a microwave-assisted fluidized bed dryer using Artificial Neural Network [J]. Food and Bioproducts Processing,2011,89 (1):15-21.
[33]J. L. Pedreno Molina, J. Monzo Cabrera, A. Toledo Moreo, D. Sanchez Hernandez. A novel predictive architecture for microwave-assisted drying processes based on neural networks [J]. International Communications in Heat and Mass Transfer,2005,32 (8):1026-1033.
[34]J. L. Pedreno Molina, J. Monzo Cabrera, D. Sanchez Hernandez. A new predictive neural architecture for solving temperature inverse problems in microwave-assisted drying process [J]. Neurocomputing,2005,64:521-528.
[35]P. Poonnoy, A. Tansakul, M. Chinnan. Estimation of moisture ratio of a mushroom undergoing microwave-vacuum drying using artificial neural network and regression models [J]. Chemical Product and Process Modeling,2007,2 (3).
[36]P. Poonnoy, A. Tansakul, M. Chinnan. Parallel dynamic artificial neural network for temperature and moisture content predictions in microwave-vacuum dried tomato slices [J]. Chemical Product and Process Modeling,2007,2 (3).
[37]A. Motevali, S. Minaei, M. H. Khoshtaghaza, M. Kazemi, A. Mohamad Nikbakht. Drying of pomegranate arils:Comparison of predictions from mathematical models and neural networks [J]. International Journal of Food Engineering,2010, 6(3).
[38]A. Sander, N. Bolf, J. Prlic Kardum. Research on dynamics and drying time in microwave paper drying [J]. Chemical and Biochemical Engineering Quarterly, 2003,17(2):159-164.
[39]P. Fito and A. Chiralt. The use of water-structure-functionality ensemble in dried food product development [J]. Food Science and Technology International,2003, 9:151-156.
[40]P. Fito, A. Chiralt and J. M. Barat, et al. Vacuum Impregnation for development of new dehydrated products [J]. Journal of Food Engineering,2001,49 (4): 297-302.
[41]J. M. Aguilera, A. Chiralt and P. Fito, et al. Food dehydration and product structure [J]. Trends in Food Science and Technology,2003,14 (10):432-437.
[42]E. M. Torringa, H. H. Nijhuis and H. J. Van Renmmen, et al. Electromagnetic energy in the drying of vegetable produce [A]. Proceedings of the Fourth International Congress on Food Industry [C].Cesme, Turkey,19-23 April 1993: 331-334.
[43]彭金辉,杨显万.微波能技术新应用[M].昆明：云南科技出版社,1997.
[44]焦李成.神经网络系统导论[M].西安：西安电子科技大学出版社,1992.
[45]吴凌云.BP神经网络学习算法的改进及其应用[J].信息技术,2003,35(3)：42-44.
[46]苏高利,邓芳萍.论基于MATLAB语言的BP神经网络的改进算法[J].科技通报,2003,19(2)：130-135.
[47]范磊,张运陶,程正军.基Matlab的改进BP神经网络及其应用[J].西华师范大学学报(自然科学版),2005,26(1)：70-73.
[48]王珏,郭其一,李维刚.基于改进BP神经网络的预测模型及其应用[J].计算机测量与控制,2005,13(1)：39-43.
[49]A. J. Adeloye, A. De Munari. Artificial neural network based generalized storage-yield-reliability models using the Levenberg-Marquardt algorithm [J]. Journal of Hydrology,2006,326 (1-4):215-230.
[50]N. B. Karayiannis, G. W. Mi. Growing radial basis neural networks:Merging supervised and unsupervised learning with network growth techniques [J]. IEEE Transactions Neural Networks,1997,8 (6):1492-1506.
[51]B. G. Kermani, S. S. Schiffman, H. T. Nagle. Performance of the Levenberg-Marquardt neural network training method in electronic nose applications [J]. Sensors and Actuators B:Chemical,2005,110 (1):13-22.
[52]G. Lera, M. Pinzolas. Neighborhood based Levenberg-Marquardt algorithm for neural network training [J]. IEEE Transactions Neural Networks,2002,13 (5): 1200-1203.
[53]H. Mirzaee. Long-term prediction of chaotic time series with multi-step prediction horizons by a neural network with Levenberg-Marquardt learning algorithm [J]. Chaos, Solitons and Fractals,2009,41 (4):1975-1979.
[54]Y. W. Li, Z. T. Yu, X. Y. Meng, W. G. Che and C.1. Mao. Question Classification Based on Incremental Modified Bayes [J]. In Proceedings of the 2008 2nd International Conference on Future Generation Communication and Networking, Hainan Island, China, December 13-15,2008,2:149-152.
[55]L. M. Fu, et al. Incremental back-propagation learning network [J]. IEEE Transactions on Neural Networks,1996,7 (3):757-761.
[56]N. B. Karayiannis and G W. Mi. Growing radial basis neural networks:Merging supervised and unsupervised learning with network growth techniques [J]. IEEE Transactions on Neural Networks,1997,8 (6):1492-1506.
[57]J. Ghosh and A. C. Nag. Knowledge enhancement and reuse with radial basis function networks [C]. Proceedings of the 2002 International Joint Conference on Neural Networks.2002,2:1322-1327.
[58]R. Parekh, J. Yang and V. Honavar. Constructive neural network learning algorithms for pattern classification [J]. IEEE Transactions on Neural Networks, 2000,11 (2):436-451.
[59]J. Zhang and A. J. Morris. A sequential learning approach for single hidden layer neural networks [J]. Neural Networks,1998,11 (1):65-80.
[60]史春朝,张国山.基于改进BP神经网络的PID控制方法研究[J].计算机仿真,2006,12(23)：156-159.
[61]H. L. Chan and A. B. Rad. Real-time flow control using neural networks [J]. ISA Transactions,2000,39 (1):93-101.
[62]L. X. Wang and F. Wan. Structured neural networks for constrained model predictive control [J]. Automatic,2001,37 (8):1235-1243.
[63]S. N. Huang, K. K. Tan and T. H. Lee. Adaptive motion control using neural network approximations [J]. Automatica,2002,38 (2):227-233.
[64]P. M. Makila. On Robustness in Control and LTI Identification:Near-Linearity and Non-Conic Uncertainty [J]. Automatica,2006,42 (4):601-612.
[65]D. S. Chen and R. C. Jain. A robust back propagation learning algorithm for function approximation. IEEE Transactions on Neural Networks,1994,5 (3): 467-479.
[66]A. N. Kolmogorov. On the representation of continuous function of many variables by superposition of continuous functions of one variable and addition [J]. American Mathematical Society Translation.1963,28:55-59.
[67]B. Irie and S. Miyake. Capacity of three-layered perceptrons [J]. Proceedings IEEE International Joint Conference on Neural Networks,1988,1:641-648.
[68]S. M. Carrol and W. Dickinson. Construction of neural nets using random transform [J]. Proceedings International Joint Conference on Neural Networks, 1998,1:607-611.
[69]G Cybenko. Approximation by superpositions of a sigmoidal function [J]. Mathematics of Control, Signal, and Systems,1989,2 (4):303-314.
[70]K. Fumabashi. On the approximate realization of continuous maing by neural networks [J]. International Joint Conference on Neural Networks,1989,2:11-23.
[71]K. Homik. Approximation capabilities of multilayer feed-forward network [J]. Neural Networks,1992,4 (2):251-257.
[72]R. H. Nielsen. Theory of the Back-propagation [J]. Proceedings IEEE International Joint Conference on Neural Networks,1989,1:593-605.
[73]张立明.人工神经网络的模型及其应用[M].上海：复旦大学出版社,1992.
[74]韩立群.人工神经网络理论、设计及应用[M].北京：化学工业出版社,2002.
[75]蒋宗礼.人工神经网络导论[M].高等教育出版社,2001.
[76]黄德双.神经网络模式识别系统理论[M].北京：电子工业出版社,1996.
[77]边肇祺,张学工.模式识别[M].北京：清华大学出版社,2000.
[78]乔双,董智红.BP网络初始值的选取方法[J].东北师大学报自然科学版,2004,36(3)：25-30.
[79]M. Kayama, et al. Modeling Generalization Capacity for a Multi-Layered Neural Network Classifier and Optimizing Its Number of Hidden Units [J]. Proceedings of International Joint Conference on Neural Networks,1992,1:429-434.
[80]E. Watanabe and H. Shimizu. Algorithm for Pruning Hidden Units in Multi-Layered Neural Network for Binary Pattern Classification Problem [J]. Proceedings of International Joint Conference on Neural Network,1993,1: 327-330.
[81]G. Mirchandani. On hidden nodes for neural nets [J]. IEEE Transactions on Circuits and Systems,1989,36 (5):661-664.
[82]S. C. Huang and Y. F. Huang. Bounds on the number of hidden neurons in multilayer perceptron [J]. IEEE Transactions on Neural Networks,1991,3 (1): 47-55.
[83]K. G Mehrotra. Bounds on the number of samples needed for neural learning [J]. IEEE Transactions on Neural Networks,1991,2 (6):548-558.
[84]F. L. Chung and T. Lee. Growth of Back-Propagation Neural Network [J]. Proceedings of International Joint Conference on Neural Network,1992,1: 230-235.
[85]S. E. Fahlman and C. Lebiere. The Cascade-Correction Learning Architecture [J]. Advances in Neural Network Using Processing System,1990,2:524-532.
[86]M. Mzard and J. P. P. Nadal. Learning in Feed-forward Layered Networks:The Tiling Algorithm [J]. Journal of Physics A:Mathematical General,1989,22 (8): 2191-2203.
[87]W. Lin and B. A. Zhang. Best Learning Algorithm of Multi-layer Feed-forward Neural Networks for Associative Memory [J]. Proceedings of International Joint Conference on Intelligent Information Processing and Systems,1992,1:475-478.
[88]王科俊,王克成.神经网络建模、预报与控制[M].哈尔滨：哈尔滨工业大学出版社,1996.
[89]高大启.有教师的线性基本函数前向三层神经网络结构研究[J].计算机学报,1998,21(2)：80-86.
[90]飞思科技研发中心.神经网络理论与MATLAB7实现[M].北京：电子工业出版社,2005.
[91]闻新,周露,李翔MATLAB神经网络仿真与应用[M].北京：科学出版社,2003.
[92]周开利,康耀红.神经网络模型及其Matlab仿真程序设计[M].北京：清华大学出版社,2005.
[93]袁亚湘,孙文瑜.最优化理论与方法[M].北京：科学出版社,1997.
[94]R. A. Jacobs. Increased rates of convergence through learning rate adaptation [J]. Neural Networks,1998,1 (1):295-307.
[95]韩力群.人工神经网络理论、设计及应用[M].北京：化学工业出版社,2007.
[96]林俊,章兢.反向传播神经网络收敛性的探讨[J].计算机与现代化,2005,7：9-12.
[97]M. T. Hagan, H. B. Demuth and M. H. Beale.神经网络设计[M].北京：机械工业出版社,2002.
[98]蔡正国,屈梁生.共轭梯度神经网络的研究[J].西安交通大学学报,1995,29(8)：72-76.
[99]周建华.共轭梯度法在BP网络中的应用[J].计算机工程与应用,1999,17-19.
[100]王永骥,涂健.神经元网络控制[M].北京：机械工业出版社,2003.
[101]哈根,戴葵.神经网络设计[M].北京：机械工业出版社,2002.
[102]S. Vijander, G Indra, H. O. Gupta. ANN-based estimator for distillation using Levenberg-Marquardt approach [J]. Engineering Applications of Artificial Intelligence.2007,20 (2):249-259.
[103]E. M. Bezerra, M. S. Bento, J. A. F. F. Rocco, K. Dia, V. L. Lourenco, L. C. Pardini. Artificial neural network (ANN) prediction of kinetic parameters of (CRFC) composites [J]. Computational Materials Science,2008,44 (2):656-663.
[104]M. A. Costa, A. D. P. Braga, B. R. D. Menezes. Improving generalization of MLPs with sliding mode control and the Levenberg-Marquardt algorithm [J]. Neurocomputing,2007,70 (7-9):1342-1347.
[105]E. D. Ubeyli, I. Guler. Multilayer perceptron neural networks to compute quasistatic parameters of asymmetric coplanar waveguides [J]. Neurocomputing, 2004,62 (1-4):349-365.
[106]G. E. P. Box, K. B. Wilson. On the Experimental Attainment of Optimum Conditions (with Discussion) [J]. Journal of Royal Statistical Society,1951, B13: 1-45.
[107]G E. P. Box, J. S. Hunter. Multifactor Experimental Design for Exploring Response Surface [J]. Annals of Mathematical Statistics,1957,28 (2):195-241.
[108]W. J. Hill, W. G Hunter. A Review of Response Surface Methodology:A Literature Survey [J]. Technometrics,1966,8 (4):571-590.
[109]G E. P Box, N. R. Draper. A Basis for the Selection of a Response Surface Design [J]. Journal of American Statistical Association,1959,54 (6):622-654.
[110]R. H. Myers, D. C. Montgomery. Response Surface Methodology [M]. New York, Wiley and Sons,1995:1-200.
[111]A. I. Khuri, J. A. Cornell. Response Surfaces,2nd Edition [M]. New York, Marcel Dekker, INC.,1996.
[112]V. V. Toropov. Simulation Approach to Structural Optimization [J]. Structural Optimization,1989,1 (1):37-46.
[113]汪仁官.试验设计与分析[M].北京：中国统计出版社,1998.
[114]张润楚,郑海涛,兰燕.试验设计与分析及参数优化[M].北京：中国统计出版社,2003.
[115]N. F. Zainudin, K. T. Lee, A. H. Kamaruddin, S. Bhatia, A. R. Mohamed. Study of absorbent prepared from oil palm ash (OPA) for flue gas desulfurization [J]. Separation and Purification Technology,2005,45 (1):50-60.
[116]S. A. Survase, P. S. Saudagar, R.S. Singhal. Production of scleroglucan from Sclerotium rolfsii MTCC 2156 [J]. Bioresource Technology,2006,98 (7): 1509-1512.
[117]R. Azargohar, A. K. Dalai. Production of activated carbon from Luscar char, Experimental and modelling studies [J]. Microporous and Mesoporous Materials, 2005,85 (3):219-225.
[118]D. Bas, I. H. Boyac. Modeling and optimisation.Ⅱ. Comparison of estimation capabilities of response surface methodology with artificial neural networks in a biochemical reaction [J]. Journal of Food Engineering,2007,78 (3): 846-854.
[119]J. K. Rao, C. H. Kim, S. K. Rhee. Statistical optimization of medium for the production of recombinant hirudin from Saccharomyces cerevisiae using response surface methodology [J] Process Biochemistry,2000,35 (7):639-647.
[120]S. J. Kalil, F. Maugeri, M. I. Rodrigues. Response surface analysis and simulation as a tool for bioprocess design and optimization [J]. Process Biochemistry,2000,35 (6):539-550.
[121]K. M. Desai, B. K. Vaidya, R. S. Singhal, S. S. Bhagwat. Use of an artificial neural network in modeling veast biomass and vield of β-glucan [J]. Process Biochemistry,2005,40 (5):1617-1626.
[122]K. Kavaklioglu, H. Ceylan, H. K. Ozturk, O. E. Canyurt. Modeling and prediction of Turkey's electricity consumption using Artificial Neural Networks [J]. Energy Conversion and Management,2009,50 (11):2719-2727.
[123]A. Azadeh, S. F. Ghaderi, S. Sohrabkhani. A simulated-based neural network algorithm for forecasting electrical energy consumption in Iran [J]. Energy Policy, 2008,36 (7):2637-2644.
[124]A. Azadeh, S. F. Ghaderi, S. Tarverdian, M. Saberi. Integration of artificial neural networks and genetic algorithm to predict electrical energy consumption [J]. IECON Proceedings (Industrial Electronic Conference),2006,32:2552-2557.
[125]王成山,杨军,张崇见.灰色系统理论在城市年用电量预测中的应用一不同预测方法的分析比较[J].电网技术,1999,23(2)：15-18.
[126]张友泉.一种基于灰色系统理论的中长期需电量预测模型[J].电网技术,1999,23(2)：47-50.
[127]陈章湖,熊岗.应用灰色系统原理进行长期电力需求预测[J].系统工程,1994,12(2)：67-71.
[128]施怀瑾,刘敏.灰色改进等维信息GM(1-1)模型在电力系统长期负荷预测中的应用[J].贵州工学院学报,1994,23(2)：43-78.
[129]李艳梅,孙薇.多元线性回归分析在用电量预测中的应用[J].华北电力技术,2003,11：40-41.
[130]曾思勇,杨文君.回归分析法在用电量预测中的应用[J].内蒙古电力技术,1999,17(6)：46-47.
[131]张杰明.试论用多元线性回归法预测城市需水量[J].城市公共事业,2004,18(4)：23-25.
[132]刘志敏.用线性回归法预测南平地区未来的年需电量[J].福建电力与电工,2001,21(1)：46-60.
[133]欧阳昌裕.正确对待电力弹性系数法[J].中国电力企业管理,2000：21-22.
[134]吴熳红,杨继旺.几种电力负荷预测方法及其比较[J].广东电力,2004,17(1)：17-21.
[135]周晖,刘万添.城乡居民生活用电状况的比较分析[J].中国能源,2002,3：34-36.
[136]潘旭红,汪小明.中长期电量的系统动力学方法预测[J].广东电力,2002,15(4)：5-8.
[137]叶舟.上海市三大产业及居民生活电力需求状况及前景分析[J].中国能源,2001,5：28-32.
[138]雷正伟,徐章遂,米东,刘美全.神经网络的预测性能的优化分析[J].计算机测量与控制,2004,12(1)：15-20.
[139]周佃民,管晓宏,孙婕,黄勇.基于神经网络的电力系统短期负荷预测研究[J].电网技术,2002,26(2)：10-18.
[140]张全,刘渺,凌振华,高敏.钢铁企业能耗预测系统的设计[J].冶金动力,2006,2：67-70.
[141]谭显东,胡兆光,李存斌,丁伟,刘达.基于改进BP神经网络的全社会用电量预测模型研究[J].华北电力大学学报,2007,34(3)：85-89.
[142]冯勇.现代计算机控制系统[M].哈尔滨：哈尔滨工业大学出版社,1996.
[143]薛定宇.控制系统计算机辅助设计-MATLAB语言与应用[M].北京：清华大学出版社,2006.
[144]W. K. Ho, C. C. Hang, L. S. Cao. Tuning of PID controllers based on gain and phase margin specifications [J]. Automatica,1995,31 (3):497-502.
[145]W. K. Ho, K. W. Lim, W. Xu. Optimal gain and phase margin tuning for PID controllers [J]. Automatica,1998,34 (8):1009-1014.
[146]W. K. Ho, K. W. Lim, C. C. Hang, L. Y. Ni. Getting more phase margin and performance out of PID controllers [J]. Automatica,1999,35 (9):1579-1585.
[147]王树青.先进控制技术[M].北京：化学工业出版社,2001.
[148]高东杰,谭杰,林红权.应用先进控制技术[M].北京：国防工业出版社,2003.
[149]P. Persson, K. J. Astrom. Dominant pole design-unified view of PID controller tuning [J]. In Preprints 4th EFAC Symposium on Adaptive Systems in Control and Signal Processing, Grenoble, France,1992,4:127-132.
[150]M. Zhuang, D. P. Atherton. Automatic tuning of optimum PID controllers [J]. IEEE Proceedings Control and Application,1993,140 (3):216-224.
[151]F. G Shinseki. Process Control System, Application, Design and Tuning [M]. New York:Mcgraw-Hill,1998.
[152]王伟,张晶涛,柴天佑.PID参数先进整定方法综述[J].自动化学报,2000,26(3)：347-355.
[153]施保华.计算机控制技术[J].华中科技大学出版社,2007.
[154]H. L. Shu, Y. G. Pi. PID neural networks for time-delay systems [J]. Computers and Chemical Engineering,2000,24 (2-7):859-862.
[155]牛建军,吴伟,陈国定.基于神经网络自整定PID控制策略及其仿真[J].系统仿真,2005,17(6)：1425-1427.
[156]M. Barletta, A. Gisario, S. Guarino. Modeling of electrostatic fluidized bed coating process using artificial neural networks [J]. Engineering Applications of Artificial Intelligence,2007,20 (6):721-733.
[157]A. Saenrung, A. Abtahi, A. Zilouchian. Neural network model for a commercial PEM fuel cell system [J]. Journal of Power Sources,2007,172 (2): 749-759.
[158]Y. C. Lin, J. Zhang, J. Zhong. Application of neural network to predict the elevated temperature flow behavior of a low alloy steel [J]. Computational Materials Science,2008,43 (4):752-758.
[159]M. K. Markey, J. Y. Lo, C. E. Floyd. Differences between computer-aided diagnosis of breast masses and that of calcifications [J]. Radiology,2002,223 (2): 489-493.
[160]A. Sharma, R. J. Roy. Design of a recognition system to predict movement during anesthesia [J]. IEEE Transactions on Biomedical Engineering,1997,44 (6): 505-511.
[161]M. H. F. Zarandi, I. B. Turksen, J. Sobhani. Fuzzy polynomial neural networks for approximation of the compressive strength of concrete [J]. Applied Soft Computing,2008,8 (1):488-489.
[162]G. K. F. Tso, K. K. W. Yau. A study of domestic energy usage pattern in Hong Kong [J]. Energy,2003,28 (15):1671-1682.
[163]Y. S. Murat, H. Ceylan. Use of artificial neural networks for transport energy demand modeling [J]. Energy Policy,2006,34 (17):3165-3172.
[164]J. H. Yang, K. W. Song, Y W. Lee, J. H. Kim, K. W. Kang, K. S. Kim, Y. H. Jung. Microwave process for sintering of uranium dioxide [J]. Journal of Nuclear Materials,2004,325 (2-3):210-216.
[165]B. Ayaz, A. N. Bilge. The possible usage of ex-ADU uranium dioxide fuel pellets with low-temperature sintering [J]. Journal of Nuclear Materials,2000,280 (1):45-50.
[166]康仕芳,陈松,赵君.制备近似球形AUC结晶的初步研究[J].原子能科学技术,2001,35(4)：351-355.
[167]裴必富,段德智.ADU工艺及其对产品性能的影响[J].核电工程与技术,1994,7(4)：41-50.
[2]D. O. Hebb. The organization of behavior [M]. New York:Wiley,1949.
[3]F. Rosenblatt. The Perceptron:A probabilistic model for information storage and organization in the brain [J]. Psychological Review,1958,65 (6):386-408.
[4]B. Widrow and M. E. Hoff. Adaptive switching circuit [M]. New York:IRE,1960: 94-104.
[5]M. Minsky and S. Papert. Perceptrons [M]. MIT Press,1969.
[6]J. J. Hopfield. Neural networks and physical systems with emergent collective computational abilities [C]. Proceedings of the National Academy of Science, 1982,79:2554-2558.
[7]D. E. Rumelhart and J. L. McClelland. Parallel Distributed Processing: Explorations in the Microstructure of Cognition [M]. Cambridge MA:MIT Press, 1986.
[8]D. E. Rumelhart, G. E. Hinton and R. J. Williams. Learning representations by back-propagation errors [J]. Nature,1986,323:533-536.
[9]M. T. Leung. Fingerprint processing using back propagation neural networks [J]. International Joint Conference on Neural Networks,1990,90:15-20.
[10]K. Fukshima. Handwritten alphanumeric character recognition by the necognitron [J]. IEEE Transactions on Neural Networks,1991.
[11]S. C. Ahalt. Performance of synthetic neural network classification of noise radar signals [J]. Advances in Neural Information Processing Systems,1989,1: 181-288.
[12]E. Wacholder. A neural network algorithm for the multiple traveling salesman problem [J]. Biology Cybernetics,1989,61 (1):11-19.
[13]K. S. Narendra and K. Parthasarathy. Identification and control of dynamical systems using neural networks [J]. IEEE Transactions on Neural Networks,1990, 1(1):4-27.
[14]K. S. Narendra and K. Parthasarathy. Gradient methods for the optimization of dynamical systems containing neural networks [J]. IEEE Transactions on Neural Networks,1991,2 (2):252-262.
昆明理工大学博士学位论文 参考文献
[15]雷蕴奇.神经元自适应控制的概念[C].中国神经网络首届学术大会论文集,1990.
[16]A. Guez, J. Eilbert and M. Kam. Neuromorphic architecture for adaptive robot control:A preliminary analysis Proc [C]. IJCNN,1987,4:567-572.
[17]沈清,胡德文.神经网络应用技术[M].国防科技大学出版社,1993.
[18]N. V. Bhat. Modeling chemical process systems via neural computation [J]. IEEE Control Systems Magazine,1990,10 (3):24-30.
[19]A. Namphol. Higher order data compression with neural network [C]. International Joint Conference on Neural Networks,1991,1:55-59.
[20]J. D. Daugman. Complete discrete 2-D Gabor transforms by neural network for image analysis and compression [J]. IEEE Transactions on ASSP,1998,36 (7): 1169-1179.
[21]B. Kosko. Adaptive bi-directional associative memories [M]. Applied Optics, 1982.
[22]J. D. Provence. Neural network implementation for maximum-likelihood sequence estimation of binary signals in Gaussion noise [C]. Proceedings of the IEEE International Joint Conference on Neural Networks,1987,3.
[23]Y. T. Zhou. Image restoration using a neural network [J]. IEEE Transactions on ASSP.1988,36(7):1141-1151.
[24]S. U. H. Qureshi. Adaptive equalization [C]. Proceedings of IEEE,1985,73: 1349-1387.
[25]R. P. Lippmann and P. Bexkman. Adaptive neural net preprocessing for signal diction in non-Gaussion noise [M]. Advances in Neural Information Processing Systems,1989,1.
[26]B. Widrow and R. Winter Neural nets for adaptive filtering and adaptive pattern recognition [J]. Computer,1988,21 (3):25-29.
[27]Q. Z. Xue. Neural network based adaptive matched filtering for QRS detection [J]. IEEE Transactions on Biomedical Engineering,1992,39 (4):317-329.
[28]Q. Zhang. Adaptive equalization using the back-propagation algorithm [J]. IEEE Transactions on CAS,1990,37 (6):848-849.
[29]朱代根.微波干燥过程的神经网络模型[J].农机化研究,2010,6(6)：94-96.
[30]I. Alibas Ozkan, B. Akbudak, N. Akbudak. Microwave drying characteristics of spinach [J]. Journal of Food Engineering,2007,78 (2):577-583.
[31]许晓春,林朝朋.佛手瓜的微波真空干燥及其神经网络模拟的研究[J].韶关学院学报(自然科学),2007,28(12)：80-84.
[32]L. Momenzadeh, A. Zomorodian, D. Mowla. Experimental and theoretical investigation of shelled corn drying in a microwave-assisted fluidized bed dryer using Artificial Neural Network [J]. Food and Bioproducts Processing,2011,89 (1):15-21.
[33]J. L. Pedreno Molina, J. Monzo Cabrera, A. Toledo Moreo, D. Sanchez Hernandez. A novel predictive architecture for microwave-assisted drying processes based on neural networks [J]. International Communications in Heat and Mass Transfer,2005,32 (8):1026-1033.
[34]J. L. Pedreno Molina, J. Monzo Cabrera, D. Sanchez Hernandez. A new predictive neural architecture for solving temperature inverse problems in microwave-assisted drying process [J]. Neurocomputing,2005,64:521-528.
[35]P. Poonnoy, A. Tansakul, M. Chinnan. Estimation of moisture ratio of a mushroom undergoing microwave-vacuum drying using artificial neural network and regression models [J]. Chemical Product and Process Modeling,2007,2 (3).
[36]P. Poonnoy, A. Tansakul, M. Chinnan. Parallel dynamic artificial neural network for temperature and moisture content predictions in microwave-vacuum dried tomato slices [J]. Chemical Product and Process Modeling,2007,2 (3).
[37]A. Motevali, S. Minaei, M. H. Khoshtaghaza, M. Kazemi, A. Mohamad Nikbakht. Drying of pomegranate arils:Comparison of predictions from mathematical models and neural networks [J]. International Journal of Food Engineering,2010, 6(3).
[38]A. Sander, N. Bolf, J. Prlic Kardum. Research on dynamics and drying time in microwave paper drying [J]. Chemical and Biochemical Engineering Quarterly, 2003,17(2):159-164.
[39]P. Fito and A. Chiralt. The use of water-structure-functionality ensemble in dried food product development [J]. Food Science and Technology International,2003, 9:151-156.
[40]P. Fito, A. Chiralt and J. M. Barat, et al. Vacuum Impregnation for development of new dehydrated products [J]. Journal of Food Engineering,2001,49 (4): 297-302.
[41]J. M. Aguilera, A. Chiralt and P. Fito, et al. Food dehydration and product structure [J]. Trends in Food Science and Technology,2003,14 (10):432-437.
[42]E. M. Torringa, H. H. Nijhuis and H. J. Van Renmmen, et al. Electromagnetic energy in the drying of vegetable produce [A]. Proceedings of the Fourth International Congress on Food Industry [C].Cesme, Turkey,19-23 April 1993: 331-334.
[43]彭金辉,杨显万.微波能技术新应用[M].昆明：云南科技出版社,1997.
[44]焦李成.神经网络系统导论[M].西安：西安电子科技大学出版社,1992.
[45]吴凌云.BP神经网络学习算法的改进及其应用[J].信息技术,2003,35(3)：42-44.
[46]苏高利,邓芳萍.论基于MATLAB语言的BP神经网络的改进算法[J].科技通报,2003,19(2)：130-135.
[47]范磊,张运陶,程正军.基Matlab的改进BP神经网络及其应用[J].西华师范大学学报(自然科学版),2005,26(1)：70-73.
[48]王珏,郭其一,李维刚.基于改进BP神经网络的预测模型及其应用[J].计算机测量与控制,2005,13(1)：39-43.
[49]A. J. Adeloye, A. De Munari. Artificial neural network based generalized storage-yield-reliability models using the Levenberg-Marquardt algorithm [J]. Journal of Hydrology,2006,326 (1-4):215-230.
[50]N. B. Karayiannis, G. W. Mi. Growing radial basis neural networks:Merging supervised and unsupervised learning with network growth techniques [J]. IEEE Transactions Neural Networks,1997,8 (6):1492-1506.
[51]B. G. Kermani, S. S. Schiffman, H. T. Nagle. Performance of the Levenberg-Marquardt neural network training method in electronic nose applications [J]. Sensors and Actuators B:Chemical,2005,110 (1):13-22.
[52]G. Lera, M. Pinzolas. Neighborhood based Levenberg-Marquardt algorithm for neural network training [J]. IEEE Transactions Neural Networks,2002,13 (5): 1200-1203.
[53]H. Mirzaee. Long-term prediction of chaotic time series with multi-step prediction horizons by a neural network with Levenberg-Marquardt learning algorithm [J]. Chaos, Solitons and Fractals,2009,41 (4):1975-1979.
[54]Y. W. Li, Z. T. Yu, X. Y. Meng, W. G. Che and C.1. Mao. Question Classification Based on Incremental Modified Bayes [J]. In Proceedings of the 2008 2nd International Conference on Future Generation Communication and Networking, Hainan Island, China, December 13-15,2008,2:149-152.
[55]L. M. Fu, et al. Incremental back-propagation learning network [J]. IEEE Transactions on Neural Networks,1996,7 (3):757-761.
[56]N. B. Karayiannis and G W. Mi. Growing radial basis neural networks:Merging supervised and unsupervised learning with network growth techniques [J]. IEEE Transactions on Neural Networks,1997,8 (6):1492-1506.
[57]J. Ghosh and A. C. Nag. Knowledge enhancement and reuse with radial basis function networks [C]. Proceedings of the 2002 International Joint Conference on Neural Networks.2002,2:1322-1327.
[58]R. Parekh, J. Yang and V. Honavar. Constructive neural network learning algorithms for pattern classification [J]. IEEE Transactions on Neural Networks, 2000,11 (2):436-451.
[59]J. Zhang and A. J. Morris. A sequential learning approach for single hidden layer neural networks [J]. Neural Networks,1998,11 (1):65-80.
[60]史春朝,张国山.基于改进BP神经网络的PID控制方法研究[J].计算机仿真,2006,12(23)：156-159.
[61]H. L. Chan and A. B. Rad. Real-time flow control using neural networks [J]. ISA Transactions,2000,39 (1):93-101.
[62]L. X. Wang and F. Wan. Structured neural networks for constrained model predictive control [J]. Automatic,2001,37 (8):1235-1243.
[63]S. N. Huang, K. K. Tan and T. H. Lee. Adaptive motion control using neural network approximations [J]. Automatica,2002,38 (2):227-233.
[64]P. M. Makila. On Robustness in Control and LTI Identification:Near-Linearity and Non-Conic Uncertainty [J]. Automatica,2006,42 (4):601-612.
[65]D. S. Chen and R. C. Jain. A robust back propagation learning algorithm for function approximation. IEEE Transactions on Neural Networks,1994,5 (3): 467-479.
[66]A. N. Kolmogorov. On the representation of continuous function of many variables by superposition of continuous functions of one variable and addition [J]. American Mathematical Society Translation.1963,28:55-59.
[67]B. Irie and S. Miyake. Capacity of three-layered perceptrons [J]. Proceedings IEEE International Joint Conference on Neural Networks,1988,1:641-648.
[68]S. M. Carrol and W. Dickinson. Construction of neural nets using random transform [J]. Proceedings International Joint Conference on Neural Networks, 1998,1:607-611.
[69]G Cybenko. Approximation by superpositions of a sigmoidal function [J]. Mathematics of Control, Signal, and Systems,1989,2 (4):303-314.
[70]K. Fumabashi. On the approximate realization of continuous maing by neural networks [J]. International Joint Conference on Neural Networks,1989,2:11-23.
[71]K. Homik. Approximation capabilities of multilayer feed-forward network [J]. Neural Networks,1992,4 (2):251-257.
[72]R. H. Nielsen. Theory of the Back-propagation [J]. Proceedings IEEE International Joint Conference on Neural Networks,1989,1:593-605.
[73]张立明.人工神经网络的模型及其应用[M].上海：复旦大学出版社,1992.
[74]韩立群.人工神经网络理论、设计及应用[M].北京：化学工业出版社,2002.
[75]蒋宗礼.人工神经网络导论[M].高等教育出版社,2001.
[76]黄德双.神经网络模式识别系统理论[M].北京：电子工业出版社,1996.
[77]边肇祺,张学工.模式识别[M].北京：清华大学出版社,2000.
[78]乔双,董智红.BP网络初始值的选取方法[J].东北师大学报自然科学版,2004,36(3)：25-30.
[79]M. Kayama, et al. Modeling Generalization Capacity for a Multi-Layered Neural Network Classifier and Optimizing Its Number of Hidden Units [J]. Proceedings of International Joint Conference on Neural Networks,1992,1:429-434.
[80]E. Watanabe and H. Shimizu. Algorithm for Pruning Hidden Units in Multi-Layered Neural Network for Binary Pattern Classification Problem [J]. Proceedings of International Joint Conference on Neural Network,1993,1: 327-330.
[81]G. Mirchandani. On hidden nodes for neural nets [J]. IEEE Transactions on Circuits and Systems,1989,36 (5):661-664.
[82]S. C. Huang and Y. F. Huang. Bounds on the number of hidden neurons in multilayer perceptron [J]. IEEE Transactions on Neural Networks,1991,3 (1): 47-55.
[83]K. G Mehrotra. Bounds on the number of samples needed for neural learning [J]. IEEE Transactions on Neural Networks,1991,2 (6):548-558.
[84]F. L. Chung and T. Lee. Growth of Back-Propagation Neural Network [J]. Proceedings of International Joint Conference on Neural Network,1992,1: 230-235.
[85]S. E. Fahlman and C. Lebiere. The Cascade-Correction Learning Architecture [J]. Advances in Neural Network Using Processing System,1990,2:524-532.
[86]M. Mzard and J. P. P. Nadal. Learning in Feed-forward Layered Networks:The Tiling Algorithm [J]. Journal of Physics A:Mathematical General,1989,22 (8): 2191-2203.
[87]W. Lin and B. A. Zhang. Best Learning Algorithm of Multi-layer Feed-forward Neural Networks for Associative Memory [J]. Proceedings of International Joint Conference on Intelligent Information Processing and Systems,1992,1:475-478.
[88]王科俊,王克成.神经网络建模、预报与控制[M].哈尔滨：哈尔滨工业大学出版社,1996.
[89]高大启.有教师的线性基本函数前向三层神经网络结构研究[J].计算机学报,1998,21(2)：80-86.
[90]飞思科技研发中心.神经网络理论与MATLAB7实现[M].北京：电子工业出版社,2005.
[91]闻新,周露,李翔MATLAB神经网络仿真与应用[M].北京：科学出版社,2003.
[92]周开利,康耀红.神经网络模型及其Matlab仿真程序设计[M].北京：清华大学出版社,2005.
[93]袁亚湘,孙文瑜.最优化理论与方法[M].北京：科学出版社,1997.
[94]R. A. Jacobs. Increased rates of convergence through learning rate adaptation [J]. Neural Networks,1998,1 (1):295-307.
[95]韩力群.人工神经网络理论、设计及应用[M].北京：化学工业出版社,2007.
[96]林俊,章兢.反向传播神经网络收敛性的探讨[J].计算机与现代化,2005,7：9-12.
[97]M. T. Hagan, H. B. Demuth and M. H. Beale.神经网络设计[M].北京：机械工业出版社,2002.
[98]蔡正国,屈梁生.共轭梯度神经网络的研究[J].西安交通大学学报,1995,29(8)：72-76.
[99]周建华.共轭梯度法在BP网络中的应用[J].计算机工程与应用,1999,17-19.
[100]王永骥,涂健.神经元网络控制[M].北京：机械工业出版社,2003.
[101]哈根,戴葵.神经网络设计[M].北京：机械工业出版社,2002.
[102]S. Vijander, G Indra, H. O. Gupta. ANN-based estimator for distillation using Levenberg-Marquardt approach [J]. Engineering Applications of Artificial Intelligence.2007,20 (2):249-259.
[103]E. M. Bezerra, M. S. Bento, J. A. F. F. Rocco, K. Dia, V. L. Lourenco, L. C. Pardini. Artificial neural network (ANN) prediction of kinetic parameters of (CRFC) composites [J]. Computational Materials Science,2008,44 (2):656-663.
[104]M. A. Costa, A. D. P. Braga, B. R. D. Menezes. Improving generalization of MLPs with sliding mode control and the Levenberg-Marquardt algorithm [J]. Neurocomputing,2007,70 (7-9):1342-1347.
[105]E. D. Ubeyli, I. Guler. Multilayer perceptron neural networks to compute quasistatic parameters of asymmetric coplanar waveguides [J]. Neurocomputing, 2004,62 (1-4):349-365.
[106]G. E. P. Box, K. B. Wilson. On the Experimental Attainment of Optimum Conditions (with Discussion) [J]. Journal of Royal Statistical Society,1951, B13: 1-45.
[107]G E. P. Box, J. S. Hunter. Multifactor Experimental Design for Exploring Response Surface [J]. Annals of Mathematical Statistics,1957,28 (2):195-241.
[108]W. J. Hill, W. G Hunter. A Review of Response Surface Methodology:A Literature Survey [J]. Technometrics,1966,8 (4):571-590.
[109]G E. P Box, N. R. Draper. A Basis for the Selection of a Response Surface Design [J]. Journal of American Statistical Association,1959,54 (6):622-654.
[110]R. H. Myers, D. C. Montgomery. Response Surface Methodology [M]. New York, Wiley and Sons,1995:1-200.
[111]A. I. Khuri, J. A. Cornell. Response Surfaces,2nd Edition [M]. New York, Marcel Dekker, INC.,1996.
[112]V. V. Toropov. Simulation Approach to Structural Optimization [J]. Structural Optimization,1989,1 (1):37-46.
[113]汪仁官.试验设计与分析[M].北京：中国统计出版社,1998.
[114]张润楚,郑海涛,兰燕.试验设计与分析及参数优化[M].北京：中国统计出版社,2003.
[115]N. F. Zainudin, K. T. Lee, A. H. Kamaruddin, S. Bhatia, A. R. Mohamed. Study of absorbent prepared from oil palm ash (OPA) for flue gas desulfurization [J]. Separation and Purification Technology,2005,45 (1):50-60.
[116]S. A. Survase, P. S. Saudagar, R.S. Singhal. Production of scleroglucan from Sclerotium rolfsii MTCC 2156 [J]. Bioresource Technology,2006,98 (7): 1509-1512.
[117]R. Azargohar, A. K. Dalai. Production of activated carbon from Luscar char, Experimental and modelling studies [J]. Microporous and Mesoporous Materials, 2005,85 (3):219-225.
[118]D. Bas, I. H. Boyac. Modeling and optimisation.Ⅱ. Comparison of estimation capabilities of response surface methodology with artificial neural networks in a biochemical reaction [J]. Journal of Food Engineering,2007,78 (3): 846-854.
[119]J. K. Rao, C. H. Kim, S. K. Rhee. Statistical optimization of medium for the production of recombinant hirudin from Saccharomyces cerevisiae using response surface methodology [J] Process Biochemistry,2000,35 (7):639-647.
[120]S. J. Kalil, F. Maugeri, M. I. Rodrigues. Response surface analysis and simulation as a tool for bioprocess design and optimization [J]. Process Biochemistry,2000,35 (6):539-550.
[121]K. M. Desai, B. K. Vaidya, R. S. Singhal, S. S. Bhagwat. Use of an artificial neural network in modeling veast biomass and vield of β-glucan [J]. Process Biochemistry,2005,40 (5):1617-1626.
[122]K. Kavaklioglu, H. Ceylan, H. K. Ozturk, O. E. Canyurt. Modeling and prediction of Turkey's electricity consumption using Artificial Neural Networks [J]. Energy Conversion and Management,2009,50 (11):2719-2727.
[123]A. Azadeh, S. F. Ghaderi, S. Sohrabkhani. A simulated-based neural network algorithm for forecasting electrical energy consumption in Iran [J]. Energy Policy, 2008,36 (7):2637-2644.
[124]A. Azadeh, S. F. Ghaderi, S. Tarverdian, M. Saberi. Integration of artificial neural networks and genetic algorithm to predict electrical energy consumption [J]. IECON Proceedings (Industrial Electronic Conference),2006,32:2552-2557.
[125]王成山,杨军,张崇见.灰色系统理论在城市年用电量预测中的应用一不同预测方法的分析比较[J].电网技术,1999,23(2)：15-18.
[126]张友泉.一种基于灰色系统理论的中长期需电量预测模型[J].电网技术,1999,23(2)：47-50.
[127]陈章湖,熊岗.应用灰色系统原理进行长期电力需求预测[J].系统工程,1994,12(2)：67-71.
[128]施怀瑾,刘敏.灰色改进等维信息GM(1-1)模型在电力系统长期负荷预测中的应用[J].贵州工学院学报,1994,23(2)：43-78.
[129]李艳梅,孙薇.多元线性回归分析在用电量预测中的应用[J].华北电力技术,2003,11：40-41.
[130]曾思勇,杨文君.回归分析法在用电量预测中的应用[J].内蒙古电力技术,1999,17(6)：46-47.
[131]张杰明.试论用多元线性回归法预测城市需水量[J].城市公共事业,2004,18(4)：23-25.
[132]刘志敏.用线性回归法预测南平地区未来的年需电量[J].福建电力与电工,2001,21(1)：46-60.
[133]欧阳昌裕.正确对待电力弹性系数法[J].中国电力企业管理,2000：21-22.
[134]吴熳红,杨继旺.几种电力负荷预测方法及其比较[J].广东电力,2004,17(1)：17-21.
[135]周晖,刘万添.城乡居民生活用电状况的比较分析[J].中国能源,2002,3：34-36.
[136]潘旭红,汪小明.中长期电量的系统动力学方法预测[J].广东电力,2002,15(4)：5-8.
[137]叶舟.上海市三大产业及居民生活电力需求状况及前景分析[J].中国能源,2001,5：28-32.
[138]雷正伟,徐章遂,米东,刘美全.神经网络的预测性能的优化分析[J].计算机测量与控制,2004,12(1)：15-20.
[139]周佃民,管晓宏,孙婕,黄勇.基于神经网络的电力系统短期负荷预测研究[J].电网技术,2002,26(2)：10-18.
[140]张全,刘渺,凌振华,高敏.钢铁企业能耗预测系统的设计[J].冶金动力,2006,2：67-70.
[141]谭显东,胡兆光,李存斌,丁伟,刘达.基于改进BP神经网络的全社会用电量预测模型研究[J].华北电力大学学报,2007,34(3)：85-89.
[142]冯勇.现代计算机控制系统[M].哈尔滨：哈尔滨工业大学出版社,1996.
[143]薛定宇.控制系统计算机辅助设计-MATLAB语言与应用[M].北京：清华大学出版社,2006.
[144]W. K. Ho, C. C. Hang, L. S. Cao. Tuning of PID controllers based on gain and phase margin specifications [J]. Automatica,1995,31 (3):497-502.
[145]W. K. Ho, K. W. Lim, W. Xu. Optimal gain and phase margin tuning for PID controllers [J]. Automatica,1998,34 (8):1009-1014.
[146]W. K. Ho, K. W. Lim, C. C. Hang, L. Y. Ni. Getting more phase margin and performance out of PID controllers [J]. Automatica,1999,35 (9):1579-1585.
[147]王树青.先进控制技术[M].北京：化学工业出版社,2001.
[148]高东杰,谭杰,林红权.应用先进控制技术[M].北京：国防工业出版社,2003.
[149]P. Persson, K. J. Astrom. Dominant pole design-unified view of PID controller tuning [J]. In Preprints 4th EFAC Symposium on Adaptive Systems in Control and Signal Processing, Grenoble, France,1992,4:127-132.
[150]M. Zhuang, D. P. Atherton. Automatic tuning of optimum PID controllers [J]. IEEE Proceedings Control and Application,1993,140 (3):216-224.
[151]F. G Shinseki. Process Control System, Application, Design and Tuning [M]. New York:Mcgraw-Hill,1998.
[152]王伟,张晶涛,柴天佑.PID参数先进整定方法综述[J].自动化学报,2000,26(3)：347-355.
[153]施保华.计算机控制技术[J].华中科技大学出版社,2007.
[154]H. L. Shu, Y. G. Pi. PID neural networks for time-delay systems [J]. Computers and Chemical Engineering,2000,24 (2-7):859-862.
[155]牛建军,吴伟,陈国定.基于神经网络自整定PID控制策略及其仿真[J].系统仿真,2005,17(6)：1425-1427.
[156]M. Barletta, A. Gisario, S. Guarino. Modeling of electrostatic fluidized bed coating process using artificial neural networks [J]. Engineering Applications of Artificial Intelligence,2007,20 (6):721-733.
[157]A. Saenrung, A. Abtahi, A. Zilouchian. Neural network model for a commercial PEM fuel cell system [J]. Journal of Power Sources,2007,172 (2): 749-759.
[158]Y. C. Lin, J. Zhang, J. Zhong. Application of neural network to predict the elevated temperature flow behavior of a low alloy steel [J]. Computational Materials Science,2008,43 (4):752-758.
[159]M. K. Markey, J. Y. Lo, C. E. Floyd. Differences between computer-aided diagnosis of breast masses and that of calcifications [J]. Radiology,2002,223 (2): 489-493.
[160]A. Sharma, R. J. Roy. Design of a recognition system to predict movement during anesthesia [J]. IEEE Transactions on Biomedical Engineering,1997,44 (6): 505-511.
[161]M. H. F. Zarandi, I. B. Turksen, J. Sobhani. Fuzzy polynomial neural networks for approximation of the compressive strength of concrete [J]. Applied Soft Computing,2008,8 (1):488-489.
[162]G. K. F. Tso, K. K. W. Yau. A study of domestic energy usage pattern in Hong Kong [J]. Energy,2003,28 (15):1671-1682.
[163]Y. S. Murat, H. Ceylan. Use of artificial neural networks for transport energy demand modeling [J]. Energy Policy,2006,34 (17):3165-3172.
[164]J. H. Yang, K. W. Song, Y W. Lee, J. H. Kim, K. W. Kang, K. S. Kim, Y. H. Jung. Microwave process for sintering of uranium dioxide [J]. Journal of Nuclear Materials,2004,325 (2-3):210-216.
[165]B. Ayaz, A. N. Bilge. The possible usage of ex-ADU uranium dioxide fuel pellets with low-temperature sintering [J]. Journal of Nuclear Materials,2000,280 (1):45-50.
[166]康仕芳,陈松,赵君.制备近似球形AUC结晶的初步研究[J].原子能科学技术,2001,35(4)：351-355.
[167]裴必富,段德智.ADU工艺及其对产品性能的影响[J].核电工程与技术,1994,7(4)：41-50.