遗传小波神经网络在机床碳排放预测中的应用
|
摘要
机床的生产加工过程中,会产生大量的碳排放,通过分析机床加工过程的碳排放相关量,预测碳排放值,从而达到降低碳排放的目的;将遗传算法对具有自适应性和函数逼近能力的小波神经网络的参数进行全局优化,来构建遗传小波神经网络模型,对机床加工过程的碳排放进行预测;并通过实验数据将遗传小波神经网络与传统小波神经网络的预测结果进行对比,结果显示,优化后的小波神经网络在机床碳排放的预测结果平均误差为0.48%,均方误差为20.5303,均优于传统神经网络,证实了在机床碳排放预测中遗传小波神经网络相对传统神经网络具有更高的逼近精度;从而能够较为准确地对机床碳排放进行预测和控制。
The process of machine production would produce a large number of carbon emissions. Forecasting carbon emissions by analyzing the value of carbon emissions related in the machining process,to achieve the purpose of reducing carbon emissions. The genetic algorithm does a global optimization on the parameters of the wavelet neural network which is adaptive and have a functional approximation,in order to construct a genetic wavelet neural network model and predict the carbon emissions in machining process. The experimental data showed:the prediction in machining carbon emission to the genetic wavelet neural network compared with the traditional wavelet neural network,the average error is 0.48% for the former,and the mean square error is 20.5303,are better than the traditional neural network. It confirms that the genetic wavelet neural network in machining carbon emission has higher approximation accuracy compared with the traditional neural network relatively;so that it does a better prediction and control on carbon emissions of machining.
The process of machine production would produce a large number of carbon emissions. Forecasting carbon emissions by analyzing the value of carbon emissions related in the machining process,to achieve the purpose of reducing carbon emissions. The genetic algorithm does a global optimization on the parameters of the wavelet neural network which is adaptive and have a functional approximation,in order to construct a genetic wavelet neural network model and predict the carbon emissions in machining process. The experimental data showed:the prediction in machining carbon emission to the genetic wavelet neural network compared with the traditional wavelet neural network,the average error is 0.48% for the former,and the mean square error is 20.5303,are better than the traditional neural network. It confirms that the genetic wavelet neural network in machining carbon emission has higher approximation accuracy compared with the traditional neural network relatively;so that it does a better prediction and control on carbon emissions of machining.
引文
[1]杜轻.基于改进灰色预测模型的分数线预测算法研究[D].天津:河北工业大学,2011.(Du Qing.Research of cutting score forecasting algorithm based on improved grey model[D].Tianjin:Hebei University of Technology,2011.)
[2]曹华军,李洪丞,宋胜利.基于生命周期评价的机床生命周期碳排放评估方法及应用[J].计算机集成制造系统,2011(11):2432-2437.(Cao Hua-jun,Li Hong-cheng,Song Sheng-li.Evaluation method and application for carbon emissions of machine tool based on life cycle assessment[J].Computer Integrated Manufacturing System,2011(11):2432-2437.)
[3]张惠萍,高栋.数控铣削加工过程碳排放量影响因素的分析[J].机械制造与自动化,2013,42(2):29-31.DOI:10.3969.(Zhang Hui-ping,Gao Dong.Analysis of influence factors of carbon dioxide emission during milling processes[J].Machine Building&Automation,2013,42(2):29-31.DOI:10.3969.)
[4]尹瑞雪.机械制造过程能耗评估模型及其在工艺规划中的应用[J].机械设计与制造,2014(9):270-272.(Yin Rui-xue.Energy consumption evalution model of the machining processes and its application in process planning[J].Machinery Design&Manufacture,2014(9):270-272.)
[5]奚婷婷,熊伟丽,张林.基于D-S算法的小波理论在温室控制中的应用研究[J].传感器与微系统,2008(11):32-34+37.(Xi Ting-ting,Xiong Wei-li,Zhang Lin.Application research of wavelet theory based on D-S algorithm in greenhouse control[J].Transducer and Micro system Technologies,2008(11):32-34+37.)
[6]张加云,张德江,李新胜.遗传小波神经网络在钢铁企业能耗预测中的应用[J].冶金自动化,2009,33(z1):849-851.(Zhang Jia-yun,Zhang De-jiang,Li Xin-sheng.Genetic algorithm optimization wavelet neural network applications in the iron and steel enterprise energy forecast[J].Metallurgical Industry Automation,2009,33(z1):849-851.)
[7]崔龙国.面向低碳制造的机械加工系统工艺优化模型及方法研究[D].重庆:重庆大学,2013.(Cui Long-guo.Research on low carbon manufacturing oriented machining process optimization model and method[D].Chongqing:Chongqing University,2013.)
[8]刘伟,邓朝晖,万林林.基于正交试验-遗传神经网络的陶瓷球面精密磨削参数优化[J].中国机械工程,2014(4):451-455.(Liu Wei,Deng Zhao-hui,Wan Lin-lin.Parameter optimization on precision grinding of ceramic sphere using orthogonal experiment and genetic neural network[J].China Mechanical Engineering,2014(4):451-455.)
[9]贾明灿.基于隐患分析的煤矿安全评价算法研究[D].邯郸:河北工程大学,2014.(Jia Ming-can.The algorithm research of coal mine safety evaluation basedonhiddendangeranalysis[D].Handan:Hebei Universityof Engineering,2014.)
[10]黄敏,王建辉,顾树生.基于遗传小波神经网络的冷轧轧制力预报研究[J].控制与决策,2004(10):1129-1132.(Huang Min,Wang Jian-hui,Gu Shu-sheng.Study on cold mill rolling force prediction based on wavelet neural network with genetic algorithm[J].Control and Decision,2004(10):1129-1132.)
[2]曹华军,李洪丞,宋胜利.基于生命周期评价的机床生命周期碳排放评估方法及应用[J].计算机集成制造系统,2011(11):2432-2437.(Cao Hua-jun,Li Hong-cheng,Song Sheng-li.Evaluation method and application for carbon emissions of machine tool based on life cycle assessment[J].Computer Integrated Manufacturing System,2011(11):2432-2437.)
[3]张惠萍,高栋.数控铣削加工过程碳排放量影响因素的分析[J].机械制造与自动化,2013,42(2):29-31.DOI:10.3969.(Zhang Hui-ping,Gao Dong.Analysis of influence factors of carbon dioxide emission during milling processes[J].Machine Building&Automation,2013,42(2):29-31.DOI:10.3969.)
[4]尹瑞雪.机械制造过程能耗评估模型及其在工艺规划中的应用[J].机械设计与制造,2014(9):270-272.(Yin Rui-xue.Energy consumption evalution model of the machining processes and its application in process planning[J].Machinery Design&Manufacture,2014(9):270-272.)
[5]奚婷婷,熊伟丽,张林.基于D-S算法的小波理论在温室控制中的应用研究[J].传感器与微系统,2008(11):32-34+37.(Xi Ting-ting,Xiong Wei-li,Zhang Lin.Application research of wavelet theory based on D-S algorithm in greenhouse control[J].Transducer and Micro system Technologies,2008(11):32-34+37.)
[6]张加云,张德江,李新胜.遗传小波神经网络在钢铁企业能耗预测中的应用[J].冶金自动化,2009,33(z1):849-851.(Zhang Jia-yun,Zhang De-jiang,Li Xin-sheng.Genetic algorithm optimization wavelet neural network applications in the iron and steel enterprise energy forecast[J].Metallurgical Industry Automation,2009,33(z1):849-851.)
[7]崔龙国.面向低碳制造的机械加工系统工艺优化模型及方法研究[D].重庆:重庆大学,2013.(Cui Long-guo.Research on low carbon manufacturing oriented machining process optimization model and method[D].Chongqing:Chongqing University,2013.)
[8]刘伟,邓朝晖,万林林.基于正交试验-遗传神经网络的陶瓷球面精密磨削参数优化[J].中国机械工程,2014(4):451-455.(Liu Wei,Deng Zhao-hui,Wan Lin-lin.Parameter optimization on precision grinding of ceramic sphere using orthogonal experiment and genetic neural network[J].China Mechanical Engineering,2014(4):451-455.)
[9]贾明灿.基于隐患分析的煤矿安全评价算法研究[D].邯郸:河北工程大学,2014.(Jia Ming-can.The algorithm research of coal mine safety evaluation basedonhiddendangeranalysis[D].Handan:Hebei Universityof Engineering,2014.)
[10]黄敏,王建辉,顾树生.基于遗传小波神经网络的冷轧轧制力预报研究[J].控制与决策,2004(10):1129-1132.(Huang Min,Wang Jian-hui,Gu Shu-sheng.Study on cold mill rolling force prediction based on wavelet neural network with genetic algorithm[J].Control and Decision,2004(10):1129-1132.)