改进非等间距GM(1,1)-BP模型的导弹退化状态预测
|
摘要
为提高导弹退化状态预测的精度,结合导弹测试数据不等时间间隔的特点,提出了一种基于改进非等间距GM(1,1)-BP模型的导弹退化状态预测方法。对传统非等间距GM(1,1)模型的背景值和初始条件进行优化,引入新陈代谢思想,在此基础上,构造灰色模型拟合值与实际值的差值序列,进而建立差值序列的BP神经网络预测模型,还原得到最终预测值,提高了预测精度。此设计方法结合了灰色模型对趋向性数据的预测优势和BP神经网络强大的非线性拟合能力,达到了取长补短、相得益彰的效果。通过导弹测试数据的预测实例,验证了方法的有效性和优越性。
In order to improve the accuracy of missile degradation state prediction,combined with the unequal interval test data of missile,a prediction method of missile degradation state based on improved unequal interval GM( 1,1) BP model is proposed. The background value and initial condition of the traditional unequal interval GM( 1,1) model are both optimized and the metabolic thought is also introduced. Based on this,the difference sequence between the fitting value of the grey model and the actual value is thus constructed. BP neural network prediction model of the difference sequence is then established,and the final prediction value is restored to improve the prediction accuracy. The designed method combines the predictive advantage of grey model to the trend data and the strong nonlinear fitting ability of BP neural network,and achieves the effect of complement each other. Prediction example of the missile test data verifies the validity and superiority of the method.
In order to improve the accuracy of missile degradation state prediction,combined with the unequal interval test data of missile,a prediction method of missile degradation state based on improved unequal interval GM( 1,1) BP model is proposed. The background value and initial condition of the traditional unequal interval GM( 1,1) model are both optimized and the metabolic thought is also introduced. Based on this,the difference sequence between the fitting value of the grey model and the actual value is thus constructed. BP neural network prediction model of the difference sequence is then established,and the final prediction value is restored to improve the prediction accuracy. The designed method combines the predictive advantage of grey model to the trend data and the strong nonlinear fitting ability of BP neural network,and achieves the effect of complement each other. Prediction example of the missile test data verifies the validity and superiority of the method.
引文
[1]徐廷学,丛林虎,董琪.基于UGM-ULSSVM的导弹制导系统状态预测方法[J].上海交通大学学报,2015,49(12):1761-1767.XU Ting-xue,CONG Lin-hu,DONG Qi. State Prediction Method for Missile Guidance and Control Systems Based on UGM-ULSSVM[J]. Journal of Shanghai Jiaotong University,2015,49(12):1761-1767.
[2] GUILLEN A J,GOMEZ J F,GRESOP A,et al. Advances in PHM Application Framework:Processing Methods,Prognosis Methods,Decision Methods[J]. The Italian Association of Chemical Engineering,2013,33:391-396.
[3] LI Q X,LIU S F,LIN Y. Grey Enterprise Input-Output Analysis[J]. Journal of Computational and Applied Mathematics,2012,236(7):1862-1875.
[4]化金,朱学昌,王立旭.基于灰色聚类理论的非合作目标交会效能评估[J].现代防御技术,2016,44(6):13-18.HUA Jin,ZHU Xue-chang,WANG Li-xu. Effectiveness Evaluation of Rendezvous with Space Non-Cooperative Target Based on Grey Clustering Theory[J]. Modern Defense Technology,2016,44(6):13-18.
[5]王大鹏,汪秉文.基于变权缓冲灰色模型的中长期负荷预测[J].电网技术,2013,37(1):167-171.WANG Da-peng,WANG Bing-wen. Medium and LongTerm Load Forecasting Based on Variable Weights Buffer Grey Model[J]. Power System Technology,2013,37(1):167-171.
[6]孔垂猛,韩印.基于灰色马尔科夫模型的波动性交通流量预测[J].森林工程,2015,31(1):92-96.KONG Chui-meng,HAN Yin. Prediction of Volatile Traffic Volume Based on the Grey-Markov Model[J]. Forest Engineering,2015,31(1):92-96.
[7]吴利丰,刘思峰,姚立根.基于分数阶累加的离散灰色模型[J].系统工程理论与实践,2014,37(4):1823-1827.WU Li-feng,LIU Si-feng,YAO Li-gen. Discrete Grey Model Based on Fractional Order Accumulate[J]. Systems Engineering-Theory&Practice,2014,37(4):1823-1827.
[8]贺照辉,申正义,徐诚.运用BP神经网络的雷达模拟训练考核评估方法[J].现代防御技术,2017,45(1):188-193.HE Zhao-hui,SHEN Zheng-yi,XU Cheng. Simulation Training Assessment of Radar-Brigade Using BP Neural Network[J]. Modern Defence Technology,2017,45(1):188-193.
[9]付晓明,王福林,尚家杰.基于多子代遗传算法优化BP神经网络[J].计算机仿真,2016,33(3):258-263.FU Xiao-ming,WANG Fu-lin,SHANG Jia-jie. Optimized BP Neural Network Algorithm Based on Multi-Child Genetic Algorithm[J]. Computer Simulation,2016,33(3):258-263.
[10]胡大红.基于背景值与初始条件优化的非等间距GM(1,1)模型[J].湖北文理学院学报,2016(11):20-22.HU Da-hong. Unequal Interval GM(1,1)Model Based on Optimization of Background Value and Original Condition[J]. Journal of Hubei University of Arts and Science,2016(11):20-22.
[11]张鑫,肖新平.非等间距GM(1,1)模型在股票预测中的优化[J].统计与决策,2012(11):85-88.ZHANG Xin,XIAO Xin-ping. Optimization of GM(1,1)Model in Stock Forecasting[J]. Statistics and Decision,2012(11):85-88.
[12]李勇,栾元重,张善廷.非等间距GM(1,1)模型背景值优化方法在沉降预测中的应用[J].北京测绘,2016,13(5):48-52.LI Yong,LUAN Yuan-zhong,ZHANG Yan-shan. Application of the Optimization of Background Value in NonEquidistant GM(1,1)Model in the Settlement Prediction[J]. Beijing Mapping,2016,13(5):48-52.
[13]许泽东,柳福林.灰色GM(1,1)模型优化研究[J].计算机科学,2016,43(S2):6-10.XU Ze-dong,LIU Fu-lin. Survey on Gray GM(1,1)[J]. Model Computer Science,2016,43(S2):6-10.
[14]张彬,西桂权.基于背景值和边值修正的GM(1,1)模型优化[J].系统工程理论与实践,2013(3):682-688.ZHANG Bin,XI Gui-quan. GM(1,1)Model Optimization Based on the Background Value and Boundary Value Correction[J]. Systems Engineering-Theory&Practice,2013(3):682-688.
[15]熊萍萍,党耀国,姚天祥.基于初始条件优化的一种非等间距GM(1,1)建模方法[J].控制与决策,2015(11):2097-2102.XIONG Ping-ping,DANG Yao-guo. Modeling Method of Non-Equidistant GM(1,1)Model Based on Optimization Initial Condition[J]. Control and Decision,2015(11):2097-2102.
[16]韩晓明,姜科,张琳,等.基于灰色神经网络武器装备研制费用预测模型[J].现代防御技术,2011(4):184-188.HAN Xiao-ming,JIANG Ke,ZHANG Lin,et al. Forecast Model of Development Cost for Weapon Equipment Based on Grey Systems Theory and Neural Network[J].Modern Defence Technology,2011(4):184-188.
[2] GUILLEN A J,GOMEZ J F,GRESOP A,et al. Advances in PHM Application Framework:Processing Methods,Prognosis Methods,Decision Methods[J]. The Italian Association of Chemical Engineering,2013,33:391-396.
[3] LI Q X,LIU S F,LIN Y. Grey Enterprise Input-Output Analysis[J]. Journal of Computational and Applied Mathematics,2012,236(7):1862-1875.
[4]化金,朱学昌,王立旭.基于灰色聚类理论的非合作目标交会效能评估[J].现代防御技术,2016,44(6):13-18.HUA Jin,ZHU Xue-chang,WANG Li-xu. Effectiveness Evaluation of Rendezvous with Space Non-Cooperative Target Based on Grey Clustering Theory[J]. Modern Defense Technology,2016,44(6):13-18.
[5]王大鹏,汪秉文.基于变权缓冲灰色模型的中长期负荷预测[J].电网技术,2013,37(1):167-171.WANG Da-peng,WANG Bing-wen. Medium and LongTerm Load Forecasting Based on Variable Weights Buffer Grey Model[J]. Power System Technology,2013,37(1):167-171.
[6]孔垂猛,韩印.基于灰色马尔科夫模型的波动性交通流量预测[J].森林工程,2015,31(1):92-96.KONG Chui-meng,HAN Yin. Prediction of Volatile Traffic Volume Based on the Grey-Markov Model[J]. Forest Engineering,2015,31(1):92-96.
[7]吴利丰,刘思峰,姚立根.基于分数阶累加的离散灰色模型[J].系统工程理论与实践,2014,37(4):1823-1827.WU Li-feng,LIU Si-feng,YAO Li-gen. Discrete Grey Model Based on Fractional Order Accumulate[J]. Systems Engineering-Theory&Practice,2014,37(4):1823-1827.
[8]贺照辉,申正义,徐诚.运用BP神经网络的雷达模拟训练考核评估方法[J].现代防御技术,2017,45(1):188-193.HE Zhao-hui,SHEN Zheng-yi,XU Cheng. Simulation Training Assessment of Radar-Brigade Using BP Neural Network[J]. Modern Defence Technology,2017,45(1):188-193.
[9]付晓明,王福林,尚家杰.基于多子代遗传算法优化BP神经网络[J].计算机仿真,2016,33(3):258-263.FU Xiao-ming,WANG Fu-lin,SHANG Jia-jie. Optimized BP Neural Network Algorithm Based on Multi-Child Genetic Algorithm[J]. Computer Simulation,2016,33(3):258-263.
[10]胡大红.基于背景值与初始条件优化的非等间距GM(1,1)模型[J].湖北文理学院学报,2016(11):20-22.HU Da-hong. Unequal Interval GM(1,1)Model Based on Optimization of Background Value and Original Condition[J]. Journal of Hubei University of Arts and Science,2016(11):20-22.
[11]张鑫,肖新平.非等间距GM(1,1)模型在股票预测中的优化[J].统计与决策,2012(11):85-88.ZHANG Xin,XIAO Xin-ping. Optimization of GM(1,1)Model in Stock Forecasting[J]. Statistics and Decision,2012(11):85-88.
[12]李勇,栾元重,张善廷.非等间距GM(1,1)模型背景值优化方法在沉降预测中的应用[J].北京测绘,2016,13(5):48-52.LI Yong,LUAN Yuan-zhong,ZHANG Yan-shan. Application of the Optimization of Background Value in NonEquidistant GM(1,1)Model in the Settlement Prediction[J]. Beijing Mapping,2016,13(5):48-52.
[13]许泽东,柳福林.灰色GM(1,1)模型优化研究[J].计算机科学,2016,43(S2):6-10.XU Ze-dong,LIU Fu-lin. Survey on Gray GM(1,1)[J]. Model Computer Science,2016,43(S2):6-10.
[14]张彬,西桂权.基于背景值和边值修正的GM(1,1)模型优化[J].系统工程理论与实践,2013(3):682-688.ZHANG Bin,XI Gui-quan. GM(1,1)Model Optimization Based on the Background Value and Boundary Value Correction[J]. Systems Engineering-Theory&Practice,2013(3):682-688.
[15]熊萍萍,党耀国,姚天祥.基于初始条件优化的一种非等间距GM(1,1)建模方法[J].控制与决策,2015(11):2097-2102.XIONG Ping-ping,DANG Yao-guo. Modeling Method of Non-Equidistant GM(1,1)Model Based on Optimization Initial Condition[J]. Control and Decision,2015(11):2097-2102.
[16]韩晓明,姜科,张琳,等.基于灰色神经网络武器装备研制费用预测模型[J].现代防御技术,2011(4):184-188.HAN Xiao-ming,JIANG Ke,ZHANG Lin,et al. Forecast Model of Development Cost for Weapon Equipment Based on Grey Systems Theory and Neural Network[J].Modern Defence Technology,2011(4):184-188.