田震, 荆双喜, 赵丽娟,等.基于粒子群优化BP神经网络的采煤机可靠性预测[J].河南理工大学学报（自然科学版）,2020,39(1):68-74.

TIAN Z , JING S X, ZHAO L J, et al.Reliability prediction of shearer based on BP neural network optimized byparticle swarm[J].Journal of Henan Polytechnic University(Natural Science) ,2020,39(1):68-74.

基于粒子群优化BP神经网络的采煤机可靠性预测

田震1,2, 荆双喜2, 赵丽娟3, 高珊1, 张成光1

1.周口师范学院机械与电气工程学院，河南周口  466000;2.河南理工大学机械与动力工程学院，河南焦作 454000;3.辽宁工程技术 大学机械工程学院，辽宁阜新 123000

摘要:为研究采煤机在截割过程中的可靠性，通过粒子群算法对BP神经网络进行优化改进， 建立采煤机可靠性预测模型。采用高斯型隶属度函数，构建材料应力-结构可靠度之间的隶属函数，通过正交仿真实验确定具有代表性工况下采煤机整机的可靠度，以实验结果建立学习样本，对预测模型的准确度进行检验，结果表明，预测结果与实验结果最大相对误差为2. 61% ， 满足精度要求。利用预测模型对采煤机在不同牵引速度和截深条件下截割不同硬度煤层的可靠度进行分析，找出采煤机可靠度随三者的变化规律：随着煤层硬度以及牵引速度增加，可靠度降低幅度变大；随着截深增大，可靠度降低幅度逐渐趋于平缓。

关键词:采煤机;可靠性预测;粒子群优化算法;BP神经网络

doi:10.16186/j.cnki.1673-9787.2020.1.9

基金项目:国家自然科学基金资助项目（51674134 ）；河南省自然科学基金资助项目（182300410250 ）；河南省科技攻关项目 （182102210606）

收稿日期:2019/04/09

修回日期:2019/05/20

出版日期:2020/01/15

Reliability prediction of shearer based on BP neural network optimized byparticle swarm

TIAN Zhen1,2, JING Shuangxi2, ZHAO Lijuan3, GAO Shan1, ZHANG Chengguang1

1.College of Mechanical and Electrical Engineering， Zhaukau Normal University， Zhaukau  466000 ， Henan， China;2.School of Mechanical and Power Engineering， Henan Polytechnic University， Jiaozuo  454000 ， Henan， China;3.College ofMechanical Engineering， Liaoning Technical University， Fuxin  123000 ， Liaoning， China

Abstract:In order to study the reliability of shearer cutting process， the reliability prediction model of shearer was established based on the BP neural network optimized by particle swarm. The membership function between the stress and structure reliability of the material was constructed by using the Gauss shaped membership function and the reliability of the shearer was determined by means of orthogonal simulation test. The test results were used as learning samples to test the accuracy of the prediction model. The maximum relative error between prediction results and experimental results was 2. 61 % ， which met the accuracy requirements. The reliability of the shearer with different hardness and cutting depth under different traction speed was analyzed by using prediction model， and the rules of reliability of shearer with the three changes was found. With the coal seam hardness and traction speed increased， the greater the degree of reliability reduction would get， and with the cut depth increased， gradually the degree of reliability reduction gradually got flat.

Key words:shearer;reliability prediction;particle swarm optimization algorithm;BP neural network

基于粒子群优化BP神经网络的采煤机可靠性预测_田震.pdf