神经网络预测算法在抽油机节能电机系统中的应用  

Application of Neural Network Prediction Algorithm in Energy-SavingMotor System of Pumping Unit

作　　者：王亚洲[1] 曹旺国 向明 曹利钢 朱永红 WANG Yazhou;CAO Wangguo;XIANG Ming;CAO Ligang;ZHU Yonghong(Chunliang Oil Production Plant,SINOPEC Shengli Oilfield Branch,Dongying 256504,China;Jingdezhen City Jingde Motor Co.,Ltd.,Jingdezhen 333000,China;School of Mechanical and Electronic Engineering,Jingdezhen Ceramic University,Jingdezhen 333001,China)

机构地区：[1]中石化胜利油田分公司纯梁采油厂,山东东营256504 [2]景德镇市景德电机有限公司,江西景德镇333000 [3]景德镇陶瓷大学机械电子与工程学院,江西景德镇333001

出　　处：《电机与控制应用》2022年第6期98-102,共5页Electric machines & control application

基　　金：国家自然科学基金项目(62063010)。

摘　　要：抽油机在工作时,每台电机的实际负载都是不相同的,如何根据工作过程中抽油机的实际负载来调整电机的实际输入电压对提高抽油机的效率有着明显的提升。本文采用了神经网络预测算法对电机实际输入电压进行预测,讨论了神经网络预测算法参数对预测结果的影响,并对神经网络预测算法获得的预测值与实测值进行了对比。结果表明本文提出的神经网络预测算法能够对电机最佳工作电压进行预测。When the pumping unit is working,the actual load of each motor is different.How to adjust the actual input voltage of the motor according to the actual load of the pumping unit during the working process can significantly improve the efficiency of the pumping unit.This paper uses the neural network prediction algorithm to predict the actual input voltage of the motor,discusses the influence of the parameters of the neural network prediction algorithm on the prediction results,and compares the predicted value obtained by the neural network prediction algorithm with the measured value.The results show that the neural network prediction algorithm proposed in this paper can predict the optimal working voltage of the motor.

关 键 词：节能电机 神经网络预测算法 算法参数 输入电压 

分 类 号：TM301[电气工程—电机]