摘要
文章首先介绍了一种发动机空燃比神经网络模型,并概述了该模型使用的批量训练算法、神经网络结构和输入输出变量。其次基于该模型构建了发动机燃空比控制器,并提出了基于权值更新算法的神经网络控制器优化策略。最后对该控制器进行了离线测试和在线仿真。仿真结果表明,该控制器具有静差较小、响应迅速等特点,实际燃空比保持在目标燃空比附近,控制效果良好。
This paper introduces a neural network model of engine air-fuel ratio,and Outlines the batch training algorithm,neural network structure and input and output variables used in the model.Then the engine fuel-air ratio controller is constructed based on the model,and a neural network controller optimization strategy based on weight update algorithm is proposed.Finally,the controller is tested offl ine and simulated online.the simulation results show that the controller has the characteristics of small static error and rapid response,and the actual burning-space ratio is kept near the target burning-space ratio,and the control eff ect is good.
作者
马尔旦·吐尔逊
陆奇志
TURSUN Mardan;LU Zhiqi(Xinjiang Vocational&Technical College of Communications,Wulumuqi 831401,China)
关键词
BP神经网络
汽车发动机
空燃比
时滞补偿
BP neural network
automobile engine
air-fuel ratio
time delay compensation