摘要
针对可变截面涡轮增压器的开度与增压柴油机参数呈非线性关系的问题,提出一种基于反向传播神经网络和量子粒子群算法的非线性模型预测控制算法,通过调节涡轮增压器的开度,控制过量空气系数,从而实现柴油机的进气量与燃油消耗量的快速匹配,使转矩快速达到期望值。仿真分析表明:该方法相比于PID控制,可使增压柴油机更加平稳地完成转矩阶跃,并使增压柴油机具备转矩跟随能力。
In this paper,a nonlinear relationship between the variable geometry turbocharger opening and the parameters of the turbocharged diesel engine is presented,and a nonlinear model predictive control algorithm based on backpropagation neural network and quantum particle swarm optimization algorithm is proposed as well.By adjusting the VGT opening degree and controlling the excess air coefficient,the intake air volume and fuel consumption of the diesel engine were realized.The rapid matching of the quantities allowed the torque to reach the desired value rapidly.The test results show that the NMPC enables the turbocharged diesel engine to complete the torque step condition more smoothly than the PID control and to possess the torque following capability by controlling the VGT opening degree.
作者
张卫波
梁昆
朱清
ZHANG Weibo;LIANG Kun;ZHU Qing(College of Mechanical Engineering and Automation,Fuzhou University,Fuzhou 350116,China)
出处
《机械制造与自动化》
2021年第1期124-127,160,共5页
Machine Building & Automation
基金
福建省客车及特种车辆研发协同创新中心基金资助项目(2016BJC011)
福建省科技计划项目(K201701)
福建省科技厅引导性基金资助项目(2016H0015)。
关键词
可变截面涡轮增压器
增压柴油机
BP神经网络
量子粒子群算法
非线性模型预测控制
过量空气系数
variable geometry turbocharger
turbocharged diesel engine
BP neural network
quantum particle swarm optimization
nonlinear model predictive control Algorithm
Excess Air Coefficient