基于非线性多输入多输出近似动态规划的发动机缸平衡智能调节算法

摘要

为解决发动机缸平衡控制问题,将标准近似动态规划(approximate dynamic programming,ADP)扩展为多输入多输出形式,给出其控制算法,并证明其收敛性.仿真结果显示,该方法能在一定范围内智能地调节各缸喷油量,用于补偿由多种不确定因素导致的各缸转速差异,从而自适应地提高缸平衡效果.该方法只需基于实时转速,不必检测和区分各缸间的转速差异,具有非线性系统的智能优化特点.该方法能直接处理各缸间的非线性多输入多输出耦合关系.%In order to solve the cylinder equilibrium control problem of engines,the standard approximate dynamic programming (ADP) is extended to be of multi-input multi-output (MIMO) form,the control algorithm is presented,and its convergence is proved.The simulation results show that,the method can intelligently adjust the injection quantity of each cylinder in a certain range so as to compensate speed differences among cylinders caused by various uncertain factors,and realize the adaptive improvement of the cylinder equlibrium effect.The method is based on the real-time speed and has no need to detect and distinguish speed differences among cylinders,so it is of intelligent optimization characteristics of nonlinear system.It can directly deal with the nonlinear MIMO coupled relationship among cylinders.