基于神经网络模型的空燃比非线性模型预测控制

采用基于径向基神经网络（RBFNN）模型的非线性模型预测控制方法，被控对象选择火花塞点火（SI）发动机的空燃比（AFR）高度非线性复杂系统，利用渐消记忆最小二乘法实现基于RBFNN的SI发动机AFR系统建模以及参数在线自适应更新。针对非线性模型预测控制中寻优问题，运用序列二次规划滤子算法对最优控制序列进行求解，并加入滤子技术避免了罚函数的使用。在相同的实验环境下，与PI控制算法和Volterra模型预测控制方法进行仿真对比实验，结果表明，所提算法的控制效果明显优于其他两种方法。

Based on the Neural Network Model of Air-fuel ratio Nonlinear Model Predictive Control

For the SI engine AFR highly complicated nonlinear system .The paper based on RBF neural network model of the nonlinear model predictive control method .Using the fading memory least square method ,complete the model of SI engine AFR systems ,and achieve the parameter adaptive update online .For nonlinear model predictive control in the optimization problem ,paper uses the SQP algorithm for solving the optimal control sequence ,meanwhile in order to avoid using penalty function paper uses Filter technology .Under the same experimental conditions ,Compare simulation results with PI control algorithm and the model predictive algorithm based on Volterra .The results show that the proposed algorithm is superior than other two methods.