【摘要】 针对天然气发动机本身的非线性特点和空燃比传输延迟的特性,提出了一种基于神经网络预测的空燃比控制策略,利用Matlab/Simulink建立控制器的算法模型,以dSPACE公司的MicoAutoBox为算法实施平台,在NQ150N型天然气发动机上进行了实验测试,实验结果表明,与普通PID控制算法相比,基于神经网络预测的控制算法稳态性能优良,能明显改善过度工况空燃比的控制效果。更多还原

【Abstract】 For air-fuel ratio signal of a CNG engine,there exist transmission delay and nonlinearities,which affects the control accuracy of air-fuel ratio using directive air-fuel ratio sensor.A neural-network predictive algorithm to air-fuel ratio is provided in this paper.The control algorithm was built under the Matlab/Simulink environment and implemented on NQ150N CNG engine with MicroAutoBox hardware of dSPACE.The actual test results in experiments show that compared with the PID algorithm,this method has better steady performance and better control effect under transient conditions.更多还原