摘要
针对目前传统PID控制对模型依赖性强,难以在线调整控制参数,具有非线性,而神经网络控制在误差控制方面又有不足之处,在变风量汽车双温区自动空调中都难以得到较好的控制效果,文章提出了将BP算法的神经网络和PID加以混合的一套控制系统,减少因为参数模糊性、非线性问题以及外界不稳定的干扰对汽车空调系统的影响,从而提高系统的鲁棒性。
The traditional PID control has strong dependence on model and is hard to adjust nonlinear parameters online; meanwhile, the Neural Network has its shortage in error control. These two control methods cannot get well ap- plication in automotive variable dual temperature zone air-conditioner, so a method compounding BP Neural-Network with PID control is proposed to reduce the influence on automated air conditioner which caused by parameters＇ fuzzi- ness, nonlinearity and the unstable disturbance in environment. It increases the robust characteristic at the same time.
出处
《仪表技术》
2013年第1期11-14,共4页
Instrumentation Technology
关键词
非线性
汽车自动空调
双温区
BP神经网络PID
模糊性
nonlinearity
automated air conditioner
dual temperature zone
BP neural network PID
fuzziness
