基于神经元PID的磁粉离合器电流控制

因为磁粉离合器的传递转矩与励磁电流呈线性关系,而且易于实现电子控制,故越来越多地应用在微型车上。针对装有磁粉离合器的电控机械式自动变速器,设计了磁粉离合器的驱动控制电路。由于驱动电路存在非线性以及汽车运行时复杂环境的干扰,应用传统的PID控制难以在控制参数整定上达到最优,为此提出利用具有自学习和自适应能力的神经元来构成磁粉离合器的自适应控制器。实验结果表明,基于该方法的磁粉离合器驱动控制动态效果较好,且系统具有较好的自适应性。

Magnetic-powder Clutch Current Control based on Neuron PID

The magnetic-powder clutch is widely used in mini-buses,because its transferred torque is linear to its field current and the electronic control strategy is easily implemented.According to AMT with magnetic-powder clutch,the drive and control circuits are designed.Because there are some non-linearity in the drive part,and some interference caused by the environment while vehicle running,it's difficult for traditional PID control to work best,then neuron which has ability of self-study and self-adaptation is applied to construct a self-adaptation controller.And then some experiments are carried out to test the control strategy.The results shows that,the controller has nice dynamic performance,and the system has nice self-adaptation.