液压机械补偿功率回收模型参考模糊神经网络控制

Model Reference Fuzzy Neural Network Control of Hydraulic Mechanical Compensation Power Recovery

介绍了液压系统试验中机械补偿功率回收的原理,建立了压力系统的数学模型。针对机械补偿功率回收系统影响压力的非线性因素多且多为缓变的特点,为满足试验要求提出了采用模型参考的模糊神经网络,提出了该网络实现的形式,设计了模糊神经网络和误差的逼近算法,根据要求确定了参考模型等。仿真结果表明:该控制方法能有效地跟踪参考模型,改变对象参数及负载输出压力无变化,能很好地满足试验要求。

The theory of power recovery with mechanical compensation in hydraulic system test was introduced. The mechanical model of hydraulic pressure was established. According to the characteristics of mechanical compensation power recovery pressure sys- tem which has many nonlinear elements and changes slowly, model reference fuzzy neural networks was employed to satisfy the demand of experiment. The form of realizing the network was put forward, fuzzy neural network and approximate algorithm were designed. Reference model were designed according to the demand of experiment. Simulation results show that this control method can track ref- erence model efficiently. When changing the object parameter and load, the output pressure is not changed, it shows that this control method can satisfy the demand of experiment.