分组自适应模糊神经网络在双摆吊车上的应用

在工业生产中,若吊钩质量较大不能忽略,吊车会呈现出双摆系统特征,即吊钩绕台车进行一级摆动,同时负载绕吊钩进行二级摆动.针对这种欠驱动非线性系统控制问题,提出一种基于双摆桥式吊车系统的自适应神经网络与分组模糊控制相结合的控制策略.具体而言,采用改进的Takagi-Sugeno型神经网络模糊控制模型,利用分组模糊控制的方法,规避多输入模糊系统规则冗杂导致的时效性问题.因模糊控制不具备自适应与迭代优化参数的能力,引入神经网络模型借助混合算法,通过前向学习与反向学习,利用二次型最优控制理论设计得出的样本数据,不断优化模糊规则和隶属度函数,从而使控制器参数高精度逼近系统要求,满足双摆吊车高温熔融金属吊运的需求.最后仿真实验结果证明该控制策略不仅使双摆桥式吊车台车与负载快速、精确到达目标位置,还能够抑制吊钩与负载的摆动,提高了控制器性能.

Application of Grouped Adaptive Fuzzy Neural Network on Double Pendulum Crane

In industrial production, if the quality of the hook is too large to be ignored, the crane will show the characteristics of double swing system, that is, the hook swings around the trolley at the first stage, and the load swings around the hook in the second stage. In order to solve the control problem of this kind of underactuated nonlinear system, a control strategy based on adaptive neural network and group fuzzy control is proposed. Specifically, the improved Takagi Sugeno neural network fuzzy control model is adopted, and the grouping fuzzy control method is used to avoid the timeliness problem caused by the redundant fuzzy rules of multi input system. Because fuzzy control does not have the ability of self-adaptive and iterative optimization parameters, the neural network model is introduced. Through forward learning and reverse learning, and using the sample data designed by quadratic optimal control theory, the fuzzy rules and membership function are continuously optimized, so that the parameters of the controller can approach the system requirements with high precision and meet the requirements of high temperature molten metal crane of double pendulum crane Transportation demand. Finally, the simulation results show that the control strategy can not only make the trolley and load reach the target position quickly and accurately, but also restrain the swing of hook and load, and improve the performance of the controller.