桥式起重机定位和防摆的RBF神经网络控制

Anti-swing and position control for bridge cranes based RBF neural network

针对起重机模型的非线性和不确定性,提出一种基于径向基函数(RBF)神经网络的起重机定位和防摆控制方法,利用2个RBF神经网络控制器对小车的位置和负载的摆动分别进行控制。RBF神经网络位置控制器参考一个简单的方波位置参考曲线,可对位置和速度进行有效地控制;RBF神经网络防摆控制器用来实现悬绳摆角和摆角速度在到达目标位置时减小到零。仿真实验结果表明:该方法能同时实现起重机系统定位控制和重物防摆功能,与线性二次型最优控制(LQR)相比,具有较好的控制性和鲁棒性。

It was proposed a radical basis function neural network（RNFNN） anti-swing and position control of overhead cranes considering the nonlinearity and uncertainty of the crane model,where two radial basis function neural network controller were designed to control the trolley position and load swing separately.The radial basis function neural network position controller was based on a simple square wave position reference trajectory to control the position and velocity of the trolley.The radial basis function neural network anti-swing controller was used to implement hanging rope swing angle and swing velocity decreases to zero when reaching the destination.Simulation results show that the presented conrtoller can realize positioning and anti-swaying simultaneously,and compared with linear quadratic regulator（LQR）,this method is of good control performance and robust.