起重机吊重摆角的自适应神经网络补偿控制

为了控制起重机吊重摆动幅度在最短时间内衰减到规定范围或平衡点附近，同时考虑到由于外界不确定性因素导致的系统模型的不确定性，设计了自适应径向基函数Radial Basis Function（RBF）神经网络补偿控制器。RBF神经网络对系统模型的未知函数进行辨识，并将辨识信息提供给控制器。实验结果表明：吊重摆角约在5s时跟踪给定幅度的正弦信号，并在参考信号发生突变时，摆角仍在给定的范围内；RBF神经网络约在5s后几乎以零误差辨识未知函数。所设计的控制器对不确定性因素具有较强的鲁棒性，这也验证了控制系统稳定性证明结论。

The Adaptive Neural Network Compensation Control for Load Swing Angle of Crane

In order to control load＇s swing angle amplitude attenuating to given range or near the stationary point in the shortest time for the crane, at the same time, system model＇s uncertainty caused by outside uncertain factors is considered, the compensation controller based on adaptive RBF neural network is designed. RBF neural network identified system model＇s uncertain function, and the identified result is provided with the controller. Experiment results show that load＇s swing angle can track sinusoidal signal with given amplitude, and the load＇s swing angle is still in given range when the given reference signal is jump signal. Uncertain function can be identified almost with zero error at about 5s and the designed controller has strong robustness for the system＇s uncertain factors which also verifies the conclusion of the control＇s stability proof.