基于自适应模糊神经网络的起重机吊摆防摇控制

Anti-swing Control of Crane Based on Adaptive Fuzzy Neural Network

以起重机吊摆防摇控制为研究对象,设计基于BP算法的自适应神经网络-模糊控制器,用BP算法来训练网络权值,学习率自适应以调整隶属度函数,,通过仿真对比实验得出,与一般模糊控制器相比,这种控制器控制过程更加平稳,振荡衰减较快,有效提高防摇控制效率。

This paper presents an anti-swing control method for crane via fuzzy neural network based on Mamdani model, in which the BP algorithm is used to train the weight of the fuzzy neural network, the algorithm of momentum and self-adaptive learning rate are used to adjust membership functions. The dynamic simulation results achieved in the environment of Matlab/Simulink show that the presented method outperforms the fuzzy logic control method.