摘要
针对具有非线性特性的控制系统,提出了一种逐级模糊神经网络控制算法。该系统控制采用了补偿模糊神经网络算法和逐级模糊控制规则。在matlab仿真环境下对简化的模型进行了仿真实验。通过仿真结果可以看出,该控制算法比传统的模糊控制具有更好的控制表面,更能适应复杂多变的非线性准确控制;补偿模糊神经网络算法在训练时,具有学习速率快、准确度高和扩展性好等优点。
A new control method is proposed by using the gradual fuzzy neural network for the Ball ＆ Plate apparatus nonlinear dynamics control system. The Compensated Fuzzy-Neural network algorithm and the rules of the fuzzy control are applied to the system. In the Matlab simulation environment, the simplified system is simulated. Referring to the result of the simulation, the algorithm of the control performs better than the sole fuzzy control algorithm. It has better control surface and will perform well in the complex nonlinear control system. The compensated fuzzy-control algorithm has rapid learning rate, high accuracy and good extensibility.
出处
《科学技术与工程》
2006年第5期540-543,共4页
Science Technology and Engineering
关键词
模糊神经网络
补偿模糊神经网络
模糊控制
fuzzy-neural network compensated fuzzy-neural network fuzzy control
