摘要
文章利用模糊神经网络的模糊推理能力以及前馈神经网络的逼近能力 ,将其与自适应控制方案结合 ,并取带有控制增量约束的广义目标函数作为优化指标 ;从而推导出一种能对非线性非最小相位系统进行有效控制的模糊神经网络间接自适应控制器。在网络学习算法上分别采用 Davidon最小二乘法和带有动量项的BP算法。仿真结果表明了该方法的有效性。
In this paper,combining the adaptive control scheme with neuro fuzzy network which has the learning and adaptive ability of neural network and the human like reasoning and thinking flair of fuzzy system,a kind of adaptive control scheme is put formard based on neuro fuzzy network.The Back Propagation algorithm with momentum term is used in training the neuro fuzzy network as controller,and the Davidon Least Squares is employed in training the multi layer feedforward network as identifier.Through the simulation,the effectiveness and rapidity of the control scheme are demonstrated.
出处
《电气传动》
北大核心
2001年第4期31-33,共3页
Electric Drive
基金
天津市自然科学基金资助 项目批准号 :99370 0 5 11
