【摘要】 提出了一种多目标系统的自适应模糊神经网络控制的方法。通过对系统的多目标量进行加权合成,采用改进的遗传算法离线调整控制器网络参数。接入控制系统后,采用在线学习方式,通过BP算法调节网络的规则权值和比例因子,达到自适应控制的目的。计算机仿真结果显示,该方法控制效果好,鲁棒性强。更多还原

【Abstract】 A method of self-adaptive fuzzy neural-network control was presented for multi-goals system here.The fuzzy(neural-network) controller was put into the control system whose input was composed by multiplying weights and network’s(parameters) were trained by improved genetic algorithm firstly.Futhermore,the(controller) adaptively controled the system by online adjusting the weights of rules and the scale fators through the BP algorithm.Computer simulation results show,the method has a good effect and robustness.更多还原