高柔结构风振AMD控制模糊神经网络预测算法研究

基于Takagi-Sugeno模型的模糊神经网络的基本理论,建立了模糊神经网络预测模型。该模型结合神经网络控制和预测控制的优点,解决了控制中的时滞问题。研究了基于聚类法产生模糊神经网络预测控制的模糊逻辑系统,该方法便捷地解决了模糊逻辑控制中模糊控制规则基于专家控制策略和经验而无自学习能力的困难。以深圳京基金融中心为算例,利用模糊神经网络预测算法控制结构在风荷载作用下的振动,仿真结果表明,模糊神经网络预测算法能够有效地减小高柔结构加速度响应。

AMD Control of Wind Response of Flexible-high Building Using Fuzzy Neural Network Forecast Algorithm

Based on a practical project,influence on control system by the classical linear quadratic optimal control(LQR)algorithm is studied to realize the reconfiguration of part of the poles in pole placement algorithm.Using those above conclusions for reference,designing scheme of expected poles is proposed and the influence on performance of control system by expected poles in pole assignment algorithm is studied.Due to the control force may be too large to realize when pole placement control is used in actual b...