【摘要】 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。更多还原

【Abstract】 Based on the theory of fuzzy neural network of Takagi-Sugeno model,a predicted model of fuzzy neural network is established.The problem of time-delay in control is solved by using this model with the advantages of the neural network control and predictive control.How the fuzzy logic system is generated is studied in fuzzy neural network based on clustering.The shortcomings of self-learning ability of fuzzy control is solved by using clustering which is easy and fast to realize the control strategy of expert and the fuzzy rules of experience.Taking Shen Zhen Jing-ji financial center as a calculating example,the vibration of the building loaded on wind load is controlled by using fuzzy neural network prediction algorithm,the simulation results indicate that fuzzy neural network prediction algorithm can effectively reduce the acceleration response of high-rise flexible building.更多还原