Study on Fuzzy Control of Structural Vibration by Genetic Algorithms and Chaos Theory

本研究的主要内容是发展一种新的主动控制方法。主动控制具有评价检测数据输入和计算控制力的处理时间短、计算编程简单、适用范围广、对风荷载和地震荷载的不规则振动有很好的适应性等特点。此外，利用模糊控制可以处理实际结构固有振动特性的不确定性。通常，在最优控制理论中，根据结构在每个时刻的状态理论上确定的控制力作用于该结构，以达到最优控制效果。然而，对于一些执行器，很难将这种控制力调节到理论上正确的精确值。模糊控制即使在控制力形式模糊的情况下也能实现有效的控制，本文以结构上的风荷载为外力，通过对风速的短期预测来提高模糊主动控制的精度。具体地说，将风速视为混沌行为下的时间序列数据，利用局部模糊重构方法根据时间序列数据预测下一时间段的风速。根据预测的风速，应用模糊控制规则确定控制力。当实际风速出现时，对结构施加控制力以抑制其振幅和加速度。通过用龙格-库塔法求解振动方程，进行了数值仿真，验证了该方法的有效性。

The main content of the present study is the development of a new method of active control. Active control features very short processing times consumed in evaluating the inputs of detected data and calculating the controlling force, quite simple programming of calculation, wide applicability, good adaptability to the irregular vibration by wind and seismic loads, and so on. Besides, the uncertainty in the natural vibration characteristics of actual structures can be dealt with by using fuzzy control. In general, in the optimal control theory, the controlling force theoretically determined in accordance with the condition of a structure at every moment is applied to it to achieve the optimal control effect. With some actuators, however, it is difficult to regulate such controlling force to the exact value theoretically correct. The fuzzy control allows up to pertform an efficient control even with controlling force in an vague form.In the present study, wind loads on a structure were taken up as its external force, and the accuracy of the fuzzy active control was improved by making the short-term prediction of wind velocity. In concrete terms, the wind velocity was considered time-series data in chaotic behavior, and the velocity in the next time period was predicted based on the time-series data by using the local fuzzy reconstruction method. The controlling force was determined based on the predicted wind velocity by applying fuzzy control rules. Upon the advent of the actual wind velocity, the controlling force was applied to the structure to curb its amplitude and acceleration. By solving the vibration equation by the Runge-Kutta method, numerical simulation was performed to verify the effectivity of this method.