摘要
PID控制算法理论成熟且在工业中应用广泛,但常规的PID(比例积分微分算法)控制器对于非线性、复杂的建筑结构的控制存在一定局限性。通过神经网络优化PID控制算法,让PID的控制参量可以自行整定,克服PID算法难以自我调节和用于复杂结构的缺点,使优化的PID算法适用于磁流变阻尼抗震结构的半主动控制。将神经网络优化PID算法用于安装有磁流变阻尼器的某3层钢筋混凝土框架结构进行的仿真研究,并对比结构在无控状态、PID控制、神经网络控制以及神经网络PID控制下结构的地震响应。结果表明:在3条不同地震波作用下,神经网络PID算法对的结构各层位移峰值、加速度峰值的控制效果明显优于PID算法与神经网络算法。
PID control algorithms has mature theory and is widely applied in industry,however,the conventional PID controller has limitations for control of nonlinear and complex building structure. The paper use neural network control to optimize PID control algorithms,and make it be suitable for magneto-rheological damping structure. Optimizing neural network PID control is applied to making numerical simulation for a three layers reinforced concrete frame structure with MRFD under conditions of the uncontrolled structure,PID control, neural network control and optimizing neural network PID control. The result shows： comparing the displacement between the layers and acceleration of structure in the conditions of three different seismic waves,the effect of optimizing neural network PID control is much better than the rest of control algorithms.
作者
张立峰
丁建国
王巍伟
Zhang Li-feng Ding Jian-guo Wang Wei-wei(School of Science, Nanjing University of Science and Technology, Nanfing 210094, China School of Energy and Power Engineering, Nanjing University of Science and Technology, Nanjing 210094, China)
出处
《工程抗震与加固改造》
北大核心
2017年第5期47-54,共8页
Earthquake Resistant Engineering and Retrofitting
关键词
结构抗震
神经网络控制
PID控制
磁流变液阻尼器
seismic resistance
neural network control
PID control
magnetorheological fluid dampers
