起重机自适应智能防摆控制方法及其仿真研究

【作者】 黄凯；

【导师】 郑加强； 徐幼林；

【作者基本信息】 南京林业大学 ， 机械设计及理论， 2007， 博士

【摘要】 消除或控制吊重的摇摆对提高起重机工作效率、减少装卸作业安全生产隐患具有重要意义，采用电子防摆装置，是减轻司机工作强度、改善司机恶劣工作环境的重要途径，也是实现装卸机械自动化的大势所趋。论文采用状态反馈、PID控制、LQR最优控制等传统控制方法对起重机防摆问题进行了仿真研究，并在对模糊控制和模糊神经网络理论进行分析研究的基础上，基于工程实际和方便司机操作，提出了“速度跟踪”模糊控制器和“速度位移双跟踪”模糊控制方法，仿真结果表明，模糊控制器既能实现小车的精确定位，又能有效地控制吊具的摆动，而在存在初始扰动的情况下小车的定位存在较大的稳态误差。因此在模糊防摆控制器的基础上，本文进一步深入研究了同时具有模糊控制和神经网络的模糊神经网络控制方法，并设计了T-S型自适应模糊神经网络控制器，该控制器继承了模糊控制器的优点，且在存在初始扰动的情况下能基本消除小车定位的稳态误差，取得了较好的防摆效果。本文的创新点主要体现在以下几个方面：(1)利用拉格朗日方法建立了起重机双向防摆的完全的、非线性动力学方程，为研究起重机双向防摆问题提供了理论依据；(2)分别采用状态反馈、PID控制、LQR最优控制方法对起重机防摆问题进行了仿真；(3)创新性地设计了“速度位移双跟踪”模糊控制器，在不存在初始扰动的情况下取得了理想的防摆效果；(4)将模糊技术、神经网络技术和LQR最优控制方法相结合，设计了T-S型自适应模糊神经网络控制器，解决了模糊控制器在存在初始扰动的情况下小车定位的稳态误差较大的问题。更多还原

【Abstract】 Eliminating and controlling the swing of loads is very important for increasing the work efficiency of crane as well as decreasing safety hazard during loading & unloading operation. It’s a main method to lighten the work strength and to improve the bad work condition for the operator by adopting electronic anti-swing device, which also the trend of actualizing the loading & unloading machinery automation.The state feedback, PID control, LQR optimal control methods were used in this study for the simulation of the crane anti-swing issues, the "speed follow-up" and the "speed & displacement double follow-up" fuzzy control methods were developed based on learning and studying fuzzy control and fuzzy neural network theory as well as considering the project practice and convenient for driver operation, the simulation result showed that: fuzzy control does either realize the accurate position of the trolley, or control the loads swing. However, there is a significant steady-state error for trolley position when initial disturbance exists. Based on fuzzy anti-swing controller, further studying fuzzy neural network control methods which have both advantage of the fuzzy control and neural network, then the T-S self-adaptive fuzzy neural network controller was designed, which succeed virtue of fuzzy controller, also mostly eliminate the steady-state error during initial disturbance. The study showed that the T-S self-adaptive fuzzy neural network controller is an effective anti-swing method.The Innovation results for this study were shown as follows:(1) Establishing the absolute and non-linear dynamics equation for crane double-way anti-swing by using Lagrange method could lay the theory foundation in studying crane double-way anti-swing problem.(2)The simulation for crane anti-swing issues was carried out by separately adopting the state feedback, PID control and LQR Optimal Control methods.(3) The "speed & displacement double follow-up" fuzzy controllers was firstly created with the ideal anti-swing efficiency without initial disturbance.(4) The fuzzy technology, neural network technology and LQR Optimal Control method were integrated to design T-S type self-adaptive fuzzy neural network controller, which can resolve the significant steady-state error for trolley position when existing initial disturbance.更多还原