Research On Predictive Control To Titling System Of Titling Passenger Car Based On Neural Network

The tilting train has great economic potential when it is used to serve as a way of high-speed railways for the situation of our diverse railways and developing country in existence. The tilting control system is the key technique of tilting train, whose safety is very important for tilting train to transport normally. The reliability of the titling control system is bound to the train safety. Not only the titling train safety but also the developing titling train intention, which speeds the velocity and improves passengers' comfort when the train passes curve, is affected by the real time of the titling control. It is the key of the titling train technology that its high reliability and real time of control is ensured. The predictive control based on the predictive acceleration signals pioneers a new way for advancing the control real time. The predictive control technology with the predictive acceleration in tilting control system is discussed in this thesis, which is the main substance of developing the tilting control system supported by the R&D Program of Ministry of Railways of China.Considering the fact that the tilting control system of tilting train is working in quite an uncertain environment and is nonlinear in internality, a conventional linear control method is hard to perform effectively. Neural network has been widely used in nonlinear control systems in many fields as a novel method or measure. However, there are very few published studies dealing in neural network used in control of the locomotive and rolling stock. Neural network has been systematically introduced into the tilting control system in tilting train for the first time in our country by this thesis. First, the control reference signal is founded through the acceleration predicted by neural network. Second, the neural network is used to identify the nonlinear system. Third, the integration of the neural network predictive control and the robust control method is put forward to grope for solving the control problem of nonlinear system. The prediction, identification and control law is studied from application based on neural network.Based on the analysis and calculation of every tache and its postpone of the reference signals of the titling control system, the research of the signal prediction, system identification and predictive control is applied by neural network. The maincontributions are as follows.(1) The lateral accelerations signal sampling from running train is filtered by the software filter. The filtered acceleration signal is predicted by diverse methods. The signal prediction method as a new compensated method is brought forward in order to replace the lag compensated function of the gyro. In certain satisfaction precision, the research result shows that the signal prediction method can replace the compensated function of the gyro.(2 ) On considering the affection of servo electromotor and mechanism system nonlinear, the servo electromotor saturation nonlinear affection is mainly took into account. According to the dynamic nonlinear relation between input and output, the dynamic system is identified in neural network and the neural network model is founded.(3) The predictive control of titling passenger car is studied with the predictive lateral acceleration. The reference signal is founded by dealing the predictive signal in some certain and the titling system is controlled to solve the nonlinear control. It shows that the real time of the titling control is attained without the gyro. The reliability is advanced and the cost of buying gyro is saved.Trial and research show that the neural network predictive control method based on prediction acceleration of the titling system is practicable in terms of theory and application of engineering. A direction to set up the neural network predictive way of the titling system of the titling passenger car by means of neural network is pointed out. The research results in this thesis lay the theoretical foundation of tilting control system of the first tilting trai...