Multi-model Based Cruise Control For High- Speed Train

In the past 10 years,China high-speed railway construction has achieved remarkable achievements.With the proposal of “The Belt and Road Initiative” and the further implementation of national innovation-driven development strategy,China high-speed railway has become a gorgeous namecard in the new era of China.At present,artificial intelligence and automatic driving are becoming research hotspots,however,the control for China high-speed train depends heavily on manual mode,therefore,how to achieve the automatic driving for high-speed train becomes future research trend,and the primary problem that automatic driving needs to solve is accurate speed tracking in the running process of high-speed train.As the highspeed train runs in a complicated environment,and running conditions will change over traction,braking and coasting,cruise operation mode is often used in the running process,which covers three kinds of running conditions above,therefore,cruise operation mode is chose as the research focus.In order to achieve the ideal speed tracking in the cruise operation of high-speed train,a multi-model based control method for high-speed train is employed.Firstly,A clustering algorithm based multi-model modeling method is utilized on the basis of the actual operating data to establish multiple models that could describe the cruise operation process of high-speed train;moreover,the adaptive fuzzy-PID controller corresponding to each model is put forward to realize the ideal speed tracking for high-speed train.Specific research works are as follows:1.In view of the complex running environment as well as changeable characteristics of high-speed train,the force in the running process of high-speed train is analyzed on the basis of studying the present research situation for modeling and control method,and obtains the dynamic equation of high-speed train.Meanwhile,the speed tracking control mechanism in cruise operation mode is explained in detail.2.Considering the obvious non-linear characteristic in running process of high-speed train,a single model is difficult to describe its running process accurately,thus,a multi-model modeling method is proposed,and fuzzy C-means(FCM)clustering algorithm is used to classify the initial data.Then,An improved clustering algorithm is designed to prevent system falling into the local optimum,and the clustering effects of the two clustering algorithms are analyzed and compared,the model parameters are identified by recursive least squares(RLS)to finish model building and establish the model switching principle subsequently,furthermore,the stability of the system in the switching process is demonstrated according to the system instability caused by frequent switching of the model.The simulation experiments are carried out to verify the validity of the proposed method.3.Based on the conventional PID control and fuzzy control theories,an adaptive fuzzyPID controller is designed to realize the accurate speed tracking in the cruise mode of high speed train according to the established multiple sub-models.The validity of the designed controller is validated with the actual running data.