Research On Iterative Learning Control Of Interior Pressure Fluctuation Of High-speed Train

China’s high-speed railway has developed rapidly in the past ten years.With the continuous growth of the operation speed of the high-speed train,the tunnel pressure fluctuation outside the vehicle becomes more intense,which induces the greater pressure fluctuation inside the carriage and affects the passenger comfort.The existing traditional passive control method of closing the air ducts for a fixed period may cause greater interior pressure fluctuation when the air ducts are reopened,which cannot effectively suppress the transfer of air pressure from the outside to the inside of the vehicle.And the air quality inside the vehicle will be affected by closing the air ducts for a long time.Thus,there is a certain contradiction between the interior air pressure comfort and the demand for fresh air amount inside the carriage under the traditional control.In respect of the issues above,the multi-factor coupling transfer model of internal and external air pressure is established,which takes the nonlinear characteristics of the gaps,the air ducts,and the car body deformation into consideration.The threshold iterative learning control algorithm of interior pressure fluctuation which matches the opening and closing time of the air ducts with the characteristics of the interior air pressure,and the threshold iterative learning control algorithm of interior pressure fluctuation based on the matching of condition and performance are designed.The control algorithm is simulated based on the transfer model of internal and external air pressure and the measured tunnel pressure wave data.Compared with the traditional control method,the interior pressure fluctuation control algorithms designed in this thesis have achieved better control results.Based on the research results of interior pressure fluctuation control algorithm,the interior pressure fluctuation test system of high-speed train is further designed.The main research contents of this thesis include:1.Considering the influence of nonlinear characteristics of gaps,air ducts and car body deformation on the internal pressure fluctuation,the multi-factor coupling transfer model of internal and external air pressure is established.The model is modified by using the measured internal and external pressure data,and the optimal delay time of the model is obtained,and the correctness of the modified transfer model with delay time is verified.2.The threshold iterative learning control algorithm of interior pressure fluctuation,which the opening and closing time of air ducts matches the characteristics of the interior air pressure,is studied and compared with the uncontrolled condition and traditional control approach.The results show that the threshold iterative learning control matching the opening and closing time of the air ducts with the characteristics of the interior air pressure,can better take into account the interior air pressure comfort and the air quality inside the carriage.3.The historical operation database including working condition information,operation state information and control information of high-speed train is established,and on the basis of threshold iterative learning control algorithm,the threshold iterative learning control algorithm of interior pressure fluctuation based on the matching of condition and performance by using the historical operation data is designed.And it is compared with the threshold iterative learning control algorithm.The results show that the threshold iterative learning control algorithm of interior pressure fluctuation based on matching of condition and performance is improved in terms of interior air pressure change rate and convergence speed.4.The interior pressure fluctuation test system of high-speed train is designed,which mainly includes the overall scheme design of the system,the construction of hardware system and the software design.And the software design includes data acquisition module,data analysis and display module,interior pressure fluctuation control algorithm module and data storage management module.