Internal pressure control for high-speed trains based on condition matching and performance iteration

Abstract

When a high-speed train passes through tunnels, tunnel pressure waves will cause pressure fluctuation inside carriage. Traditional control strategy of shutting down air ducts for a fixed period may fail to consider both riding comfort and air quality. The similarity of tunnel pressure waves when the same train passes through the same tunnel provides a possibility to solve the problem by iterative learning control (ILC) algorithm. However, the varying amplitude and scale limit the application of conventional ILC. Thus in each iteration, the control inputs of the nearest condition in the historical database will be matched and applied to the control process, after which the control error will be gained and then the control inputs will be updated by the error. Next, the performance will be evaluated to refresh the database to make the control inputs in the database to be optimal. Feasibility analysis and simulations show the feasibility of controlling the fluctuation inside the carriage by matching the database records.

Internal pressure iterative learning control for high-speed trains based on condition and performance matching under morphologically similar tunnel pressure waves

“…These provide a new perspective and possibility for the quasi-periodic repetitive internal pressure fluctuation control tasks under the excitation of morphologically-similar tunnel pressure waves. He et al 9 proposed an ILC based on condition matching and performance iteration algorithm by utilizing the similarity of morphologically-similar tunnel pressure waves. However, the algorithm only considers the matching of historical conditions through a condition related to the time scale, and does not consider the use of performance indicators for matching.…”

Iterative learning control of train internal pressure under extreme tunnel conditions

“…This method works well in short tunnels, but is not suitable for long tunnel conditions [3] . In response, Chen's team studied the pressure transfer mechanism inside and outside the train , established a pressure model in the train and proposed an iterative learning control for internal pressure fluctuations in long tunnels [4][5][6][7][8] . All the above controls are proposed for flat tunnels at low altitudes, and no research has been seen on the control of internal pressure fluctuations under extreme tunnel conditions.…”