A Soft Measurement Method for Air Supply Systems of Railway Vehicles Based on Improved Multivariate Support Vector Regression

Abstract

Aiming at the problem that the air supply system of railway vehicles lacks sensor data and most of the measurement points are not easy to measure directly, a soft measurement method based on improved multivariate support vector regression (IMSVR) was proposed. By analysing the structure composition and working principle, intake temperature, intake pressure and exhaust pressure of the air supply system were selected as auxiliary variables of the variable to be measured. In order to make full use of the acquired data information, the phase space reconstruction technology was introduced, and a soft measurement model between the variables to be measured and auxiliary variables of the air supply system was established based on the improved multivariate support vector regression (IMSVR) algorithm, and the particle swarm optimization (PSO) algorithm was used to optimize the kernel parameter g and the penalty parameter c. By installing pressure and temperature sensors on a typical air supply system and carrying out performance tests and fault injection tests on the modified air supply system test bench, the air supply system experimental data set was obtained. Finally, taking the fuel injection temperature as an example, the validity and accuracy of the method proposed in this paper were verified based on the experimental data set. The research result provides a reference for the fault early warning, diagnosis and maintenance of the air supply system.