Neuro-adaptive fault-tolerant control of high speed trains under traction-braking failures using self-structuring neural networks

Abstract

This paper develops an adaptive control scheme for position and velocity tracking control of high speed trains under uncertain system nonlinearities and actuator failures. Neural networks with self-structuring capabilities are integrated into control design, where the number of the neurons can be adjusted online automatically, so that not only the problem inherent in the NN with fixed structure is avoided, but also the negative impacts arising from nonlinear in-train forces, traction/braking uncertain dynamics as well as the unknown actuation faults are effectively attenuated. It is shown that the resultant control algorithms are able to achieve high precision train speed and position tracking under varying operation railway conditions, as validated by theoretical analysis and numerical simulations.