Adaptive Active Fault-tolerant Control For Train Control System Of High-speed Railway

Epitomised by fast velocity,high efficiency and low energy consumption,high-speed railway plays an increasingly significant role in modern transportation system as an indis-pensable ground transport service.The normal operation of high-speed trains(HSTs)not only affects the safety of passengers'lives and properties,but also influences the growth of economics and the stabilities of society.Safe operation of HSTs is the fundamental prerequisite for the development of high-speed railway.The train control system is the core equipment to ensure the safe and efficient operation of trains.Faults are the key factors affecting the safe operation of trains and reducing control system performances.However,fault-tolerant control(FTC),as an effective means to compensate faults and en-hance system reliability and security,has become a research hotspot at home and abroad in recent years and has been widely used in the compensation research of system actuator faults.If actuator failures occur in the process of trains operation,the train control system with the original driving strategy can not ensure that the train still maintains the desired tracking performances.Therefore,it is necessary to carry out research on FTC algorithm for train control system of high-speed railway subject to actuator failures.Starting from three aspects of curve tracking,traction/braking restraint and state stabilization,this paper mainly focus on the adaptive active FTC algorithm for train control system of high-speed railway with actuator faults.The main works are as follows:Firstly,the back-stepping adaptive active FTC problem for HSTs with unknown ac-tuator faults is handled.Aiming at the operation control problem of HSTs,an adaptive active FTC strategy based on back-stepping technique is proposed.Furthermore,the pro-posed control scheme is extended to high-order strict feedback non-linear systems sub-ject to unknown actuator faults and multiple disturbances.Given the characteristics of the strict feedback nonlinear system,the single-mass model of HSTs under the unknown actuator failure is analyzed.The good approximation ability of neural networks(NNs)is used to approximate the unknown bounded additional resistance.Hence,a neural adap-tive back-stepping active FTC strategy based on the tuning function is proposed for the high-order strict feedback non-linear systems,which can avoid the over-parameterization phenomenon resulting from designing n times unknown parameter adaptive laws under a single back-stepping adaptive control strategy.Secondly,the neural adaptive active FTC problem for HSTs with traction/braking restraint is handled.Based on proportional integral differential-like(PID-like)sliding surface,a neural adaptive active FTC strategy is proposed for HSTs subject to unknown parameters and actuator faults.Based on the single mass-point model of HSTs with actuator faults and input constraints,a novel PID-like sliding surface with time-varying parameters is designed.Thus,a robust FTC scheme is proposed.The feasibility and effectiveness of the proposed control schemes are proved by simulation results which are based on three train operation conditions(ideal operation conditions,only unknown actuator faults,both unknown actuator faults and input constraints).Next,the sliding mode FTC problem for HSTs under time-varying uncertain op-erating conditions is handled.Aiming at the input saturation characteristics of trains during operation,a neural adaptive active FTC strategy based on multi-dimensional slid-ing mode surface is proposed.Based on the multiple mass-point model of HSTs subject to unknown actuator faults,the influence resulting from coupler force between adjacent carriages is taken into consideration.Considering the inaccuracy and limitation of the constant Davis'equation describing the basic running resistance,two cases for known and unknown parameters boundary are discussed,respectively.On the basis of the multi-dimensional sliding surface,an appropriate Lyapunov function is constructed in order to prove the stability of the system.Simulation results show that the proposed control strategy is feasible and effective.Then,the linear matrix inequality(LMI)-based adaptive FTC problem of HSTs un-der unknown actuator faults is handled.Aiming at the time-varying characteristics of coupler elastic coefficient,a composite active FTC strategy based on LMI is proposed for HSTs.Considering the modeling uncertainty,the multiple mass-point model for HSTs subject to under the actuator faults is linearized via the equilibrium point linearization method.However,the influence of higher-order terms on the system performance is ig-nored.Hence,the tracking control problem of HSTs can be transformed into the state stabilization problem of the error system.Based on this,an adaptive active FTC method based on hyperbolic tangent function is proposed.Simulation results verify the feasi-bility and effectiveness of the proposed control strategy which can avoid the chattering phenomenon caused by the introduction of sign function.Lastly,the adaptive H? active FTC problem for HSTs subject to unknown actuator faults and multi-source disturbances is handled.Aiming at the influence of multi-source disturbances and actuator faults on the operation safety of HSTs,an adaptive H? active fault-tolerant control strategy based on disturbance observer is proposed.Furthermore,the presented control scheme is extended to general systems.Through an improved ob-server,effects resulting from modeled disturbances with known amplitude and unknown frequency and phase can be solved.During the procedure of train operation,the gust is modeled as a harmonic signal whose unknown state can be estimated by a linear dis-turbance observer.The adaptive technique can be applied to eliminate or attenuate the effects of actuator faults.The sufficient conditions of the system stability is given in the form of LMI.Furthermore,the harmonic signal with known amplitude and unknown frequency and phase is taken into consideration.To deal with the problem,an adaptive active FTC strategy which is based on an improved disturbance observer is presented,which feasibility and effectiveness of is proved by simulation experiments.