On Adaptive Iterative Learning And Data Driven Iterative Learning Control Methods In Subway Train Operation Control

In this thesis,novel adaptive iterative learning control and data driven iterative learning control schemes are investigated for single/multiple subway trains operation control.The practical issues of modeling difficulty,speed constraint,actuator faults,and multi-train cooperative fault-tolerant operation control are addressed.Taking advantage of its repetitive running characteristics,the precise tracking control of the subway train position and speed is achieved,and the safe and reliable operation of the train is ensured.The main contributions of this thesis are as follows:1.Firstly,based on the characteristic of the subway train system repetitive operation pattern and the multiple-point-mass dynamic model,a novel adaptive iterative learning control algorithm is proposed in this thesis for subway trains.A composite energy function technique is utilized to obtain the asymptotic convergence of tracking error in the iteration axis for the proposed controller.Secondly,a speed constraint adaptive iterative learning control algorithm is also designed based on the subway train multiple-point-mass dynamic model to avoid over speed,derailment and collision of the subway train for the subway train over-speed protection.Then,a barrier composite energy function method is used to ensure that the subway train speed and position tracking errors asymptotically converge to zero along the iteration axis.Finally,two simulation examples are given for the subway train system to verify the effectiveness and robustness of the proposed control method.2.A radial basis function neural network-based adaptive iterative learning fault-tolerant control(RBFNN-AILFTC)algorithm is developed for subway trains subject to the time-iteration-dependent actuator faults and speed constraint by using the multiple-point-mass dynamic model.First,the RBFNN is utilized to approximate the time-iteration-dependent unknown nonlinearity of the subway train system.Then,the iterative learning mechanism is used to tackle the outstanding repetitive operational pattern of a subway train,and the adaptive mechanism is designed to deal with the unknown factors caused by time-iteration-varying nonlinearity of the subway train.Second,a barrier composite energy function technique is exploited to obtain the convergence property of the proposed RBFNN-AILFTC scheme for subway train system,which can guarantee that the tracking error is asymptotic convergence along the iteration axis,meanwhile keep the speed profile of the subway train system satisfies the constraint.Finally,compared with the existing train control algorithm,a subway train simulation is shown to verify the effectiveness and advantages of the proposed fault-tolerant control algorithm.3.A cooperative adaptive iterative learning fault-tolerant control(CAILFTC)algorithm with the radial basis function neural network(RBFNN)is proposed for multiple subway trains subject to the time-iteration-dependent actuator faults by using the multiple-point-mass dynamic model.A composite energy function(CEF)technique is applied to obtain the convergence property of the presented CAILFTC,which can guarantee that all train speed tracking errors are asymptotic convergence along the iteration axis.Meanwhile,the headway distances of neighboring subway trains are kept in a safety range.Finally,the effectiveness and superiority of the presented CAILFTC method is verified through a subway train simulation.4.An adaptive iterative learning fault-tolerant control(AILFTC)algorithm is presented for state constrained nonlinear systems with randomly varying iteration lengths subjected to actuator faults.First,the modified parameters updating laws are designed through a new defined tracking error to handle the randomly varying iteration lengths.Secondly,the RBFNN method is used to deal with the time-iteration-dependent unknown nonlinearity and a barrier Lyapunov function is given to cope with the state constraint.Finally,a new barrier composite energy function(BCEF),consisting of barrier Lyapunov function with a new defined tracking error,parametric learning error and the information of actuator faults,is constructed to handle the concurrently randomly varying iteration lengths,constraints and actuator faults,which can achieve the tracking error convergence of the presented AILFTC algorithm along the iteration axis with the state constraint,and then followed with the extension to the high order case.A simulation for a single-link manipulator is given to illustrate the effectiveness and advantage of the proposed AILFTC algorithm compared with P-type iterative learning control(PILC)and adaptive iterative learning control(AILC).5.For the problems of modeling difficulty,large amount of calculation,high cost and measurable/unmeasurable disturbances in the conventional model based control algorithm of subway trains,two data driven point-to-point iterative learning control(ILC)methods are proposed.Firstly,for the measurable disturbances,a robust data driven optimal point-to-point ILC algorithm is designed by only utilizing input/output(I/O)inforamtion of the train system.The tracking task requires that the control input is updated according to the prespecified measured multiple-point tracking error values rather than the complete output trajectory,which can reduce computational cost.Then,rigorous analysis is developed which demonstrates that the train tracking error is monotonic bounded convergence.Secondly,the design and convergence analysis of point-to-point ILC controller with unmeasurable disturbances are also given.Finally,a simulation is conducted for train system to verify the effectiveness and correctness of the presented control method.6.For the problems of the difficulty of establishing accurate subway train dynamic model,tracking accuracy of position and speed curves,and input and output constraints,a modular data-driven control algorithm is designed for the subway train operation control system.Firstly,under the speed and traction/braking force constraints,a double loop controller(PFDL-MFAC-ILC)is designed for subway trains,which consists of PILC feedforward control loop and partial form dynamic linearization based model free adaptive control(PFDL-MFAC)feedback control loop.The main function of the feedback PFDL-MFAC is to stabilize the train control system,and the feedforward ILC is used to improve the control performance of the train system which can achieve the train accurate tracking control.It is a modular design method to realize the complementary advantages of MFAC and ILC.Secondly,for the influence of the time-iteration-dependent measurement disturbances on the subway train system,a double-loop data-driven robust controller is also investigated.Finally,the effectiveness of the proposed data-driven control algorithm is verified by a subway train system simulation.