Jiateng Yin

Intelligent Train Operation Algorithms for Subway by Expert System and Reinforcement Learning

Current research in automatic train operation concentrates on optimizing an energy-efficient speed profile and designing control algorithms to track the speed profile, which may reduce the comfort of passengers and impair the intelligence of train operation. Different from previous studies, this paper presents two intelligent train operation (ITO) algorithms without using precise train model information and offline optimized speed profiles. The first algorithm, i.e., ITOe, is based on an expert system that contains expert rules and a heuristic expert inference method. Then, in order to minimize the energy consumption of train operation online, an ITOr algorithm based on reinforcement learning (RL) is developed via designing an RL policy, reward, and value function. In addition, from the field data in the Yizhuang Line of the Beijing Subway, we choose the manual driving data with the best performance as ITOm. Finally, we present some numerical examples to test the ITO algorithms on the simulation platform established with actual data. The results indicate that, compared with ITOm, both ITOe and ITOr can improve punctuality and reduce energy consumption on the basis of ensuring passenger comfort. Moreover, ITOr can save about 10% energy consumption more than ITOe. In addition, ITOr is capable of adjusting the trip time dynamically, even in the case of accidents.

Efficient Real-Time Train Operation Algorithms With Uncertain Passenger Demands

The majority of existing studies in subway train operations focus on timetable optimization and vehicle tracking methods, which may be infeasible with disturbances in actual operations. To deal with uncertain passenger demands and realize real-time train operations (RTOs) satisfying multiobjectives, including overspeed protection, punctuality, riding comfort, and energy consumption, this paper proposes two RTO algorithms via expert knowledge and an online learning approach. The first RTO algorithm is developed by a knowledge-based system to ensure the multiple objectives with a constant timetable. Then, by considering uncertain passenger demand at each station and random running time errors, we convert the train operation problem into a Markov decision process with nondeterministic state transition probabilities in which the aim is to minimize the reward for both the total time delay and energy consumption in a subway line. After designing policy, reward, and transition probability, we develop an integrated train operation (ITO) algorithm based on Q-learning to realize RTOs with online adjusting the timetable. Finally, we present some numerical examples to test the proposed algorithms with real detected data in the Yizhuang Line of Beijing Subway. The results indicate that, taking the multiple objectives into account, the RTO algorithm outperforms both manual driving and automatic train operations. In addition, the ITO algorithm is capable of dealing with uncertain disturbances, keeping the total time delay within 2 s and reducing the energy consumption.

Data-driven train operation models based on data mining and driving experience for the diesel-electric locomotive

Traditional control methods in automatic train operation (ATO) models have some disadvantages, such as high energy consumption and low riding comfort. To alleviate these shortcomings of the ATO models, this paper presents three data-driven train operation (DTO) models from a new perspective that combines data mining methods with expert knowledge, since the manual driving by experienced drivers can achieve better performance than ATO model in some degree. Based on the experts knowledge that are summarized from experienced train drivers, three DTO models are developed by employing K-nearest neighbor (KNN) and ensemble learning methods, i.e., Bagging-CART (B-CART) and Adaboost.M1-CART (A-CART), into experts systems for train operation. Furthermore, the DTO models are improved via a heuristic train parking algorithm (HPA) to ensure the parking accuracy. With the field data in Chinese Dalian Rapid Rail Line 3 (DRRL3), the effectiveness of the DTO models are evaluated on a simulation platform, and the performance of the proposed DTO models are compared with both ATO and manual driving strategies. The results indicate that the developed DTO models obtain all the merits of the ATO models and the manual driving. That is, they are better than the ATO models in energy consumption and riding comfort, and also outperform the manual driving in stopping accuracy and punctuality. Additionally, the robustness of the proposed model is verified by a number of experiments with some steep gradients and complex speed limits.

A flexible and robust train operation model based on expert knowledge and online adjustment

Most existing automatic train operation (ATO) models are based on different train control algorithms and aim to closely track the target velocity curve optimized offline. This kind of model easily leads to some problems, such as frequent changes of the control outputs, inflexibility of real-time adjustments, reduced riding comfort and increased energy consumption. A new data-driven train operation (DTO) model is proposed in this paper to conduct the train control by employing expert knowledge learned from experienced drivers, online optimization approach based on gradient descent, and a heuristic parking method. Rather than directly to model the target velocity curve, the DTO model alternatively uses the online and offline operation data to infer the basic control output according to the domain expert knowledge. The online adjustment is performed over the basic output to achieve stability. The proposed train operation model is evaluated in a simulation platform using the field data collected in YiZhuang Line of Beijing Subway. Compared with the curve tracking approaches, the proposed DTO model achieves significant improvements in energy consumption and riding comfort. Furthermore, the DTO model has more advantages including the flexibility of the timetable adjustments and the less operation mode conversions, that are beneficial to the service life of train operation systems. The DTO model also shows velocity trajectories and operation mode conversions similar to the one of experienced drivers, while achieving less energy consumption and smaller parking error. The robustness of the proposed algorithm is verified through numerical simulations with different system parameters, complicated velocity restrictions, diverse running times and steep gradients.

Parallel Control and Management for High-Speed Maglev Systems

This paper puts forward a systems approach for the parallel control and management of the high-speed maglev system (HMS). An artificial HMS is first established by using a multiagent-based technique, and we demonstrate its consistence with the actual HMS. We then conduct some computational experiments and summarize some operational rules for the artificial HMS. Finally, the parallel control and management for the HMS are achieved by parallel execution of the artificial and actual HMSs with parallel interactions between them. We evaluate our approach overall by ensuring the safety and reliability of the HMS through parallel control and management. The solutions and recommendations for the safety control and effective management of the HMS can be provided by the proposed approach.

Online adjusting subway timetable by q-learning to save energy consumption in uncertain passenger demand

Current researches in subway train operation concentrate on timetable optimization and real-time tracking methods, which may be infeasible with disturbances in actual operation. To overcome stochastic negative effects caused by disturbances and realize energy-efficient train operation, we propose a comprehensive model to integrate train operation with real-time rescheduling. The proposed model focuses on minimizing the reward for both the total time-delay and energy-consumption with intelligent decision support. After designing policy, reward and transition probability, we develop an Intelligent Train Operation (ITO) algorithm based on Q-learning to calculate the optimal decisions. Simulation results with field data in Beijing subway Yizhuang Line demonstrate the effectiveness and efficiency of ITO approach.

High-speed maglev parallel control and management system-overview and framework

Firstly, this paper puts forward the framework of high-speed maglev parallel control and management system, then builds a high-speed maglev artificial system model on the basic of multi-agent which is consistent with actual high-speed maglev system. Secondly, we do computing experiments and summarize the law of the system on the basic of high-speed maglev artificial system. Lastly, we can achieve parallel controlling and managing actual high-speed maglev system through the parallel interaction between artificial and actual high-speed maglev system. Ensuring the safety and reliability of the high-speed maglev system, we evaluate the high-speed maglev system from overall, management and implementation, then provide solutions and recommendations for the safety control and effective management of the high-speed maglev system.

Constraint local principal curve

Existing principal curve algorithms have some drawbacks such as time consuming and narrow application scope in practice, since these algorithms are mainly based on global optimization. In this paper, we present the concept of Constraint Local Principal Curve (CLPC), which uses local optimization methods and restricts the principal curve with two fixed endpoints to reduce the computational complexity. In addition, we propose three CLPC algorithms by Local Optimization and Adaptive Radius to expand the range of applications and increase the solution quality. The first algorithm, i.e., CLPCg is based on greedy thinking. The second algorithm, i.e., CLPCs uses one dimensional search and the last algorithm CLPCc combines the greedy thinking and one dimensional search. Then, we define six performance indices to evaluate the performance of the CLPC algorithms. Finally, we present some numerical experiments with three simulation data sets and two GPS measured data sets in both highway and railway. The results indicate that all of the three CLPC algorithms can obtain high-accuracy data from multiple low-accuracy data efficiently. The CLPC algorithms can improve the accuracy and computational speed compared with the existing K-segment principal curve (KPC) algorithm. In addition, CLPCc outperforms CLPCg and CLPCs according to the comprehensive experiments while CLPCg runs much faster than other ones. We present the concept of Constraint Local Principal Curve (CLPC) to reduce the computational complexity.We propose three CLPC algorithms that combine local optimization and adaptive radius to expand the range of applications and increase the solution quality.We present some numerical experiments with three simulation data sets and two measured GPS data sets in highway and railway.The CLPC algorithms can improve the accuracy and computational speed compared with the existing KPC algorithms.The features of the each CLPC algorithm are analyzed according to the comprehensive experiments.