Modelling And Real-time Algorithm On Energy-Efficient Operation For Freight Train With Diesel Locomotive

Reducing the energy consumption and emissions of rail vehicles is one of the main concerns of today's railway industry. However, train operation optimization is a typical time-varying problem with such characteristics as multi-objective, multi-constrictions, and non-linear. Under the existing railway condition of China, its performances largely depend on drivers'operational proficiency. Consequently, it is essential to explore the real-time energy-efficient operation algorithms under dynamic conditions sheds light on an energy-saving, safe, and punctual running of the train. This also sets a foundation for Auto Train Operation study.On the basis of domestic and international achievements, this study probes into reducing energy consumption by means of improved train control. A real-time optimization algorithm for train operation is proposed, and an on-board driving aided system is presented. The following components are involved:1. The forms of train energy consumption are explained by work analysis of locomotive traction, which includes work against running resistance and the kinetic energy loss caused by apply the brake. The core of energy-efficient operation is illustrated with theoretical analysis, such as ensuring running speed equilibrium and avoiding unnecessary brake. The simulation result also demonstrates that 6.8% energy consumption can be reduced by a reasonable control of upper and lower limit of running speed in the section; the 9% energy consumption decrease with a 0.5% time increase are obtained by reducing unnecessary brake and extending the coasting distance.2. A fuzzy predictive model for energy-efficient running is developed with fuzzy inference and predictive control theory. The fuzzy rule set is designed considering running speed, target speed, and modified gradient. An on-line optimization algorithm is presented based on moving horizon strategy, energy consumption and schedule delays are selected as objectives, and the constraints include safety, longitudinal impulse, and operating regulations factors. The case study indicates that the proposed algorithm performs well under different railway conditions. Running on the long sharp slope, the applied high power traction is capable of a high-speed pass, which greatly guarantees speed equilibrium. In the area of heavy down slope and front of speed-limited section, the kinetic energy loss can be effectively reduced by replacing the handle to coasting position.3. The key point of speed regulating braking is to determine the initial point and braking releasing point. The optimal strategy is to reduce kinetic energy loss, satisfying the constraints of lowest releasing speed and avoiding second brake regulating. With illustration of the air-braking crucial factors and its constraints, a new bi-level fuzzy neural network model of train stop braking is formulated. The initial control variable associated with initial speed and braking distance are first provided by sample training. Considering the influences from calculation error and gradients in the forward section, the correction values of the control variable are obtained by fuzzy inference. In which, the approach of "four position and three step braking" is presented improve the braking accuracy. According to the results of simulation case, the fuzzy network control method is proved to be more effective, especially that it ensures the safe, stable, accurate and energy-saving train brake with the real operational constraints.4. The designing procedure and crucial techniques of the on-board system are addressed for energy-efficient train operation. Sampling feedback compensation is able to diminish the impact of calculation errors on the effectiveness of recommend proposal. Serial communication module is designed to share the train dynamic information, and the control approaches for shielding the noisy data and kilometer post conversion are presented, which makes a good basic-data fusion for railway lines data.5. The on-board system for train energy-efficient operation with diesel locomotive is developed in accordance with the railway condition of China, and its effectiveness is illustrated by wide field tests. Referring to the strategies like reaching steep uphill slope with high speed, extending coasting distance in steep downhill section, and avoiding unnecessary brake, the system provides real-time optimized operating schemes for train drivers and insures an efficient, safe, punctual, and stable running. The energy consumption is cut by 5.88%, and the safe and relatively low-cost operation for freight train with diesel locomotive is realized.