Train Behavior Optimization Of Urban Rail Transit System Considering Energy Saving

Urban rail transit system is an important component of urban transportation. Both the operating mileage and the passenger volume of China’s urban rail transit system have been growth rapidly. How to build energy-saving urban rail transit systems in large passenger carriers conditions is worthy of further study.To effectively reduce the energy consumption of urban rail systems as the main traction energy, this research focuses on energy-efficient operation skills of urban rail transit system. Based on the review related studies abroad to train behavior, a neural network energy consumption calculation model of urban rail transit system is presented. Based on the utilization of regenerative braking kinetic energy, a schedule optimization model is proposed to maximum the overlapping time. Considering the account of off-peak time passengers, an optimization model of train operation adjustment is proposed. Finally, traction energy and transport efficiency evaluation model are raised for application. The main contents and conclusions are as follows:(1) The defects of three existing energy calculation models of urban transit system which are:the regression analysis model based on data (MBD), the calculation model based on electric power (MBE), and the calculation model based on kinematic methods (MBK) are systematic analyzed. Based on AHP gray correlation traction on the existing line analysis of factors affecting energy consumption, a neural network energy consumption calculation model (MBNN) is represented. The application of AHP model in an urban rail transit line demonstrates the average daily passenger traffic (0.86)> technical speed value (0.72)> monthly average temperature (0.68). The application of neural network calculation model in an urban rail transit line demonstrates the model predicted value and the actual value of the relative error of up to8.61%, the minimum is0.01%and the average relative error is3.12%, the model accuracy is acceptable.(2) The operation of train is divided into three phases including the traction process, coasting process and braking process. Through a detailed ananlysis of train operation process, the overlapping time is defined clearer as the common stages of train traction process and regenerative braking process, which is an effective way to improve energy efficiency. Combined with safe driving demand for urban rail transit system in line with proposed urban rail transport constraints operating characteristics, in order to maximize the overlap time as the objective function, build optimization based on the use of regenerative braking can schedule model of urban rail transit system, and a case subway line optimization results and compare the difference between the original timetable. The results showed that:when the minimum tracking interval assignment is90seconds, the total overlap time is optimized for119seconds,95seconds than the original schedule of overlap time increase compared to25.3%, indicating a better applicability. The model deepens the existing studies analyzing the braking process, more in line with the actual operation.(3) By train stops to analyze specific components division, and off-peak periods in the passenger analysis on the basis of the time required to construct a divided decision variables between division and station stops running, the arrival time of a single minimum deviation trains to optimize traction and minimum energy consumption target optimization model train operation adjustment, and a subway line to case studies to optimize the degree of optimization potential train stops and the traction energy division. The results showed that:when the weights of both the deviation of arriving time and the energy consumption are0.5, the total stop-time decreases from325s to189s, the decline in the136s; which add into all stations running-time, therefore, the traction energy consumption are decreased by62.07kWh.(4) Two traction energy and transport efficiency evaluation models are proposed based on the traction energy and transport efficiency rate changing mechanism with different technical conditions. The models is capable to calculate the traction energy consumption (EC) per100vehicle-kilometer, the traction energy consumption (EC) per10000person-kilometer, the time consumption (TC) per100vehicle-kilometer, the time consumption (TC) per10000person-kilometer, the operation cost (OC) per vehicle-kilometer, the operation cost (OC) per person-kilometer. The application of the approach in two metro trains demostrates that when the technology speed is lower than25km/h, traction energy consumption decreases slightly with increasing of technology speed; when the technology faster than35km/h, the traction energy consumption growth rapidly with increasing of technology speed. Time consumption decreases with increasing of technology speed. Consumption per vehicle kilometer time using the measure of time, the same technology speeds, different load factor corresponding to the percentage of time the car was no significant difference in consumption kilometers. While the use of time-consuming indicators million kilometers, the same technique speeds, lower load factor, the greater the value of the unit of time consuming workload.