Energy-saving Optimization And Energy Evaluation Of Train In Urban Rail Transit

In urban rail transit, the problem of energy consumption is getting more and more serious, especially that in train traction. Nowadays, a target speed curve is tracked by train when operating during most of urban rail transit systems, and so, it is one of the most effective ways of reducing the energy consumption of train traction to study energy-saving target speed curve to guide the train operation. But the calculation of target speed curve must have a better real-time performance, which is determined by the characteristic of short distance between stations and frequent acceleration and deceleration in urban rail transit. Therefore, this thesis is to research and explore a fast and effective real-time algorithm for optimization of train energy saving.Two stage optimizations is proposed to design train energy-saving optimization algorithm in this thesis by using discrete combinatorial optimization model with ant colony optimization of MAX-MIN Ant System (MMAS) as the core algorithm. With low density discretization, the first stage optimization could seek energy-saving target speed curve widely and provide optimized guidance for the second. And with high density discretization, the second stage optimization could adequately use information of the first and optimize the result fast to realize the real-time algorithm. Moreover, the operation experience of train driver has been considered to design the optimization algorithm, which makes the algorithm more effective.The MMAS optimization algorithm is realized by using MATLAB and also proceeded simulation and checking. The optimized result shows that train energy is lower than before by tracking the optimized target speed curve, which explains that the algorithm has a good energy-saving effect. And a better real-time performance is reflected with the fast optimization of the second stage. Moreover, Beijing metro Yizhuang line data from Rongjing East Street to Wanyuan Street has been taken and simulated, and the result is better than train test operation. Therefore, the optimized result of algorithm could have a major reference value for the operation of a train in the real line.