Optimal Sizing of Stationary Energy Storage System for Traction System

Loo, Huan Min (2020) Optimal Sizing of Stationary Energy Storage System for Traction System. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Loo Huan Min.pdf Restricted to Registered users only Download (8MB)

Abstract

Regenerative braking energy (RBE) is used to improve the energy efficiency in electrical traction system. In order to store the energy recovered from RBE, many different type of energy storage systems (ESS) are used. One of the design issues is the sizing of the ESS. The implementation of the same capacity of ESS in all the substations may cause the ESS in some station to experience overvoltages or under-voltages due to the total regenerative energy produced between each station is different and not constant. Besides that, the implementation of large capacity will increase the capital cost and the cause under-voltage in some station, especially in the low demand substation. This project has determined the most suitable sizing for a particular substation. This paper aimed to determine an efficient method to optimise the size of ESS for RBE in railway system without affecting the performance. The sizing model is developed in MATLAB/Simulink for the storage of regenerative energy produced between the stations. In this paper, genetic algorithm was used to determine the optimum sizing of HESS. The fitness equation of genetic algorithm has form to guide the simulation toward minimum cost of power for HESS based on the running cost of the traction system. The limitation of the traction system such as catenary (third-rail voltage) and limitation of battery chemistry have considered into the optimization process. The total power absorbs by the HESS is same as the maximum regenerative energy from the traction system in order to make the cost of installation of HESS is reasonable and minimize the energy losses. There are three operators that has considered in genetic algorithm which are selection, crossover and mutation to produce a better result compare to previous result. This process will keeps on iterating until the fitness result is found. The result from genetic algorithm has been validated by applying into the sizing model in MATLAB/Simulink. Besides that, the result from optimization has compare with other sizing of HESS using the same speed profile of train to analysis the energy efficiency of the HESS. It is found that the energy efficiency obtained from the optimal sizing of HESS is above 80% whereas the random sizing of HESS is below 80%.