摘要
The optimization of the control strategy of a plug-in hybrid electric bus(PHEB) for the repeatedly driven bus route is a key technique to improve the fuel economy. The widely used rule-based(RB) control strategy is lacking in the global optimization property, while the global optimization algorithms have an unacceptable computation complexity for real-time application. Therefore, a novel hybrid dynamic programming-rule based(DPRB) algorithm is brought forward to solve the global energy optimization problem in a real-time controller of PHEB. Firstly, a control grid is built up for a given typical city bus route, according to the station locations and discrete levels of battery state of charge(SOC). Moreover, the decision variables for the energy optimization at each point of the control grid might be deduced from an off-line dynamic programming(DP) with the historical running information of the driving cycle. Meanwhile, the genetic algorithm(GA) is adopted to replace the quantization process of DP permissible control set to reduce the computation burden. Secondly, with the optimized decision variables as control parameters according to the position and battery SOC of a PHEB, a RB control is used as an implementable controller for the energy management. Simulation results demonstrate that the proposed DPRB might distribute electric energy more reasonably throughout the bus route, compared with the optimized RB. The proposed hybrid algorithm might give a practicable solution, which is a tradeoff between the applicability of RB and the global optimization property of DP.
The optimization of the control strategy of a plug-in hybrid electric bus （PHEB） for the repeatedly driven bus route is a key technique to improve the fuel economy. The widely used rule-based （RB） control strategy is lacking in the global optimization property, while the global optimization algorithms have an unacceptable computation complexity for real-time application. Therefore, a novel hybrid dynamic programming-rule based （DPRB） algorithm is brought forward to solve the global energy optimization problem in a real-time controller of PHEB. Firstly, a control grid is built up for a given typical city bus route, ac- cording to the station locations and discrete levels of battery state of charge （SOC）. Moreover, the decision variables for the energy optimization at each point of the control grid might be deduced from an off-line dynamic programming （DP） with the historical running information of the driving cycle. Meanwhile, the genetic algorithm （GA） is adopted to replace the quantiza- tion process of DP permissible control set to reduce the computation burden. Secondly, with the optimized decision variables as control parameters according to the position and battery SOC of a PHEB, a RB control is used as an implementable control- ler for the energy management. Simulation results demonstrate that the proposed DPRB might distribute electric energy more reasonably throughout the bus route, compared with the optimized RB. The proposed hybrid algorithm might give a practicable solution, which is a tradeoff between the applicability of RB and the global optimization property of DP.
基金
supported by the National Natural Science Foundation of China(Grant No.51275557,5142505)
the National Science-Technology Support Plan Projects of China(Grant No.2013BAG14B01)
关键词
全局优化算法
电动公交车
混合动力
能源优化
动态编程
实时应用
规则基
实时控制器
plug-in hybrid electric bus （PHEB）, control strategy optimization, dynamic programming （DP）, genetic algorithm （GA）,city bus route
