Research On Energy Management Strategy For Parallel Hybrid Bus Based On Dynamic Programming

Designing energy management strategies is a major task when developing hybridelectric vehicles, a “good” strategy can effectively reduce fuel consumption byimproving the power system efficiency. However, doing this work is faced with manychallenges: On the one hand,due to strong-nonlinearity of power system,the optimalcontrol actions are very complicated as driving conditions changed; on the other hand,the fuel consumption can be influenced both by the strategy quality and thehybridization ratio (HR),this means strategy optimization and HR value selection areinherently and intimately coupled. In this paper,a parallel hybrid bus is selected andmany typical driving cycles are employed,then a powerful method is proposed toconduct optimization work based on dynamic programming(DP),aiming at exploringthe energy distribution rule and mining the potential of fuel-saving, providingpractical guidance and theoretical basis for strategies design and HR selection. Themain contents include:1, In view of the energy distribution rule of hybrid power system iscomplicated due to the uncertain driving conditions, carry out the research on globaloptimization for energy management strategy based on DP: on the premise ofknowing driving cycles, a process design is completed for optimizing control strategywhich can minimize the cumulative fuel consumption;With the help of MATLABsoftware platform, a longitudinal dynamics vehicle model is set up based on a hybridelectric bus, and then the DP algorithm is developed to find the optimal controlactions which can minimize the cumulative fuel consumption while sustaining theSOC charged balanced (the initial SOC is equal to the final SOC during a drivingcycle). 2, In view of developing energy management strategy and selecting HR areinherently and intimately coupled, carry out the research on optimizing hybridizationratio based on DP: through analyzing the power system characteristics of theprototype vehicle, a variable hybridization ratio with range of (0,1) vehicle model isdeveloped,which the variable parameters are motor peak power, battery capacity andengine displacement;In order to find the optimal control actions under different HR,the completed DP algorithm is applied for the Chinese City Bus Cycle (CCBC), theUSA Urban Dynamometer Driving Schedule for Heavy Duty Vehicle (UDDSHDV)and the British City Bus Cycle (UKBUS6) respectively.By this way, not only thefuel-saving potential is discovered,but also the coupled problem between developingenergy management strategy and selecting HR is solved;Finally, the performanceevaluation under different HR is performed which respected to the three drivingcycles.The evaluation index is the cost per100km,including the fuel consumptioncost and the power system cost (the power system cost is got by converting the totalcost to100km using the life driving distance in proportion) the results show that forthe three cycles,the optimal HR is0.34,0.27and0.40respectively.3, In view of the energy management strategy developed by DP cannot beapplied online, carry out the research on designing online strategy by extracting rulefrom DP results for the three cycles with respective optimal HR, then the onlineenergy management strategies are developed.Verification results show that the SOC,engine and motor torque track DP results well while the fuel consumption per100kmincreased by5.74%,6.02%and3.26%than DP. Above results indicate that the onlinestrategies have considered the complexity of energy distribution.