基于模型预测控制的双机组混合动力船舶能量管理研究.pdfVIP

基于模型预测控制的双机组混合

动力船舶能量管理研究

摘要:［目的］船舶柴电混合动力系统合理分配柴油机和电机输出功率，可大幅降低油耗和排放。针对传

统混合动力能量管理策略的性能最优与运算实时的矛盾，提出采用模型预测控制（MPC）进行能量管理瞬时

优化。［方法］首先，利用反向建模法搭建由双柴油发电机组、储能系统、岸电组成的客渡船混合动力系统

能量流模型；然后，提出以油耗和电能消耗的总温室气体（GHG）排放为目标函数，在系统约束条件下可在线

滚动优化求解的MPC能量管理算法，最终进一步进行了不同预测时域长度的灵敏度分析。［结果］仿真结

果表明，MPC较传统规则控制方法可分别降低4.85%的燃油消耗量和3.54%的温室气体总排放。［结论］相

比于传统的规则控制策略，MPC油耗低，温室气体排放少，且计算负荷较小，具有良好的实船应用潜力。

关键词：双机组混合动力；模型预测控制；温室气体排放

Studyonenergymanagementofdual-dieselgeneratorsetshybridpowerships

basedonmodelpredictivecontrol

Abstract:［Objectives］Themarinediesel-electrichybridsystemreasonablydistributestheoutputpowerof

thedieselengineandmotor,whichcansignificantlyreducefuelconsumptionandemissions.Aimingatthe

contradictionbetweentheoptimalperformanceandreal-timeoperationoftraditionalenergymanagement

strategiesappliedtohybridpowersystems,thisstudyproposestoimplementmodelpredictivecontrol(MPC)

toachievetheinstantaneousoptimizationofenergymanagement.［Methods］First,anenergyflowmodelof

apassenger-ferryhybridsystemconsistingofdualdieselgeneratorsets,energystoragesystemsandshore

powerisestablishedbyemployingthereversemodelingmethod.AnMPCenergymanagementalgorithmthat

canbesolvedonlinebyrollingoptimizationundersystemconstraintsisthenproposed,takingthetotalgreen-

housegas(GHG)emissionsoffuelconsumptionandelectricenergyconsumptionastheobjectivefunction.Fi-

nally,thesensitivityanalysisofthevariablepredictionhorizonlengthsiscarriedout.［Results］Thesimula-

tionresultsshowthattheMPCmethodcanreducefuelconsumptionby4.85%andtotalcarbondioxideemis-

sionsby3.54%respectivelycomparedwiththetraditionalrule-basedcontrolmethod.［Conclusions］The

MPCmethodachieveslowerfuelconsumption,lowercarbon