Customized Method And Operation Strategy Of Virtual Power Plant Based On Hvac System’s Regulation Model

In recent years,there has been huge demand of peak shaving for power system due to the increasing growth of urban peak load.Meanwhile,distributed energy trading has been promoted with the great penetration of renewable generation.Heating,ventilation,and air conditioning(HVAC)system with numerous controllable variables and great flexibility potential,as a typical flexible load,can participate in peak shaving through demand response,which relieve operation stress of power system.Virtual power plant(VPP),as an aggregator of distributed energy resources(DERs)such as distributed generator,flexible load,energy storage devices and so on,provides an effective approach for DERs’ participation in power system dispatch and energy trading.Thus,VPP has been a research hot point in power system field.Nevertheless,most research associated with thermostatically controlled loads inside VPP focuses on split air conditioners;DERs are aggregated according to their characteristics and locations to form a VPP;centralized control method is used in VPP’s optimal operation.All of these mentioned above cause neglect of HVAC system’s flexibility and DER owners’ willingness,as well as huge pressure of VPP’s control center.These problems are addressed from transaction-driven and distributed optimal control aspects in this paper,which focuses on VPP’s customized formation and operation strategies based on regulation characteristics of HVAC system.Firstly,physical model of HVAC system including chiller system,water system,air system and users’ zones is established according to operating principles.Thermal dynamic process of chilled water pipes is modeled based on thermal inertia of water system.Users’ comfort index is defined from air temperature and quality aspects.Then dynamic regulation characteristics of HVAC system are analyzed in three control modes,including chilled water temperature control,air volume control and comprehensive coordination control,which provides theoretical basis for HVAC systems’ participation in VPP’s formation and operation.Besides,transaction-driven and customized formation mechanism of VPP in distributed energy trading market is proposed,with day-ahead bilateral contracts trading as the use case.Conditional value at risk(CVa R)is utilized to evaluate uncertainty risk caused by DERs’ characteristics.The trading negotiation between VPP operator and DER owners is modeled as a Stackelberg game,which contributes to the VPP’s customized formation method.Afterwards,alternating direction method of multipliers(ADMM),as a distributed optimization method,is introduced to execute power sharing among electric devices in HVAC system according to dayahead trading results,which provides optimal real-time operation points of each device.It is shown that distributed optimization can promote calculation efficiency compared with central optimization method.Lastly,a virtual energy storage model of HVAC system is proposed based on thermal inertia of both water system and the building.Internal coordinated control mechanism of HVAC system is proposed simultaneously on the basis of multi-agent system.Then the flexibility potential of HVAC system is evaluated from adjustable capacity and duration time aspects according to its virtual energy storage characteristics.HVAC virtual generation unit(VGU)is innovatively established in this paper with comprehensive coordinated control method,which realizes normalized and unified modeling of HVAC system.Next,VGUs’ invocation strategy and VPP’s operation strategy in peak shaving are proposed based on local autonomy optimization.Results of case study show that the strategies proposed in this paper can achieve local distributed optimal control of VPP’s internal resources,alleviate communication,calculation and data storage burden of VPP operator,which is of great significance for VPP’s participation in power system operation.