Research On The Operation Strategy Of A Virtual Power Plant Formed By Multiple Demand-side Entities In Power Market

In recent years,distributed generation,storage and demand response resources are prevailing on demand-side in response to the demands on more reliable and economical power supply,as well as low-carbon economy.The demand-side entities who own or operate those distributed resources have been emerging with the strong support of the government policies.However,problems of few profit channels and long investment return cycle are bothering those demand-side entities,as the subsidy of distributed resources decreases.Therefore,it is urgent for the demand-side entities to participate in power market for profit.According to the market regulations,the distributed resources of the demand-side entities are nevertheless not allowed to bid in power market because of their trivial scale.Furthermore,the fact that distributed resources used to be installed in a “fit and forget” way,and renewable resources are accused of their volatility and uncertainty.These technically deterred demand-side entities from power market.Hence,virtual power plant was proposed,which can widely aggregate resources of demand-side entities to realize multi-energy complement and increase the total scale so as to participate in the power market and offer demand response support to the grid.The National Grid has endeavored the construction of power internet of things,one important application of which is the virtual power plant.Advanced sensor,smart meter,edge computing and smart grid technology will technically support the realization of virtual power plant.In this paper,uncertainty is modeled by interval optimization,and price maker virtual power plant bidding strategy model is proposed.The profit is optimized considering the game among the market entities.The formed virtual power plant is obtained by the game among the demand-side entities with the proposed demand-side trading mechanism.The major contributions are as follows:Firstly,considering the uncertainty modeling problem,this paper proposes a comprehensive contract and day-ahead market bidding strategy optimization model based on interval optimization.Due to the aggregated multiple flexible resources,virtual power plant can provide flexible reserve and ramping interval.Therefore,the uncertain variables and adjustable interval are presented by interval numbers and pessimism degree based interval optimization is applied.Meanwhile the flexibility characteristics of available reserve and ramping interval are evaluated.Compared to conventional scenario tree generation in stochastic optimization,interval optimization curtails the computational burden and has no assumption for probability distribution function.The results show that the contracted energy and bid in day-ahead market of virtual power plant are optimized based on the market price and uncertainty acceptance level,and the virtual power plant’s flexibility can be properly evaluated.Secondly,since the demand-side entities are more widely aggregated,virtual power plant may influence the market clearing price.Accordingly,a two-level bidding strategy model of virtual power plant is proposed,which is considered as price maker.The upper level problem includes the trading gap mitigation mechanism to realize the coordination operation of demand-side resources,while the lower level problem optimizes the market clearing process.The model is transformed into one level mathematical problem with equilibrium constraints by Karush-Kuhn-Tucker conditions.The problem is solved after linearization.Via coordinating various controllable resources,this mechanism can effectively eliminate the deviation of traded power caused by uncertainty of renewables,leading to the optimized profits.Thirdly,given that there are multiple strategic entities in the day-ahead market,a market equilibrium model is proposed,involving virtual power plant.Each entity forms a mathematical problem with equilibrium constraints and all the problems jointly become an equilibrium problem with equilibrium constraints.The problem is solved by strong stationary conditions in the target of either generating company profit maximization or social welfare maximization.As such,the profit of virtual power plant is obtained under market equilibrium The results indicate that the impact of virtual power plant upon market clearing can not be neglected,as the scale of virtual power plant increases with more demand-side resources integrated.Lastly,considering the problem of profit allocation and the virtual power plant formation of demand-side entities by coalitional game,A Shapley value based profit interval allocation approach is proposed to calculate the pessimism degree profit of each demand-side entity.The results reveal that the total profit of virtual power plant formed by demand side entities is higher than the accumulated profit of the individually operating entities.This approach can also serve to describe the marginal contribution of each entity to the midpoint and uncertainty of coalitional profit.Nonetheless,due to the problem of the assumptions in application of Shapley value allocation approach,a trading mechanism based on dynamic coalitional game is proposed according to the characteristics of demand-side entities,which includes an output and uncertainty contribution based allocation method.This approach allows the demand-side entities to repeatedly change their coalitional strategy until the stable coalition status is achieved.Having analyzed the domination relationship,effective relationship and preference relationship of dynamic coalition game theory,the virtual power plant structure is deduced by the game among the demand-side entities.To summarize,this study explored the bidding strategy,profit allocation and coalition formation problem of virtual power plant.It is found that the demand-side entities can optimize the coordination operation of demand-side resources and exercise market power to improve profit by forming virtual power plant,and application mechanism is designed for this approach.