Research On Bidding Strategies For Electric Vehicle Aggregators Coordinating Conditional Value At Risk

With the constantly development of social economy, the contradiction between the growth of electricity demand and environmental pollution,the energy supply is increasingly outstanding. Although the renewable energy efficient utilization is beneficial to ease the tension of energy supplying, but the randomness of the renewable energy, which lead to a series of problems such as electric grid internal power unbalance, the voltage flicker etc after renewable energy access to electric grid, and make pollution to the power quality. In order to ease the contradiction between power supply and demand, developing and using the renewable energy at the same time, through controllable load aggregators aggregate geographical dispersion, large numbers controllable load resources, implement the demand side management, there is potential research value for make full use of power resources and maintain the power system reliability. Electric vehicles as a controllable load, but also a wide prospect application green transportation. With the constantly expansion of electric vehicle applications, as well as the constant improvement of advanced metering infrastructure,electric vehicle aggregators(EVAs) as an independent load in electric power, by integrating the geographical dispersion, large number of electric vehicles load resources in grid,participating in electric market, then achieve maximum economic benefits, not only beneficial demand side to participate in the market through bidding to achieve the optimal allocation of power resources, but also has important to research significance for the electricity market reform im China to setup the reasonable bidding rules.At first, this paper analyses the electric market transaction mode and type, compared the pros and cons of market bidding mechanism. Based on analyzing the principle and model of electric vehicles charging and discharging, Put forward the bidding mechanism of electric vehicles participating in market through the aggregators agent the dispersed electric vehicles in system. The realization of the trading mechanism including day-ahead market and real-time market. EVAs according to the prediction of rival bidding strategies, and the declared information of electric vehicles, with the goal of maximizing the electric vehicles charging and discharging efficiency to make bidding strategies.Under the background framework of EVAs participating in the competitive market.Based on the flexible bidding of EVAs, this paper put forward the Bi-level optimization model of participating in electricity market. The Upper-lever model is to maximize the social benefits, the lower-lever model is to maximize the electric vehicles charging and dischargingefficiency. The mixed-integer bi-level programming formulation is transformed into an equivalent single-level mixed-integer program by its Karush–Kuhn–Tucker optimality conditions. By using the nonlinear complementary function, the multi-layer optimization model can be transformed into a set of semi-smooth nonlinear algebraic equations, and through primal-dual interior point algorithm to solve the model. Finally, an example is to conducted to demonstrate the feasibility and effectiveness of the proposed model, it is provided theoretical basis for electric vehicles to participate in the competitive market.Based on the theory of EVAs participating in the competitive market, as well as the existing day-ahead market and real-time market trading mechanism. Coordinates the inflexible biddings in day-ahead market and bidding adjustment in real-time market, and proposes a bidding strategy model to minimize the expected electricity purchasing cost based on the probabilistic relationship between the bidding price and market clearing price. The model thoroughly considers the uncertainties of nodal prices in both day-ahead and real-time markets, takes into account the penalty cost due to deviations between bidding quantities in day-ahead and real-time markets in the cost function, hence, the unbalanced issues resulting from the market arbitrage by EVAs in order to pursuit of bidding profits can be avoided.Furthermore, the study converts the conditional expectation optimization problem into a convex optimization problem with condition value at risk(CVaR) constraints such that the optimal solutions are computed. Numerical simulation results validate the effectiveness and feasibility of the proposed model and algorithm, and quantify the relationship among biddings,confidence level of successful biddings, and expected costs.