| The renewable energy source(RES),as an important way to maintain development sustainability and realize the goal of carbon peaking and carbon neutralization,has received more and more attention from the world.Compared with traditional power generation sources,distributed RES has advantages in prioritizing local consumption and avoiding power loss caused by long-term transmission,which demonstrates the desired application in the active distribution network(ADN)and micro-grids(MG).Besides,under the background of the new electricity reform,ADN and MGs may belong to different stakeholders in the future.The access of RES enables MGs to sell electricity to ADN,which changes MG’s or ADN’s market role from traditional buyers or sellers to prosumers and realizes the energy transactions among multi-stakeholders.Aiming at the collaborative dispatch and planning of MGs and AND within a market environment,this paper establishes the reliability-constrained planning model for ADN,and collaborative dispatch and planning method considering bilateral transactions for multi-agents.The research focus includes the transaction mechanism,privacy issues,maximizing social welfare,fairness,the existence of Nash equilibrium,etc.Following work has been accomplished.(1)An reliability-constrained ADN optimal planning model is proposed.An improved K-means clustering algorithm based on peak density method is used to generate typical scenes of wind,photovoltaic,and load in ADN,which achieves a balance between uncertainty and solution time.The optimal planning model of AND takes the sum of the annual investment cost,annual power purchase cost and RES maintenance cost as the goal,considering the RES configuration and operation constraints,ADN power flow constraints,and reliability constraints.The convex optimization method is further used to convert the model into a mixed-integer second-order cone program for solving(2)A day-ahead coordinated dispatch model of multiple MGs and AND considering bilateral trading is proposed,where AND indirectly participates in the bilateral transaction market by charging network fees.A two-stage market has been established in the dispatch model,where the first stage maximizes the social welfare to clear the transaction power and the second stage adopts the Nash bargaining solution(NBS)to distribute the profit and ensure the fairness of multi-agents.The analytical target cascading(ATC)method and closed-form alternating direction multiplier method(CF-ADMM)is used to solve the two-stage problem to protect the privacy of each stakeholder.(3)A collaborative planning model of multiple MGs and AND based on generalized Nash bargaining and alternating optimization is proposed.To solve the non-convexity,the alternating optimization procedure transforms the planning model into a bi-level iterative problem.In each iteration,the upper level utilizes the generalized Nash bargaining solution(GNBS)to encourage investors participating in the bilateral trading market to make profits and determines the transaction power and payments.The lower level is the local optimization problem of each investor to determine the 0-1 variables for the upper level.It is proved that the proposed bi-level iterative framework takes into account fairness and social welfare.Similarly,distributed algorithms are used to deal with privacy issues in the solution process. |
