Research On Stochastic Optimization And Decentralized Dispatch For Integrated Electricity-Natural Gas System

Integrated Electricity-Gas System(IEGS)is not only the manifestation of modern smart grid,but also the core technology supporting Energy Internet.While the interaction between electricity and gas system improving system safety level,it also poses challenges to the analysis and scheduling of this newborn complex system.Traditionally,IEGS adopts deterministic and centralized methods for scheduling.Deterministic methods are deficient in situational awareness and risk management due to the neglect of multiple random factors.Affected by unified regulation pattern,centralized methods are inevitably weak in information protection and autonomous decision-making.Under this background,this thesis starts from the decentralized coordinated risk-based dispatch method of traditional power system,and then carries out theoretical research on the reliability assessment of integrated energy system connecting transmission-distribution levels,robust optimization of IEGS with energy storage equipment and the distributed dispatch of multi-area interconnected IEGS.The main conclusions are given as follows:(1)Facing to electricity market reform,a decentralized coordinated dispatch method for interconnected power systems is proposed,which takes economy,safety and risk into account.By introducing risk price into traditional LMP,an improved Risk-based Locational Marginal Price(RLMP)model is proposed to reflect the level of operating risk.Based on the production-consumption attributes of regional systems,a distributed risk-based dispatch framework for interconnected power systems without upper-level control center is designed.Corresponding dispatch method is also proposed.Each regional dispatch center adjusts the power flow on tie-lines through"autonomous dispatch and collaborative optimization " mechanism to maximize the economic benefits of the whole interconnected power systems.(2)Through energy hub,integrated model for IEGS connecting transmission and distribution levels is established considering both transmission-level energy system and distribution-level terminal unit.This thesis puts forward the reliability evaluation method for IEGS and designs the multi-dimensional index system including operation economy,system reliability,environmental friendliness and renewable energy accommodation.Random scenarios such as equipment failure and wind fluctuation are also considered.Mix scattered monte-carlo(MS-MC)sampling method is proposed to improve sampling efficiency.Incremental linearization method is applied for the complex natural gas flow to speed up evalutation procedure.(3)To handle the impact of multiple random contingencies on IEGS operation,a day-ahead robust optimal scheduling method considering the worst scenario is proposed,which can improve the depth and breadth of risk defense.A three-layer two-stage min-max-min mathematical optimization framework is established.In the first stage,day-ahead scheduling is formulated for normal state.In the second stage,the day-ahead startegy is adjusted according to the actual contingency scenario.In thems of methodology,Column and Constraint Generation technique(C&CG)is applied to detect the worst contingency scenario with largest natural gas adjustment,and Big-M method is applied to solve the dual lower-level bilinear problem.In addition,distributed gas storage is introduced to enhance the flexibility of regulation.(4)The optimization model and solution algorithm of distributed multi-energy flow are proposed for multi-area interconnected IEGS.Framework of decentralized coordination and the mechanism of region decomposition are also provided.In terms of modeling,a tight second-order cone reformulation for steady-state natural gas flow based on Sequential Cone Programming(SCP)is presented.In terms of methodology,the thesis puts forward Nested Alternating Direction Method of Multipliers(N-ADMM)to guarantee the convergence of distributed optimization for the primal nonconvex problem with integer variables.Finally,the feasibility and effectiveness of proposed distributed optimal multi-energy flow for multi-area interconnected systems are verified by simulations.