Font Size: a A A

Research On Security-Constrained Economic Dispatch Of Power System With Volatile Wind Power Generation Considering Synergies Between Generation And Demand Side Management

Posted on:2015-09-19Degree:DoctorType:Dissertation
Country:ChinaCandidate:N YangFull Text:PDF
GTID:1312330428475153Subject:Power system and its automation
Abstract/Summary:PDF Full Text Request
The deployment of non-fossil renewable energy has significant impact on china's energy structure. Wind power, as a typical renewable green energy, has advantages of abound resources,mature technology, strong competitiveness and broad prospect. It is expected that the total installed wind power capacity will reach to200,000MW in2020and wind power will play a leading role instead of a complement. In the future, china's power system will evolve as a wind-hydro-thermal interconnected large power grid. However, owing to the volatility and intermittency of wind, it's hard to consume wind power within local area and adds operation costs. Consequently, it's of great importance to make a uniform power system development plan considering exploitation, transmission and consumption to build a feasible large-scale wind power dispatch strategy.This article focuses on power system dispatch strategy considering wind power. Electricity price response model with controllable load?cost function and wind power output probability density function are built and studied. Based on the proposed model, we regard controllable loads as a peak load regulation resource and modify power system dispatch strategy considering wind power. With volatility and intermittency of wind and security requirement, we propose a security economic wind power dispatch strategy considering source and load interactive scheduling.1) According to different demandside control mode, the responses and control characteristics are studied respectively. This paper builds a mathematical model to describe the demandside price response characteristics. Based on beneficial cost theory, this article proposes a cost model to evaluate controllable load and the relative parameters is evaluated, a wind farm power output forecasting method based on SVM and chaos particle swarm algorithm are proposed. In case study, the algorithm using realistic data is carried out and validated to be effective and outperform the classical algorithm in accuracy perspective.2) This chapter focuses on the impact of large-scale wind farm connecting with power grid on peak load regulation and its regulation strategy. Peak regulation balance model?related calculation algorithm and its optimized peak regulation model is proposed. In case study, the simulation based on the proposed method is carried out using realistic power grid data. The result indicates that:the deployment of large-scale wind power has dramatically negative effects on the stability of power system and the impact depends heavily on the grid-connected mode. The optimized peak load regulation method proposed can effectively decrease the regulation loss.3) Controllable load, as a peak load regulation resource, is adopted in power system dispatch strategy. Based on the uncertainty of wind and risk constraint theory, this chapter develops a large-scale wind power random dispatch model considering controllable load. The model can be effectively calculated by double-layer search algorithm. The model is verified with10simulation samples. The result indicates that the proposed model can extremely exploit the potential benefits of controllable load and overcome the uncertainty of wind power output. Compared with the current dispatch model, the proposed model not only has higher economic benefits, but makes a better use of wind power.4) This chapter focuses on the impact of fluctuating wind power on the security of transmission grid. A security economic wind power dispatch strategy is built by integrating security checking system into the wind power random dispatch model. The proposed model is decoupled into two subproblems: unit commitment problem and line security verification problem. The former problem can be solved by Imperialist Competitive Algorithm, the latter problem can be solved by simplex linear programming Algorithm. Using benders cut as a tie, the two subproblems can be solved by alternating iterative. Simulation based on standard samples indicates that the proposed dispatch, though may decrease the economic benefits, can effectively avoid line outraging.
Keywords/Search Tags:Demand Response, Marginal Cost, Least Square Support VectorMachine, Chaos Particle Swarm Optimization, Short-Term Wind Power Prediction, Load Peaking, Chance Constraints, Security Constraints Combination Unit, BendersDecoupling
PDF Full Text Request
Related items