基于二阶段随机规划风-水-火动态经济调度策略-wind - water - fire dynamic economic dispatch strategy base on two-stage stochastic programming.docxVIP

例，表明引入随机场景后，所提策略可行且提高了发电公司的总收益。在均衡解法下优化了水火电源收益，验证了合理构建多源共济动态经济调度 策略对创造共赢格局的重要性。不平衡电价取值较高时，总收益下降并趋 于恒定值，此时发电公司可以通过实际收益期望选择是否接纳风电并网调 度，提高发电主动性和预见性。关键词：动态经济调度 随机规划 风水火 场景树 利益重配WIND-HYDRO-THERMAL DYNAMIC ECONOMIC DISPATCH BASED ON TWO-STAGE STOCHASTIC PROGRAMMINGABSTRACTWind power energy has a renewable growth rate and tremendous economic benefits, but its randomness and intermittent increases the technical difficulty of the power system dynamic economic dispatch.This paper following the principles of modern electric market competition schedule is proposed a wind-hydro-thermal dynamic economic dispatch strategy based on two-stage stochastic programming with compensation. This strategy is committed to optimal economic profit by using hydro and thermal power to eliminate the intermittent impact of wind power, and the two-stage stochastic programming model is to reflect the deviation between the scheduling situation and the actual situation caused by the wind power intermittently.Model building requires scenario tree theory as support, taking real-time scheduling and dispatching before trading mechanism, using the value and probability of the classified price and wind power to make the randomness into deterministic problem. In this Model, the first stage decision is made before the actual market to get the most cost-effective; second stage is the feedback of the first stage to make penalties when devision appeared. Compromise between the interests of a balanced solution allocation principles and risk assessmentindicators to determine the final revenue for each power plant, the system is re- allocated to maintain the expected benefits and risks, and increase power generation company initiative in scheduling. By calling GAMS software DICOPT solver to solve the high-dimensional, nonlinear, multi-constrained mixed integer programming problems, a combination of different examples verify the feasibility of the method and model.When the load level is consistent, whet