摘要
考虑火电厂的发电机组出力、能耗成本及污染排放三者之间、及梯级电站出力与水力耦合参数之间,以及水火电机组之间综合效益基础上,建立了梯级水电站弃水量最小、火电能耗成本最低、和污染排放最小的多目标综合节能调度模型,并提出了计及梯级水电站的水火电节能发电调度策略;利用标度转换法,将梯级水电站弃水量最小、火电机组能耗成本最低和污染排放最小的多目标综合节能调度模型转化为不同权重系数下的单目标函数;运用混合惩罚函数优化算法将问题的求解转化为无约束优化问题,并利用Hessian矩阵改进了该优化算法,克服了缩减系数选择问题;通过对含4个梯级水电站、3个火电厂的水火电系统进行仿真,优化结果验证了该算法的可行性,并具有较明显的节能效果。
Considering the overall benefit among generating unit output, energy consumption cost and pollution emission of thermal power plant and that between output of cascaded hydropower stations and hydraulic coupling parameters as well as that between hydropower generating units and thermal power generating units, a comprehensive multi-objective energy-saving scheduling model, in which the minimized surplus water quantity of cascaded hydropower stations, the lowest energy consumption cost of thermal power plant and minimized pollution emission are taken as the objectives, is established, and an energy-saving generation scheduling strategy of hydropower and thermal power plants, in which the cascaded hydropower stations are taken into account, is proposed. Using scale conversion method, the built multi-objective scheduling model is turned into single objective function with different weight coefficients, and utilizing the mixed penalty function optimization algorithm the solution of the model is turned into unconstrained optimization problem and this optimization algorithm is modified by Hessian matrix, thus the selection of reduction coefficient is solved. Through the simulation of a hydropower-thermal power system containing four cascaded hydropower stations and three thermal power plants, the feasibility of the modified optimization algorithm is validated by optimization results, besides simulation results also show that using the proposed method an obvious energy-saving effect can be achieved.
出处
《电网技术》
EI
CSCD
北大核心
2014年第3期616-621,共6页
Power System Technology
基金
国家自然科学基金项目(510670001)~~
关键词
节能发电调度
梯级水电站
内点法
外点法
混合惩罚函数法
energy-saving generation scheduling
cascaded hydropower stations
interior point method
external point method
mixed penalty function method