地区供电公司日前购售电计划的联动优化模型

将非线性灰色Bernoulli模型用于到中长期电力负荷预测，提出了优选模型参数的粒子群优化算法。该模型是将GM(1,1)模型与Bernoulli微分方程相结合的一种灰色模型，适用于对不同发展趋势曲线的预测。通过粒子群优化算法，以模型预测平均绝对百分误差最小为目标，选择最优的模型参数。采用不同测试数据以及实际电网负荷数据进行了验证，结果表明上述模型有很好的适应性及较高的预测精度。

Linkage Optimization Model for Day-Ahead Power Purchasing and Selling Scheduling of Regional Power Supply Company

The optimal decision-making of power purchasing and selling linkage for regional power supply company is researched. By means of considering the market factor, policy factor and basic distribution network constraint, the day-ahead power purchasing and selling linkage scheduling is constructed as a nonlinear programming problem. Taking a 5-node radial distribution network for example, the situations of this network in a certain day in dry season and wet season are simulated respectively. Through quantitative and qualitative analysis on simulation results, the practicality and novelty of above-mentioned model is illustrated, and it is also shown that whether the distribution network constraint is considered or not will impact the drawing up of the day-ahead power purchasing and selling scheduling. An auxiliary decision-making and analysis system of power purchasing and selling for a certain regional power supply company in Southwest China that is developed based on the proposed model is in the practical stage.