基于大数据技术的交直流混合主动配电网规划模型构建

Construction of planning model for AC/DC hybrid active distribution network based on big data technology

为了提升交直流混合主动配电网规划综合效益,使规划后的配电网具备较好的电力负荷指标,构建一种基于大数据技术的交直流混合主动配电网规划模型。根据网损灵敏度确定交直流混合主动配电网中的分布式电源安装容量和位置,并将此作为目标构建上层规划模型;选取最佳分布式电源出力作为控制变量,将电压偏差指标、年投资运行效,益作为目标构建下层规划模型。结合两个模型构建交直流混合主动配电网双层规划模型,选取Spark大数据平台作为模型求解过程中的信息处理平台,利用随机森林算法实现交直流混合主动配电网规划问题的最终求解。实验结果表明,该模型规划效果更好,相比其他模型建设成本更低,规划后的配电网电力负荷指标较好,综合效益更优,在以后的工程项目中可以推广及应用。

In order to improve the comprehensive benefits of AC-DC hybrid active distribution network planning, so that the planneddistribution network has better power load indicators, a big data-based AC-DC hybrid active distribution network planning model isconstructed. According to the sensitivity of network loss, the installation capacity and location of distributed generation in AC / DChybrid active distribution network are determined, and the upper level planning model is constructed with this as the target;the optimaldistributed generation output is selected as the control variable, and the voltage deviation index and annual investment and operationbenefit are taken as the objectives to construct the lower level planning model. Combined with the two models,a two-level programmingmodel for AC-DC hybrid active distribution network is constructed. Spark big data platform is selected as the information processingplatform in the process of model solving. The final solution of AC / DC hybrid active distribution network planning problem is realizedby using random forest algorithm. The experimental results show that this model has better planning effects and lower construction coststhan other models. The planned distribution network has better power load indicators and better overall benefits. It can be promoted andapplied in future engineering projects.