基于ELM优化模型的用户短期负荷研究

SHORT-TERM LOAD OF USERS BASED ON ELM OPTIMIZATION MODEL

随着售电侧的逐步开发以及用电大数据时代的到来,短期负荷预测更加复杂,必须综合考虑实时电价、用户历史用电行为以及预测模型的精度和时间复杂度。在分析各种短期负荷影响因素的基础上,利用K-means聚类方法对用户历史用电行为进行聚类,再利用兼具有自动寻找隐层节点数和在线学习功能的I-OS-ELM学习机进行负荷预测。实例预测结果证明,该模型能够有效地解决实时电价机制下短期负荷的预测问题。

With the gradual development of power selling side and the arrival of big data era,short-term load forecasting is more complex.It is necessary to consider the real-time electricity price and users historical electricity usage behavior,as well as the accuracy and time complexity of the prediction model.Based on the analysis of various short-term load influencing factors,we used K-means clustering method to cluster users historical electricity usage behavior.Then we used I-OS-ELM learning machine,which has the function of searching hidden layer nodes automatically and learning online,to forecast the load.The forecasting results show that the model can effectively solve the short-term load forecasting problem under the real-time electricity price mechanism.