论文部分内容阅读
为了提高短期负荷的预测准确度,提出一种聚类分析和改进贝叶斯算法的短期负荷预测模型。首先收集短期负荷的历史样本,并进行归一化处理,加快建模速度;然后采用模糊均值聚类算法对短期负荷历史样本进行分类,构建贝叶斯算法的学习样本;最后采用贝叶斯算法建立短期负荷预测模型,并针对贝叶斯算法的不足进行相应改进。采用具体短期负荷历史数据序列对模型的有效性进行仿真测试,结果表明,聚类分析和改进贝叶斯算法的短期负荷预测模型提高了短期电力负荷的预测准确度,加快了模型的训练速度,预测结果更加可靠,可以为电力管理部门科学决策提供参考。
In order to improve the prediction accuracy of short-term load, a short-term load forecasting model based on clustering analysis and improved Bayesian algorithm is proposed. Firstly, the historical samples of short-term load are collected and normalized to speed up the modeling. Secondly, the fuzzy mean clustering algorithm is used to classify the short-term load history samples to construct the learning samples of Bayesian algorithm. Finally, Bayesian algorithm The short-term load forecasting model is established and corresponding improvements are made to the shortage of Bayesian algorithm. The simulation results show that the short-term load forecasting model based on clustering analysis and Bayesian algorithm can improve the prediction accuracy of short-term load and speed up the training speed of the model, The forecast result is more reliable, which can provide reference for scientific decision-making of power management department.