中文摘要：

多级负荷预测的基本协调模型和关联协调模型能够较好地解决各种具有直接加和特性的多级预测协调问题。文中在分析其模型特征的基础上,针对原有的加权最小二乘可信度容易出现对预测值调整幅度过大的不合理现象,提出了约束最小二乘可信度优化模型,使得由此形成的多级协调模型可以更加有效地提高电力需求预测的精度。文中还分析了基于约束最小二乘可信度的多级协调模型提高预测精度的机理,并通过实例验证了约束最小二乘可信度的有效性及其对多级预测效果的改善。

英文摘要：

The basic coordinating model and correlative coordinating model are effective tools to solve the coordinating problem of multilevel load forecasting.The characteristics of the multilevel coordinating model for load forecasting are analyzed.The optimization model of least-squares credibility with constraints is proposed to avoid the possible excessive adjustment phenomenon under the credibility of weighted least squares.Under this optimized combination of credibility,the accuracy of load forecasting can effectively be improved.The mechanism of improving the forecasting accuracy for the multilevel coordinating model based on least-squares credibility with constraints is discussed.In conclusion,a numerical example is given to validate the effectiveness of the least-squares credibility with constraints and to illustrate its effect on the accuracy improvement of multilevel load forecasting.