基于多变量相空间重构和卡尔曼滤波的冷热电联供系统负荷预测方法

发布时间：2018-06-19 08:32

本文选题：能源互联网 + 冷热电联供系统　；参考：《中国电机工程学报》2016年02期

【摘要】：文中设计一种新型的基于多变量相空间重构和卡尔曼滤波的冷热电联供系统负荷预测方法。首先选择冷热电负荷及与负荷密切相关的天气因素的历史时间序列组成多变量时间序列,然后运用混沌理论和C-C方法重构多变量相空间,最后建立多变量相空间的自回归模型并采用卡尔曼滤波方法预测冷热电负荷。以中国北方某医院冷热电联供系统的8月份历史负荷数据和天气数据验证该冷热电负荷预测方法。结果表明,与采用单变量相空间重构和卡尔曼滤波预测方法相比,文中设计的负荷预测方法充分考虑冷热电负荷中多个变量的相互耦合关系,可有效提高负荷的预测精度。算例分析验证了该冷热电负荷预测方法的可行性和有效性。
[Abstract]:A new load forecasting method based on multivariable phase space reconstruction and Kalman filter for combined cooling and heat supply system is designed. Firstly, the multivariable time series is composed of the historical time series of the cold, hot and electric loads and the weather factors closely related to the load, and then the multivariable phase space is reconstructed by using the chaos theory and the C-C method. Finally, an autoregressive model of multivariable phase space is established and Kalman filter method is used to predict the thermal and cold load. Based on the historical load data and weather data of a combined cooling and heat supply system in a hospital in northern China, the method is verified. The results show that compared with the single-variable phase space reconstruction and Kalman filter forecasting method, the load forecasting method designed in this paper fully takes into account the coupling relationship of many variables in the cold and hot electric load, and can effectively improve the forecasting accuracy of the load. The feasibility and effectiveness of this method are verified by an example.
【作者单位】：山东大学控制科学与工程学院;
【基金】：国家自然科学基金重大国际(地区)合作研究项目(61320106011);国家自然科学基金项目(61573224) 国家863高技术基金项目(2014AA052802)~~
【分类号】：TM611.3

【相似文献】