适于暖通空调系统节能/储能的负荷预报新方法的建立与应用

项目名称： 适于暖通空调系统节能/储能的负荷预报新方法的建立与应用

项目编号： No.51207035

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 电能储存与节电技术

项目作者： 张永明

作者单位： 哈尔滨工业大学

项目金额： 24万元

中文摘要： 暖通空调系统的节能潜力巨大，探求相关的能量传输及使用共性理论与关键技术，系电气工程学科- - 电能储存与节电技术领域的重要基础性研究问题。 鉴于负荷预报对暖通空调系统节能/储能的关键作用，针对暖通空调系统负荷大惯性/大滞后的自身非线性特性和外扰的随机性特征，本项目拟建立一种基于信息粒化及支持向量机理论的负荷预报新方法。新方法的主要构成要素为：1）依据负荷构成及影响因素选择样本数据；2）利用信息粒化适应负荷外扰的随机性；3）利用支持向量机适应负荷的自身非线性。 理论上，确定出负荷构成因素的相互作用关系，掌握暖通空调系统的主要动力学特征；技术上，构建一种具有广泛适应性的集合预报技术框架，提高负荷预报的准确性/可靠性；应用上，给出适于暖通空调系统节能/储能应用的工程化算法，提高方法的实用性/易用性。为暖通空调系统负荷优化调度、节能控制及电力需求侧管理储能提供一种重要的理论/技术支撑。

中文关键词： 暖通空调系统；节能减排；负荷预报；需求侧/供应侧；建筑能源微网

英文摘要： Based on the fact that Heating Ventilating and Air Conditioning (HVAC) system has tremendous energy conservation potential, the exploitation of universal theory and critical technology can give a good service for energy transmission and utilization in the field of energy storage and conservation affiliated to the electrical engineering discipline. Load prediction plays a critical role in energy conservation/storage, and a novel integrated set approach to HVAC load prediction applying load composition and impact factors is proposed aiming at its intrinsic nonlinearity with high inertia & delay and external disturbances with randomness, in which Support Vector Machine (SVM) and Information Granulation (IG) are employed to adapt the nonlinear and random characteristics in HVAC load, respectively. In this project, we first determine the relationships among load composition factors, and furthermore analyze the dynamic characteristics of HVAC in theoretical aspect. Secondly, a set prediction technology framework with extensive adaptability is constructed to enhance the accuracy and reliability of load prediction. Thirdly, to improve the flexibility and feasibility of the new approach, an engineering algorithm is employed for energy conservation/storage in application. Therefore the proposed approach is expected to l

英文关键词： Heating Ventilation and Air Conditioning(HVAC)；Energy Conservation and Emission Reduction；Load Prediction；Demand Side/Supply Side；Building Energy Micro-grid