绿色建筑系统节能运行自适应动态规划研究

项目名称： 绿色建筑系统节能运行自适应动态规划研究

项目编号： No.61273326

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 汪明

作者单位： 山东建筑大学

项目金额： 81万元

中文摘要： 绿色建筑系统节能运行是支撑建筑可持续发展的热点与难点问题之一，具有广阔的应用前景，目前尚缺乏有效的解决手段。本项目针对设备总系统提出采用Spiking神经网络自适应动态规划方法来研究和寻找其解决办法。首先，研究多源人员信息融合问题，提出基于传感映射方法进行Spiking神经网络信息融合建模，解决建筑内人员分布时空信息的检测问题并对建筑空间进行分区；其次，基于操作行为研究数据建模、动态环境与人体舒适及运行能耗间的相关性；最后，开展Spiking神经网络自适应动态规划研究，依据能效标准和建筑空间分区等提出一系列优化控制策略，研究自适应动态规划算法的收敛性、稳定性，并通过人、动态环境、设备总系统的协调作用达到环境舒适，建筑高效运行的节能效果。预期成果可帮助绿色建筑、智能建筑提高运行效率，降低运行能耗，提高服务水平，也可应用到环境监测、能耗监测、复杂系统优化控制等领域。

中文关键词： 自适应动态规划；Spiking神经网络；信息融合；绿色建筑；建筑节能

英文摘要： The energy-saving operation of green buildings is one key issue in supporting building sustainable development, which has been shown of broad application prospects in modern constructions. However, there are still some technical issues lack of feasible solutions in practice. To this end, the proposed project presents a green building system based on the Spiking Neural Network and adaptive dynamical programming approach. First, the multi-source personnel information fusion is considered. The sensor projection method together with spiking neural networks is applied to model the spatial and temporal distribution of human beings inside the building. The building can be divided into several regions to make the energy optimization more feasible. Second, by inducing human behaviors, the correlation factor among data model, dynamical environment, human comfort and energy consumption is investigated. At last, to satisfy the Energy Efficiency Standards and certain spatial division, some optimal control strategies based on the spiking neural network and adaptive dynamic programming are planned to be proposed. At the same time, the convergence and stability of the strategies are investigated as well. The efficient operation of buildings, the comfort of human, and the reduction of energy consumption will be achieved simultan

英文关键词： adaptive dynamic programming；Spiking neural networks；information fusion；green building；building energy-saving