物联网环境下数据驱动舒适性建模与控制研究

发布时间：2018-03-25 02:40

本文选题：数据驱动　切入点：室内环境舒适性　出处：《山东建筑大学》2016年硕士论文

【摘要】：建筑为人们日常生活、学习、工作提供了干净、整洁的环境。随着社会的进步、发展,建筑不仅讲究外在艺术美,更讲究内部的舒适性和节能性。但舒适性与能耗有时存在一定冲突,需要用相应的策略进行协调优化。为此,本文提出了一种物联网环境下数据驱动的舒适性建模与控制策略。该研究主要涉及以下内容：第一,采用统计方法对建筑用电设备物联网系统收集的环境数据进行预处理,预处理之后的数据作为整个课题研究的数据支撑。第二,提出了一种数据驱动的二型模糊集模型的构建方法。与传统模糊方法相比,该方法中的二型模糊集理论处理不确定性的能力显著增强。将该方法应用于室内环境舒适性控制中,提出了基于热舒适性模型的室内环境调控策略。通过仿真实验证明了该策略对于特定房间内的热舒适性建模的有效性。第三,提出了一种数据驱动的多变量模糊系统模型的构建方法,该方法可以有效地解决传统模糊系统通常面对的规则爆炸问题。应用提出的单输入规则模块连接的神经模糊系统对特定房间内的冷热负荷预测,仿真结果证明了所提出的方法对于能耗预测问题的有效性和优越性。第四,针对高舒适与低能耗的追求目标,提出了一种多目标优化策略。该策略采用具有强搜索能力的遗传算法试图得到一组最优解集(又称Pareto最优解集),最后根据先验知识及实际应用背景提供一个最优解用于平衡相互矛盾的目标。
[Abstract]:Architecture provides a clean and tidy environment for people's daily life, study and work. With the progress and development of society, architecture not only stresses the beauty of external art, More attention is paid to internal comfort and energy saving. However, there are some conflicts between comfort and energy consumption, which need to be coordinated and optimized with corresponding strategies. This paper presents a data-driven comfort modeling and control strategy in the Internet of things environment. The main contents of this study are as follows: firstly, the environmental data collected by the Internet of things system of building electrical equipment are preprocessed by statistical method. The pre-processed data is used as the data support of the whole subject. Secondly, a data-driven fuzzy set model is proposed. The ability of the second type fuzzy set theory to deal with uncertainty is greatly enhanced. This method is applied to indoor environment comfort control. The indoor environment control strategy based on thermal comfort model is proposed. The simulation results show that the strategy is effective for the thermal comfort modeling in a specific room. Third, In this paper, a data-driven multivariable fuzzy system model is proposed. This method can effectively solve the problem of rule explosion in traditional fuzzy system. The proposed neurofuzzy system with single input rule module is used to predict the heat and cold load in a particular room. The simulation results show that the proposed method is effective and superior to the problem of energy consumption prediction. Fourth, aiming at the goal of high comfort and low energy consumption, A multi-objective optimization strategy is proposed, in which a set of optimal solutions (also known as Pareto optimal solution set) is obtained by using a genetic algorithm with strong searching ability, and an optimal solution set is provided according to prior knowledge and practical application background. Solutions are used to balance conflicting goals.
【学位授予单位】：山东建筑大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TU111.195;TP391.44;TN929.5

【参考文献】