可持续学习的室内设备调控系统设计与实现

Design and Implementation of Sustainable Learning Indoor Equipment Control System

室内设备调控系统可以构建舒适的室内环境,进而提升室内人员的工作效率。但由于不同人对不同环境下的舒适标准不统一,因此需要具有学习能力的系统来学习室内人员的使用习惯,进而满足室内人员多元化的需求。选择百叶窗设备调控方向,研究开发一套可持续学习的室内设备调控系统,可检测室内环境数据并分析得出最适宜的百叶窗设备运行状态,同时可对设备状态进行反馈,满足了多元化的需求。本系统基于LSTM神经网络搭建,加入Attention注意力模块使控制结果更精确,解决了现有室内设备调控系统与用户无法反馈等问题,为建立个性化设备调控系统提供了技术基础。测试结果表明,本系统在搭建的仿真场景中可根据室内环境信息变化实时且正确地调整百叶窗设备,表明了本系统具有准确性、可行性与实用性。

Indoor equipment control system can build a comfortable indoor environment,so as to improve the work efficiency of indoor personnel.However,because different people have different comfort standards for different environments,a system with learning ability is needed to meet diversified demands.The regulation direction of louver equipment is selected to research and develop a set of indoor equipment regulation system of sustainable learning,which can detect indoor environmental data and analyze the most appropriate operating state of louver equipment.Meanwhile,it can give feedback to the state of equipment,meeting the diversified needs of users.The data analysis module is built based on LSTM(long short-term memory)neural network,and the Attention module is added to make the control result more accurate.This system solves the problems of the existing indoor equipment regulation system,such as the single function and the user's inability to feedback,and provides the technical basis for the establishment of personalized indoor equipment regulation system.The test results show that the system can adjust the louver equipment in real time and correctly according to the change of indoor environmental information in the simulation scene,which indicates that the system is accurate,feasible and practical.