Research On Operation Strategy Of Hvac Equipment Based On Multidimensional Data

With the popularization of building automatic control systems in China,the application research based on the multidimensional data has garnered more and more attention.However,the research and application of multidimensional data started late.Research on equipment operation strategies is not perfect.In this paper,a combination of theoretical analysis,mathematical derivation and experimental research is carried out to study the identification and recommendation of equipment operation strategies based on itemized energy consumption data on HVAC equipment operation strategies,and build an online room temperature prediction system to assist staff in adjusting the equipment strategy.This paper first analyzes the influencing factors of building energy consumption and indoor temperature in theory.One is to search out the main equipment that needs to be identified by strategy,analyze the main components of commercial building energy consumption through analysis,analyze the important energy consumption components from the historical data of classified energy consumption,and summarize the important structure of the total energy consumption of the building.Subsequently strategy identification research content provides data support and guidance.The second is to explore the necessary and available factors for establishing a temperature prediction model.On the one hand,it analyzes the importance of some influencing factors for establishing a prediction model.On the other hand,the available factors for establishing the room temperature prediction model are screened from the perspective of monitoring and controllability of the room temperature influencing factors.Then the research on the identification method of HVAC equipment operation strategy given the itemized energy consumption data can be divided into two parts:the basic identification method of the equipment strategy and the self-correction method of the identification parameters.In the basic identification part,this paper mainly uses mathematical derivation and model simplification methods to study the identification of equipment operating conditions under the condition of a single power line carrying multiple devices.The purpose of the self-correction part of the identification parameter is to correct the power parameters of the related equipment,thereby improving the accuracy of the policy identification.The multi-device strategy identification is realized through programming,and the strategy identification is carried out on the circuit including 7 devices.Finally,the strategy identification is realized by using the correction parameters after 6 iterations.Then,data crawler,data cleaning and data integration technologies are used to integrate historical indoor temperature data,equipment strategy recognition results and meteorological data in order to establish a historical strategy database,and at the same time,the similar day method suitable for commercial buildings is studied to optimize the strategy,and the equipment strategy recommendation method based on historical data is realized by combining the historical strategy library.Finally,the research on the indoor temperature prediction method that can be operated online is carried out to guide the operators to adjust the operation strategy when the equipment is running.Firstly,a temperature rate prediction model based on XGBoost is established.Then,an on-line scheduler mechanism is constructed to set up schedulers with different functions in order to ensure prediction accuracy and reduce calculation consumption.Finally,the operational data and indoor temperature data of actual commercial buildings are used to verify the effectiveness and accuracy of the indoor temperature prediction algorithm,and The root mean square error of temperature prediction in the test data reaches 0.16℃.