Optimization Of Operating Parameters For Chiller Plant Of Central Air Conditioning System Based On Data Mining

As the main energy-consuming equipment of large-scale public buildings,the central air-conditioning system accounts for more than 40% of the energy consumption of public buildings.And the energy consumption of the chiller plant accounts for more than 50% of the energy consumption of central air-conditioning,which is the key to central air-conditioning energy saving to optimize control of the chiller plant.The optimization of operation performance and energy efficiency of the chiller plant is of great significance for realizing the economic operation of the central air conditioning system and reducing the energy consumption of large public buildings.The development of the Internet of Things and information technology has accumulated a large amount of historical operational data for the central air-conditioning system.Taken the central air-conditioning cold source system of a shopping mall in the hot summer and warm winter area as the research object,a data mining method for the optimization of operating parameters of the cold source system is proposed based on the actual operation data of a large number of the chiller plant.This method does not need to establish mathematical models of the chiller plant,which is simple and easy to implement for engineering application.The main research work of this paper includes:(1)For the chiller plant,there are many operating parameters,some of which are redundancy and uncontrollable.By analyzing the main influencing factors of the operating efficiency of the chiller plant,based on the principle that parameter must be controllable and have strong correlation with the energy efficiency of the chiller plant,the data mining target containing nine main operating parameters is preliminarily determined.Then the attribute reduction method of the positive domain based on the rough set theory is used to reduce the 9-dimensional operating parameters to 4 dimensions,which reduces the difficulty of mining strong association rules.(2)Aiming at the problem that the non-steady-state operation data exists in the startup process of the chiller plant and affects the optimization effect of the target value of the operating parameters,the normal distribution identification and cleaning method of the unsteady operation data is proposed.And the missing and abnormal,logical and outlier data are identified and processed to provide effective data for data mining.(3)Since the algorithm for association rules is applicable to discrete data,and the inappropriate discretization method is easy to cause the information loss after data discretization.In this paper,the hypothesis test method is used to establish the distribution model of different operating parameters,then the comprehensive evaluation index is defined.According to the data distribution characteristics of each operating parameter,different optimal discretization methods are selected to preserve the integrity of the original data information as much as possible,which improves the efficiency of mining and improves the reliability of strong association rules.(4)Aiming at the problem of uneven distribution of data volume under various operating conditions,a multi-minimum support threshold is adopted in the process of association rule mining,which solved the problem that there were no strong association rules under some operating conditions with the single minimum support threshold.Then,combined with different operating modes of the chiller plant,the Apriori algorithm is used to mine the association rules under the typical operating conditions,and the strong association rules that guide the optimal operation of the chiller plant are obtained.(5)Finally,through the simulation experiment and practical application,the energy-saving effect of strong association rules under different operation modes is verified.The simulation shows that the energy saving rate of the chiller plant after optimized are 12.13%,8.83% and 5.21% respectively in mode 1,mode 2 and mode 4 of chiller plant;Actual tests show that the chiller plant system saves 9.27% and 11.98% respectively in 2 tests after optimization.