Water chilling unit fault diagnosis method based on support vector regression model

Abstract

The invention relates to a water chilling unit fault diagnosis method based on a support vector regression model. The fault type of a water chilling unit is determined; sensitive characteristic parameters and fault diagnosis rules of faults of the water chilling unit are determined; training a support vector regression model of each sensitive characteristic parameter by using the operation data of the water chilling unit, wherein parameters required to be selected by the support vector regression model are obtained through a wolf pack algorithm; determining a fault threshold value of each sensitive characteristic parameter; and predicting each current sensitive characteristic parameter value by using the support vector regression model, and judging the fault of the water chilling unit according to the sensitive parameter values and the fault threshold. According to the invention, the fault diagnosis prediction precision and accuracy can be effectively improved.

Description

technical field [0001] The invention relates to a fault diagnosis method for a chiller based on a support vector regression model. Background technique [0002] The chiller is the core component of the central air conditioning system. Its main function is to provide cooling capacity for the air handling equipment. The failure of the chiller will cause problems such as increased energy consumption, reduced equipment service life, and reduced ability to control the indoor environment. Therefore, it is of great significance to study the fault diagnosis of chillers. [0003] At present, the methods for fault diagnosis of chillers mainly include: analytical model method, data-driven method, and semi-empirical model method. The analytical model method relies on the establishment of the physical model of the chiller, the use process is complex and the universality is poor, and it is difficult to realize in practice; data-driven methods, such as principal component analysis (PCA), ...

AI Technical Summary

Problems solved by technology

MPR model is the earliest regression model applied to chiller fault diagnosis, but in some cases, its prediction accuracy is poor; BP-ANNR model is suitable for nonlinear and high-dimensional fault diagnosis problems, and has been widely used in recent years, but often There will be over-fitting problems; the support vector regression model (SVR model) has a simple modeling process, takes less time, and shows excellent performance for small samples, and has become a research hotspot in recent years

However, the prediction results of the SVR model will be affected by the model parameters. At present, the grid search method, particle swarm algorithm, genetic algorithm, etc. are mainly used to optimize the model parameters, but these algorithms are easy to fall into the local optimal solution and cannot obtain the optimal solution. Model parameters

Images