Intelligent mining of air-conditioner field failure data based on the fusion of GRU and CAP for after-sales service text
The after-sales service text of air-conditioner contains extremely rich field failure data that reflects its operational reliability.To solve the low efficiency and poor quality assurance based on human data mining,this study proposes an intelligent mining method based on deep learning.The method is based on the field failure data mining process of air-conditioner failure criteria,its intelligent mining model integrates two deep learning techniques for extracting feature information in parallel,namely gate recurrent unit(GRU)and capsule neural network(CAP),and a dynamic fusion mechanism of GRU and CAP algorithm results is designed according to the loss function value.The application results of this method in the field failure data mining of three types of household air-conditioners show that the minimum values of the accuracy and harmonic mean of data mining results are 97.78%and 97.69%,respectively.
field failure datatext mininggated recurrent unitcapsule neural networkair-conditioner
NETL
NSTL
万方数据
