Dynamic performance prediction of vehicle variable speed air conditioner based on LSTM recurrent neural network

Abstract

Accurate prediction of air conditioner's dynamic operation is very important for advanced control and fault diagnosis method. The prediction of vehicle air conditioners' performance faces many challenges like unstable working conditions and frequent "on-off" operation. This research proposes a long short-term memory (LSTM) recurrent neural network-based method to tackle this tricky issue. The proposed model is trained and tested with field operation data to prove its capability.