Survey on the application of machine learning in elevator fault diagnosis

Real time state detection and fault diagnosis are very important to the normal use and safe operation of elevators. With the development of intelligent diagnosis technology, the powerful computing power and diagnosis ad-vantages of machine learning are increasingly prominent in elevator fault detection and diagnosis. This paper summarizes the application of four typical algorithms of artificial neural network (ANN), support vector machine (SVM), Bayesian network (BN) and deep learning (DL) in elevator fault diagnosis in machine learning technology in recent years, and analyzes the method expansion and application based on its basic theory and the problems encountered in practice, The advantages and disadvantages of each method in the diagnosis process are discussed, and the improvement measures proposed at this stage are analyzed. Finally, the future research and development direction are proposed.