The characteristic of the tribology system is complex so that the conventional methods cannot solve the tribological complex problems accurately.In order to fully realize the methods of tribological research and make a quantitative study of complex tribological problems,the latest development of applications of artificial neural network theory in wear prediction,friction coefficient prediction and wear particles identification were summarized respectively.The development trends were prospected,including the establishment of a new neural network model and a high integration research model coupled with a variety of analytical methods,as well as the innovation of research objects.