Abstract: The emergence of a new generation of weather radar increases the complexity of radar control system and the difficulty of troubleshooting. In order to ensure the reliability and maintenance performance of weather radar systems, a weather radar intelligent fault diagnostic expert system based on improved case-based reasoning (CBR) is designed. The radar fault diagnosis principle,fault diagnosis method based on similar distance function and the BP neural network matching method,and how to join the hardware replacement adjustment parameter database are described in this paper. Focusing CINRAD / CC radar antenna servo system hardware characteristics, the CBR method of motor drive and failure in radar servo system, the experts mentoring program of parameter detection and setting after hardware replacement are introduced in detail. Three CINRAD / CC radar fault diagnosis cases show the effectiveness and practicality of the fault diagnosis expert system.