This study introduces a new diagnostic framework that combines improved particle swarm optimization, K-means clustering algorithms, support vector machine (SVM), and learning vector quantization neural networks to provide a comprehensive fault diagnosis and pre-diction model for. This study introduces a new diagnostic framework that combines improved particle swarm optimization, K-means clustering algorithms, support vector machine (SVM), and learning vector quantization neural networks to provide a comprehensive fault diagnosis and pre-diction model for. Selectivity is a mandatory requirement for all protection, but the importance of it depends on the application. For example, unselective protection operation during a medium voltage network fault will cause an outage for an unnecessarily large number of consumers. While this is bad, It's not a. To promptly detect the faults of the relay protection system and the circuit breakers in time and to ensure the operational reliability of these protective devices, this paper proposes a fault tracing method for a relay protection system–circuit breaker based on improved Random Forest. Developing and applying intelligent relay protection systems has become an important way. With the development of the power industry, people's demand for electricity is growing, there is a contradiction between the current power resources and user demand for electricity, the main reason is that the substation operation there are some problems, causing power resources hard work. However, in actual operation, the relay protection device may cause failure due to hardware failure, software problems or external.