To help students understand the theory and simulation methods of actual circuit fault diagnosis, a neural network-based diagnostic model in the simulation circuit experiment and conduct corresponding case analysis is proposed. It adopts the method of simulation and circuit board verification, and uses an independent neural network to implement primary diagnosis of the circuit based on improved BP network. Then the genetic algorithm is used to calculate the optimal weight and threshold of the BP neural network. Through circuit diagnosis examples in MATLAB, the possibility of the diagnosed circuit belonging to different fault states based on various test information is also acquired. The experimental test results show that the BP neural network optimized by GA has a high ability to process information in parallel, and it has the characteristics of fast diagnosis speed, high accuracy, and strong real-time performance when diagnosing analog circuit faults. Therefore, this method can enable students to intuitively understand and master the process of power electronic device fault diagnosis, and promote their practical understanding of circuit fault diagnosis technology.