Fault diagnosis method for circuit board based on BP neural networks

This paper presents a new fault diagnosis method for circuit boards based on BP neural networks. First, a description of traditional fault diagnosis methods is given and the shortcomings are demonstrated. To improve the accuracy of fault diagnosis, a new method based on BP neural networks is proposed, which takes the temperature data of chips as well as voltage data as the basis of diagnosis. Finally, an example is provided, which shows that the proposed method performs better in fault diagnosis of the circuit board.