The quality of internal plunge grinding process is affected by the grinding performance of different grinding wheels. In order to online monitor the grinding performance of different grinding wheels under the same experimental parameters during the internal grinding process. A particle swarm optimization-back propagation (PSO-BP) neural network-based grinding performance monitoring method for different grinding wheels was proposed. Firstly, the feature parameters of acoustic emission signal, power signal, vibration signal, displacement signal and current signal were extracted. Then, according to the eigenvalue data samples of each sensor and the global optimization function of BP neural network by particle swarm optimization algorithm, the PSO-BP online monitoring model was established by using PSO algorithm to optimize the initial weights and thresholds of BP neural network to accurately monitor the grinding performance of different grinding wheels. Finally, the BP neural network model and the PSO-BP model were analyzed and compared with the experimental data. The results show that the PSO-BP monitoring model has higher monitoring accuracy than the BP neural network model, with an average correct rate as high as 97.6%, and the validity of PSO-BP is verified through a large number of experiments, which is able to effectively monitor the grinding performance status of different grinding wheels.