（3）由于BP神经网络故障诊断时普遍存在收敛缓慢、容易出现局部极小值的

(3) Because the BP neural network generally has slow convergence and prone to local minimums in the fault diagnosis of the BP neural network, a particle swarm algorithm is proposed to optimize the BP neural network to realize the fault diagnosis of the plunger pump. Two models of BP neural network and PSO-BP neural network are used to classify and identify the faults of the plunger pump, realize the diagnosis of three typical faults of the plunger pump, and achieve the purpose of fast and accurate fault diagnosis. It is proved that both of the two network models can be used in the fault diagnosis of the plunger pump. By comparing the results of the fault diagnosis of the BP neural network, it is concluded that the PSO-BP neural network has more advantages in the fault diagnosis of the plunger pump.

(3) Because of the situation of slow convergence and easy local extreme values in BP neural network fault diagnosis, it is proposed that the particle group algorithm optimizes BP neural network to achieve plunger pump fault diagnosis. Using two models, BP neural network and PSO-BP neural network, the fault classification and identification of plunger pump are used, and the diagnosis of three typical faults of plunger pump is realized, which achieves the goal of rapid and accurate fault diagnosis. It is proved that both network models can be used for the fault diagnosis of plunger pump, and by comparing the results of BP neural network fault diagnosis, PSO-BP neural network has more advantages in plunger pump fault diagnosis.

(3) Due to the slow convergence and local minimum in BP neural network fault diagnosis, particle swarm optimization (PSO) algorithm is proposed to optimize BP neural network for piston pump fault diagnosis. BP neural network and PSO-BP neural network are used to classify and identify the faults of piston pump. The three typical faults of piston pump are diagnosed, and the purpose of fast and accurate fault diagnosis is achieved. It is proved that the two kinds of network models can be used in fault diagnosis of piston pump. By comparing the results of BP neural network fault diagnosis, it is concluded that PSO-BP neural network has more advantages in fault diagnosis of piston pump.