High Quality Diesel injection pump Plunger 2425-988 Fuel Plunger Element Plunger 2425 Series

The axial piston pump with traditional triangular groove plate can not completely eliminate the pressure pulsation caused by the commutation in the transition zone, and will produce obvious flow backflow phenomenon due to the influence of throttling. In the absence of subversive design of cylinder block or valve plate, a pressure control reciprocating valve structure is proposed to be connected in series between each plunger cavity, so as to buffer the pressure pulsation in the transition zone, reduce the vibration in the transition zone, and eliminate the triangular groove structure of the valve plate, so as to reduce the flow backflow. Given the diameter and other parameters of the reciprocating valve, it is found through simulation that the pressure pulsation generated by the structure in the transition zone is only 2.5%, which can effectively reduce the pressure pulsation of the axial piston pump in the process of high and low pressure transition zone compared with the triangular groove valve plate.

Aiming at the problems such as insufficient sample number and weak fault characteristics of audio signal in fault diagnosis of piston pump, A Fault Diagnosis of Plunger Pump Based on MTL (McL-pafd) based on audio signals combined with Meta-transfer Learning (MTL) was proposed. In this method, the audio signal of the plunger pump is taken as the sample, and the signal is processed by the Gammatone filter bank under the condition of a single sensor, which can effectively improve the characterization ability of the audio signal under strong noise interference. Then, combined with the meta transfer learning, the fault diagnosis of the plunger pump under the condition of small sample can be realized. At the same time, according to the actual needs of fault diagnosis of plunger pump, the test method of meta-transfer learning in fault diagnosis application is improved, and the unknown fault class can be adaptive dealt with. The experimental results show that the accuracy of McL-pafd diagnosis can reach 91.41% only for known fault classes, but after fast adaptive learning, the accuracy of McL-pafd diagnosis can reach 89.64% when identifying unknown fault classes.