PID Algorithm Application in the Mechanical Sealing Based on BP Neural Network

Based on BP neural network algorithm, combined with PID linear control theory, we establish neural network residual training model of mechanical seal device, and design the simulation system of mechanical seal device with FLUENT simulation software. In order to verify the availability and reliability of the system, we design two different mechanical sealing devices, and do simulation calculation on the two kinds of devices. We get the sealing pressure field and velocity field distribution curve of device 1 and device 2. Through calculation we get sealing quality change curve with time. Through comparing the two different sealing performances we found the sealing effect of device 2 is better, the sealing efficiency reaches 98.2%, which meets the design requirements of the mechanical seal.