Estimation of centrifugal pump operational state with dual neural network architecture based model

The sensorless estimation of the operational state of a centrifugal pump system provides new opportunities for the monitoring and control of pumping systems. In this paper, the disadvantages of traditional estimation methods based on pump characteristic equations are investigated and analyzed. A novel estimation method based on dual neural network architecture is advised and trained offline to determine the operation state of the pump drive. Simulation studies are performed on ISG150-400 centrifugal pump using the proposed estimation method. The test results are provided to validate the effectiveness.