Investigation of the Pump Unit Control System With the Neural Network Productivity Estimator

This article deals with the novel approach to evaluate the technological coordinates of a pump unit. This method is is based on the use of the theory of neural networks. The expanded methodology of design a productivity estimator using signals of pump pressure and motor power consumption is developed. It is shown that for the development of the estimator, is required preliminary information on the behavior of the desired coordinate when the task signal and disturbing actions in the system vary. The research of developed estimator by a method of mathematical modeling of pressure stabilization system is carried out. A comparison of the results of the system performance with the estimator and the calculated productivity value and the recommendations for the possible application of the proposed methodology are presented.