THE NEURAL NETWORK APPLICATION IN A CONTROL SYSTEM FOR PERFORMANCE OPTIMIZATION OF A GROUP OF ROD DEPTH PUMPS WITH A MECHATRONIC DRIVE WITH UNEQUAL CYCLES WITH LIMITED PEAK POWER CONSUMPTION

Larin Mikhail Yurievich, Volkov Andrey Nikolaevich

For the development of low- and medium-rate wells, the most commonly used installations are sucker-rod pumps. The group of sucker-rod pumps connected to the same power supply system is installed on the wells of one oilfield. There is a mathematical relation between the maximum power consumption of a group of cycle drives and the performance of this group. In this paper, a strategy for managing a group of sucker-rod pumps with a mechatronic drive has been developed, focused on reducing peak power consumption while maintaining the productivity of the entire group and based on the use of a neural network.

Keywords:Simulation modeling, control system, sucker-rod pumps, oil production, mechatronic drive, energy efficiency, peak power, optimization, adaptation, neural networks