Intelligent energy-efficient control system for parallel pumps

Abstract

Relevance: today, pump stations used in water supply and wastewater systems are among the main consumers of electrical energy. In particular, the unbalanced operation of parallel pump units leads to increased energy consumption and higher operational costs. In situations where traditional control methods fail to address these challenges, the implementation of modern intelligent control systems becomes a pressing issue. One of the modern control methods is the use of artificial neural networks (ANN) integrated with PLC controllers. These systems enable continuous monitoring of pump station operating modes, adaptation to varying loads, and the formation of optimal control strategies, significantly enhancing energy efficiency.

Aim: to develop an optimized intelligent control system for modeling the electric drive of parallel pumping units and improving the efficiency of their control.

Methods: the study constructed mathematical models of a centrifugal pump, an induction motor, and a frequency converter. The control system was simulated in MATLAB/Simulink and tuned using neural network-based adaptive algorithms.

Results: the use of individual frequency control reduced energy consumption by an average of 25–35%. Pressure stability and operation of pumps at their optimal operating point were ensured. Hydraulic shock and cavitation phenomena were prevented. The model was evaluated as being close to real operating conditions and ready for practical implementation.