Development of an Optimized Compressor Station Control System to Minimize Energy Consumption

The article presents an intelligent control system for a compressor station (CS) based on a hybrid application of a genetic algorithm (GA) and gradient descent (GD). Two key process parameters are optimized: the compressor rotor speed (ω, rpm) and the discharge pressure (P_out, bar). The model automatically adapts its operating modes to changes in gas flow, ambient temperature, and load factor, reducing energy consumption without compromising performance. Experimental studies on a 5 MW digital twin of a power plant showed that the hybrid GA+GD provides energy savings of 12.1% to 17.7% compared to the classical PID controller and 5-7% better than using only GA. The adaptation time to the new mode does not exceed 40 seconds, and the power fluctuations are reduced to ±1.3%. The proposed system can be implemented on the basis of industrial controllers with limited computing resources.