Багатопотоковий електрогідравлічний привод з адаптивною системою регулювання на базі штучного інтелекту

Abstract

The article addresses the problem of improving the efficiency of hydraulic systems through the implementation of multi-flow electrohydraulic drives with adaptive control systems based on artificial intelligence. The relevance of developing intelligent control algorithms for hydraulic systems to reduce energy consumption, improve positioning accuracy, and adapt to variable operating conditions is substantiated. A mathematical model of a multi-flow electrohydraulic drive has been developed, taking into account the nonlinear characteristics of hydraulic components, the interaction of parallel fluid flows, and the dynamics of variable loads. The architecture of a neural network for adaptive control of an electrohydraulic drive and a methodology for its training, including a preliminary stage on experimental data and adaptive retraining during operation, are proposed. The results of experimental studies are presented, confirming the significant advantages of the neural network controller compared to traditional control systems. It is shown that the implementation of an adaptive control system based on artificial intelligence provides reduced energy consumption, improved positioning accuracy, and reduced transition process times. Particular attention is paid to the stability of characteristics when changing operating conditions and temperature regimes. Practical aspects of implementing a neural network controller on a microcontroller platform with limited computing resources are considered. The economic efficiency of implementing multi-flow electrohydraulic drives with adaptive control systems based on artificial intelligence in various industries is substantiated.