METHODS AND ALGORITHMS FOR PROACTIVE CONTROL OF DIGITAL SUBSTATION IN ACTIVE-ADAPTIVE ELECTRIC POWER SYSTEMS

Abstract

The paper considers methods and algorithms for proactive control of energy facilities in active-adaptive electric power systems. Proactive control is presented as an organizational task requiring a new approach to the distribution of decision-making functions. It is shown that digital substations are transformed into active agents of the control system, and digital transformers act as information agents ge¬nerating data to support decisions at all levels of the hierarchy. Aim. Development of methods and algorithms for proactive control that improve the validity and efficiency of dispatch personnel decisions based on predictive analysis of digital transformer data and redistribution of functions between hierarchy levels. Materials and methods. The research is based on the theory of organizational systems, the theory of active systems and the theory of information processes and systems. Methods of system analysis, mathematical and simulation modeling, artificial intelligence (hybrid neural network architectures CNN + Bi-GRU with attention), and experimental methods on the Avacha platform were applied. Results. A concept of distribu-ted proactive control with the transfer of intelligent support functions to the field level has been developed, reducing response time to 1–2 ms. A decision support method based on a neural network architecture with 96.1 % accuracy and 0.8 MFLOPS efficiency is proposed. The principle of function distribution with an asynchronous lock-free synchronization algorithm is substantiated. A methodology for integration into the digital substation infrastructure compatible with IEC 61850 has been developed. Experiments on the Avacha platform confirmed 95.7 % accuracy and 1.2 ± 0.3 ms response time. The results are implemented at Chelenergopribor LLC. Conclusion. The developed methods implement a new organizational structure for digital substation management corresponding to the strategy of creating intelligent active-adaptive systems and increase the survivability of power systems through early detection of emergency modes.