Abstract
The growing number of electricity consumers and unstable generation by renewable energy sources (RESs) decrease the efficiency of traditional techniques for modeling large power systems. The diakoptics method is based on structuring a system into subsystems and enables reducing the amount of computation by 2.9 times. However, its classical implementations require successive (iterative) expansion of a scheme, which increases computational time for scaling. This method decomposes a complex system into interacting subsystems with different generation and load characteristics. Generalized parameters are independently calculated for each subsystem and then integrated into the total solution. This significantly reduces the amount of computation, simplifies simulation, optimizes control with allowance for local characteristics, and analyzes intersystem connections to minimize losses and improve stability. This paper suggests not just another Kron calculation but a modification of the diakoptics method, which enables avoiding multiple step-by-step redesigns of a grid by generating generalized parameters based on the complete graph of a system, which significantly reduces computational costs when analyzing large power systems. The method simultaneously considers the entire grid topology. This eliminates successive iterations: if a classical approach requires 13 iterations for a network with 11 nodes and 21 branches, the suggested method performs computation in a single computational cycle. The results show high computational efficiency and practical applicability of the method for large power systems. The suggested approach opens up possibilities for reliability analysis, including redundancy assessment and identification of vulnerable elements, which contributes to an increase in the power system stability under conditions of RES integration.