Grid structure optimization using slow coherency theory and holomorphic embedding method

• Slow coherency theory’s relevance matrix establishes three types of nodes. • Employs the pure embedding method to solve the sigma indices. • Sigma indices constraints improved node voltage stability margins. • The discrete particle swarm algorithm can improve the speed of solving. This paper addresses the issue of complex fault oscillation modes and weak voltage points in large power systems by proposing a network structure optimization method that balances system synchrony and node voltage stability. The method uses slow synchrony clustering theory to establish node classification criteria and a comprehensive synchrony indicator for quantitative description of network synchrony performance. Simultaneously, it employs the holomorphic embedding method to solve the voltage sigma indicator for quantitative assessment of voltage stability. A model that considers both system synchrony and node voltage stability is then developed and optimized using a discrete particle swarm algorithm in simulations with 13-node, 118-node, and 2383-wp systems, compared to other classical algorithms. Simulation results show that the proposed optimization method effectively improves the synchrony clustering performance and voltage stability of the test systems, offering faster optimization speed and better results compared to other classical algorithms.