Integrated Energy Multi-Agent Collaborative Optimization Regulation Method for Typhoon Events
Abstract
Typhoon disasters pose severe challenges to the safe and stable operation of integrated energy systems. Traditional single energy regulation models are difficult to cope with the dynamic uncertainties in multi-energy flow coupled systems. In this study, a regulation method based on multi-agent collaborative optimization is proposed to achieve the dynamic balance of the electricity-gas-thermal multi-energy system under typhoon events by constructing a distributed decision-making framework. Experiments show that in the simulated line fault scenario caused by a typhoon, multi-agent cooperative regulation reduces the average voltage deviation of key nodes from 0.17 p.u. to 0.05 p.u., verifying the effect of the cooperative strategy on improving voltage stability. For the extreme scenario of a 24-hour continuous typhoon, this method improves the power supply reliability of the system from 82.3% to 95.6%, and the renewable energy consumption rate by 20.7 percentage points (from 68.5% to 89.2%). At the same time, through multi-energy complementation and demand-side response, comprehensive operating costs are reduced by 21.1%. Iterative analysis reveals that the five types of agents exhibit significant differentiation characteristics in the regulation process. Notably, the peak regulation amount of the S3 agent reaches 38 units, and the system prediction error MAE (Mean Absolute Error) remains stable at below 0.08 after 20 iterations. The load fluctuation experiment further shows that collaborative optimization can narrow the normalized load fluctuation range to ± 0.15 and increase the unit load resource utilization rate by 12%, confirming the rapid convergence ability and dynamic balance advantages of this method under extreme disasters.