Research on Adjustable Resource Allocation Methods Based on Sac Algorithm

With the increasing variety of energy storage types, traditional optimization algorithms face significant challenges in solving adjustable resource allocation models due to exponential growth in computational complexity, slow convergence rates, and tendency to become trapped in local optima. This paper proposes a solution method using the Soft Actor-Critic (SAC) algorithm to address these limitations. First, a comprehensive adjustable resource allocation model is constructed incorporating electric, thermal, and hydrogen storage systems with their respective operational constraints. The SAC algorithm is then configured with carefully designed state space, action space, and reward function specifically tailored for energy storage optimization. Comparative simulation analysis against Proximal Policy Optimization (PPO) and Deep Deterministic Policy Gradient (DDPG) algorithms demonstrates that our SAC-based method achieves 8.3 % higher net revenue (7429k¥ vs. 6861k¥) and better satisfies operational constraints. The results validate that the proposed method can obtain optimal configurations of adjustable resources more efficiently and accurately while ensuring safe and stable power system operation, offering an effective solution for increasingly complex multi-energy storage systems.