Various physical phenomena in nature provide a wealth of heuristic inspiration for optimization algorithms. This paper proposes a new heuristic optimization algorithm based on the stellar flare phenomenon, the Stellar Flare Optimization Algorithm (SFOA). This algorithm achieves efficient search for complex optimization problems by simulating the sudden energy release, neighborhood cooperative perturbation, and energy recovery mechanism of stellar flares. The algorithm introduces flare cooperation and energy recovery mechanisms to form a unique search strategy, and achieves a balance between exploration and utilization by adaptively controlling the flare trigger probability and perturbation amplitude. This paper elaborates on the algorithm design principles, mathematical modeling, pseudocode flow, and complexity analysis, and discusses its potential application value in multi-peak complex optimization problems.