Concrete thermal parameters during construction and laboratory tests can differ, so the particle swarm optimization (PSO) algorithm and optical fiber temperature monitoring data were used to identify the thermal parameters of the concrete for super high arch dams at low winter temperatures. Traditional PSO algorithms can easily fall into local extrema, so a swarm intelligence-hybrid particle swarm optimization (HPSO) model was developed in this study using a concave function weight decreasing strategy to avoid local extrema. HPSO combines PSO with genetic algorithm cross and mutation operations. These more effectively balance the algorithm global and local search abilities. HPSO was used for thermal parameter inversion searches of various strength concretes and concretes with various gradations with cooling water and a specified environmental temperature. The model considers the influences of multi-stage water flows and water temperature variations along the cooling water pipe. Examples show the effectiveness of the intelligent identification system and the quick convergence of HPSO. The inversion results clarify the relationship between the thermal parameters and the temperature changes.