Application of Particle Swarm Optimization-Based Intelligent Algorithms in Temperature Prediction for Changchun City

Temperature changes have extensive and profound impacts on the natural environment and human society. The temperature in Changchun City is a non-stationary series with significant variations and instability. To achieve accurate temperature prediction in Changchun City, this paper proposes an improved particle swarm optimization (PSO) algorithm to optimize the support vector machine (SVM) temperature prediction model. The improved PSO algorithm introduces adaptive inertia weights on the basis of traditional PSO to optimize the SVM's penalty factor and kernel function parameters, enhancing the algorithm's global search ability and local search accuracy. Using meteorological observation data from Changchun City, the improved PSO-SVM model is trained and tested and compared with the traditional SVM model. The experimental results show that the improved PSOSVM model significantly outperforms the traditional SVM model in terms of prediction accuracy and convergence speed, further validating the feasibility of the model and providing a reference for other prediction applications.