Fuzzy time series (FTS) models have demonstrated significant potential in forecasting nonlinear and uncertain data; however, their accuracy heavily depends on the appropriate selection of fuzzy intervals and cluster centers. This study offers an enhanced FTS forecasting model that integrates a Hybrid Particle Swarm Optimization–Grey Wolf Optimizer (PSO-GWO) to optimize the cluster centers emanated from fuzzy c-means clustering. The ideal number of clusters is first gauged through the elbow method based on the sum of squared errors (SSE). Gaussian membership functions are employed to fuzzify the input data, and a weighted average method utilizing the cluster centers is used for forecasting. By improving the cluster centers in the normalized space, the suggested hybrid optimization approach considerably lowers the forecast's root mean square error (RMSE). The outperformance of the model in comparison to the alternative models pertaining to the lower RMSE value is illustrated using the TAIEX dataset.