Under the trend of energy reform, traditional power systems are also facing reforms towards smart grids, and power load forecasting and dispatch work in the grid is naturally of utmost importance. This article improved and optimized these two tasks through big data (BD) technology to drive changes in traditional power systems. It compared the prediction accuracy of the power system with and without BD support and the personnel ratings during mobilization work. The average prediction accuracy using BD was 89.71%. The average prediction accuracy without the use of BD was 79.28%, while in terms of scheduling efficiency, management efficiency, and operational status, the system scores using BD were 8.14, 8.34, and 8.57, respectively. The system scores for not using BD were 6.48, 7.02, and 6.94, respectively, indicating that BD has a comprehensive impact on the improvement of the power system. Even compared to advanced cloud computing technology, BD technology still has certain advantages. Therefore, this study believed that BD technology has the ability to significantly improve power load forecasting and power dispatch work.