With the deepening reform of the power market, great progress has been made in informatization. Blockchain can improve the reliability of power management system (PMS) data processing. PMS informatization has become the basis for improving the quality and efficiency of project management and maximizing the social and economic benefits of the project. Due to the requirement of safe and stable power production, PMS attaches great importance to the application and implementation of information in power management, but does not attach enough importance to the informatization of power production management. Therefore, this article analyzes the current situation, characteristics, and existing problems of PMS through a machine learning algorithm, then constructs the design principles, and finally proposes the optimization path of PMS according to the principles. The information collection ability and system control ability of the optimized PMS were better than the original PMS. The information collection ability was 14.2% higher than the original, and the system control ability was 9.8% higher than the original. In general, both blockchain and machine learning can improve the data reliability of PMS.