In daily production life, heat generation and dissipation of electronic devices are important issues concerning the safety of the devices. To enhance the prediction and analysis of heat production of electronic devices, this paper analyzes and studies the heat production and dissipation of electronic systems of different electronic devices by collecting their historical operating power, hot spot temperature, ambient temperature, and other data, and iteratively optimizes these data using genetic algorithms to seek the best temperature fitting curve according to the research progress of existing artificial intelligence algorithms. The experimental results show that the population genes are sufficiently optimized as the number of iterations increases. The prediction model established by the genetic algorithm has a global optimization-seeking ability, high prediction accuracy, relatively small absolute and relative errors, and a fast convergence rate. This model has practical feasibility and can play a good role in the operation and maintenance of electronic devices.