Abstract: In order to improve the dynamic scheduling of industrial tasks and enhance long-term returns, a new industrial task scheduling algorithm based on genetic algorithm and task priority(GATP) is proposed. Firstly, the industrial task scheduling problem was modeled, and then the task was processed according to the task priority, and the preemption mechanism was designed to ensure the completion ratio of important tasks. Finally, the genetic algorithm was used to pre-allocate the tasks waiting to be scheduled to obtain higher task satisfaction and improve long-term revenue. The experimental results show that compared with IRSA, LLF and FCFS, when the number of task arrivals is high, the important task completion ratio of GATP is increased by 7.1%-31.9%, and the average task satisfaction is increased by 4.13%-18.62%. The algorithm performs better in terms of important task completion ratio and long-term revenue improvement.