Research on Production Scheduling of Industrial Big Data for Internet of Things Based on Dynamic Planning Algorithm

Abstract

Abstract In this paper, the optimized Dual Heuristic Dynamic Programming (DHP) algorithm is studied in depth, compared and analyzed with the other three algorithms through experimental simulation, and the DHP algorithm is applied to different industrial accurate scheduling tests to explore the loads of the industrial equipment during scheduling operation. The simulation results show that the dual-heuristic dynamic programming DHP algorithm has a significant decrease in both the mean waiting time (MWT) and the mean response time (MRT), where the highest peak value of the mean waiting time is only 0.06901 μs, and the highest peak value of the mean response time is 0.24493 μs. In industrial accurate production scheduling, the improved DHP algorithm performs the best for textile industry’’s task scheduling with the best performance. Specifically, the task scheduling completion time for the textile industry is 4522 seconds when the number of automated guided vehicles (AGVs) is the minimum value of 3. In comparison, the completion time is 8541 seconds when the number of AGVs is the maximum value of 15. The textile industry shows the fastest task completion time compared to the other industry types for different settings of the number of AGVs. In addition, the analysis of the operation of the equipment engines shows that although there are high and low torques during operation, the Torque of most of the engines stays within the range of 29 Nm to 30.7 Nm, which is generally stable and has little impact on the performance of the equipment. The improved dynamic programming algorithm proposed in this paper has effective application potential in industrial extensive data production scheduling, which can improve the efficiency and responsiveness of production scheduling. It has important practical significance for industrial production management.