Improvement of Scheduling Optimization of Cyber-Physical Systems Based on Petri Net and Intelligent Algorithm

Cyber-physical systems need more intelligent decision-making methods. To address this issue with respect to incomplete process models and inefficient scheduling, we have previously proposed a new method called Petri-nets-adaptive ant colony optimization (PN-AACO). This method targets small-scale job shops with shared resource limits. These shops require symmetric job designs for resource sharing but have asymmetric job processing times. PN-AACO uses Petri net symmetry at edge nodes but faces a problem. Its marking–transition pheromone index mechanism causes state space explosion from Petri nets. This leads to a decrease in the computational speed of the algorithm in the face of an increase in scale or state, which results in a longer overall manufacturing process time that impacts productivity. Thus, we propose the improved PN-AACO (iPN-AACO). The improved method uses transition–transition pheromone recording to control pheromone amounts. It also adds pheromone-based initial selection and best-known-paths-based probability rules. Tests show this approach speeds up computations up to 92% in more-states models while keeping scheduling effective.