Abstract: Online scheduling is usually referred to as a closed loop that includes a feedback after scheduling execution. The whole system depends on data collection fed to the receivers of the reactive layer from sensors and various input devices. The practical success of the scheduling system in reacting to practical applications depends entirely on having the data ready for execution for the algorithm on time. This Paper introduce a methodology for enhancing the rate of data collection from the input devices, which acts as a front end to the reactive scheduling system and will have a direct impact on scheduling horizon, frequency, time limit selection and reasoning reuse. We've got the solution for driving the multi-purpose machines and the way of control and enhance scheduling performance within the context of a dynamic production scheduling problem by using our Multi-Agent (MA) program model for the reactive scheduling methods. We've got the solution for driving the multi-purpose machines and get the way of controlling and enhancing the scheduling performance within the context of a dynamic production scheduling problem by using our Multi-Agent (MA) program model for the reactive scheduling methods, Multi-Agent (MA) using the Short Processing Time (SPT) that the rule of shorting time of the jobs in its queue. In this Paper, we compare between two methods (Multi-agent vs. fuzzy) to minimize the total processing time and improve the performance of the production lines. The results indicate that the Multi-agent approach has a better performance compared to the Fuzzy methodology.

Keywords: scheduling;Multi-Agent;Short Processing Time (SPT);fuzzy