This paper deals with assumptions for creation and implementation of the universal model for integration Petri nets and genetic algorithms, whose primary goal is to constantly control all jobs in a discrete event system and to guide the system to the targeted destination. In the first phase, the method of simulation discrete event system using Petri nets (PN) was described. Hereafter the basis for determining work schedules, methods of evolutionary computation, with special emphasis on genetic algorithm were presented. The aim of the second phases was to construct the integration method of genetic algorithm (GA) and matrix models MRF1 Petri nets into a comprehensive system of assigning jobs. It is necessary to determine the job schedule in a multi-project system with shared resources using heuristic rules which are priorities, delays and availability of jobs defined by genetic algorithm. The emphasis of the third phase was the verification of proposed algorithm at two systems. The first system is maritime traffic canal system in which to complete deadlock can be reached by taking improper channels by ships passing in opposite directions. The second system is a container terminal. We consider the problem of distribution jobs for unmanned vehicles to transport containers within the terminal. To increase the efficiency of container terminals, the proposed algorithm involves the procedure for forecasting and avoiding the conflicts and deadlocks. Also, the algorithm minimizes the waiting time for resources and minimizes the time that a ship spends in port. The algorithm was evaluated with different population sizes to examine the influence of this parameter on the convergence of results towards the final solution. The test results show the effectiveness and robustness of the proposed integration of GA and PN to determine the job schedule in a discrete event system.