The Scheduling Model Of Active Network Project With Dividable Task And Its Hybrid Co-evolutionary Algorithm

In research of traditional active network collaborate, it restricted the processing time and the due date. But in actually, due to the various completion of the project, there are so many stochastic and uncertain factors in the real product environment, such as machine broke-down, processing ability decreased, the production of the optimization, the core technical staff leave of emergent events, the most wok will not hit the product target just at the delivery date, Therefore, the product collaboration is badly in need of a more flexible plan form, change the traditional rigid control into a flexible control system. Simultaneously, because of the cooperation between enterprises more inseparable, collaboration tasks in the detachable makes small and medium enterprises to obtain greater living space, to create greater customer value. As the establishment and development in a variety of related subjects and optimization technology, there are more methods to resolve the collaborative plans. Using the superiority of one algorithm to make up for the deficiency of another algorithm or various kinds of hybrid algorithm enhance the performance of the algorithm. However, although different kind of evolution algorithm has its irreplaceable advantage, it also has its own weakness and insufficiency. Thus co-evolution of the hybrid mode became a new pattern to get a more optimization algorithm, a lot of researches show that using good global optimization algorithm and good local search algorithm combining, algorithm exhibits faster convergence speed and obtained the better the quality of the solution.1. This paper using co-evolution of the swarm intelligence algorithm, research collaborative planning based on active network in project management, explore scientific collaborative planning, combined with the actual production and the influence of environment, establish the activity network collaborative planning model, and as the object, design the corresponding algorithm for solving the problem. The main research work is as follows.2. Construct the co-evolution hybrid algorithm frame; analyze the hybrid pattern among each evolutionary algorithm; elaborate the current situation of co-evolution hybrid algorithm at home and abroad; describe the procedure of the hybrid evolutionary algorithm frame which based on the genetic algorithm, the ant colony algorithm, the particle swarm optimization, and the harmony search algorithm and the differential evolution algorithm.3. Study the fuzzy scheduling model. In the consideration of fuzzy factors during the real production processing, and construct the mathematical model of fuzzy scheduling, describe the exact operating procedure of fuzzy numbers; propose the co-evolutionary algorithm based on the idea of co-evolution, the co-evolutionary algorithm combines the advantages of several evolutionary algorithms, and abandon the inferiority among them, solve the fuzzy scheduling problems in a high effectiveness.4. In the actual implementation process, collaborative planning due date uncertainty and the task of reparability constructed with time windows task can be divided into active network collaborative planning model, based on hybrid evolutionary algorithm, combined with the evolutionary algorithm advantages, abandon its disadvantages, design a hybrid particle swarm optimization algorithm, the solution of the model is efficient an example.