Research On Optimal Path Problem On Migrating Workflow Instances Based On Immune Algorithm

As a concept originated to manage those fixed activities of businesses, workflow can execute and monitor these activities by dividing them into well-defined tasks, roles, rules and procedures. As a result, workflow can achieve the aim at improving the organization level of business and work efficiency, and provide an advanced method for enterprises to carry out their business goals more effectively.The migrating workflow is a new technology which applies mobile computing to the traditional workflow managent area.In migrating workflow system that based on the mobile computing paradigm,the three elements are migrating workflow managent engine,migrating instance and workstation. Migrating workflow management engine is in charge of the definition of workflow processes, the generation of MI and the coordination of migrating instances.Workstation which refers to the anchorage server and working-machine network is the workspace of migrating instance.MI is defined as the executor of tasks. It is generated by management engine and can migrate within favorable workstations,using local resources and services provided by workstation executing one or many tasks.When MI finds current workstation can not statisfies the requirement of executing task,MI can migrate to another workstation that can statisfies it's requirement and continue executing.The path planning research of migrating workflow is a an important research direction within the workflow managent area. It optimizes the migrating decision of migrating instance. The so-called path planning refers to find a optimal or sub-optimal path from the start state to target state on the basis of some factor.Optimal path of MI is decided by a number factors as follows: service,safety, hardware and software environment of network and other constraints. Two factors are considered generally in the most current research on migrating workflow instances which are resourcess and services. Because of the intelligence, dynamic characteristics and autonomy of migrating workflow system, we need consider the dynamic changes of the environment from an overall perspective to propose solutions of path planning.On the basis of above research background, we consider that method of path planning should apperceive the information of environment and reflect the dynamic change of system resources. The article proposed a new method combining the global path planning based on series of tasks and the local path planning based on preferential choice of service based on the frame of migrating workflow system The optimal path problem is transformed into a multi - objective optimization problem.The planning process and framework of optimal path problem on migrating workflow instances are introduced in this paper.The safety factor is quantified and described based on fuzzy mathematics. The steps of local path planning based on the immunity algorithm are proposed in this paper and the important operators and functions are introduced into it.CLONALG and evolution are applied in the new algorithm. This algorithm has much better performance in the aspects of robustness adaptability. The experimental results show that the new algorithm can improve the efficiency of path planning and provide a new research direction.