Algortihm for problems rezolving in distributed system using agent technology.

Popa, Cicerone Laurentiu ; Parpala Ciobanu, Lidia Florentina ; Aurite, Traian 等

Abstract: A cooperative system represents a system where different users or agents work together at the same project from different locations. Agent technology has been recognized as an important approach for developing distributed intelligent design and manufacturing systems. In this paper we will present an algorithm that agents can use to resolve a problem occurred in a department of an enterprise part of a cooperative system.

Key words: agent technology, distributed system, algorithm, influence agent

1. INTRODUCTION

Enterprises in today economy context must consider collaboration with others to realize the product development in short time, with a lower cost, a short delivery time and a product that satisfy the client demand. Actual tend is to realize multidisciplinary cooperative systems for collaboration in real time between all professions involved in product development at the enterprise level. This collaborative environment actually constitutes a virtual enterprise. (Dragoi, 2003)

Actors that work in a cooperative system have access to common files, can exchange information and verify the project evolution. (Farley, 1998)

In agent-based programming, the agent is the basic element of distribution. Each agent serves as an independent component with its own local state and execution model. The agent designer can choose to assign a particular set of functionalities to an agent, specify the types of events and messages the agent may invoke or respond to, and implement those triggers and/or responses.

In building an agent-based system, each agent behaves independently, interacts with other agents through the events and messages that are communicated from one agent to another. (Weiss, 1999).

[FIGURE 1 OMITTED]

2. TYPES OF AGENTS USED IN OUR CASE STUDY

The environment has a direct influence to agent's action, so an agent approach from the environment perspective is very useful (Popa, 2005):

* Influence agent: is an agent with many privileges. The results of his action made for his own purpose realization will change the environment condition. Because of these changes other agents will be affected.

* Accessible agent: is the agent with a few privileges. When the environment will change he must wait for another agent to realize the purpose and his action will be affected.

* Mixed agent: in his action the agent become accessible or influence agent as a result of the action of some elements.

The agents involved must be able to take decisions, to offer possibilities of cooperation and to negotiate the relations between different domains involved in the project. (Popa&Hadar 2006).

For our case study we will suppose a manufacturing department part of a cooperative system based on agent technology. In this department more parts from the product assembly must be manufactured. We have "n" machine tools robots and conveyors; every machine is supervised by an agent. The supervisor agent we consider to be an influence agent and the agents that manages the machines are mixed agents.

3. CASE STUDY

We will suppose a problem that appear on machine 3 from the manufacturing department. To identify the problem, the agent that manages the machine activity will make a hierarchical verification and will eliminate the possible problem of interruption of machine functioning step by step. First the machine3's agent w ill announce the system that a problem occurred in the activity.

[FIGURE 2 OMITTED]

The supervisor agent must manage efficiently the crisis; he will communicate to the others agents involved in the process that a problem occurred in the system. The supervisor agent can decide: to interrupt all the activities from the system, or to overlying the operations. The supervisor agent will offer a solution by planning-programming-reprogramming the activities affected. (Figure 1)

After the machine3's agent announces the problem to the supervisor agent he will make a check of the general situation of the machine (Figure 2):

* First he will have an exchange of information with the agent that manages the delivery of electrical energy in the system. If the delivering of electrical energy for this machine is interrupted the machine3's agent will ask for the reason and will try to find a way to resolve the conflict. If he is not able to do it he will ask to the supervisor agent to manage this conflict because he is an influence agent in the system and has more privileges.

* Machine3's agent with the s upervisor agent's help will verify the workpiece and part supply. If the supply is interrupt maybe the previous machine from the system is damaged (the one that makes a previous machining of the part) or the conveyor is damaged. Also delay estimation is needed. In all this cases the machine3's agent will wait for new indications and solutions from the supervisor agent

* The agent will make verification for all the functions of the machine to find the operation that can't be made on this machine because of the damaged.

* Machine3's agent will make a part integrity verification using the data that come from the supervisor agent. If the workpiece is damaged or there is a machining error for this part made on a previous machine the agent will decide the part dismissing. Also a material error may occur (if the workpiece is made from other material that was design). The agent will verify the machine' s tool changer, if there is any problems with the tool transfer or tool assemble on the machine. For example to replace a tool from the machine tool changer the machine3's agent will contact the storage agent that can tell if there is a tool available, the time needed to transport the tool to the machine.

* The supervisor agent will also send a message to the troubleshoot agent that manage the human resource involved in the activity. The agent will generate a report and will supervise directly the human resource intervention and will give real time information and will make time estimation for fixing the problem.

4. CONCLUSIONS

All this problems can be avoided with a careful programming of the system or a part of them can be prevented by the agent anticipation for a problem appearance. However the system must be able to react, to find the problems and to resolve the conflicts. In the future we intend to study the agent interaction and analyze their behavior in the whole system.

5. REFERENCES

Dragoi, G. (2003). Intreprinderea integrata: metode, modele, tehnici si instrumente de dezvoltare si realizare a produselor, Politehnica Press, ISBN 973-8449-25-1, Bucuresti

Farley, J. (1998). Java[TM] Distributed Computing, O'Reilly Media, ISBN 1-56592-206-9, Sebastopol, USA

Popa, C.L. (2005). An Agent Classification From the Environment Perspective, Proceedings of 16th International DAAAM Symposium, Katalinic, B. (Ed.), pp. 309-310, ISBN 3-901509-46-1, Opatija-Croatia, Published by DAAAM International, Vienna

Popa, C.L. & Hadar, A (2006). A cooperative system design based on agent technology, Proceedings of 17th International DAAAM Symposium, Katalinic, B. (Ed.), pp. 311-312, ISBN 3-901509-57-7, Vienna-Austria, Published by DAAAM International, Vienna

Weiss G. (1999). Multiagent Systems: A Modern Approach to Distributed Artificial Intelligence, The MIT Press, ISBN 026223-203-6, Cambridge