Multi-agent Optimization Research Articles

Polymorphic optimization

The vast majority of engineering optimization literature focuses on one of two main aspects of the optimization process: algorithmic development and formulation development. Although researchers have recognized the interplay among different algorithms through mechanisms like hybrid algorithms, blackboard systems, and most recently agent-based optimization systems, relatively little attention focuses on the relationships among problem formulations. In this paper, we present a new view of the optimization process that distinguishes between model development and formulation development. Recognizing that each potential formulation represents a different form of the underlying model and that the model serves as a point of commonality among all formulations allows us to develop a framework wherein multiple formulations can coexist and seamlessly share problem and solution information. We term such an approach Polymorphic Optimization and present a sample implementation as part of a multi-agent optimization system. We illustrate this approach through the multi-objective optimization of microfluidic separation systems.

A Multi-Agent Model and Tabu Search Optimization to Manage Agricultural Territories

This paper proposes a multi-agent model to manage agricultural territories in order to reduce damage caused by runoff and floods processes. The proposed approach consists in optimizing the distribution of crops in a territory based on a multi-agent optimization approach. A group of agents represents each type of crop. The objective of an agent group is to occupy a total surface of parcels to achieve its objectives of production while reducing the value of risk for the whole watershed. Some agronomic, economic and hydrologic constraints restrict the behavior of agents and must be satisfied during the optimization process. The value of runoff risk is evaluated by a spatial hydrological model at the level of the Digital Terrain Model. Three optimization strategies based on Tabu search are developed and tested using real data. They provide satisfactory solutions in various computation times.

Bounded optimization of resource allocation among multiple agents using an organizational decision model

Multi-agent System (MAS) can be used to dispose bounded optimization problems with dynamically changing resources because its autonomous distributed management model aspect is fitted to dealing with such external disturbances. One of the problems in multi-agent optimizations is that it is difficult to rigorously define the optimization criteria with respect to the global optimization in advance. Rather, it may depend much on more situated factors such as temporal availability of resources and coexistence of current conflicts, conflicts among what has been already scheduled and what is to be scheduled. In this paper, an organizational model called Garbage Can Model (GCM) is introduced. In GCM, through its three decision-making strategies and the fluidities of problems and resources, solutions made by an individual agent are concerned with several agents that are co-existing in the environment. The problems allocated to each agent are solved not only by an agent's own efforts, but also by the change of problem solving status of other agents. Our simulation experiment shows that GCM is a preferred framework for multi-agent optimization problems in dealing with the above difficulties.

Multi-agent approach and stochastic optimization: random events in manufacturing systems

We propose a method to solve industrial problems and to take into account random events. It is called the triple coupling. It is based on stochastic algorithms, a simulation model and the multi-agents model of artificial intelligence. The method we propose is easy to use and allows us to take into account most of the constraints found in manufacturing systems. Experts look for solutions to increasing the capacity of production. But the production can be disturbed by random events experienced by the system. Industrial experts need schedules which prevent the consequences of random events. Minimizing such consequences is very important to increasing system delivery. Capital investment is often very high in factories and the cost of the investment goes on regardless of whether the resources are running or not. The multi-agent approach is used to determine schedules for which the consequences of random events are low, and a stochastic algorithm is proposed which permits us to optimize a random variable. We prove that this algorithm finds, with probability one, the schedule of the production for which the consequences of random events are the lowest. We propose to measure the consequences of random events using an influence ratio. Our approach has been used to study the consequences of random events in Peugeot sand foundries of Sept-Fons (France). A benchmark test is presented to prove the efficiency of our solution. For the Peugeot sand foundry of Sept-Fond, random events increase the production time by about 20% compared with the production time without any random events occurring. We have determined schedules of production for which the consequences of random events are about 0.5%.