Non-cooperative intelligent control Agents (ICAs) with dedicated cost functions, can lead the System to poor performance and in some cases, closed-loop instability. A robust solution to this challenge is to place the ICAs at the feedback Nash equilibrium point (FNEP) of the differential game between them. This paper introduces the designation of a robust decentralized infinite horizon LQR control System based on the FNEP for a linear time-invariant System. For this purpose, two control strategies are defined. The first one is a centralized infinite horizon LQR (CIHLQR) problem (i.e. a supervisory problem), and the second one is a decentralized control problem (i.e. an infinite horizon linear-quadratic differential game). Then, while examining the optimal solution of each of the above strategies on the performance of the other, the necessary and sufficient conditions for the equivalence of the two problems are presented. In the absence of the conditions, by using the least-squares error criterion, an approximated CIHLQR controller is presented. It is shown that the theorems could be extended from a two-Agent control System to a Multi-Agent System. Finally, the results are evaluated using the simulation results of a Two-Area non-reheat power System.

The increasing importance of transportation in urban spaces, as well as the numerous problems it has created, have prompted decision-makers to encourage citizens to use public transportation as much as possible. Therefore, scientists and urban decision-makers are looking for a comprehensive System for more connections between transportation Systems. On the other hand, the use of public transportation is very important and will create spatial justice, will bring proper access for all people. In this paper, a factor-based Multiprocessor routing including taxis, buses and agencies was designed. Agents appear in the role of travelers who have the opportunity to choose any possible means of transportation to cross from origin to destination and also fill the gaps in their path by walking. The main criteria are choosing the path of time and cost along the way. Field data were collected to evaluate the results of the model. The results of analyzing the field data by Multicriteria methods revealed that the priorities of people’s travelling mode are internet taxi, agency (called by telephone), bus and taxi (with predetermined rout and taking more than one passenger) with values of 0.4, 0.27, 0.19 and 0.14 respectively. The observations differed somewhat from the model output, which in fact represents optimal choices. This seems to be due to complex human decision-making in which many factors, including risk-taking and personality traits and habits and other factors, play a role.

After occurrence of a fault and outage of loads in a distribution network, it is tried to reconnect as much as possible loads in the minimum time by a process which is generally known as restoration. Computational and decision-making tools can play an important role to have a successful restoration. One of these tools is Multi-Agent System, which is a software-based System that includes components or Agents. These Agents have the ability to exchange information and make local decisions. In this paper, a Multi-Agent System is proposed for restoration of radial distribution networks. The Multi-Agent System brings about decentralized management in the restoration. Also, the proposed System considers priority for zones in distribution feeders. The priority is in accordance with type of loads in the zones. Implementation of the proposed method using Java Agent Development Framework (JADE) and its evaluation on a test distribution network, demonstrates an effective restoration with consideration of load priorities.

In this paper a finite-time flocking of Multi-Agent System with input saturation is investigated. By introducing a new control protocol, it is show that speed and position of Agents with double-integrator dynamics converge to the speed and position of leader in finite time. It is assumed that the graph between the Agents is connected and that at least one Agent is informed of the leader speed and position. It is shown that the Agents do not collide under this control protocol and the graph remains connected. Further, conditions are provided by which the control signal level remains less than the limiting saturation range.

Environmental monitoring for extraction of learner’s features and accurately learner modeling is a time and cost consuming task. This paper proposes a flexible framework using Agent architecture capabilities in learners’ monitoring and also in optimization of Intelligent Tutoring subSystems interactions. In this research learner’s learning approaches have been recognized by learning style theory then their uncertainty has been reduced by Bayesian Network. This framework prepares recommendations by combination of learners’ learning style and learner’s abilities (that are computed by Item Response Theory (IRT)) for three groups: learners, tutors and System designer. The architecture of proposed tutoring System has been presented in three layers. In middle layer there are four Agents that monitor learners, create learner’s model, update it and prepare some recommendations based on courseware features, learner’s abilities and his/her learning style. A study was conducted on 30 Computer Engineering students during one semester. The implementation of the proposed System on participants indicates an increase in all evaluation factors, for example it doubles educational success rate and learner’s satisfaction rate.

Bayesian Networks (BN) provides a robust probabilistic method of reasoning under uncertainty. They have been successfully applied in a variety of real-world tasks but they have received little attention in the area of load-frequency control (LFC). In practice, LFC Systems use proportional-integral controllers. However since these controllers are designed using a linear model, the nonlinearities of the System are not accounted for and they are incapable to gain good dynamical performance for a wide range of operating conditions in a Multi-area power System. A strategy for solving this problem due to the distributed nature of a Multi-area power System, is presented by using a BN Multi-Agent System. This method admits considerable flexibility in defining the control objective. Also BN provides a flexible means of representing and reasoning with probabilistic information. Efficient probabilistic inference algorithms in BN permit answering various probabilistic queries about the System. Moreover using Multi-Agent structure in the proposed model, realized parallel computation and leading to a high degree of scalability. To demonstrate the capability of the proposed control structure, we construct a BN on the basis of optimized data using genetic algorithm (GA) for LFC of a three-area power System with two scenarios.

This paper presents a new architecture for implementing intelligent Agents cooperating in a special Multi Agent environment, namely the RoboCup Rescue Simulation System. This is a three layered architecture which customized for cooperating three kinds of Agents to minimize damages, caused by a big earthquake in rescue simulation environment. Structural decision making algorithms are combined with heuristic ones in this model. One of advantage of this architecture is supporting both Centralized and Distributed Multi Agent Systems. At the end comparing this new architecture with those which be used till now could show the strength of new architecture.