Wenjun Liu, Azlan Mohd Zain, Mohamad Shukor Talib
et al.
Chicken swarm optimization (CSO) is a new metaheuristic algorithm inspired by biologically inspired metaheuristic algorithms that imitate the behavior of chicken flocks. CSO has the advantages of strong global search ability, high stability, and strong multi-subgroup collaborative search ability, and has important research potential. It is widely used in various optimization problems in real life. However, CSO also has some problems, including slow convergence, insufficient local search ability, and easy to fall into local optimality. Therefore, various variants of CSO have been proposed. This article presents a structural review of Chicken Swarm Optimization (CSO) and its variants published between 2014 and 2025. The review systematically organizes and synthesizes existing research by outlining the fundamental principles of CSO, categorizing improvement strategies, and analyzing the characteristics and performance of different variants. Instead of conducting an empirical comprehensive study, this structural review focuses on mapping algorithmic developments, identifying methodological trends, and summarizing application domains such as engineering design, energy systems, image processing, and intelligent diagnosis. Furthermore, the review highlights current limitations of CSO, discusses theoretical considerations, and provides future research directions. The presented structural framework offers clearer insights into the evolution of CSO and serves as a reference for researchers seeking to understand or extend CSO-based methods.
In this paper, we analyze the behavior of a multi-agent system driven by the interactions of agents within a competitive environment. To achieve this, we describe the transition probabilities that underlie the system's stochastic nature. We also derive the Fokker-Planck equations for the density distribution of the number of agents in the system and solve these equations for specific cases.
This paper demonstrates the use of the Multi-Agent MicroServices (MAMS) architectural style through a case study based around the development of a prototype traffic simulation in which agents model a population of individuals who travel from home to work and vice versa by car.
In this paper we examine the effectiveness of several multi-arm bandit algorithms when used as a trust system to select agents to delegate tasks to. In contrast to existing work, we allow for recursive delegation to occur. That is, a task delegated to one agent can be delegated onwards by that agent, with further delegation possible until some agent finally executes the task. We show that modifications to the standard multi-arm bandit algorithms can provide improvements in performance in such recursive delegation settings.
Existing protocols for multilateral negotiation require a full consensus among the negotiating parties. In contrast, we propose a protocol for multilateral negotiation that allows partial consensus, wherein only a subset of the negotiating parties can reach an agreement. We motivate problems that require such a protocol and describe the protocol formally.
This study examined a simulated confined space modelled as a hospital waiting area, where people who could have underlying conditions congregate and mix with potentially infectious individuals. It further investigated the impact of the volume of the waiting area, the number of people in the room, the placement of them as well as their weight. The simulation is an agent-based model (ABM).
Abstract Software defect prediction can detect whether there are defects in the program module so as to effectively reduce the unnecessary cost of software development and maintenance. In this paper, the limitation of the traditional BP neural network in the field of defect prediction leads to the inaccuracy of the prediction results. By using the global optimization ability of cuckoo search, the BP neural network is improved, the important initial parameters of the network are optimized, and the software defect prediction method of CS-BP is proposed. The experimental results show that compared with traditional machine learning algorithms such as BP neural network, J48 and SVM, CS-BP method has a better effect on the prediction of software defects.
The 2019 Multi-Agent Programming Contest (MAPC) scenario poses many challenges for agents participating in the contest. We discuss The Requirement Gatherers' (TRG) approach to handling the various challenges we faced -- including how we designed our system, how we went about debugging our agents, and the strategy we employed to each of our agents. We conclude the paper with remarks about the performance of our agents, and what we should have done differently.
The main contribution of this paper is a novel method allowing an external observer/controller to steer and guide swarms of identical and indistinguishable agents, in spite of the agents' lack of information on absolute location and orientation. Importantly, this is done via simple global broadcast signals, based on the observed average swarm location, with no need to send control signals to any specific agent in the swarm.
We are considering a scenario where a team of bodyguard robots provides physical protection to a VIP in a crowded public space. We use deep reinforcement learning to learn the policy to be followed by the robots. As the robot bodyguards need to follow several difficult-to-reconcile goals, we study several primitive and composite reward functions and their impact on the overall behavior of the robotic bodyguards.
Complex systems have interested researchers across a broad range of fields for many years and as computing has become more accesible and feasible, it is now possible to simulate aspects of these systems. A major point of research is how emergent behaviour arises and the underlying causes of it. This paper aims to discuss and compare different methods of identifying causal links between agents in such systems in order to gain further understanding of the structure.
In this note we study a problem of fair division in the absence of full information. We give an algorithm which solves the following problem: n $\ge$ 2 persons want to cut a cake into n shares so that each person will get at least 1/n of the cake for his or her own measure, furthermore the preferences of one person are secret. How can we construct such shares? Our algorithm is a slight modification of the Even-Paz algorithm and allows to give a connected part to each agent. Moreover, the number of cuts used during the algorithm is optimal: O (n log(n)) .
Tristan Charrier, François Schwarzentruber, Eva Soulier
We study the so-called dynamic coverage problem by agents located in some topological graph. The agents must visit all regions of interest but they also should stay connected to the base via multi-hop. We prove that the algorithmic complexity of this planning problem is PSPACE-complete. Furthermore we prove that the problem becomes NP-complete for bounded plans. We also prove the same complexities for the reachability problem of some positions. We also prove that complexities are maintained for a subclass of topological graphs.
The integration of multiple viewpoints became an increasingly popular approach to deal with agent-based simulations. Despite their disparities, recent approaches successfully manage to run such multi-level simulations. Yet, are they doing it appropriately? This paper tries to answer that question, with an analysis based on a generic model of the temporal dynamics of multi-level simulations. This generic model is then used to build an orthogonal approach to multi-level simulation called SIMILAR. In this approach, most time-related issues are explicitly modeled, owing to an implementation-oriented approach based on the influence/reaction principle.
In this paper, we develop a variational method to track and make predictions about a real-world system from continuous imperfect observations about this system, using an agent-based model that describes the system dynamics. By combining the power of big data with the power of model-thinking in the stochastic process framework, we can make many valuable predictions. We show how to track the spread of an epidemic at the individual level and how to make short-term predictions about traffic congestion. This method points to a new way to bring together modelers and data miners by turning the real world into a living lab.
This paper presents a simulation model based on the general framework of Multi-Agent System (MAS) that can be used to investigate construction project bidding process. Specifically, it can be used to investigate different strategies in project bidding management from the general contractors' perspective. The effectiveness of the studied management strategies is evaluated by the quality, time and cost of bidding activities. As an implementation of MAS theory, this work is expected to test the suitability of MAS in studying construction management related problems.
This paper develops a dynamic agent-based model for rural-urban migration, based on the previous relevant works. The model conforms to the typical dynamic linear multi-agent systems model concerned extensively in systems science, in which the communication network is formulated as a digraph. Simulations reveal that consensus of certain variable could be harmful to the overall stability and should be avoided.