Fig 4 - uploaded by Alfredo Weitzenfeld
Content may be subject to copyright.
Beta wolf hunting behavior. Consists of five states: Wander, Formation, Stalk, Attack and Eat. Each state is activated by the corresponding conditions.
Source publication
A great amount of work has been made in biologically-inspired robotic systems on single and multiple animal behavior models. These studies have advanced the understandings of animal behavior and have provided at the same time inspiration in the design of single and multiple robotic architectures. Additionally, applications in the real word domain h...
Contexts in source publication
Context 1
... Wolves. The beta wolf behaviour is described by five states, Wander, Formation, Stalk, Attack and Eat, as shown in Figure 4: ...
Context 2
... wolf behaviors have additional states from that of the alpha wolf. In addition to the basic alpha states, the beta wolf includes a formation state in response to the presence of the alpha wolf, as previously shown in Figure 4. While the alpha wolf only considers the presence of a prey (or predator), the beta wolf also responds to the presence of its leader, the alpha wolf. ...
Similar publications
![Figure 1.1: Comparison of wolf body size to that of prey [17].](profile/Amy-Briggs/publication/264885172/figure/fig1/AS:295910555832328@1447561982542/Comparison-of-wolf-body-size-to-that-of-prey-17_Q320.jpg)
![Figure 2.2: Mormon cricket with radio transmitter [18].](profile/Amy-Briggs/publication/264885172/figure/fig3/AS:295910555832330@1447561982661/Mormon-cricket-with-radio-transmitter-18_Q320.jpg)
![Figure 3.1: Pack formation with alpha and beta wolves [29].](profile/Amy-Briggs/publication/264885172/figure/fig4/AS:295910555832331@1447561982687/Pack-formation-with-alpha-and-beta-wolves-29_Q320.jpg)
Citations
... In this algorithm, there are two types of wolves: alpha wolves and beta wolves . The alpha wolves are responsible for exploring the search space and finding potential solutions, while the beta wolves follow the alpha wolves and refine the solutions found by the alpha wolves (Weitzenfeld & Vallesa, 2006). ...
Because most attacks target computers, intrusion detection has emerged as a key component of network security. This is a result of the widespread expansion of internet connectivity and information system accessibility on a global scale. The Wolf Sheep Predation Algorithm (WSPA), evolved from the Wolf Pack Algorithm. It models how wolves hunt in packs. This paper focused on the Lotka-Volterra predator-prey model. Due to its global convergence and computational strength, it has mostly been applied in a variety of engineering optimization issues. The method, however, has numerous flaws, including slow convergence and a tendency to quickly reach the local optimum. To address the above-mentioned flaws, this research developed the Modified Wolf Sheep Predation Algorithm (MWSPA) to reduce network threats. the algorithm models the wolves and sheep, where the wolves in this study represent the network security agent while the sheep represent network threats. The model suggests a better strategy to address the problem of slow convergence and quickly reach the local optimum by making sure that there is a balanced ecosystem at any point in time. This is achieved by ensuring that the network security agents(wolves) are not outnumbered by threats(sheep) and they do not become extinct when there is no food source. So in the absence of food, the MWSPA ensures the wolves can survive on grass and maintain their strength to hunt their next prey. This idea prevents the algorithm from crashing if the wolves die while the prey grows to infinity and consumes all the available grass. This therefore solves the problem of rapidly failing into a local optimum. This study aimed to identify the most pertinent features employed by wolves (network security agents) while hunting the sheep (network threats). We therefore established that sense of hearing and smell, splitting prey, encircling prey, assisting the hunter with the best chance of success, and looking for alternative prey as the most outstanding attributes used by wolves while hunting. The study further evaluated the MWSPA, and the outcomes demonstrate that the suggested algorithm outperforms its predecessor approach in a variety of search environments. Therefore, this shows that the MWSPA may possess the necessary qualities for creating a solution that will completely eradicate network threats and might provide leads in solving growing cybersecurity concerns globally.
... The leader-follower mode was first proposed by Desai J P [27] and is a widely used algorithm, but the model relies too much on the navigator and is difficult to deal with navigator crashes and lacks sufficient robustness. Weitzenfeld A et al. [28] proposed an algorithm with a backup navigator to improve the leader-follower model's robustness. Lewis M A [29] et al. proposed the virtual navigator concept, in which there is no real navigating UAV, but a virtual navigator instead, which further improves the stability compared to the leader-follower model. ...
... The closer the distance, the stronger the repulsive force. Let UAV be subjected to the repulsive force field generated by other UAVs besides itself, and let the repulsive force field of UAV be as shown in Equation (28). ...
The current challenge in drone swarm technology is three-dimensional path planning and adaptive formation changes. The traditional A* algorithm has limitations, such as low efficiency, difficulty in handling obstacles, and numerous turning points, which make it unsuitable for complex three-dimensional environments. Additionally, the robustness of drone formations under the leader–follower mode is low, and effectively handling obstacles within the environment is challenging. To address these issues, this study proposes a virtual leader mode for drone formation flight and introduces a new Theta*–APF method for three-dimensional space drone swarm path planning. This algorithm optimizes the A* algorithm by transforming it into an omnidirectional forward Theta* algorithm. It also enhances the heuristic function by incorporating artificial potential field methods in a three-dimensional environment. Formation organization and control of UAVs is achieved using speed-control modes. Compared to the conventional A* algorithm, the Theta*–APF algorithm reduces the search time by about 60% and the trip length by 10%, in addition to the safer flight of the UAV formation, which is subject to artificial potential field repulsion by about 42%.
... c [5][6][7], among others, and it is a prevalent natural phenomenon. The trapping phenomenon has provided valuable insights into swarm robotics technology and intelligent systems [8][9][10][11]. A range of methods for multi-robot cooperative trapping, akin to behaviors observed in the natural world, have been developed. These methods find application in various domains, including military and security [12,13], search and rescue [14][15][16], engineering and transportation [17], healthcare and medicine [18], and more. ...
In swarm robotics, multi-target trapping usually relies on global information or explicit communication, posing a challenge for robots to autonomously self-organize and trap multiple targets with only local perceptual data. We present a self-regulated density-based approach for self-organized multi-target trapping. This method employs density-based negative feedback control and density-distance weighted target selection. Our approach leverages distributed perception, where robots perceive nearby peers within their sensory range, eliminating the need for explicit information exchange and enabling autonomous trapping. During task execution, negative feedback control adjusts swarm density, ensuring it reaches an optimal level. Building on this, individual robots use density fields and relative target positions to select suitable targets, achieving self-regulated dispersed selection and multi-target trapping. We validate our approach through numerical simulations and real-world robot experiments. Results demonstrate stable self-organization, efficient self-regulated dispersed target selection, and successful target trapping, even with a limited number of trapping robots in low-redundancy scenarios.
... Multi-robot systems are widely deployed in a large range of robotic fields, such as multi-body formation, logistics and transportation, cooperative operation and modular robotics [1]- [8]. Cooperative hunting, originated from the hunting behavior of wolves in the wild, is a typical and significant scene to study multi-robot behaviors, where the predatory robots exchange local information with each other to conduct the cooperative hunting [9]- [12]. Most of the current research are based on classical control theory to construct control strategies according to the mathematical models of agents, which are difficult to build accurately in real life. ...
... Biological systems have evolved such innovative ways of cooperating that they often serve as the inspiration for optimization algorithms [235], robotic control algorithms [78,227,265], and methods for designing controllers (e.g. via evolutionary algorithms) [8,78,122]. Regarding the design of controllers, the difficulty often lies in the dimensionality of either the state space or the control space. ...
Supremacy in armed conflict comes not merely from superiority in capability or numbers but from how assets are used, down to the maneuvers of individual vehicles and munitions. This document outlines a research plan focused on skirmish-level tactics to militarily relevant scenarios. Skirmish-level refers to both the size of the adversarial engagement -- generally one vs. one, two vs. one, and/or one vs. two -- as well as the fact that the goal or objective of each team is well-established. The problem areas include pursuit-evasion and target guarding, either of which may be considered as sub-problems within military missions such as air-to-air combat, suppression/defense of ground-based assets, etc. In most cases, the tactics considered are comprised of the control policy of the agents (i.e., their spatial maneuvers), but may also include role assignment (e.g, whether to act as a decoy or striker) as well as discrete decisions (e.g., whether to engage or retreat). Skirmish-level tactics are important because they can provide insight into how to approach larger scale conflicts (many vs. many, many objectives, many decisions). Machine learning approaches such as reinforcement learning and neural networks have been demonstrated to be capable of developing controllers for large teams of agents. However, the performance of these controllers compared to the optimal (or equilibrium) policies is generally unknown. Differential Game Theory provides the means to obtain a rigorous solution to relevant scenarios in the form of saddle-point equilibrium control policies and the min/max (or max/min) cost / reward in the case of zero-sum games. When the equilibrium control policies can be obtained analytically, they are suitable for onboard / real-time implementation. Some challenges associated with the classical Differential Game Theory approach are explored herein. These challenges arise mainly due to the presence of singularities, which may appear in even the simplest differential games. The utility of skirmish-level solutions is demonstrated in (i) the multiple pursuer, single evader differential games, (ii) multi-agent turret defense scenarios, and (iii) engage or retreat scenarios. In its culmination, this work contributes differential game and optimal control solutions to novel scenarios, numerical techniques for computing singular surfaces, approximations for computationally-intensive solutions, and techniques for addressing scenarios with multiple stages or outcomes.
... Beta wolves are secondary wolves that assist the alpha to make various decisions associated with the pack. Beta wolves could be both male and female and are the strongest alpha candidates in case an alpha wolf dies or gets too old [15,33]. The lowest level among grey wolves is the omega. ...
In the 19th and 20th centuries, social networks have been an important topic in a wide range of fields from sociology to education. However, with the advances in computer technology in the 21st century, significant changes have been observed in social networks, and conventional networks have evolved into online social networks. The size of these networks, along with the large amount of data they generate, has introduced new social networking problems and solutions. Social network analysis methods are used to understand social network data. Today, several methods are implemented to solve various social network analysis problems, albeit with limited success in certain problems. Thus, the researchers develop new methods or recommend solutions to improve the performance of the existing methods. In the present paper, a novel optimization method that aimed to classify social network analysis problems was proposed. The problem of stance detection, an online social network analysis problem, was first tackled as an optimization problem. Furthermore, a new hybrid metaheuristic optimization algorithm was proposed for the first time in the current study, and the algorithm was compared with various methods. The analysis of the findings obtained with accuracy, precision, recall, and F-measure classification metrics demonstrated that our method performed better than other methods.
... Biological systems have evolved such innovative ways of cooperating that they often serve as the inspiration for optimization algorithms [5], robotic control algorithms [10], [11], [12], and methods for designing controllers (e.g. via evolutionary algorithms) [12], [13], [14]. Regarding the design of controllers, the difficulty often lies in the dimensionality of either the state space or the control space. ...
In this paper a scenario is considered in which a group of predators cooperate to maximize the number of prey captures over a finite time horizon on a two-dimensional plane. The emphasis is on developing predator strategies, and thus the behavior of the prey agents is fixed to a Boids-like flocking model which incorporates avoidance of nearby predators. At each time instant, the predators have control authority over their heading angle; however, we restrict the headings to be governed by one of five different pre-specified behaviors. No communication occurs between the predator agents-each agent determines its own control without knowledge of what the other predators will implement; thus, the predator strategies are fully decentralized. The full state space of the system is collapsed to a set of features which is independent of the number of prey. An evolutionary algorithm is used to evolve an anchor point controller wherein the anchor point lies in the feature space and has a particular predator behavior associated, thus providing a candidate solution to the question of "what to do when". The two predator case is the focus in this work, although the ideas could be applied to larger groups of predators. The strategies resulting from the evolutionary algorithm favor aiming at the nearest prey mostly, and also avoiding having the predators getting too close and then pursuing the same prey. Thus useless behaviors are generally not present among the elite at the end of the evolutionary process.
... Put together, our model in this paper along with recent related work (see [52]) helps towards advancing our understanding of how a functional embodied and situated AI that can operate in a multi-agent social environment. For this purpose, we plan to extend this model to study other aspects of cooperation such as in wolf-pack hunting behavior [69,70], and also aspects of competition within agent populations as in predator-prey scenarios. In ongoing work, we are developing a setup in which embodied cognitive agents will have to compete for limited resources in complex multi-agent environments. ...
What is the role of real-time control and learning in the formation of social conventions? To answer this question, we propose a computational model that matches human behavioral data in a social decision-making game that was analyzed both in discrete-time and continuous-time setups. Furthermore, unlike previous approaches, our model takes into account the role of sensorimotor control loops in embodied decision-making scenarios. For this purpose, we introduce the Control-based Reinforcement Learning (CRL) model. CRL is grounded in the Distributed Adaptive Control (DAC) theory of mind and brain, where low-level sensorimotor control is modulated through perceptual and behavioral learning in a layered structure. CRL follows these principles by implementing a feedback control loop handling the agent’s reactive behaviors (pre-wired reflexes), along with an Adaptive Layer that uses reinforcement learning to maximize long-term reward. We test our model in a multi-agent game-theoretic task in which coordination must be achieved to find an optimal solution. We show that CRL is able to reach human-level performance on standard game-theoretic metrics such as efficiency in acquiring rewards and fairness in reward distribution.
... According to the hierarchical structure of the wolves, Zhou et al. [10] divides wolves into three logical layers by defining each level's control range and mimicking the mutual influence among them, to get a stable aero-formation when flying of cluster system. Weitzenfeld et al. [11] implements the basic wolf pack algorithm using three robotic dogs to copy the hunt and avoidance behaviors of wolves. However, without effective communications between different levels, when the target distance is far and the search space is large the hunting wolf will waste a lot of time and energy. ...
Swarm intelligence inspired algorithms have so many profound natural advantages in solving large-scale and distributed problems. This paper systematically analyzes the characteristics of wolves’ behaviors such as cooperative searching, hunting and attacking, and further abstracts those behaviors into four basic ways, that is, wandering, summoning, lurking and besieging, in accordance with the different roles of wolves. Then, we formulate a cluster cooperative rule based on the principle of Dynamic Wolf Head Alternation and Real-time Role Assignment, and propose a fatigue-rendering tactics based on interception strategy in two teams. Finally, the clustering cooperative rule enlightened by the group’s behavior is established, and the convergence of the algorithm is proved with the Markov asymptotic convergence theory. Experiments show that the model can effectively guarantee the efficiency of solving large-scale complex optimization problems and the operational effectiveness of distributed cluster cooperative attack problems.
... Put together, our model in this paper along with recent related work (see [44]) helps towards advancing our understanding of a functional embodied and situated AI that can operate in a multi-agent social environment. For this purpose, we plan to extend this model to study other aspects of cooperation such as in wolf-pack hunting behavior [58,59], and also aspects of competition within agent populations as in predator-prey scenarios. In ongoing work, we are developing a setup in which embodied cognitive agents will have to compete for limited resources in complex multi-agent environments. ...
In order to understand the formation of social conventions we need to know the specific role of control and learning in multi-agent systems. To advance in this direction, we propose, within the framework of the Distributed Adaptive Control (DAC) theory, a novel Control-based Reinforcement Learning architecture (CRL) that can account for the acquisition of social conventions in multi-agent populations that are solving a benchmark social decision-making problem. Our new CRL architecture, as a concrete realization of DAC multi-agent theory, implements a low-level sensorimotor control loop handling the agent's reactive behaviors (pre-wired reflexes), along with a layer based on model-free reinforcement learning that maximizes long-term reward. We apply CRL in a multi-agent game-theoretic task in which coordination must be achieved in order to find an optimal solution. We show that our CRL architecture is able to both find optimal solutions in discrete and continuous time and reproduce human experimental data on standard game-theoretic metrics such as efficiency in acquiring rewards, fairness in reward distribution and stability of convention formation.
