One, the natural world of group intelligence traceability
In the natural world of many biological groups, such as ant swarms, bee colonies, bird flocks, etc., in the process of their swarming life, even if there is not a unified, clear leadership core, but can show a harmonious and orderly state. These groups of organisms are composed of simple individuals with limited intelligence, but when they come together, they can perform complex intelligent behaviors that are beyond the reach of individuals, such as ant colonies that search for and transport food according to established routes, bee colonies that work together to build elaborate nests, and migratory birds that form a stable queue on their way to migrate.
The behavior of these individual organisms is not "cooperative" in the sense of human society, and they have only limited, simple responses to group activities, so what is it that makes these collective efforts possible? In the case of ants, for example, research has shown us that ants release a special chemical to induce other ants to adopt directed behavior during the search for the shortest path to the nest, and that it is this simple signal that gives coordination to complex systems. The similarly intelligent behavior exhibited by these swarming creatures is group intelligence.
Second, the characteristics of group intelligence
The characteristics of group intelligence can be summarized as follows: first, the control is distributed, there is no central control, and therefore it is more adaptable to the current network environment in the state of work, and will not be due to the failure of one or several individuals to affect the group of the entire problem solving; second, each individual in the group is able to change the environment. Secondly, each individual in the group can change the environment, and can transmit information and cooperate through non-direct communication, which is a kind of indirect communication between individuals, and is called "*** knowledge initiative"; again, the ability of each individual in the group or follow the rules of behavior is very simple, which makes the group intelligence simple and easy to realize; finally, the group has self-organization, that is, the group exhibits a self-organization, that is, the group exhibits a self-organization, that is, the group exhibits a self-organization. organization, i.e., the complex behaviors shown by the group are the intelligence that emerges through the interaction process of simple individuals, which is different from the intelligence of individuals.
On the other hand, the group intelligence must follow the following principles in the process of actually generating the effect:
1, the principle of proximity, the group is able to carry out simple spatial and temporal calculations.
2, the principle of quality: the group can correspond to the quality factors in the environment.
3, the principle of diversity of response, the group should not be too narrow range of action.
4, the principle of stability: the group should not change its behavior every time the environment changes
5, the principle of adaptability: the group is able to change its behavior at the right time when the required cost is not too high.
The representative algorithms of group intelligence include the ant colony algorithm, which makes use of the collective optimization characteristics of the biological ant colony to search for the shortest path from the ant colony to the food through simple information transfer between individuals; and the particle swarm optimization algorithm, whose basic concept originated from the study of flocking of birds in group movement behavior, and searches for the optimal solution through the collaboration of individuals, which was initially designed to graphically simulate the graceful and unpredictable behavior of flocks of birds in a two-dimensional space. Simulating the graceful and unpredictable movements of bird flocks, it was later used to solve optimization problems.
Third, the application of group intelligence
Group intelligence as an emerging discipline, its algorithms and ideas can be applied to many fields, such as for the traveler's problem, the quadratic assignment, the workshop scheduling, the sequence of sequential ordering, graphic coloring, connection-oriented network routing as well as the connectionless network routing and other combinatorial optimization problem solving; the development and manufacture of group robots; by simulating the real ants The development and manufacture of colony robots; by simulating the behavior of real ants in sorting their young or foraging collectively, retrieving Web information, filtering out the content needed by users, and classifying Web documents; and applying it to classification and prediction, cluster analysis, etc., to construct new algorithms for data mining. In addition, with the development of society and technological progress, group intelligence in the field of Internet of Things, computer industry, metallurgical automation, and power systems and other aspects of the application of space is also being developed.
An example here is one of the applications of group intelligence - the manufacture of group robots. Desktop miniature mobile robot ROBO-MAS is one of the representatives, which not only can carry out group collaboration of a large number of intelligent robots in a limited space, but also has a leading edge in the study of dynamic assignment of tasks in known environments, the study of task prediction and task decomposition, as well as the study of heterogeneous large-scale multi-coupled tasks. The operational platform that enables this robot to function is the ROBO-MAS Multi-Intelligent Body Autonomous Collaboration Experimental Platform. The experimental platform consists of a robot system, a high-frequency projector system, a wireless communication control system, and a group intelligent decision-making software system, which are jointly composed of four major support systems, and with the help of the relevant technical contents, it can provide an experimental model and a theoretical validation for the realization of the relationship between large-scale activities and individuals in life.
The specific embodiment of group intelligence on ROBO-MAS mainly lies in the group intelligent decision-making software system, which is a software platform system for multi-intelligent body autonomous collaborative simulation as well as physical verification, and can be used for the simulation of multi-robot group behavioral decision-making, robot path planning, robot status display, etc., and the realization of communication management and information interaction among multiple micro mobile robots. Under the unified operation of this system, multiple micro-robots interact and collaborate together to realize the intelligent behavior of the robot as a whole in terms of autonomous perception, autonomous dynamic decision-making and path planning. The collaboration of the micro-robots in information acquisition, target perception, decision-making and planning allows ROBO-MAS to create a virtual robot dynamic target trajectory planning on the simulator, as well as the display of the robot's body state parameters, including position, direction, speed, etc., and to carry out path planning and strategy design of the target location in the process of moving to ensure the precise control of the robot as a whole in the course of its actions. The intelligent response of these micro-robots and their ability to The intelligent response of these miniature robots and their harmonious cooperation in composing different patterns under instructions have the shadow of group intelligence. The application of group intelligence in the field of AI is thus evident.
Four, the development of group intelligence vision
Group intelligence has collaborative decision-making, classification and prediction, automated screening and other outstanding advantages, in the economy and society has a great deal of space for use. Especially in today's Internet of Things (IoT), artificial intelligence (AI), and big data (BD) industries, which are getting more and more attention from all walks of life, the correct and effective utilization of group intelligence can promote the research and development of related products, and provide more convenience for people. In the financial sector, for example, technologies that incorporate group intelligence are expected to play a huge role in making optimal decisions and analyzing various types of data.Group Intelligence originates from people's exploration of natural phenomena, and this process of exploration and understanding has a long history. In the continuous progress of science and technology and means of understanding, Group Intelligence will have more new forms of expression as well as research direction, and the research and development of Group Intelligence has a great scientific research value and strategic value. Looking ahead, we expect that group intelligence can play its advantages in the vision of AI development, realize the breakthrough of key technologies, and become a powerful booster for the development of a new round of artificial intelligence.
References:
1. Exploring the Wonder of Group Intelligence - Multi-Intelligence Body Collaboration - Knowledge
2. Wang Mei, Zhu Yunlong, He Xiaoxian. A review of group intelligence research[J]. Computer Engineering, 2005, No. 22
3. Group Artist in Robotics - ROBO-MAS Robot - Smart Jia Robot
4. An Introduction to Group Intelligence - An Important Direction for the New Generation of AI- Author shang-/s534347795