The continuous progress of productive forces promotes the progress and innovation of science and technology in order to establish a more reasonable production relationship. Since the industrial revolution, human labor has been gradually replaced by machines. This change has created great wealth for human society and greatly promoted the progress of human society. Today, electromechanical integration, mechanical intelligence and other technologies are constantly emerging, becoming the main theme of the times. Human beings give full play to their subjective initiative, further improve the utilization efficiency of machinery, make it create more and more huge productivity for us, and maintain social harmony to a certain extent. The emergence of industrial robots is a milestone in the history of human social production by using machinery. In developed countries, complete sets of equipment for industrial robot automation production line have become the mainstream and future development direction of automation equipment. Foreign automobile industry, electronic and electrical industry, construction machinery and other industries have used a large number of industrial robot automatic production lines to ensure product quality, improve production efficiency and avoid a large number of industrial accidents. The application practice of industrial robots in many countries in the world for nearly half a century shows that the popularization of industrial robots is an effective means to realize automatic production, improve social production efficiency and promote the development of enterprises and social productive forces. The history of robots is not long. The history of robots really began in 1959 when engelberg and Devall in the United States made the world's first industrial robot. De Waal invented a system in 1946, which can "repeat" the recorded machine actions. 1954, Dewar obtained the patent of programmable manipulator. This kind of manipulator works according to the program and can make different programs according to different work needs, so it has universality and flexibility. Both engelberg and Dewar are studying robots, and think that robots are most suitable for the automobile industry, because it works with heavy machines and the production process is relatively fixed. 1959, engelberg and Dewar cooperated to produce the first industrial robot. 1993 The number of robots installed in China is about 1000, accounting for only 0. 16% of the world, which is negligible, and the proportion of domestic robots is even lower. At present, the total number of robots in China is small, but the domestic robot market is in great demand and is on the rise. China's robot technology has made great progress under the impetus of the National Seventh Five-Year Plan, the Eighth Five-Year Plan and the 863 Plan. The research of intelligent robot has made progress, and remarkable achievements have been made in robot technology model, robot application engineering and robot basic technology research, which has tracked the international advanced robot technology and shortened the gap with the international advanced level. 1993 The number of robots installed in China doubled compared with 199 1 year, with a great relative growth rate. Although some people are still vague about the development of robot technology in China, more and more people have realized that advanced robots (including industrial robots and intelligent robots) are one of the key technologies of automation and an indispensable technology in China's modernization. This high-tech involves flexible manufacturing system (FMS), computer integrated manufacturing system (CIMS), intelligent manufacturing system (IMS), flexible automation (FA) and automated factory (AF). Robot is an interdisciplinary product, which combines advanced theories and technologies in the fields of kinematics and dynamics, mechanical design and manufacturing, computer software and hardware, control and sensors, pattern recognition and artificial intelligence. At the same time, it is a typical automaton, which is the extension and development of special automaton and numerical control machine. At present, social demand and technological progress put forward new requirements for the intelligent development of robots. With the acceleration of social progress, the demand for automation is expanding from manufacturing to engineering, society, life and other fields. In the past, automatic machines or industrial robots that worked in the factory structured environment were suitable for large-scale, inflexible and changeable production environments, and did not require too much intelligence. In order to meet the different needs in different unstructured environments, the automation machines needed in a wide range of fields must have the ability of comprehensive integration and autonomy, and develop into intelligent robots characterized by technical integration. Information technology needs carriers, transforming traditional industries and all walks of life with informationization, and finally realizing the materialization of information with automatic machines, and robots are one of its carriers. On the other hand, the development of information technology, especially the progress of high-performance computers, communication networks and electronic devices, pattern recognition and signal processing, software and other technologies, can promote the "intelligence" and "physique" of robots and create conditions for the intelligent and diversified development of robots. Today, with the rapid development of information technology, the interactive development between robot technology and information technology is more prominent, which makes the high-tech content of robots constantly improve and is always in high-tech research. Because of its infinite imagination, robots have always been the source of conceptual innovation and technological innovation. Whether in space, underwater, disaster relief, service, medical care, entertainment and other fields. You can imagine intelligent robots with corresponding functions according to your needs. This imagination space is endless from low to high. At present, the concept of automation is rapidly expanding to a wide range of fields. The development of information technology has greatly improved the intelligence of robots, which makes it necessary and possible to expand this imagination space, thus stimulating conceptual innovation and technological innovation around robots, which contains the possibility of producing various pre-competitive core technologies, which will surely be an important competitive point of international scientific and technological innovation. Robots are the product of many disciplines. However, with the complexity of robot application environment and tasks, the difficulty and influence of information comprehensive processing and planning and coordination of complex tasks in unstructured and complex environment are increasingly prominent, and the theory, method and technology of information feedback, optimal control and coordination and integration need to be adopted to solve it. The advantages of control discipline in system optimization and comprehensive integration will play a leading role in intelligent robots more and more. As an automation system, intelligent robot is incomparable to any other automation system in the coverage and frontier of theory and technology, the combination with various advanced information technologies, and the diversity of physical realization. Therefore, the representativeness and position of robots in automation science and technology will be further recognized with the expansion of its application scope, the update of information technology and the improvement of intelligence. In the process of intelligent robot development, multi-robot cooperation system is a kind of technology integration platform with coverage. If the intelligence of a single robot only makes a single person smarter, then many robot cooperation systems should not only have a group of smart people, but also require them to cooperate effectively. Therefore, it not only embodies the individual intelligence, but also embodies the collective intelligence, which is an imagination and innovative exploration of human social production activities. Multi-robot cooperation system has a wide application background, which is closely related to the expansion of automation to non-manufacturing fields. With the application environment turning to unstructured, multi-mobile robot system must have a high degree of decision-making intelligence if it can adapt to the change of tasks and the uncertainty of the environment. Therefore, the research on multi-mobile robot cooperation is not only the coordination of control, but also the coordination and cooperation of the whole system. Here, the organization and control mode of multi-robot system largely determines the effectiveness of the system. Multi-robot cooperation system is also a mode to realize distributed artificial intelligence. The core of distributed artificial intelligence is to divide the whole system into several intelligent and autonomous subsystems, which are physically and geographically dispersed and can perform tasks independently. At the same time, they can exchange information and coordinate with each other through communication, so as to complete the whole task together. This is undoubtedly attractive for completing large-scale complex tasks, so it has quickly gained widespread attention in military, communication and other application fields. Multi-machine cooperation system is the concrete realization of this idea, in which each robot can be regarded as an autonomous agent. Mars (Multi-Agent Robotics Systems) has become a new research hotspot in robotics. Multi-mobile robot system can complete complex tasks in unstructured environment because of its mobile function. It is the most typical and widely studied system in multi-robot cooperation system. Architecture is the logical and physical information relationship and control relationship between robots in the system, as well as the distribution mode of problem solving ability. It is the basis of cooperative behavior of multi-mobile robots. Generally speaking, the architecture of multi-mobile robot cooperation system can be divided into centralized and distributed. The centralized architecture can be planned by a single master robot, which has all the information about the system activities. There is no such robot in the distributed architecture, and all robots are controlled equally in the distributed architecture. Although centralized architecture can achieve the global optimal solution, people actually prefer distributed architecture because of the influence of uncertainty. In recent years, in the distributed architecture, in order to overcome the difficulty of environment modeling in the actual environment and improve the robustness and working ability of the multi-mobile robot cooperation system, some scholars have adopted the reactive control system based on behavior, and some scholars have adopted the reactive control system based on behavior, which establishes the cooperative behavior in a reactive mode, accelerates the reaction of mobile robots to the outside world and avoids complicated reasoning, thus improving the real-time performance of the system. Perception is the basis of intelligent robot action, including the fusion and utilization of "feeling" (perception) and "knowledge and understanding" information. The most important perception problem in mobile robot is positioning and environmental modeling [7]o Although there are many positioning methods, such as odometer calculation, vision-based landmark recognition, global positioning based on map matching, gyro navigation, GPS, etc., only GPS can achieve practical global positioning in the unknown unstructured environment. But GPS is limited by accuracy, safety and other factors. How to improve the ability of positioning and environment modeling through cooperation between robots is an important content of studying the system intelligence of multi-mobile robots. In recent years, people have put forward a variety of synchronous processing methods for the establishment and positioning of environmental maps [8], in which the process of environmental modeling and positioning is accompanied by each other and gradually becomes clear in the process of mutual iteration, but it often requires harsh environmental conditions. In addition, in many cooperative tasks, only the relative position information between partners, such as formation and local collision avoidance, is needed, so sensor-based local positioning is also concerned. Robots detect each other through sensors such as ultrasonic, infrared, laser or vision, and then fuse information through statistics, filtering and other algorithms, so as to obtain the relative position of each robot in the system. China's research work in this field has begun. With the support of the 863 Program and the Natural Science Foundation, after years of continuous research, a number of domestic units have reached a higher level of research in local areas, and the experimental research situation has also improved significantly, but they have also encountered many difficulties, especially in the fields of complex system control and distributed intelligence, which are obviously insufficient and lack of strong theoretical and technical support. Moreover, most technologies have strict requirements on the environment, which limits the development of multi-mobile robot systems and the transformation to practical systems. Facing the unstructured and dynamic characteristics of the real world, highly adaptable and flexible cooperative theories, methods and technologies will be the focus of future research.