Study Of Autonomous Mental Developmental Robot Based On Cloud Computing

With the development of robot technology, the robot has brought profound influence to human society, and go deep into the various fields of human life. Robot technology is developing fast, especially in the field of intelligent robot, raise a hot wave research. But at present, how to realize the real robot intelligence is still a problem to be solved. Although the traditional method, including method based on knowledge, method based on learning, and the evolution method, has made a series of abundant results in the field of intelligent robot research. But these methods all can’t make robot to complete task like humans in unknown environment. In order to overcome the shortcomings of traditional methods, inspired by new discoveries in neuroscience, in 1996, Rutgers University professor Weng proposed the concept of autonomous mental development robots, and then a series of theories and algorithms for autonomous mental development robots was introduced in detail.Compared with traditional robot, autonomous mental development robot is not one aimed at a specific task, which enables online learning just like humans and deals with unknown tasks. At present, the improvement of autonomous mental development robots task complexity, and computation intensive tasks involved in autonomous mental development algorithm, ask the robot platform to carry more and more computing and storage resources, this not only increases the manufacturing costs and design complexity of robot platform, also restricts the practicability and flexibility. In addition, autonomous mental development robots often need to train many times to master a new task, so how to avoid the repetition when facing the same task is also a problem.In order to solve these problems, this paper apply cloud computing to the study of autonomous mental development robots, build a cloud robot system based on china telecom cloud service platform and Amigo robot. In the system, intensive computation and data storage tasks are transferred to the cloud computing and storage resources, robots only focus on the task execution. In the process of autonomous mental development, the learning knowledge will be stored in the cloud resources, can be used by different robots, so as to realize knowledge sharing. In this paper, we design three experiments based on the robot system platform, and analyze the experimental data, results show that this system and effectively transfer robot intensive computing tasks, and can realize the knowledge sharing between different robot platforms, thus verify the significance of cloud computing in autonomous mental development robot.Firstly, the paper analyzes the research background and significance of autonomous mental development cloud robot, sum up the research and development of autonomous mental development robot and cloud computing at home and abroad, as well as the main problems. Then the main content and chapter framework of this paper are described brief.Secondly, this part present the autonomous mental development IHDR algorithm in detailed, including the basic concept, principle and method. Besides, this section present the cloud computing’s influence on the industrial robot in detail.The third part introduce the component module of autonomous mental development cloud robot system, and the operation mode and operation process of the system. In the system, Amigo robot only focus on information acquisition and task execution, while information processing, storage and autonomous development are completed in the cloud server.The fourth part introduce the design of experiment according to autonomous mental development robot system, the experiment is conducted on Amigo robot platform, and we design a kind of industrial platform to compare. Results show that the robot system effectively transfer the intensive computation tasks, and realize the knowledge sharing.Finally, conclusions are given with recommendation for future work.