Agv Steering Of Vision-based Model And Control

With the developments of Factory Automation, and Computer Integrated Manufacture System, and the extensive application of the Flexible Manufacture System, and the Automatic Stereoscopic Warehouse, the field of the application and the level of the technology for Automatic Guided Vehicle (AGV), which link and adjust the discrete logistics system, and become the necessary tools for the automation transportations, loads and unloads, have been greatly enlarged and improved. Path tracking control is the key technology in the researching fields of the AGV, and it is very important to design an AGV steering control system, which has the following characteristics:smaller tracking error, quickly dynamic response, the adaptability to complicated circumstances, the robustness, and so on.First of all, based on the literatures of the Chinese and international related to the research contents in detail, this paper summarizes the developmental history and researching conditions of AGV and its steering control.Then, through understanding of the basic structure and operating principles of the Vision-based Automatic Guided Vehicle, the model has been built by human-simulated intelligent control based on Sensorimotor Schema. It mainly included the models of the straight line, the right angle, the S line and the evading bonds.Thirdly, based on analyzing the steering control strategy of AGV, designed the multimoding intelligent controller designed based on the principle of schema. Through the result of characteristic identification the corresponding control algorithm is chosen, including the human-simulated intelligent control (HSIC), the fuzzy control and so on.Fourthly, this paper has studied the combining fuzzy-control with neural network control according to their characteristics and application situations for each kind of intelligent method. That is the comprehensive superiority of neural network's parallel computing, the distributional information storage, the fault-tolerant ability strong as well as the fuzzy system suits expresses these fuzzy or the qualitative knowledge, the inference way is quite similar to person's ideological mode.Fifthly, the traditional control method has complete system of theory, but its system's analysis and design need the accurate mathematical model. The present intelligent control methods need the few models, but it still lacks systematic theory, and it mostly depends upon the experience and tries on to designers.It is obviously well to combine the respective superiority.Finally, the results of simulation and the experiment indicated that the steering model and the multimoding control algorithm can make the system track the guidance path smoothly faster, the control performances is obviously more advantageous than the ones of traditional controller, and the software realizations is convenient, showing that the V-AGV system will of the market application prospects widely.