Research On AGV Guided System Based On Vision Identifier Localization

With the rapid development of social science and technology, traditional warehouse lo-gistics system is facing the pressure of modern upgrade. AGV(Automated Guided Vehicle) is a kind of highly efficient, stable, secure automation transportation means, which plays an important role in the process of replacing traditional human-powered transportation. This di-rectly affects the enterprise logistics cost and core competitiveness, and has very great practical guiding significance. According to the requirements of modern warehouse logistics system, this paper designs a set of automatic guided and automatic obstacle avoidance AGV navigation system based on vision identifier.First, the overall AGV design scheme is put forward. The modules of AGV navigation system and the function of each module is detailedly illustrated from three levels which are collecting layer, calculation layer and executive layer.Next, the positioning technology is an important part of AGV navigation system. Based on digital image processing technique, this paper put forward a kind of AGV navigation identi-fier recognition algorithm, which contains the design of two-dimensional navigation identifier, visual sensor calibration, exposure and white balance automatic adjustment, morphological pro-cessing, identifier content recognition. The pose of AGV is identified through the above process.Then, this paper gives the control strategy by setting the virtual target, which converts two deviation variables(position and angle) to one variable. The control algorithm is realized using single PID control system. The test results show the great performance of this control strategy. In order to improve the performance of control algorithm, fuzzy control theory is introduced into the system. The design process of fuzzy controller is demonstrated from fuzzy language variables, membership functions and fuzzy rules. The simulation results and application results indicate that the control effect of fuzzy PID controller is much better than traditional PID con-troller. Kalman filter is used to improve the accuracy of positioning and the stability of AGV system.Finally, K/N information fusion is applied to fuse the multi-cycle detection results of in-frared sensor to improve the accuracy of the infrared sensor. According to the sensor detection results, the obstacles are abstracted into different digital models, and each model has corre-sponding obstacle avoidance strategy.