Improved Coding Landmark-Based Visual Sensor Position Measurement and Planning Strategy for Multiwarehouse Automated Guided Vehicle
Yindong Lian,Wei Xie,Qifan Yang,Yanru Liu,Yibin Yang,Ai-Guo Wu,Toshio Eisaka
DOI: https://doi.org/10.1109/tim.2022.3162293
IF: 5.6
2022-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:The application of vision-guided multiwarehouse automated guided vehicles (AGVs) is becoming more extensive. This article mainly studies the position measurement and corresponding path planning of AGV guided by visual sensors. An improved vehicle-mounted visual sensor model is proposed, which effectively reduces the cost of the sensor and achieves position prediction. The proposed manual coding landmark significantly improves the decoding speed and accuracy while ensuring the information coding capacity. An improved heuristic path planning algorithm combined with the visual sensor system is implemented, considering the planning effectiveness and directionality. Finally, simulations and experiments verify the reliability of the visual guidance method in position measurement and the advancement of the planning strategy. Compared with the traditional 2-D code and bottom visual model, the improved visual sensor position measurement effectively reduces motion blur and corresponding pose deviation. The decoding speed and accuracy are improved by more than 41.21% and 53.85%, respectively. Planning performance comparison is achieved between the same type of heuristic algorithms (A*, D*, and algorithm deployment). The significant improvement in core system indicators includes task efficiency (up to 70.02% improvement), parking time (up to 72% reduction), deadlock (up to 85.05% reduction), and planning time (up to 89.39% reduction).
engineering, electrical & electronic,instruments & instrumentation