Accurate location of AGV based on subjective Bayesian multi-sensor data fusion

Aiming at the problems of large deviation and low accuracy of indoor Automated Guided Vehicle (AGV) positioning, a sensor data fusion localization method based on subjective Bayesian network is proposed. The method first combines the Karman filtering model to filter the data from different sensors and calculates the information gain magnitude by the subjective Bayesian network model to select the sensor data autonomously, then fuses the selected sensor data and finally updates the AGV position status based on the fused sensor information to obtain more accurate AGV position information. The simulation experiments show that in the laboratory experimental environment, the subjective Bayesian network fusion algorithm can achieve data complementarity of multiple sensors, and compared with the RUKF algorithm, the root mean square error of the method is reduced to 0.17 m, the positioning accuracy is improved by 43.6%, the positioning time is shortened by 0.071 s, the efficiency is improved by 4.5%, the data stability is improved by 47.8%. The positioning deviation is significantly smaller than that of single sensor, proving the effectiveness of the method.