Handling Of The Uncertainty Of Information Perception And Behavior In Obstacle Avoidance Of Warehousing Robots In Unstructured Environment

Robots in smart factory and logistics warehousing industry are facing multiple uncertainties such as unstructured environment changes,complex machine technology composition,dynamic operating conditions and human factors.It is difficult to model and evaluate perceptual information accurately.At the same time,the control of autonomous behavior is complex,which can not meet the safety requirements of high-intensity and high-reliability operation.In order to ensure the accuracy and reliability of information perception of warehousing robots in unstructured environment,and to improve the working ability and environmental adaptability,it is of great significance to deal with the uncertainty of information perception and behavior in obstacle avoidance of warehousing robots in unstructured environment.The paper takes the warehousing robot with omnidirectional wheel movement as the research object on the basis of clarifying the structure composition and function.Firstly the paper identifies and analyzes the source,interference factors,forms of expression and the transmission process of uncertain information in unstructured environment,meanwhile,considering the influence of different environmental factors on the measurement performance of the sensor to optimize sensor observation model and estimating the location of environmental obstacles.Then the paper fuses the multi-sensor measurement data under different measurement conditions to eliminate the error,redundancy and contradiction of measurement information,so as to reduce the uncertainty of information,and provide more accurate ranging information for obstacle avoidance behavior control.Finally,the paper studies the obstacle avoidance behavior control of warehousing robot in unstructured environment.The main contents of this paper are as follows:(1)From the perspective of external environment and robot system,the paper analyze and summarize the representation and propagation of perceived uncertainty information,and carry out ultrasonic sensor observation modeling under the influence of different environmental factors.Combined with the operating environment and working characteristics of the warehousing robot,the interference factors and propagation processes that affect the performance of the sensor are summarized.Focusing on the random uncertainty of sensor measurement information,we use probability theory to quantify it,and we establish sensor observation models under typical measurement conditions,and the model sampling provides data for the information fusion.Finally,we consider the location estimation of environmental obstacles under measurement uncertainty.(2)Based on the analysis of uncertain information in environmental perception,the paper improves the multi-sensor measurement data fusion and verifies the algorithm in dynamic and static environment.Aiming at the uncertainty and inconsistency of measurement data of multi-sensor system in unstructured environment,the multi-sensor measurement data is preprocessed and fused under the distributed information fusion framework.Aiming at the inconsistency of measurement data,the data preprocessing is performed by the distribution graph method based on robust statistical theory,and the calculation process of the intermediate order statistic of the algorithm is optimized;Based on the optimization of membership function,the improved fuzzy closeness is used to describe the degree of support between the same type of sensors.At the same time,the difference between the measured data before and after Kalman filter is used to optimize the fusion value.Finally,the verification of the improved algorithm is carried out under dynamic and static observation conditions.(3)On the basis of measurement data fusion,the paper consider the distribution characteristics of environmental obstacles,fuzzy logic is used to control the obstacle avoidance of warehousing robots.Aiming at the complexity of unstructured environment modeling and the uncertainty of obstacle avoidance behavior control of warehousing robots,the warehousing robots uses fuzzy control to avoid obstacles,because fuzzy control can handle uncertain information and has strong environmental adaptability.Firstly,the distribution characteristics of environmental obstacles are analyzed,and the kinematics model of the Mecanum wheel chassis is established.Then,based on the classification of environmental obstacles,the design and optimization of fuzzy controller are carried out.Finally,the obstacle avoidance is simulated in different environment.