Search

Picking robots have been widely used for citrus harvesting in recent years. However, the rest citrus during continuous picking can be irregularly disturbed by the wind, robot force, and the load weight of bearing branches under the natural environment. The citrus in the disturbed state cannot be rapidly and accurately detected, and then localized online, leading to the low efficiency of automatic robotic picking. In this study, online target detection and rapid localization were proposed using improved YOLOv5s+DeepSORT. The position of citrus at rest was predicted using the motion-tracking trajectory of disturbed citrus within a short period of time. The coordinates of the citrus were then obtained rapidly. Firstly, the CBAM (Convolutional Block Attention Module) attention mechanism was added to the YOLOv5s network, in order to detect the small and occluded targets. The SIoU loss function was used to enhance the direction matching between the prediction and the calibration frame, in order to improve the convergence speed of regression. Secondly, the target re-identification network was improved in the DeepSORT more suitable for the feature extraction of citrus targets. The feature extraction of the network was enhanced to improve the tracking performance on the disturbed citrus; The Count counter was used to accumulate the number of tracking frames in each citrus for an optimal target. Since the disturbance of the rest citrus was progressively propagated over time, the localization prediction and picking were only for targets with optimal tracking trajectories at a time. The real-time updating was realized in real time. Finally, the values of the depth camera were combined within the critical distance range, excluding the influence of background citrus on the detection speed. The number of tracking targets each time was limited to effectively improve the tracking speed of disturbed citrus. The experimental results show that the P (precision) and mAP (average detection accuracy) of improved YOLOv5s were improved by 3.9 and 1.1 percentage points, respectively, with a detection rate of 69.3 frames per second. The MOTA (Multi-Object Tracking Accuracy) and MOTP (Multi-Object Tracking Precision) of the improved DeepSORT were improved by 9.2 and 5.4 percentage points, respectively, whereas, the average number of ID (identity) switching times of targets was reduced by 32 times. Grasping experiments were conducted in the laboratory, in which the citrus was randomly swung along different orientations with an amplitude of about 10 cm. When the predicted localization time was 1, 2, 3, 5, 7, and 10 s, the average precision values of disturbed citrus localization were 21.3%, 53.0%, 81.9%, 83.7%, 86.1%, and 94.9%, respectively. The citrus picking test was conducted with the citrus localization time of 3 s. The average grabbing time for each citrus was 12.8 s, which was 5.6 s shorter than that without the optimization. The efficiency was improved by 30.4%. This finding can provide technical support and references for citrus picking in disturbed states. [ABSTRACT FROM AUTHOR]

BACKGROUND: Traditional rehabilitation training has limited potential to improve functional recovery from chronic injuries. Increasing attentions have been paid to the application of rehabilitation robots in rehabilitation training of chronic injuries. OBJECTIVE: To analyze and track the hot spots and frontiers of related research in the field of rehabilitation robot in the past 10 years, thereby providing some guidance for future research. METHODS: The literatures regarding rehabilitation robots published from 2010 to 2020 were retrieved in the Web of Science core database by computer. Search strategy was TS=(rehabilitation robot OR rehabilitation robotics). CiteSpace 5.7 visualization software was used to analyze the hot spots and frontiers of rehabilitation robot research from the aspects of high-influence countries/regions, institutions, authors, high-frequency keywords and burst words. RESULTS AND CONCLUSION: A total of 3 194 literatures were included. In recent years, the research on rehabilitation robots has been growing steadily. The number of relevant literatures published annually shows a steady increase trend. The number of articles published by the United States ranks first, followed by China. Northwestern University in the United States ranks first with 161 articles published. Professor Riener R, from the University of Zurich, Switzerland, has published 48 articles, ranking the first among the authors. Nine representative keyword clusters have formed, and the research emphasis and direction change over time. In recent 10 years, the relevant research focuses on intelligent control of rehabilitation robots, task analysis and learning, performance and reliability, and how to realize the natural and accurate interaction between robot and human through the integration of machine intelligence and biological intelligence. At present, rehabilitation robots are the most widely used in the field of upper limb and nerve injuries, and the interest in brain-machine interface, virtual reality, flexible wearable, task analysis and the exoskeletons has increased in the research of rehabilitation robot, which are the frontiers and hot issues in recent years, representing the development trend of future studies, and referring to the future research direction. Future research in the field of rehabilitation robots should focus on the following aspects: First, focus on the functional needs of patients to improve the quality of life. Secondly, strengthen multidisciplinary cooperation and build a cooperation network among regional countries. [ABSTRACT FROM AUTHOR]

Aiming at the problem that the local path planning algorithm is only applicable to single-vehicle robots, this paper proposed a dynamic window algorithm for trailer-type mobile robots. Firstly, it used the equation of multi-vehicle structure to realize the motion control of the trailer mobile robot. Then, it used the evaluation function to evaluate both the tractor and the trailer and add them according to the weights. Finally, for the structural characteristics of the trailer, it added the angle constraint between the tractor and the trailer during the movement to ensure the stability of the motion track. The simulation experiments show that the motion control of the trailer-type mobile robot can satisfy the convergence. At the same time, the proposed algorithm realized the task of the local path planning of the trailer-type mobile robot, and the variation of the angle was beyond the limit during the motion. This research has great reference value for the autonomous navigation of trailer-type mobile robots. [ABSTRACT FROM AUTHOR]

The reliability method was introduced into the prediction of modal coupling chatter in robotic milling. Firstly， the reliability model was established and the limit state equation was obtained. Secondly， the fourth-order moment method was used to calculate the reliability and sensitivity to analyze the influence of various parameters on the milling stability. Finally， an example was given based on the engineering background of robotic milling bone. The results showed that the two-degree-freedom stiffness， cutting force coefficient and feeding direction of the robot will affect the stability of the milling process. Within a certain range， increasing the stiffness in X divection，reducing the stiffness in Y direction， reducing the cutting force coefticient and choosing the feed divection resonably are helpful to avoiding modal coupling chatter. The system is more sensitive to the change of cutting force coefficients than to the stiffness in X and Y directions， and has more significant benefit influence on stability. [ABSTRACT FROM AUTHOR]