Please wait a minute...
Most Download
Most Cited
Total visitors:
Visitors of today:
Now online:

Current Issue

    15 June 2022, Volume 4 Issue 2
    Review Intelligence
    Crypto management: a novel organizational management model based on blockchain
    Juanjuan LI, Ge WANG, Xiao WANG, Junqing LI, Yong YUAN, Fei-Yue WANG
    2022, 4(2):  145-156.  doi:10.11959/j.issn.2096-6652.202232
    Asbtract ( 489 )   HTML ( 67)   PDF (1141KB) ( 349 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Aim to deal with the problem of data, trust and timeliness asymmetry faced by modern organizational management from the root, a novel organizational management model towards Web 3.0 namely crypto management was proposed.It was enabled by blockchain technology and smart contracts based on it, supported by the federated data, organized in the form of DAO (decentralized autonomous organization), and driven by the incentive mechanism with NFT (non-fungible token) as the core.The primary goal of crypto management was to realize trustable, reliable and usable real-time management decision-making under the premise of data security and privacy protection.The framework of crypto management was formulated, its core components and implementation mode were discussed, and its operation process using the example of personnel performance management was also introduced.Towards the end, the potential future works in this emerging new area were discussed.

    Surveys and Prospectives
    Overview of intelligent game:enlightenment of game AI to combat deduction
    Yuxiang SUN, Yihui PENG, Bin LI, Jiawei ZHOU, Xinlei ZHANG, Xianzhong ZHOU
    2022, 4(2):  157-173.  doi:10.11959/j.issn.2096-6652.202209
    Asbtract ( 1378 )   HTML ( 121)   PDF (869KB) ( 2775 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    The field of intelligent game has gradually become one of the hotspots of AI research.A series of research breakthroughs have been made in the field of game AI and intelligent wargame in recent years.However, how to develop game AI and apply it to the actual intelligent combat deduction is still facing great difficulties.The overall progress of research in the field of intelligent games in domestic and overseas were explored, the main attribute requirements of intelligent combat deduction was tracked, and it was summarized with the latest advancements in reinforcement learning.The feasibility of developing game AI into intelligent combat deduction were comprehensively analyzed from three dimensions: mainstream research technology in the field of intelligent game, relevant intelligent decision technology and technical difficulties of combat deduction, and finally, some suggestions for the development of future intelligent combat deductiongives were given.This paper can introduce a clear development status and provide valuable research ideas for researchers in the field of intelligent game.

    Application of intelligent optimization algorithms in supply chain network
    Xin ZHANG, Zhihui ZHAN
    2022, 4(2):  174-183.  doi:10.11959/j.issn.2096-6652.202202
    Asbtract ( 402 )   HTML ( 62)   PDF (1042KB) ( 549 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Supply chain network connects members by the relationship of demand and supply and facilitates the coordination and cooperation among these members, which is particularly important in the global competition environment.The optimization and improvement of supply chain network can reduce the operating cost of enterprises, increase the income of enterprises and the customer satisfaction, and then improve the competitiveness of enterprises.Firstly, the optimization problems of supply chain network were analyzed, and these problems from several different aspects were classified, such as modeling characteristics, decision variable types, and scene features, so as to introduce the optimization problems in the existing research of supply chain network more clearly.Then, three frequently used intelligent optimization algorithms and their applications in supply chain network optimization were introduced and analyzed, such as genetic algorithm, ant colony optimization algorithm and particle swarm optimization algorithm.Finally, the future research of the optimization of supply chain network was prospected.

    Special Topic: Autonomous Agent Learning for Dexterous and Accurate Manipulations
    A survey of visuotactile sensing technologies for robotic manipulation
    Shaowei CUI, Shuo WANG, Jingyi HU, Chaofan ZHANG
    2022, 4(2):  186-199.  doi:10.11959/j.issn.2096-6652.202222
    Asbtract ( 962 )   HTML ( 134)   PDF (13817KB) ( 515 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Thanks to the high spatial resolution and multi-mode tactile sensing, visuotactile sensing technology has been widely applied to various robotic manipulation tasks, such as robotic active perception, pose estimation, and in-hand manipulation.Firstly, the current mainstream visuotactile sensing technologies based on sensing principles were summarized, which could be mainly divided into three categories: GelSight-type visuotactile sensors, binocular (multi-view) visuotactile sensors, and other types.Meanwhile, the sensing methods of different tactile sensing modes were further summarized, including contact surface 3D geometry, force/torque, and sliding.Furthermore, focusing on the field of robot operation, the specific application scenarios of visuotactile sensors were discussed.Finally, future work of the visuotactile sensing technology and how they can be further applied to robotic dexterous manipulation tasks were given.

    Research development on automated robotic peg-in-hole assembly
    De XU, Fangbo QIN
    2022, 4(2):  200-211.  doi:10.11959/j.issn.2096-6652.202223
    Asbtract ( 522 )   HTML ( 45)   PDF (1312KB) ( 757 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Peg-in-hole assembly is a typical operation task in manufactory.The research of peg-in-hole assembly based on industrial robots is valuable for the application of robots in the automated assembly area.For the peg-in-hole components with high precision or complex shapes, the efficient and reliable assembly is still very challenging.The development of automated robotic peg-in-hole assembly of peg-in-hole was reviewed from the view of control.First, the process of robotic peg-in-hole assembly was introduced.Secondly, the assembly control methods based on the traditional models were described.The newly emerged intelligent assembly methods based on learning mechanism were discussed, especially the applications of imitation learning and reinforcement learning in the automated robotic assembly.The combination of the traditional methods and the artificial intelligent methods will provide new energy for the automated robotic assembly, which will be one of the important developing tendencies in future.

    A survey on applications of ontology knowledge representation in robotics
    Yueguang GE, Shaolin ZHANG, Yinghao CAI, Tao LU, Dayong WEN, Haitao WANG, Shuo WANG
    2022, 4(2):  212-222.  doi:10.11959/j.issn.2096-6652.202224
    Asbtract ( 293 )   HTML ( 46)   PDF (1371KB) ( 644 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    The technology about knowledge representation plays an increasingly important role in the autonomous operation of robots facing the complex and unstructured working environment.Knowledge representation focuses on the model of knowledge symbols and how to realize knowledge processing through reasoning procedures automatically.The robot knowledge representation framework and the latest application progress based on ontology representation and reasoning were reviewed.The technical background, realization methods of knowledge representation and reasoning, and recent research progress in the robotics field were summarized from deterministic knowledge and uncertain knowledge.And the future research direction of knowledge-enabled robots was predicted.

    Research on the manipulator intelligent trajectory planning method based on the improved TD3 algorithm
    Qiang ZHANG, Wen WEN, Xiaodong ZHOU, Weihui LIU, Xiaoyu CHU
    2022, 4(2):  223-232.  doi:10.11959/j.issn.2096-6652.202225
    Asbtract ( 375 )   HTML ( 33)   PDF (3862KB) ( 554 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    An intelligent trajectory planning and obstacle avoidance method based on the improved twin delayed deep deterministic policy gradient algorithm (TD3) was proposed to solve the trajectory planning problem for a 4-DOF manipulator mounted on a satellite.The training strategy had 2 periods.In the pre-training stage, the target position was always guided combining with the output of the strategy network to optimize the trajectory.After the pre-training, the algorithm can autonomously output the velocity trajectory while the initial position and the target were specified randomly in the joint space of the manipulator.This target-guided mechanism decreased the unnecessary explorations and improved the learning efficiency in high dimensional action space.In the second training stage, a collision-free safety reference trajectory was firstly obtained by demonstration, and then this trajectory was constantly learned during the training process until the final output trajectory has the ability to avoid obstacles.

    3D edge reconstruction method based on tactile sensing and servo
    Jingyi HU, Shaowei CUI, Chaofan ZHANG, Boyue ZHANG, Shuo WANG
    2022, 4(2):  233-245.  doi:10.11959/j.issn.2096-6652.202226
    Asbtract ( 313 )   HTML ( 34)   PDF (13020KB) ( 163 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    The dexterous manipulation of robots has always been a hot topic in the field of robotics.In many dexterous manipulation tasks, robots need not only visual feedback, but also tactile feedback at the fingertips, especially in poor lighting conditions.Currently, 3D shape sensing of objects using tactile servo is still very challenging.The flat GelSight tactile sensors are not suitable for 3D shape sensing of objects.Therefore, the design of GelSight tactile sensor was improved to make it more suitable for the acquisition of edge information of 3D objects.At the same time, an object edge tracking control strategy based on tactile servo and a 3D object edge reconstruction method based on tactile point cloud were proposed to support the generation of dexterous manipulation strategies for robots.For a variety of complex 3D edge models, the effectiveness of the proposed method was verified by the tactile servo tracking and 3D edge reconstruction experiments.

    Category-level object pose estimation from depth point cloud
    Renwu LI, Lingxiao ZHANG, Lin GAO, Chunpeng LI, Hao JIANG
    2022, 4(2):  246-254.  doi:10.11959/j.issn.2096-6652.202227
    Asbtract ( 390 )   HTML ( 45)   PDF (2469KB) ( 300 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Aiming at the problem of category-level object pose estimation, a method was proposed to accurately estimate the pose of the target object by only taking the point cloud scanned by the depth camera as the input, with knowing the category of input point cloud only.The method did not reply on a huge amount of labeled dataset, but used virtual data produced by simulation instead, which achieved better accuracy on real-world dataset.This method first filtered the background noise of the input point cloud.Then standardized the point cloud through the well-designed center prediction module.After that, the normalized object coordinate space would be estimated through a shape template deformation module.Finally, the pose would be obtained from least squares.Experiments on real-world dataset demonstrates that the method achieve higher accuracy and better generalization ability.

    Papers and Reports
    Automatic path planning program generation system based on swarm intelligence results
    Yuqian WANG, Rong DING
    2022, 4(2):  255-263.  doi:10.11959/j.issn.2096-6652.202228
    Asbtract ( 258 )   HTML ( 29)   PDF (1437KB) ( 488 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Path planning algorithms are widely used in various motion planning tasks, such as robot motion and autonomous driving.So far, many excellent path planning algorithms have been proposed for applications in different fields.For a specific task environment, choosing the appropriate path planning algorithm can plan a better path that satisfies the constraints more efficiently.Based on the results of swarm intelligence, the adaptability and path planning efficiency of rapidly-exploring random tree (RRT) path planning algorithm and its variants RRT-Star path planning algorithm and RRT-Star-Smart path planning algorithm under different task environments were studied.Using genetic programming algorithm as a framework to design a system, which could automatically analyze the map features of the current environment and combine the characteristics of RRT path planning algorithm and its variants to generate new path planning algorithms that were more suitable for the current environment.The generated path planning algorithm can efficiently plan a feasible path from the starting point to the target point.

    Path planning for unmanned surface vehicle in complex dynamic environment based on improved RRT*-Smart
    Lu DONG, Ailing XIONG
    2022, 4(2):  264-276.  doi:10.11959/j.issn.2096-6652.202229
    Asbtract ( 595 )   HTML ( 51)   PDF (2790KB) ( 386 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Aiming at the path planning problem of unmanned surface vehicle (USV) in complex dynamic environment with moving multi-obstacle ships, a path planning method based on improved RRT*-Smart (RTSNew) was designed.Firstly, the sampling mode of nodes was optimized, nodes were sampled in the polar coordinate system with USV as the origin, an elliptical sampling range constraint was adopted to avoid invalid sampling, and a historical path buffer pool was used to make full use of the historical path.The optimization greatly reduced the amount of calculation and improved the speed of path planning.Secondly, the expansion mode of nodes was improved.In order to avoid treating dynamic obstacles as static obstacles, time information was added to each node to realize dynamic collision detection and the full use of dynamic obstacles motion information greatly improves the feasibility of the planned path.At the same time, considering the maneuverability of USV, the angle constraint was added in expansion of nodes to ensure smooth path.Finally, virtual obstacles were designed to mobile obstacle ships to make the planned path comply with International Regulations for Preventing Collisions at Sea (COLREGS).Based on VREP platform, the USV navigation simulation experiments and comparative experiments were carried out.The results show that RTSNew can make USV reach the destination efficiently and safely, and it performs better in planning efficiency, path optimization and path security than traditional methods in complex dynamic environment with multi-obstacle ships.RTSNew ensures that the motion path complies with COLREGS, and solves the problems of traditional methods: treating the dynamic obstacles as static obstacles, ignoring COLREGS, large amount of calculation and low efficiency, not suitable for the complex dynamic environment with moving multi-obstacle ships.

    TD3-based energy management strategy for hybrid energy storage system of electric vehicle
    Jiacheng LIU, Xiangwen ZHANG
    2022, 4(2):  277-287.  doi:10.11959/j.issn.2096-6652.202230
    Asbtract ( 229 )   HTML ( 34)   PDF (1674KB) ( 269 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    Combining batteries and super capacitors into a composite power system (CPS) with an effective energy management strategy can significantly improve the energy utilization, and increase the service life of the energy storage system.To minimize the energy loss of the system, an energy management strategy based on the twin delayed deep deterministic policy gradient (TD3) algorithm was designed.Compared with the deep deterministic policy gradient (DDPG) algorithm, TD3 algorithm solved the problem of overestimation of Q value and less energy loss.A MATLAB/Simulink simulation model based on the TD3 algorithm was built, and tested with the electric vehicle driving equation and the equivalent circuit model of CPS.The outcomes indicate that the proposed energy management strategy can effectively reduce the impact of high current on the battery, and compared with DDPG algorithm, the energy utilization efficiency is improved by 1.36%, the peak of output current of the battery is reduced by 14.68%, the temperature rise of the battery is reduced by 3.52%, the total energy consumption of the system is reduced by 2.17%.

    Distributed agricultural organization based on federated learning
    Mengzhen KANG, Xiujuan WANG, Dong LI, Xuwei WANG, Haoyu WANG, Menghan FAN, Yulin XU, Fei-Yue WANG
    2022, 4(2):  288-297.  doi:10.11959/j.issn.2096-6652.202231
    Asbtract ( 339 )   HTML ( 41)   PDF (8691KB) ( 187 )   Knowledge map   
    Figures and Tables | References | Related Articles | Metrics

    At present, small-scale agriculture is dominating in China.How to develop appropriate smart agriculture for agricultural management with small farmers and small plots is quite challenging.A distributed agricultural AI framework combining federal learning and blockchain technology was proposed, which can achieve the purpose of the training model and establish the incentive mechanism for participants without data sharing.This framework helped make full use of agricultural multi-source heterogeneous data, reducing user data requirements, developing decision-making models according to local conditions, and promoting the connection of production and marketing of small-scale agriculture.