陈慧岩 , 张玉 . 军用地面无人机动平台技术发展综述 [J ] . 兵工学报 , 2014 , 35 ( 10 ): 1696 - 1706 . DOI: 10.3969/j.issn.1000-1093.2014.10.026 http://doi.org/10.3969/j.issn.1000-1093.2014.10.026 地面无人机动平台对发展高机动地面无人战斗系统具有重要的战略意义，是当前各国国防科技领域的研究热点。文中综述了军用地面无人机动平台的发展历程与最新进展，分别阐述和分析了其基本组成和发展特点，然后从环境感知、运动规划、跟踪控制等方面总结了军用地面无人机动平台发展中的关键技术，并对军用无人机动平台的研究方向和研究重点进行了展望。

CHEN H Y , ZHANG Y . An overview of research on military unmanned ground vehicles [J ] . Acta Armamentarii , 2014 , 35 ( 10 ): 1696 - 1706 . (in Chinese)

CLAUSSMANN L , REVILLOUD M , GRUYER D , et al . A review of motion planning for highway autonomous driving [J ] . IEEE Transactions on Intelligent Transportation Systems , 2019 , 21 ( 5 ): 1826 - 1848 . DOI: 10.1109/TITS.6979 http://doi.org/10.1109/TITS.6979 https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6979 https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6979

HELMICK D , ANGELOVA A , MATTHIES L . Terrain adaptive navigation for planetary rovers [J ] . Journal of Field Robot , 2009 , 26 ( 4 ): 391 - 410 . DOI: 10.1002/rob.v26:4 http://doi.org/10.1002/rob.v26:4 https://onlinelibrary.wiley.com/toc/15564967/26/4 https://onlinelibrary.wiley.com/toc/15564967/26/4

MENG X R , CAO Z Q , LIANG S , et al . A terrain description method for traversability analysis based on elevation grid map [J ] . International Journal of Advanced Robotic Systems , 2018 , 15 ( 1 ): 1 - 12 .

SINGH S , SIMMONS R , SMITH T , et al . Recent progress in local and global traversability for planetary rovers [C ] // Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings(Cat. No. 00CH37065).Washington,D.C.,US:IEEE , 2000 , 2 : 1194 - 1200 .

THORESEN M , NIELSEN N H , MATHIASSEN K , et al . Path planning for UGVs based on traversability hybrid A [J ] . IEEE Robotics and Automation Letters , 2021 , 6 ( 2 ): 1216 - 1223 . DOI: 10.1109/LSP.2016. http://doi.org/10.1109/LSP.2016. https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=7083369 https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=7083369

KREBS A , PRADALIER C , SIEGWART R . Comparison of boosting based terrain classification using proprioceptive and exteroceptive data [C ] // Proceedings of the Experimental Robotics: The Eleventh International Symposium . Berlin,Germany : Springer , 2009 : 93 - 102 .

SURBLYS V , ŽURAULIS V , SOKOLOVSKIJ E . Estimation of road roughness from data of on-vehicle mounted sensors [J ] . Eksploatacja i Niezawodność , 2017 , 19 ( 3 ): 369 - 374 . DOI: 10.17531/ein http://doi.org/10.17531/ein http://www.ein.org.pl/ http://www.ein.org.pl/

OLIVEIRA F G , SANTOS E R S , NETO A A , et al . Speed-invariant terrain roughness classification and control based on inertial sensors [C ] // Proceedings of the 2017 Latin American Robotics Symposium and 2017 Brazilian Symposium on Robotics. Washington,D.C.,US:IEEE , 2017 : 1 - 6 .

TCHAPMI L , CHOY C , ARMENI I , et al . SEGCloud: Semantic segmentation of 3D point clouds [C ] // Proceedings of the 2017 International Conference on 3D Vision . Washington,D.C., US : IEEE , 2017 : 537 - 547 .

DOUILLARD B , UNDERWOOD J , KUNTZ N , et al . On the segmentation of 3D LiDAR point clouds [C ] // Proceedings of the 2011 IEEE International Conference on Robotics and Automation.Washington,D.C., US:IEEE , 2011 : 2798 - 2805 .

袁夏 , 赵春霞 . 一种激光雷达可通行区域提取算法 [J ] . 兵工学报 , 2010 , 31 ( 12 ): 1702 - 1707 . 基于模糊模型和时空关联分析，提出一种在激光雷达获取的点云数据中寻找可通行区域的方基于模糊模型和时空关联分析，提出一种在激光雷达获取的点云数据中寻找可通行区域的方法。首先使用基于最大熵的模糊预测模型将激光雷达获得的单条扫描线数据聚类分段；然后寻找符合可通行特征的分段作为初始可通行区域；最后利用时空关联模型在多条扫描线数据中进一步优化可通行区域提取结果。算法通过利用不同方向三维到二维的映射避免了三维空间中计算量大的缺陷，同时又较好地保留了三维数据包含的信息，不涉及迭代计算和在线机器学习过程。实验表明，在城市环境中，算法能够较好的提取可通行区域。

YUAN X , ZHAO C X . Traversable area extraction using LADAR sensor [J ] . Acta Armamentarii , 2010 , 31 ( 12 ): 1702 - 1707 . (in Chinese) An algorithm to detect traversable area from ladar point cloud based on

SHARMA S , BALL J E , TANG B , et al . Semantic segmentation with transfer learning for off-road autonomous driving [J ] . Sensors , 2019 , 19 ( 11 ): 2577 . DOI: 10.3390/s19112577 http://doi.org/10.3390/s19112577 https://www.mdpi.com/1424-8220/19/11/2577 https://www.mdpi.com/1424-8220/19/11/2577 Since the state-of-the-art deep learning algorithms demand a large training dataset, which is often unavailable in some domains, the transfer of knowledge from one domain to another has been a trending technique in the computer vision field. However, this method may not be a straight-forward task considering several issues such as original network size or large differences between the source and target domain. In this paper, we perform transfer learning for semantic segmentation of off-road driving environments using a pre-trained segmentation network called DeconvNet. We explore and verify two important aspects regarding transfer learning. First, since the original network size was very large and did not perform well for our application, we proposed a smaller network, which we call the light-weight network. This light-weight network is half the size to the original DeconvNet architecture. We transferred the knowledge from the pre-trained DeconvNet to our light-weight network and fine-tuned it. Second, we used synthetic datasets as the intermediate domain before training with the real-world off-road driving data. Fine-tuning the model trained with the synthetic dataset that simulates the off-road driving environment provides more accurate results for the segmentation of real-world off-road driving environments than transfer learning without using a synthetic dataset does, as long as the synthetic dataset is generated considering real-world variations. We also explore the issue whereby the use of a too simple and/or too random synthetic dataset results in negative transfer. We consider the Freiburg Forest dataset as a real-world off-road driving dataset.

黄庭鸿 , 聂卓赟 , 王庆国 , 等 . 基于区块自适应特征融合的图像实时语义分割 [J ] . 自动化学报 , 2021 , 47 ( 5 ): 1137 - 1148 .

HUANG T H , NIE Z Y , WANG Q G , et al . Real-time image semantic segmentation based on block adaptive feature fusion [J ] . Acta Automatica Sinica , 2021 , 47 ( 5 ): 1137 - 1148 . (in Chinese)

刘忠泽 , 陈慧岩 , 崔星 , 等 . 无人平台越野环境下同步定位与地图创建 [J ] . 兵工学报 , 2019 , 40 ( 12 ): 2399 - 2406 . DOI: 10.3969/j.issn.1000-1093.2019.12.002 http://doi.org/10.3969/j.issn.1000-1093.2019.12.002 为了满足无人平台在越野环境下碰撞检测以及定位的需求、解决三维点云地图处理计算资源耗费大的问题，设计一种适用于越野环境的同步定位与地图创建方案。提出一种拓扑层次地图，将整个地图划分为多个具有层级结构的三维体素子地图，并以概率方法表示每个体素的状态，进而提取可通行区域。基于三维激光雷达和惯性测量单元，提出一种基于分支定界法的旋转直方图最近邻匹配实时闭环检测方法，并使用Ceres优化稀疏位姿图实现6自由度全局位姿实时优化。越野环境下的实验结果显示：该地图的创建以及可通行区域的提取效果良好，全局定位误差在1 m以内，姿态基本与参考高精度惯性导航系统保持一致；基于该方案提取的可通行区域以及位姿优化结果可满足无人平台实时运动的需求。

LIU Z Z , CHEN H Y , CUI X , et al . Real-time LiDAR SLAM in off-road environment for UGV [J ] . Acta Armamentarii , 2019 , 40 ( 12 ): 2399 - 2406 . (in Chinese) DOI: 10.3969/j.issn.1000-1093.2019.12.002 http://doi.org/10.3969/j.issn.1000-1093.2019.12.002 Simultaneous localization and mapping (SLAM) plays a more and more important role in the environment perception system of unmanned ground vehicle (UGV). A hierarchical global topological map is proposed to solve the low real time capability problem during processing of 3D lidar point cloud map, which divides the entire map into many submaps consisting of large numbers of tree-structure-based voxels, and the submaps are organized with the help of topology. The probabilistic methods are used to represent the state of each voxel being occupied/null. A transitable area is extracted based on the submaps for the UGV's path planning. With the help of an inertial measurement unit (IMU) and two 3D lidars, a branch and bound search (BBS)-based loop detection algorithm called RHM-ICP algorithm is proposed to realize a real time 6-degrees-of-freedom global pose optimization with the help of Ceres, which is to process the sparse pose adjustment (SPA). Experimental results obtained from a real large-scale off-road environment shows an effective reduction of lidar odometry pose accumulative error with a global location error of less than 1 m and a good performance of 3D mapping. Key

彭晓燕 , 谢浩 , 黄晶 . 无人驾驶汽车局部路径规划算法研究 [J ] . 汽车工程 , 2020 , 42 ( 1 ): 1 - 10 .

PENG X Y , XIE H , HUANG J . Research on local path planning algorithm for unmanned vehicles [J ] . Automotive Engineering , 2020 , 42 ( 1 ): 1 - 10 . (in Chinese)

周孝添 , 任宏斌 , 苏波 , 等 . 基于微分平坦的分层轨迹规划算法 [J ] . 兵工学报 , 2023 , 44 ( 2 ): 394 - 405 . DOI: 10.12382/bgxb.2021.0756 http://doi.org/10.12382/bgxb.2021.0756 为充分考虑横纵向耦合和汽车运动学特性对轨迹规划的影响,提出一种分层优化的轨迹规划算法框架。利用贝塞尔曲线的凸包性设计安全走廊约束,以轨迹平滑性为目标函数得到一个基于贝塞尔曲线节点的下层规划器。在上层规划器中,基于下层规划器求解得到的横纵向贝塞尔曲线和车辆运动学模型的微分平坦输出进行三维耦合,构建满足车辆乘坐舒适性、高效性和安全性的目标函数,利用粒子群优化算法对贝塞尔轨迹初始参数进行二次优化得到综合性能最优的行驶轨迹。仿真结果表明:新算法在保证安全性的同时,具有良好的乘坐舒适性和可跟踪性;由于二次规划与粒子群优化算法的求解效率高,此框架实时性强,具有概率完备性。

ZHOU X T , REN H B , SU B , et al . Hierarchical trajectory planning algorithm based on differential flatness [J ] . Acta Armamentarii , 2023 , 44 ( 2 ): 394 - 405 . (in Chinese) DOI: 10.12382/bgxb.2021.0756 http://doi.org/10.12382/bgxb.2021.0756 To fully consider the influence of transverse and longitudinal coupling and vehicle kinematics on trajectory planning, a hierarchical optimization-based trajectory planning algorithm framework is proposed. The safe corridor constraint is designed with the convex hull of a Bezier curve. Taking the trajectory smoothness as the objective function, we obtain a lower planner based on Bezier curve nodes. In the upper planner, based on the transverse and longitudinal Bezier curves solved by the lower planner and the differentially flat output of the vehicle kinematics model, the objective function meeting the vehicle ride comfort, efficiency and safety requirements is constructed, and quadratic optimization is applied to the initial parameters of the Bezier trajectory by particle swarm optimization algorithm to obtain the driving trajectory with the best comprehensive performance. The simulation results show that: the algorithm has good ride comfort and traceability while ensuring safety; due to the high efficiency of quadratic programming and particle swarm optimization, this framework has strong real-time and probabilistic completeness.

邓海鹏 , 麻斌 , 赵海光 , 等 . 自主驾驶车辆紧急避障的路径规划与轨迹跟踪控制 [J ] . 兵工学报 , 2020 , 41 ( 3 ): 585 - 594 . DOI: 10.3969/j.issn.1000-1093.2020.03.020 http://doi.org/10.3969/j.issn.1000-1093.2020.03.020 为减少道路突发事故，提高车辆通行效率，需要研究车辆的紧急避障以实现自主驾驶。基于车辆点质量模型，设计了非线性模型预测控制（MPC）路径规划器；基于车辆动力学模型，设计了线性时变MPC轨迹跟踪器。在路径规划层引入避障功能函数，通过车辆与障碍物的距离调节函数值大小，综合避障函数权重和路径偏差权重，规划出一条既能避开障碍物又使路径偏差最小的临时轨迹。在轨迹跟踪层，利用该临时轨迹和航向角偏差作为车辆主动转向控制参考量，将线性时变MPC优化问题转化为二次规划问题，计算满足车辆动力学约束的前轮转向角最优解。结果表明：所设计的双层MPC紧急避障控制策略对低速(60 km/h)、中速(80 km/h)、高速(100 km/h)行驶车辆有很强的适应性，高速行驶时最大质心侧偏角不超过1.0°，最大航向角偏差不超过2.5°，车辆横向稳定性良好，随着车速增大，车辆避障响应时刻提前；在多车连续避障场景中，自主驾驶车辆的质心侧偏角和航向角偏差均能控制在较小范围内，在多目标连续避障的路径规划和轨迹跟踪问题上同样具有很好的控制效果。

DENG H P , MA B , ZHAO H G , et al . Path planning and tracking control of autonomous vehicle for obstacle avoidance [J ] . Acta Armamentarii , 2020 , 41 ( 3 ): 585 - 594 . (in Chinese) DOI: 10.3969/j.issn.1000-1093.2020.03.020 http://doi.org/10.3969/j.issn.1000-1093.2020.03.020 Emergency obstacle avoidance is one of the key points for autonomous driving system. A path planning controller based on non-linear model predictive control and a path tracking controller based on linear time-varying model predictive control are designed.In path planning controller，an obstacle-avoiding function is used to adjust the distance between the intelligent vehicle and obstacles by calculating the value of obstacle-avoiding function.A new route is supposed to be planned，which can not only keep away from obstacles but also decrease the deviations from the global course，by considering the different weights of obstacle-avoiding function and the planning error. In path tracking controller，the solution of linear time-varying model predictive control is converted into a positive-definite quadratic program.And then the desired steering angles of front wheel are calculated by using the reference trajectory and orientation angle of vehicle center. The results show that the obstacle-avoiding controller has strong robustness and the path tracking controller has better performance in controlling accuracy and vehicle dynamics stability when a vehicle travels at 60 km/h，80 km/h，or 100 km/h. The maximal side-slip angle is no more than 1°，and the maximal deviation of yaw angle is less than 2.5°.Obstacle avoiding action occurs in advance with the increase in velocity of autonomous vehicle. In addition，the dual-MPC controller is adaptive to the circumstance of multi-obstacles avoidance as well，which can be easily demonstrated by the small deviation of yaw angles and small side-slip angles.Key

LI X H , SUN Z P , CAO D P , et al . Development of a new integrated local trajectory planning and tracking control framework for autonomous ground vehicles [J ] . Mechanical Systems and Signal Processing , 2017 , 87 : 118 - 137 . DOI: 10.1016/j.ymssp.2015.10.021 http://doi.org/10.1016/j.ymssp.2015.10.021 https://linkinghub.elsevier.com/retrieve/pii/S0888327015004811 https://linkinghub.elsevier.com/retrieve/pii/S0888327015004811

YANG T , XIONG G M , ZHANG Y , et al . A practical planning framework and its implementation for autonomous navigation in off-road environment [C ] // Proceedings of the 2019 IEEE Intelligent Vehicles Symposium.Washington,D.C.,US:IEEE , 2019 : 1571 - 1576 .

刘龙龙 , 陈慧岩 , 刘海鸥 , 等 . 基于拓扑路网的多挡无人履带平台路径重规划 [J ] . 兵工学报 , 2023 , 44 ( 1 ): 279 - 289 . DOI: 10.12382/bgxb.2021.0827 http://doi.org/10.12382/bgxb.2021.0827 针对越野环境中无人履带混合动力平台重规划鲁棒性差、场景适应性弱的问题，提出一种通过 融入多目标函数模型进行规划路径重构，采用三次螺旋线对重构路径进行多阶段采样优化来生成重规划路径的方法。该方法主要研究无人履带平台运动过程中行驶路径出现局部动态无解、全局规划被动触发重规划功能、重新生成可达目标点的参考轨迹的策略，重点解决重规划时回退道路行驶的成本最优问题。平台搭载两挡行星自动机械变速器，无人平台在行进过程中可根据行驶速度切换到不同挡位，以应对不同路面条件下的速度需求，保证平台具有较强的可通过性。通过实车平台验证新方法针对不同越野场景，根据通行时间、能量消耗和换挡次数的三参数目标代价函数模型，得到最佳的重规划策略并生成一条最优可通行路径。研究结果表明：基于拓扑路网的多挡无人履带平台路径重规划方法降低了重规划过程的时间和能量成本，在兼备考虑时间、能量以及不同转弯半径和速度时选择适当的挡位，确保平台纵向运动动力性的同时，也充分考虑了时效性和经济性。

LIU L L , CHEN H Y , LIU H O , et al . Path replanning of multi-speed unmanned tracked platforms based on topological road network [J ] . Acta Armamentarii , 2023 , 44 ( 1 ): 279 - 289 . (in Chinese) DOI: 10.12382/bgxb.2021.0827 http://doi.org/10.12382/bgxb.2021.0827 To solve the problem of poor robustness and weak scene adaptability of replanning for unmanned tracked hybrid platforms in the off-road environment, a method that incorporates a multi-objective function model and uses a three-order spiral of multi-stage sampling optimization to reconstruct the planned path is proposed. This method regenerates a reference trajectory that can reach the target when no local path is found in dynamic driving. It focuses on solving the cost optimization problem during the back-off road driving of a unmanned tracked platform. The platform is also equipped with a two-speed planetary automatic mechanical transmission. The unmanned platform can switch to different gears according to the driving speed during the traveling process to meet the speed requirements under different road conditions, and thus it has high adaptability. The real vehicle platform verifies that the method proposed in this study can be used for different off-road scenarios. According to the objective cost function model of the three parameters of travel time, energy consumption, and number of shifts, the best replanning strategy can be obtained, as well as an optimal and passable route. The results show that the path replanning method reduces the time and energy costs of the planning process. The appropriate gear is selected while both time and energy, different turning radiuses and speeds are considered. Thus, the longitudinal movement performance, effectiveness, and economy of the platform are guaranteed.

王博洋 , 关海杰 , 龚建伟 , 等 . 面向异构履带车辆的统一运动规划方法 [J ] . 兵工学报 , 2022 , 43 ( 2 ): 241 - 251 .

WANG B Y , GUAN H J , GONG J W , et al . Unified motion planning method for heterogeneous tracked vehicles [J ] . Acta Armamentarii , 2022 , 43 ( 2 ): 241 - 251 . (in Chinese) DOI: 10.3969/j.issn.1000-1093.2022.02.001 http://doi.org/10.3969/j.issn.1000-1093.2022.02.001 A motion planning method based on the offline generation and online extension and selection of motion primitives is proposed for planning the motion of heterogeneous tracked vehicles under a unified algorithm framework. The heterogeneous tracked vehicle studied refers to a tracked vehicle equipped with three different steering mechanisms. The optimized generation of motion primitives takes trajectory smoothness as the goal. A motion primitive library that meets the motion planning requirements of heterogeneous tracked vehicles is generated by comprehensively considering the differentiated behavior constraints of heterogeneous tracked vehicles, platform kinematics model constraints, smooth transition constraints, etc. The extension and selection of motion primitives are based on a unified algorithm framework, which comprehensively considers the attributes of the motion primitives, reference lines, expected velocity, obstacles, and other factors. By selecting the appropriate primitives from the motion primitive library to form the primitive sequences, the coordinated motion at the level of trajectory and velocity is planned. The results show that the proposed unified motion planning method for heterogeneous tracked vehicles, with the help of differentiated behavior constraints when primitives are generated, and unequal selection weights set according to primitive attributes, generates a motion planning result that reflects the difference in characteristics of heterogeneous tracked vehicles and meets real-time requirements.

田洪清 , 王建强 , 黄荷叶 , 等 . 越野环境下基于势能场模型的智能车概率图路径规划方法 [J ] . 兵工学报 , 2021 , 42 ( 7 ): 1496 - 1505 .

TIAN H Q , WANG J Q , HUANG H Y , et al . Probabilistic roadmap method for path planning of intelligent vehicle based on artificial potential field model in off-road environment [J ] . Acta Armamentarii , 2021 , 42 ( 7 ): 1496 - 1505 . (in Chinese)

KOIDE K , MIURA J , MENEGATTI E . A portable 3D LIDAR-based system for long-term and wide-area people behavior measurement [J ] . International Journal of Advanced Robotic Systems , 2019 , 16 ( 2 ): 1 - 16 .

SHI X J , CHEN Z R , WANG H , et al . Convolutional LSTM network:a machine learning approach for precipitation nowcasting:arXiv:1506.04214 [R ] . Ithaca,NY,US:Cornell University , 2015 :1506.04214.

STELLATO B , BANJAC G , GOULART P , et al . OSQP: an operator splitting solver for quadratic programs [J ] . Mathematical Programming Computation , 2020 , 12 ( 4 ): 637 - 672 . DOI: 10.1007/s12532-020-00179-2 http://doi.org/10.1007/s12532-020-00179-2