自主式水下机器人全局路径规划的基于案例的学习算法研究

本文讨论了基于案例的学习方法在水下机器人全局路径规划中的应用问题．基于案例的学习方法是一种增量式的学习过程，它根据过去的经验进行学习及问题求解．本文对基于案例的学习方法在规划中的应用框架进行了初步研究，对案例属性的提取，案例的匹配和择优，以及案例库的更新等问题提出了相应的算法．最后给出了几组仿真结果．     

一种不确定环境下移动机器人的避障规划算法

本文提出了在不确定的环境中，移动机器人的一种全局路径规划算法．将全局路径 规划分解为局部路径规划的组合．为提高规划的效率，在局部规划中，采用了基于案例的学 习方法．以ART-2神经网络实现案例的匹配、学习和扩充，满足了规划的实时性要求，仿真 结果说明了本算法的有效性．     

基于案例推理的服务组合模板的设计及应用

走向工业化的服务组合之路，是一种基于“组合套餐”一服务组合模板的途径．在适于业务用户使用的基于模板的大粒度服务组合方法中组合模板的设计是一个面向目标的、有约束的决策、探索和学习活动．本文有效地利用专家的知识和经验，给出了一种基于案例推理的模板设计方法，并通过在面向城市应急联动的服务组合平台原型系统中的一个应用，对其使用效果进行了评价．     

医疗事故辅助鉴定系统的设计与实现

本文以典型的医疗事故(事件)鉴定问题为应用背景，提出一类PRS案例的推理问题及其知识结构和表达方法，以及基于相似度的PRS案例检索算法和PRS案例的混合推理算法。根据医疗事故鉴定问题的特点．应用理论方法研究结果，设计实现了一个计算机辅助医疗鉴定系统(CAMCS)，并且对实际发生的几个医疗事故鉴定的典型案例进行分析，验证了基于案例混合推理方法的正确性，以及CAMCS系统的实际应用效果。     

智能规划研究综述——一个面向应用的视角

智能规划是智能系统研究的重要领域．在分析了智能规划中典型问题类型特征的基础上，从应用的视角对智能规划研究的前沿问题以及近年来的主要理论与方法进行了评述．结合若干工业领域的应用实例，讨论了工程应用中的规划问题特征及研究现状，并在此基础上结合智能工程理论对智能规划的应用研究方向进行了展望．     

神经网络的规划学习算法

本文利用二次规划方法，讨论对神经网络训练样本的吸引半径的优化问题，并借用二交规划中的ＲＰＡ算法求该优化解，得到一种新的神经网络基于规划的学习算法，其次，将规划学习算法与现有的几种常见算法进行比较，指出该算法的特点。     

基于强化学习的自动驾驶汽车路径规划方法研究综述

路径规划作为自动驾驶的关键技术,具有广阔的应用前景和科研价值。探索解决自动驾驶车辆路径规划问题的方法,着重关注基于强化学习的路径规划方法。在阐述基于常规方法和强化学习方法的路径规划技术的基础上,重点总结了基于强化学习和深度强化学习来解决自动驾驶车辆路径规划问题的算法,并将算法按照基于值和基于策略的方式进行分类,分析各类算法的特点、优缺点及改进措施。最后对基于强化学习的路径规划技术的未来发展方向进行了展望。     

基于反射的规划方法

本文针对机器人在实际环境中执行任务时所面临的不确定性和实时性提出了一种基于反射的规划方法．该方法在较好地利用全局信息的基础上，充分发挥出机器人自身的能动的反射作用．我们首先讨论该方法的基本原理，然后着重讨论它在移动机器人规划和导航中的应用．     

多移动机器人协作任务的分布式控制系统

本文以多移动机器人协作执行围捕任务为背景，探讨了多机器人系统行为层规划及任务层中的任务分解和分配问题，提出了一种能有效降低计算量的编队中子任务的分配方法．为多移动机器人提供了一种在不确定环境中，动态变化的协作任务下进行基于传感器信息和通信信息的在线路径规划的方法．     

基于学习机制的移动机器人动态场景自适应导航方法

针对在单一学习机制中,移动机器人自主导航一般只适用于静态场景,适应性差的问题,提出一种动态场景自适应导航方法．该方法通过激光测距仪（LRF）获取周围环境的距离信息,在基于增量判别回归（IHDR）算法的单一学习机制导航的基础上,提出了最远距离优先机制的局部避障环节．该导航方法克服了传统导航方法对环境模型的过度依赖,并且本文提出的基于最远距离优先机制的局部避障算法,解决了基于单一学习机制的导航方法对动态场景适应能力不足的问题．本文将动态场景自适应导航方法应用到了MT-R机器人中,与基于单一学习机制的导航方法进行了对比实验,并且运用提出的局部避障算法,对实验中的激光数据进行了算法性能分析．实验结果证实了该方法的可行性,并显示了该方法在动态场景下的良好表现．     

Path planning approach in unknown environment

This paper presents a new algorithm of path planning for mobile robots, which utilises the characteristics of the obstacle border and fuzzy logical reasoning. The environment topology or working space is described by the time-variable grid method that can be further described by the moving obstacles and the variation of path safety. Based on the algorithm, a new path planning approach for mobile robots in an unknown environment has been developed. The path planning approach can let a mobile robot find a safe path from the current position to the goal based on a sensor system. The two types of machine learning: advancing learning and exploitation learning or trial learning are explored, and both are applied to the learning of mobile robot path planning algorithm. Comparison with A* path planning approach and various simulation results are given to demonstrate the efficiency of the algorithm. This path planning approach can also be applied to computer games.     

A new path planning method for bevel-tip flexible needle insertion in 3D space with multiple targets and obstacles

In this paper, a new bevel-tip flexible needle path planning method based on the bee-foraging learning particle swarm optimization (BFL-PSO) algorithm and the needle retraction strategy in 3D space is proposed to improve the puncture accuracy and shorten the puncture distance in the case of multiple puncture targets. First, the movement of the needle after penetrating the human body is analyzed, and the objective function which includes puncture path error, puncture path length, and collision function is established. Then, the BFL-PSO algorithm and the needle retraction strategy are analyzed. Finally, medical images of the tissue to be punctured are obtained by medical imaging instruments, i.e., magnetic resonance (MR), and the 3D model of the punctured environment is constructed by 3D Slicer to obtain the environment information on targets and obstacles, and the path of flexible needle is carried out based on the BFL-PSO optimization algorithm and the needle retraction strategy. The simulation results show that, compared with other path planning methods in the related literature, the new path planning method proposed in this paper has higher path planning accuracy, shorter puncture distance, and good adaptability to multi-target path planning problems.     

深度强化学习及在路径规划中的研究进展

路径规划的目的是让机器人在移动过程中既能避开障碍物,又能快速规划出最短路径。在分析基于强化学习的路径规划算法优缺点的基础上,引出能够在复杂动态环境下进行良好路径规划的典型深度强化学习DQN（Deep Q-learning Network）算法。深入分析了DQN算法的基本原理和局限性,对比了各种DQN变种算法的优势和不足,进而从训练算法、神经网络结构、学习机制、AC（Actor-Critic）框架的多种变形四方面进行了分类归纳。提出了目前基于深度强化学习的路径规划方法所面临的挑战和亟待解决的问题,并展望了未来的发展方向,可为机器人智能路径规划及自动驾驶等方向的发展提供参考。     

Cooperative and Geometric Learning Algorithm (CGLA) for path planning of UAVs with limited information

In this paper, we propose a new learning algorithm, named as the Cooperative and Geometric Learning Algorithm (CGLA), to solve problems of maneuverability, collision avoidance and information sharing in path planning for Unmanned Aerial Vehicles (UAVs). The contributions of CGLA are three folds: (1) CGLA is designed for path planning based on cooperation of multiple UAVs. Technically, CGLA exploits a new defined individual cost matrix, which leads to an efficient path planning algorithm for multiple UAVs. (2) The convergence of the proposed algorithm for calculating the cost matrix is proven theoretically, and the optimal path in terms of path length and risk measure from a starting point to a target point can be calculated in polynomial time. (3) In CGLA, the proposed individual weight matrix can be efficiently calculated and adaptively updated based on the geometric distance and risk information shared among UAVs. Finally, risk evaluation is introduced first time in this paper for UAV navigation and extensive computer simulation results validate the effectiveness and feasibility of CGLA for safe navigation of multiple UAVs.     

PROCASE: a case-based process planning system for machining of rotational parts

This paper introduces a case-based process planning system PROCASE which generates new process routines through learning from existing process routines. In contrast to traditional rule-based systems, the process planning knowledge of the PROCASE is represented in terms of cases instead of production rules. The planning basically comprises case retrieving and case adaptation rather than chaining applicable rules together to form process plans. The advantages are, first, the system is cheaper to build as it saves the expense of knowledge acquisition. Second, the system is able to advance its knowledge automatically through planning practice. Third, it is robust, because the reasoning is not based on pattern matching but similarity comparison. PROCASE has three modules: the retriever, the adapter and the simulator. It is supported by a feature-based representation scheme which naturally serves as the case indices for case retrieving and adaptation. The retriever uses a similarity metric to retrieve an old case which is the most similar case, among all old ones, to the new case. The adapter is then activated to adapt the process plan of the retrieved case to fit the needs for the new case. The simulator is used to verify the feasibility of the adapted plan. PROCASE is implemented on a Silicon Graphics IRIS workstation using C++ . An example is given to demonstrate how the process routine is generated by the system proposed by the authors.     

Optimum motion planning in joint space for robots using genetic algorithms

The optimum motion planning in joint space (OMPJS) for robots, which generally consists of two subproblems, optimum path planning and optimum trajectory planning, was considered as a whole in the paper. A new method for optimum motion planning problem based on an improved genetic algorithm is proposed, which is more general, flexible and effective. This approach incorporates kinematics constraints, dynamics constraints, and control constraints of robotic manipulator. The simulation results for a two and a three degrees of freedom robots are presented and discussed. The simulations are based on genetic algorithm class library WGAClass 1.0 developed by us with Borland C++ 3.1.     

基于深度学习的多机械手轨迹规划系统设计

目前提出的多机械手轨迹规划系统路径规划精准度低,避障能力差。基于深度学习对多机械手的规划系统进行设计,通过研究传统系统中存在精确度、智能性不足的缺点,在设计的系统分别引入了相应的解决条件,在硬件结构的设计中本文应用ISL-320型号的伺服电机提升多机械手的动力功能,应用SKT64系列的芯片提升多机械手的路径精准度;在应用程序设计上应用拟合算法与叠加算法对规划路径中的节点精准运算,在提升系统整体精准度的同时提升了系统的智能程度。实验结果表明,基于深度学习的多机械手轨迹规划系统路径与标准路径十分接近,说明该方法的规划精准度较高,避障能力得到有效增强。