Multi-Object Tracking Based on Segmentation and Collision Avoidance

An approach to track multiple objects in crowded scenes with long-term partial occlusions is proposed. Tracking-by-detection is a successful strategy to address the task of tracking multiple objects in unconstrained scenarios,but an obvious shortcoming of this method is that most information available in image sequences is simply ignored due to thresholding weak detection responses and applying non-maximum suppression. This paper proposes a multi-label conditional random field（ CRF） model which integrates the superpixel information and detection responses into a unified energy optimization framework to handle the task of tracking multiple targets. A key characteristic of the model is that the pairwise potential is constructed to enforce collision avoidance between objects,which can offer the advantage to improve the tracking performance in crowded scenes. Experiments on standard benchmark databases demonstrate that the proposed algorithm significantly outperforms the state-of-the-art tracking-by-detection methods.

Effects of Introducing Ship Handling Training for Collision Avoidance in Anchoring Training-Effects of Ship Handling Training for Collision Avoidance by Group Work

In shipping,which is one of the drivers of the world’s economy,many marine accidents continue to occur,such as ship collisions and grounding.To reduce marine collision accidents,seafarers’skills must be improved through training.Therefore,the authors propose a ship handling training for collision avoidance(hereinafter referred to as“T for CA”)in which a group of several people discusses the ship handling for collision avoidance,assuming the situation of the collision avoidance.After T for CA implementation,anchoring training was done and the effect of T for CA was verified through comparison with a group where T for CA was not applied.Two instructors evaluated the anchoring training conducted with and without“T for CA”.The anchoring training experiment showed a difference of 27.5%in the achievement rate between the proposed training and previous training.T for CA maximises the effects of group work and resulted in good evaluations in the anchoring training experiments.The training was effective because the students themselves set the scenarios and devised ship handling strategies for collision avoidance.In addition,group work discussions helped deepen students’knowledge and skills.

Collision-avoidance path planning for multi-ship encounters considering ship manoeuvrability and COLREGs

Ship collision prevention has always been a hot topic of research for navigation safety.Recently,autonomous ships have gained much attention as a means of solving collision problems by machine control with a collision-avoidance algorithm.An important question is how to determine optimal path planning for autonomous ships.This paper proposes a path-planning method of collision avoidance for multi-ship encounters that is easy to realize for autonomous ships.The ship course-control system uses fuzzy adaptive proportion-integral-derivative(PID)control to achieve real-time control of the system.The automatic course-altering process of the ship is predicted by combining the ship-motion model and PID controller.According to the COLREGs,ships should take different actions in different encounter situations.Therefore,a scene-identification model is established to identify these situations.To avoid all the TSs,the applicable course-altering range of the OS is obtained by using the improved velocity obstacle model.The optimal path of collision avoidance can be determined from an applicable course-altering range combined with a scene-identification model.Then,the path planning of collision avoidance is realized in the multi-ship environment,and the simulation results show a good effect.The method conforms to navigation practice and provides an effective method for the study of collision avoidance.

基于改进樽海鞘群算法的船舶避碰决策及仿真实验

文中针对多船会遇避碰决策中过渡依赖单一寻优决策的问题采用了加入自适应权重的樽海鞘群优化算法(weight salp swarm algorithm, WSSA),在算法中融入国际海上避碰规则(convention on the international regulations for presenting collisions at sea, COLREGs)和良好船艺的要求.使用速度障碍法判断船舶的碰撞危险度并将多船会遇避让的过程中避让的安全性、经济性以及船舶领域侵入程度作为建立避碰决策的目标函数.算法测试的结果中,WSSA与原始樽海鞘群算法(SSA)以及经典粒子群算法(partide swam optimization, PSO)相比较,WSSA算法在收敛的精度和速度方面都明显优于SSA和PSO算法.结果表明:WSSA在寻找最优碰撞路线的过程中迭代的次数更少,精度更高.

山区河流典型桥梁船撞概率分析模型比较

船撞桥概率是多学科交叉融合、多因素影响的复杂研究,目前山区河流中梁桥和拱桥在计算船撞桥概率选取规范模型时不太相同。现有船撞桥概率模型及适用条件有所不同,采用美国AASHTO规范模型、LARSEN模型以及三概率参数积分模型,开展船舶对山区河流典型桥梁撞击概率及规律分析。通过分析连续钢构梁桥、上承式拱桥两种桥型的桥在最高、最低通航水位下涉水桥墩的年撞击概率,分析水位变化对年撞击概率的影响规律,得到山区河流梁桥和拱桥的模型适用性。结果表明,三种规范模型下的年撞击概率皆随着通航量的增加而不断增加,AASHTO模型与三概率参数模型,对于梁桥的概率计算结果变化趋势大致相似;对于拱桥,三概率参数积分模型对拱桥撞击横向、纵向位置的变化较为敏感,更适用于山区水位变化较大的拱桥。成果可为不同桥型在船撞概率模型选取提供参考。

具有避碰和连通性保持的多USV分布式事件触发编队控制

[目的]为解决水面无人艇(USV)编队控制中的实际需求,提出一种具有避碰和连通性保持功能的分布式事件触发编队控制方法。[方法]针对海洋环境扰动和系统模型的不确定性等因素设计扰动观测器;针对编队实现过程中USV间可能发生碰撞和丢失通信连接的问题,基于人工势场函数法设计具有避碰和连通性保持功能的编队控制器;为减少USV控制器的执行频率,设计分布式的事件触发机制,使每个USV根据局部信息来独立决定其控制器的执行时刻。[结果]理论分析证明了编队误差的一致最终有界性,且事件触发机制不存在Zeno行为。仿真结果验证了所设计的控制方法能够显著减少USV控制器更新次数并可实施避碰操作及保持连通性。[结论]研究成果可保证USV编队控制的顺利安全完成,同时减轻USV在复杂海洋环境下的控制机构损耗。

基于碰撞概率的GEO业务卫星空间安全规避策略

针对GEO业务卫星空间规避问题,提出一种基于碰撞概率优化的GEO卫星空间规避策略.建立脉冲推力控制下的位保模型,解释碰撞概率的假设条件与计算原理,针对GEO卫星对原有碰撞概率计算方法进行改进,提出了一种考虑了速度误差的碰撞概率计算方法,以脉冲速度大小为变量,以消耗速度冲量最小和碰撞概率最小为优化目标,基于非支配排序多目标遗传算法(NSGA-Ⅱ)进行多目标优化,对碰撞概率较大的问题求出其帕累托最优解.考虑GEO卫星定点保持的业务要求,得到满足其位保精度的空间规避策略,并进行仿真验证了空间规避策略的有效性.研究结果可为在轨业务卫星实现满足位保要求的空间安全规避提供思路.

基于海事规则和船舶操纵性的运动规划方法研究

文中提出了一种基于海事规则和船舶操纵性的运动规划方法.根据海事规则的要求,限制船舶转艏可行域.根据船舶推进器布局分析船舶操纵性能,并通过人工势场求取船舶转艏方向,采用计算势能的方式取代传统方法中计算合力的方式,计算船舶所需推力和转矩.将所求推力和转矩作用于船舶,得到一条包含速度和艏向信息的船舶运动轨迹.将并提出的方法与传统人工势场法进行仿真对比.结果表明:所提出的运动规划方法所规划的轨迹能够满足海事规则的要求,与传统人工势场法所规划的轨迹相比更加平滑,且满足船舶操纵性的要求,控制效果更好.

自主船舶与有人驾驶船舶动态博弈避碰决策

[目的]为实现船舶的高效避碰与安全航行,针对自主船舶与有人驾驶船舶混合航行环境下的交互避碰问题展开研究,提出一种多智能体交互的船舶动态博弈避碰决策方法。[方法]依据驾驶实践,理解并分析混行环境下的船舶避碰问题,基于《国际海上避碰规则》量化船舶会遇态势及碰撞危险,引入动态博弈理论,将存在碰撞危险的船舶个体建模为博弈中具有独立思想的参与者,并以船舶的航向改变量为博弈策略,在船舶安全性收益、社会性收益及经济性收益约束下求解船舶的最优行动序列,以及将船舶驾驶员风格的差异化引入仿真实验中以验证避碰决策的有效性。[结果]结果显示,所提方法能够在自主船舶与有人驾驶船舶混行场景下实现多船的安全会遇;各船在面对不同驾驶风格的目标船舶时均能类人地调整自身的行为策略,从而实现安全避让。[结论]所做研究可为自主船舶及有人驾驶船舶的避碰决策提供参考。

Multi-agent技术在海上船舶避碰决策系统开发中的应用

计算机控制技术的发展给船舶行业带来了新的发展,为了最大限度保障船舶的航行安全,通过计算机通信技术和智能算法,人们已经能够实现船舶的自动避碰,这极大地降低了船舶的事故率。但是传统的避碰决策系统并不能同时处理大量的船舶航行数据,因此必须借助Multi-agent技术,提高船舶的避碰数据处理速度,在显著提高避碰效果的同时,有效降低了航运成本。本文通过建立基于Multi-agent技术的船舶避碰模型,分析了该技术的应用效果。

Multi-agent在船舶避碰决策中的研究与应用

船舶避碰决策问题由于受到诸多因素的影响,使得避碰决策模型必须能够适应复杂多变的环境。本文给出一种基于Multi-Agent的船舶避碰决策方法,对于复杂问题,Multi-Agent模型性能更优于单Agent模型。通过对多目标船舶避碰实验的仿真,能够得到与人工分析一致的结果表明,本文方法具有一定的合理性。

基于NMPC的智能汽车变道避撞控制研究

针对无人驾驶车辆仅采用转向主动避撞控制策略难以应对复杂交通环境的问题,提出了一种基于非线性模型预测控制的变道轨迹规划与跟踪控制方法,实现了无须预先定义轨迹的实时轨迹规划。首先,分析了在高速公路上需要进行变道的场景,针对车辆在变道过程中不仅涉及横向位置的变化而且还与纵向速度有关的问题,建立了车辆横向和纵向联合非线性预测车辆模型;然后,基于模型预测控制算法可以考虑各种复杂约束、非线性等前提以实现系统性能最优控制的优势,设计了基于非线性模型预测控制(NMPC)的变道避撞控制器;最后,利用Matlab和Carsim联合仿真平台对所提方法进行了验证。仿真结果表明:所提出的方法不仅可以使车辆在复杂交通环境下成功地进行变道,而且能在变道过程中保持车身稳定、变道轨迹平滑。

船舶避碰时机曲线图研究

船舶最佳避碰时机表现为两船之间距离变化中的某一段,是船舶值班驾驶员最关心的问题.通过分析不同情况下船舶避让行动与时机曲线图,给出了船舶避让最佳时机的起始点D_(act1)和终止点D_(act2),以及最佳避碰行动点,得到了这些避让行动时机点与最小安全会遇距离(D_(s))、船速比(k_(V))、目标船相对方位(Q)、目标船最近会遇距离(D_(cpa))的关系.船舶避让行动与时机曲线图表明,最佳避让时机是存在的,可以用数值表示出来,能够用于指导海上船舶避碰实践.

基于改进蚁群算法的船舶路径规划算法与避碰仿真实验

近年来,船舶正逐渐向智能化和自主化方向发展,船舶路径规划作为实现船舶智能化的基础,已成为学术界的研究热点。蚁群算法作为最常用的元启发式算法之一,在解决路径规划问题上取得了不错的效果,但仍然存在若干缺陷。为解决蚁群算法实现船舶路径规划时存在的迭代速度慢、路径安全性较低等问题,采用启发式和融合式策略改进算法:在蚁群算法迭代初期,引入人工势场,提高算法迭代效率;将路径长度、路径安全性和路径平滑性约束函数融入信息素更新规则,保障船舶航行路径安全;构建混合蚁群静态路径规划算法,加入船舶避碰方法,搭建船舶动态路径规划算法。为验证路径规划算法的可行性和稳定性,设计各种仿真环境并进行了对比分析,结果表明,改进后的蚁群算法收敛速度快,且路径规划更加贴合实际。

基于Avoid-auv行为的多AUV系统避碰仿真

为了解决多AUV系统编队中存在动障碍物的避碰问题,尤其是编队成员间的避碰和协调,在Avoid-obstacle行为的基础上设计了一种基于良好通信条件下的Avoid-auv行为,它通过建立多AUV系统之间的避碰规则和策略,使AUV编队能在比较复杂的环境中较快地规划出一条无碰路径,而且建立的避碰规则也能协调AUV系统成员之间有序地运动。仿真研究表明,基于Avoid-auv行为能较好地解决多AUV协调避碰问题,但必须更加重视多AUV系统之间的通信问题。

Deliberative Collision Avoidance for Unmanned Surface Vehicle Based on the Directional Weight

The purpose of this study is to realize the intelligent deliberative obstacle avoidance for unmanned surface vehicle(USV), based on the International Regulations for Preventing Collisions at Sea called COLREGS.Firstly, a three-level system architecture is designed for the ship intelligent collision avoidance system. Secondly,an intelligent collision avoidance algorithm is presented based on the calculation of motion parameters and the rules of COLREGS. Thirdly, according to COLREGS, the marine environment is modeled from the electronic chart data and radar information, and a deliberative collision avoidance algorithm is written. Finally, the deliberative collision avoidance simulation is carried out for the USV. The feasibility and reliability of intelligent collision avoidance algorithm are verified by the vivid simulation results.

Max-Min Adaptive Ant Colony Optimization Approach to Multi-UAVs Coordinated Trajectory Replanning in Dynamic and Uncertain Environments

Multiple Uninhabited Aerial Vehicles (multi-UAVs) coordinated trajectory replanning is one of the most complicated global optimum problems in multi-UAVs coordinated control. Based on the construction of the basic model of multi-UAVs coordinated trajectory replanning, which includes problem description, threat modeling, constraint conditions, coordinated function and coordination mechanism, a novel Max-Min adaptive Ant Colony Optimization (ACO) approach is presented in detail. In view of the characteristics of multi-UAVs coordinated trajectory replanning in dynamic and uncertain environments, the minimum and maximum pheromone trails in ACO are set to enhance the searching capability, and the point pheromone is adopted to achieve the collision avoidance between UAVs at the trajectory planner layer. Considering the simultaneous arrival and the air-space collision avoidance, an Estimated Time of Arrival (ETA) is decided first. Then the trajectory and flight velocity of each UAV are determined. Simulation experiments are performed under the complicated combating environment containing some static threats and popup threats. The results demonstrate the feasibility and the effectiveness of the proposed approach.

一种基于深度强化学习的船舶智能避碰方法

文中提出了一种基于深度强化学习(deep reinforcement learning,DRL)的船舶智能避碰方法.该方法利用D3QN(double deep q-learning network with dueling architecture)算法与船舶领域模型,结合《国际海上避碰规则》(COLREGs)的避碰操作规范设计奖励函数,通过时序差分法实现优先经验回放,构建自主避碰的智能体.通过ROS-gazebo搭建仿真环境,构建神经网络处理环境中的视觉与雷达数据,快速有效地获取环境特征信息.结果表明:对比传统DQN算法,该方法具有更好的决策能力,训练时间更短;在避碰过程中可以对会遇局面做出正确的判断,选择符合COLREGs规范的避碰动作,最终可以准确并及时的避让目标船.

一种有效顾及复杂海域避碰的路径规划方法

针对现有舰船避碰路径规划方法难以适用于复杂海域的现状,本文提出一种有效顾及复杂海域避碰的路径规划方法。首先,通过K-means方法将舰船避碰阶段选用合理的危险度判别指标进行划分;然后,依据Nomoto模型进行航迹推演,预测舰船在一定时间分辨率后的航迹;接着,在避碰阶段中剔除具有碰撞危险和不符合此阶段避碰规则的预测航迹,并在余下的预测航迹中选取危险度最小的航迹作为最优预测航迹;最后,在正常航行阶段建立遴选函数,选取最优预测航迹,将所有最优预测航迹合成以生成路径。实验结果表明:相对于当前先进的改进人工势场方法来说,所提方法在只有点状碍航区的较简单环境下可有效提升规划效率;所提方法在有各类碍航区的较复杂海域,能顾及来船的航向变化,规划出遵循国际海上避碰规则的避碰路径。

Multi-telerobot collaboration based on coordinated controller

A coordinated controller used for multi-telerobots collaboration was presented based on the strategy of shared control. First, it can overcome the effect of time delay. And, it combines the intelligence of the master side and the slave side, which cannot only increase the efficiency and the safety of the system but also relieve the burden and requirements of the operator. The controller can also cope with the collision between two telerobots. A simulation experiment was carried out to verify the validity of the controller for two slave robots.