SIA OpenIR  > 信息服务与智能控制技术研究室
基于跟随-领航结构的联合收割机群协同导航控制策略及方法研究
其他题名Research on Collaborative Navigation Control Strategy and Method Based on Leader-followers Structure for Combine Harvester Group
白晓平1,2
导师胡静涛
分类号TP273
关键词收割机群 跟随-领航结构 路径规划 协同导航 协同定位
索取号TP273/B16/2016
页数122页
学位专业机械电子工程
学位名称博士
2016-05-30
学位授予单位中国科学院沈阳自动化研究所
学位授予地点沈阳
作者部门信息服务与智能控制技术研究室
摘要本文面向收割机群协同导航作业需求,以跟随-领航结构为编队队形,深入研究了多机协同导航路径规划方法、协同导航定位方法和协同导航控制方法。为了验证本文所提方法的有效性和可行性,在收割机群协同导航实验平台上进行了方法的验证实验。本文的主要研究内容包括以下几个部分:第一,基于跟随-领航结构的收割机群协同导航路径规划方法研究。针对收获区域机群路径规划问题,提出了一种基于虚拟结构的路径规划方法。该方法采用跟随-领航几何结构对多台收割机进行编队,并将编队后的收割机群等价为一台虚拟的收割机,将收获作业地块划分为多个作业单元,并采用AB线偏移的方法一次性生成领航者导航路径,跟随者路径则以领航者路径为基准进行规划。针对地头转向区域机群路径规划问题,提出了一种分层地头转向路径规划方法。分层地头转向路径规划方法将地头转向路径规划分为轨迹层和次序层两层,轨迹层不考虑多收割机间的碰撞问题,机群中各收割机独立的规划地头转向轨迹,次序层不考虑地头转向轨迹规划问题,完全根据编队队形生成地头转向次序序列,分层地头转向路径规划方法简化了地头转向路径规划问题。第二,基于视觉导引激光测距的收割机群协同导航定位方法研究。针对收割机群中跟随者的定位问题,提出了一种基于相对位姿的跟随者定位方法。针对相对位姿测量需解决的靶标识别问题,提出了一种基于矩形度的靶标识别方法,可根据靶标的颜色和形状特征快速识别出靶标。该方法采用YUV颜色模型对图像进行分割,可有效去除与靶标颜色区别较大的干扰;基于靶标的形状和姿态特征,采用改进的旋转目标法提取连通区域的外接矩形,通过比较多个连通区域的矩形度,可快速提取出靶标投影矩形区域。针对相对位姿测量需解决的对靶控制问题,提出了一种基于视觉伺服的对靶控制方法。该方法通过在图像中设置内外波门,缩小了靶标投影矩形区域的搜索范围,降低了视觉反馈延迟;采用变尺度变论域模糊控制方法设计对靶控制器,通过引入变论域因子和变尺度因子,提高了控制器的控制精度和鲁棒性。在搭建的收割机群相对定位实验平台上进行了验证实验,实验结果表明本文所提定位方法与收割机群运动速度基本无关,能够满足收割机群协同导航定位要求。第三,基于跟随-领航结构的收割机群协同导航控制方法研究。针对收获作业区域机群协同导航控制问题,提出了一种基于跟随-领航结构的机群协同导航控制策略。建立了多机协同导航运动学模型,将机群协同导航控制问题分解为纵向距离保持控制和路径跟踪控制两个子问题,克服了跟随-领航法存在误差叠加的缺点;采用滑模控制理论和反馈线性化理论,设计了渐进稳定的队形保持控制律和路径跟踪控制律,保证了路径跟踪控制精度和队形保持控制的稳定性。针对地头转向区域机群协同导航控制问题,提出了一种基于行为的多机协同导航控制方法。该方法将复杂的地头转向任务分解为多个简单的感知行为,并为每种行为匹配开启条件和控制策略,多个行为根据优先级和任务要求进行切换,实现无碰撞地头转向。在搭建的收割机群编队控制实验平台上进行了路径跟踪实验和编队控制实验,验证了本文所提方法的有效性。第四,联合收割机群协同导航系统设计与集成。简要介绍了基于跟随-领航结构的收割机群协同导航控制系统的总体结构、所需研发的核心装置。重点对协同导航系统的主要软件核心算法进行了详细阐述。最后,以CF806为实验样机集成了联合收割机群协同导航系统,并开展了联合收割机群协同导航系统示范应用,进一步验证了本文所提方法的可行性及实用性。
其他摘要For navigation operation requirement of combined harvester group, with the structure of leader-follower, this paper deeply studied path planning method, positioning method and control method. In order to verify the effectiveness and feasibility of these methods, the experiment and application were taken on the experimental platform of the cooperative navigation experimental platform. The main contents of this paper can be generalized as follows: First of all, a collaborative navigation path planning method based on leader-follower structure was proposed. In order to solve the path planning problem of harvest area, a path planning method based on virtual structure was proposed. Based on the leader-follower structure, the combined harvester group can be seen as a virtual harvester, and the harvest area can be divided into multiple units. Then, the leader navigation path can be generated in one time by adopting AB lines offset method, and the follower navigation path can be generated based on the leader navigation path. Secondly, in order to solve the path planning problem of headland turning area, a path planning method based on hierarchical structure was proposed. This method divided the path planning problem into two layers: track layer and order layer. In track layer, the problem of the collision of multiple harvesters was ignored and any of the harvester group was seen as an individual. In order layer, the problem of the headland turning is ignored, and the headland turning order sequence is generated completely according to formation. Therefore, we can see that the headland turning problem is simplified. Secondly, a collaborative navigation positioning method based on vision-based guided laser ranging was proposed. In order to solve the problem of target recognition, a method based on the rectangular degree was proposed. This method can quickly identify the target based on color and shape features of the target. In order to solve the problem of the target measurement, a method based on the visual servo was proposed. In order to reduce the visual feedback delay, a motion target auto identify method based on dual window was proposed. This target identify method filter out most of the identified area by setting dual window, and search target point only in a small area of the image. In order to improve target control accuracy, a variable scale variable universe fuzzy control method was proposed. This method introduces contraction-expansion factor into the antecedents of fuzzy rules, and improve granularity of fuzzy rules without increasing the number of control rules. At the same time, this method introduces variable scale factor into the target rules, and improves the adaptability to changes in the distance between master and slave. Finally, relative position and orientation measurement apparatus was designed based on the above theory and collaborative positioning experiments were carried out. The experimental results show that the collaborative positioning accuracy is independent of the forward speed, and can meet the slave positioning requirement. Thirdly, the collaborative navigation control method based on the leader-follower structure was proposed. In order to solve the collaborative navigation control problem of the harvest area, a method based on the leader-follower method was proposed. By establishing the multi-machine collabarative navigation kinematic model, the collaborative navigation control problem was converted into two sub-problems: formation keeping problem and path tracking problem, which overcomes the leader-follower method has the problem of error stack. Considering the multi-machine collaborative navigation kinetatics, formotion control law and path tracking law were designed based on feedback linearization and sliding mode control theory. In order to solve the collaborative navigation control problem of the headland turning area, a method based on the behavior was proposed. The complex headland turning control problem was decomposed into several simple perception-action behavior equipped with open conditions and control strategy for each behavior. By a plurality of switching behavior, the multi-machine can realize headland turning control without collision. In order to verify the feasibility of the proposed method, formation experiment and path tracking experiment are performed on the harvester group platform. Finally, harvester group collaboration navigation control systems application integration technology was studied. Above all, the harvester group collaboration navigation control system based on the follow-lead structure is brief introduced. It contains the overall structure, the core device, and the main software algorithms. Then, this paper studied the harvester group collaboration navigation control system integration technology in the following two aspects: hardware integration, and software integration. At last, in order to verify the feasibility and effectiveness of the proposed method in this paper, formation control experiments were carried out.
语种中文
产权排序1
文献类型学位论文
条目标识符http://ir.sia.cn/handle/173321/19655
专题信息服务与智能控制技术研究室
作者单位1.中国科学院沈阳自动化研究所
2.中国科学院大学
推荐引用方式
GB/T 7714
白晓平. 基于跟随-领航结构的联合收割机群协同导航控制策略及方法研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2016.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
基于跟随-领航结构的联合收割机群协同导航(4011KB)学位论文 开放获取CC BY-NC-SA请求全文
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[白晓平]的文章
百度学术
百度学术中相似的文章
[白晓平]的文章
必应学术
必应学术中相似的文章
[白晓平]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。