CASIA OpenIR  > 毕业生  > 博士学位论文
仿生金枪鱼巡游和机动控制研究
杜晟
Subtype博士
Thesis Advisor谭民 ; 喻俊志
2021-05-26
Degree Grantor中国科学院大学
Place of Conferral北京
Degree Discipline控制理论与控制工程
Keyword仿生机器鱼 金枪鱼 运动控制 路径规划 路径跟踪
Abstract

海洋是生命起源之地,同时蕴藏有丰富的资源。在开发利用海洋资源的同时保护其生态环境免遭破坏,不单是可持续发展的需要,更是关系到人类自身的生存。常规螺旋桨推进的水下航行器容易给海洋生物造成伤害,而仿生型水下航行器能在保护海洋生态环境的同时完成探索作业,其研究和应用具有重大的科学价值和现实意义。
本文围绕仿生金枪鱼的系统设计与建模、转向和定向控制、高速游动和速度控制以及定速巡航控制等问题展开研究。主要研究工作总结如下:
一、针对现有仿生机器鱼平台无法同时兼顾游速和转向性能的问题,以金枪鱼为仿生对象,研制了一种机动型仿生金枪鱼系统,通过对推进机构的创新设计实现了游速和转向性能的平衡。两主动关节的推进机构采用驱动电机前置、动力通过齿轮和连杆机构传递的方式,可在对关节角实现精确控制的同时获得一个较高摆频。在上述新颖结构的基础上,设计了配套的控制系统,研制开发了仿生金枪鱼样机平台。之后,针对仿生金枪鱼的三维运动,构建了全状态动力学模型,并在开环控制下进行了直游和转向两项基本运动的实验,实验结果验证了动力学模型的准确性和有效性。
二、针对仿生金枪鱼转向性能较差的问题,提出了一种基于拍动模式和参数优化的控制方法,实现了 0.16 m 的最小转向半径以及约 80°/s 的连续转向速度。随后针对仿生金枪鱼的定向控制问题,给出了一种基于模型的自适应调整算法。用实际航向角代替实时航向角,以减少头部晃动带来的不利影响;结合转向运动的研究结果,实现航向角快速调整的同时降低对速度的影响。仿真结果验证了所提控制方法的有效性。
三、针对仿生金枪鱼高速游动的问题开展研究,结合仿生关节驱动电机实际工况,从底层电机驱动角度出发,提出了一种适用于电机连续正反转的转速预测控制算法。所提算法提高了电机的有效功率,实现了最大游速从 0.62 m/s 至0.78 m/s 的提升。随后针对仿生金枪鱼的游速控制问题,提出了一种基于模型的游速控制方法,并结合航向角控制实现了给定方向的游速控制。基于摆动周期的速度估计给出了给定方向的平均速度;基于模型的控制方法解决了速度估计的滞后性。仿真结果表明,所提控制方法能较好实现对给定方向的速度控制。
四、针对仿生金枪鱼的定速巡航问题,从巡航和定速两方面展开研究。首先,结合仿生金枪鱼的动力学特性将路径规划问题转换成寻找关键路径点的问题,并以此为基础提出了一种基于 Bresenham 算法的改进 A* 算法。所提算法在障碍物稀疏的情景下表现良好,基本能满足实时性规划的要求。随后针对仿生金枪鱼的路径跟踪问题,提出一种基于视线导航法和滑模控制思想的路径跟踪策略,利用横向误差和给定直线的方向角构造目标航向角,以实现对给定直线进行跟踪。针对定速的要求,给出了一种在跟踪给定路径时,估算所需摆动频率的方法。最后结合路径跟踪算法和游速控制方法,提出了面向定速巡航任务的控制策略。仿真结果表明,所提算法较好地实现了仿生金枪鱼在给定时间到达给定地点的任务。

Other Abstract

As the cradle of life, the immense ocean is rich in resources. However, not only the sustainable development of ocean but also the survival of humanity, needs the protection of the ecological environment. Conventional propeller-driven underwater vehicles are likely to hurt marine organisms, and by contract, the bionic underwater vehicle can complete the task with no damage to others. Hence, the research and application of bionic underwater robots have great scientific value and practical significance.
This dissertation mainly concerns system design and dynamic modeling, turning and course angle control, high-speed implementation and speed control, and constant-speed cruising of a bionic robotic tuna. The main contributions are summarized as follows.
Firstly, a mobile bionic robotic system modeled after a thunnus thynnus has been developed. Driven by an innovative design of the propulsion mechanism, it has a balanced performance of swimming speed and steering. By putting the motors in front and transmitting the power through gears and four-bar linkage, the proposed two-active-joint propulsion mechanism can achieve both precise control of joint angles and a high oscillating frequency. On the basis of this innovative structure, the bionic tuna prototype is developed with a supporting control system. Subsequently, a full-state dynamic model is constructed to describe the three-dimensional motion of the bionic tuna, and experiments of straight swimming and turning are carried out to validate the accuracy and effectiveness of the dynamic model.
Secondly, a turning control is proposed to improve the steering performance of the bionic tuna, which is based on the change of flapping mode and parameter optimization. A minimum turning radius of 0.16 m with a continuous turning speed of about 80°/s is achieved in the experiments. Next, an adaptive control algorithm derived from the simulation model is proposed. The actual course angle is used instead of the instant one, to reduce the negative effect of inevitable head shaking. With the research results of open-loop steering, the proposed control method can realize the rapid adjustment of course angle with little influence on speed. The effectiveness of proposed algorithm is verified by simulation.
Thirdly, in pursuit of a high swimming speed, a speed prediction control algorithm for the continuous reciprocating rotation of the motors is proposed, from the perspective of the underlying motor driving. The given algorithm improves the effective power of the motors, and achieves an increase of the maximum speed from 0.62 m/s to 0.78 m/s. Subsequently, the swimming speed control is realized by combining the course angle control and a model-based speed control method. The average velocity in the given direction is estimated on the basis of the oscillation period, and the lag of velocity estimation is solved by the model-based method. The simulation results indicate that the proposed strategy can effectively control the speed in a given direction.

Fourthly, the research of fixed-speed cruise problem is carried out from two aspects: cruising and fixed-speed control. For the cruise, the path planning problem is converted into finding critical path points, due to the dynamic characteristics of bionic tuna. And then an improved A* algorithm modified by Bresenham algorithm is proposed to implement path planning. It performs well in the scenario of sparse obstacles, and can meet the requirements of real-time planning. Next, in order to track a straight line, a path tracking method inspired by line-of-sight (LOS) navigation and sliding mode control (SMC) is proposed. By using the cross-track error and the angle of given line to construct the reference head angle, the control of linear path tracking is substantially simplified. For the requirement of a fixed-speed, an estimation of the oscillation frequency for tracking the given path is presented. Finally, combining the path tracking algorithm and the speed control method, a strategy for the cruise control task is proposed. The simulation results reveal that it can make the bionic tuna reach the destination exactly in a given time.

Pages136
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/44713
Collection毕业生_博士学位论文
Recommended Citation
GB/T 7714
杜晟. 仿生金枪鱼巡游和机动控制研究[D]. 北京. 中国科学院大学,2021.
Files in This Item:
File Name/Size DocType Version Access License
毕业论文.pdf(9091KB)学位论文 限制开放CC BY-NC-SA
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[杜晟]'s Articles
Baidu academic
Similar articles in Baidu academic
[杜晟]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[杜晟]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.