CASIA OpenIR  > 复杂系统管理与控制国家重点实验室  > 先进机器人
面向仿生机器鱼的容错控制方法研究
杨越麒
Subtype硕士
Thesis Advisor喻俊志
2019-05-16
Degree Grantor中国科学院大学
Place of Conferral中国科学院自动化研究所
Degree Name工学硕士
Degree Discipline控制理论与控制工程
Keyword仿生机器鱼 动力学模型 容错控制 Cpg模型 反馈控制 多层感知机
Abstract

海洋占据了地球表面积的71%,蕴含着丰富的能源和生物资源。探索海洋是21世纪的主题,海洋经济具有无限的发展潜力。建设海洋强国,必须进一步关心海洋、认识海洋、经略海洋,加快海洋科技创新步伐。目前为止,根据人类所绘制的海底地图,只有不到5%的海洋区域得到了探索。要提高海洋探索和资源开发能力,需要强大的水下科技支持。自主水下航行器(Autonomous Underwater Vehicle,AUV)作为海洋探索和开发的主要工具被广泛应用于搜索和救援,海洋生物监测以及水环境检测等水下作业任务。提升水下航行器的运动性能,使其能效率更高地应用在更广泛的水下任务环境,有助于提升人类探索和利用海洋的能力,推动海洋科技进步,具有重要的现实意义。 鱼类在漫长的自然进化中获得了优异的生理结构和卓越的水下运动性能。依靠其肌肉组织和躯干的协调运动,鱼类可以在水中实现快速起动、高速游动和灵活转向从而完成捕食或逃避天敌的任务。相比之下,AUV的运动性能和效率与真实鱼类高低悬殊。近年来,随着仿生学的兴起,如何利用生物通过长期进化获得的天赋作为启发,探索其高超运动性能的奥秘用于研究和设计高性能的AUV,逐渐成为了AUV领域的研究热点。Triantafyllou等人于1994年首次研发了仿生机器鱼RoboTuna,在此之后,仿生机器鱼作为一种新型水下航行器,以其优异的性能受到了广泛关注,国内外科研人员围绕其推进机理、结构与外型以及运动控制等方面展开了研究。仿生机器鱼使用机器鱼尾和仿生鳍面作为运动和转向装置,通过模仿真实鱼类的外形和游动方式,具备高效率、高机动等性能优势,可以被应用于复杂、狭窄的水下环境作业。除此之外,区别于常规水下航行器主要采用螺旋桨式驱动进行水下运动,仿生机器鱼具备噪声小,不影响水下生态环境等优点,可以在开发海洋资源的同时保护海洋生态系统。 另一方面,随着海洋资源探索能力需求的增长,针对水下航行器系统的自主性、抗干扰能力和可靠性的要求越来越高。水下航行器的故障发生率和故障之后的表现是评价其系统稳定性的重要指标。除了从机械结构和电路等硬件层面降低故障发生频率以外,越来越多的研究人员致力于设计具有鲁棒性、抗干扰能力的容错控制系统保障水下航行器发生故障后的运动性能。虽然仿生机器鱼表现出优异的运动能力,但因结构复杂、驱动机构负载高等特点为其系统稳定性带来严峻挑战。此外,水下环境复杂且危险,仿生机器鱼执行复杂任务时容易发生故障。除此之外,仿生机器鱼具有驱动装置单一且耦合性强等特点,这使其游动性能对故障极其敏感。一旦故障发生,不仅无法保持原有机动性优势甚至连基本运动能力也会受到严重影响。目前对于仿生机器鱼的容错控制方法研究仍然欠缺,由于与传统水下航行器结构和驱动方式上的差异,面向AUV设计的容错控制方法很难适用在故障机器鱼系统。因此,面向仿生机器鱼设计容错控制算法,有利于提升控制系统稳定性和故障发生后的运动能力,是仿生机器鱼实际应用在更复杂的水下环境的重要保障,具有重要的现实意义。

Other Abstract

As a new kind of underwater vehicle, bionic robotic fish is characterized by excellent swimming performance. Fault tolerance is critical to the practicle applications and survival ability of robotic fish in complex underwater environments, and it is an important guarantee for maneuverability of robots. Research on fault-tolerant control of bionic robotic fish can not only effectively guarantee its swimming performance, but also promote its application in dynamic and complex operating environments, which has important theoretical significance and application value. This thesis focuses on the 3D dynamic modeling of multi-link bionic robotic fish, yaw fault-tolerant control, and speed fault-tolerant control. The main technical contributions are summarized as follows.

 

Firstly, a 3D dynamic model is established for multi-link bionic robotic fish, and a spatial trajectory simulation method based on the model is given. The kinematic analysis of the fish body and pectoral fins is carried out respectively, and the recursive relationship between velocity and acceleration is established. Then, the hydrodynamics of the bionic robotic fish body link are analyzed by the Morrison equation and the Strip method. By defining the transitional coordinate system and the velocity coordinate system, the hydrodynamic forces of the pectoral fins are calculated by the quasi-steady state lift resistance model. Thereafter, by applying the Newton-Euler method, the explicit expression of the dynamic model is obtained by inverse recursion according to the dynamic equations of the pectoral fin and the body link. Furthermore, in order to describe the robotic fish attitude, a 3D trajectory simulation method is given by establishing an inertial coordinate system and using Euler angles. Finally, the effectiveness of the proposed model is analyzed and verified by simulation.

 

Secondly, a yaw fault-tolerant control method is proposed to solve the problem that the caudal joint failure of the multi-link robotic fish will seriously affect its swimming yaw. The effects of faults and motion control parameters on yaw are analyzed. Thereafter, a mean value filter is designed for the problem that the swing of the bionic robotic fish affects the yaw measurement during the swimming process. Then, by employing the filtered yaw angle as feedback, a feedback yaw control system based on the Central Pattern Generator (CPG) is built. Furthermore, in order to optimize the performance of the feedback system, a feedforward compensator based on the dynamic model is incorporating by introducing the mathematical description of the fault information into the two-dimensional dynamic model. The feedback system and the feedforward compensator are combined to form the yaw fault-tolerant system, and the performance of each component of the system is analyzed through simulation. Finally, aquatic experiments are carried out on the bionic robotic fish to validate the effectiveness of the proposed method.

 

Thirdly, in order to solve the problem that the tail joint fault affects the swimming speed of the robotic fish, a speed fault-tolerant control method for the bionic robotic fish is proposed based on the yaw fault-tolerant control. The coupling phenomenon of CPG model in speed and yaw adjustment is analyzed, and the stable judgment switch is designed. Speed feedback control is implement    by using the swimming speed as feedback. Based on it, a speed fault-tolerant control method on the basis of multi-layer perceptron is presented. The multi-layer perceptron takes the expected swimming speed and fault information as input, and adopts the motion control parameters as the output. The data sets collected by the dynamic model are taken to train the multi-layer perceptron. Further, the combination of speed feedback, multi-layer perceptron, and yaw fault-tolerant control constitutes a bionic robotic fish speed-yaw fault-tolerant control system. Finally, the performance and effectiveness of each component of the system are analyzed and verified by simulation.

Pages104
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/23942
Collection复杂系统管理与控制国家重点实验室_先进机器人
Recommended Citation
GB/T 7714
杨越麒. 面向仿生机器鱼的容错控制方法研究[D]. 中国科学院自动化研究所. 中国科学院大学,2019.
Files in This Item:
File Name/Size DocType Version Access License
Thesis.pdf(13042KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[杨越麒]'s Articles
Baidu academic
Similar articles in Baidu academic
[杨越麒]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[杨越麒]'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.