Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Department of Mechanical Engineering | en_US |
| dc.contributor.advisor | Jing, Xingjian (ME) | en_US |
| dc.creator | Hameed, Imran | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/14302 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | en_US |
| dc.rights | All rights reserved | en_US |
| dc.title | Learning swimming gaits and navigation on a biomimetic marine robot | en_US |
| dcterms.abstract | Marine ecosystem contains enormous resources for mankind in the form of food, fuel, oxygen, minerals etc. It also contains answers to a lot of questions related to the inception of life and subsequent evolution. Ocean exploration is thus very beneficial to mankind. Marine robotics is one way to explore aquatic environments. | en_US |
| dcterms.abstract | Biologists have been studying marine life for centuries. They have classified organisms into several categories. One ubiquitous category is Body and Caudal Fin (BCF) swimmers. They are abundantly found in all regions (oceans, rivers, ponds etc.). They have further subtypes which swim with different swimming gaits. All the gaits have their own features and benefits. | en_US |
| dcterms.abstract | In bio-inspired robotics, inspiration is drawn from nature for design, and locomotion strategies. We find a lot of studies to develop fish-like robots in which a certain organism is mimicked in terms of design and with respect to swimming gait. Most studies revolve around a particular organism to develop and study its gait and characteristics. However, multiple gait patterns can be utilized together on a single biomimetic design to employ collective benefits. In literature, we find that there is a lack of a unified control scheme that can be used to optimize and mimic undulatory patterns observed among different organisms in the Body and/or Caudal Fin (BCF) category. | en_US |
| dcterms.abstract | Therefore, in this study, a generic biomimetic marine robot BCFbot (Body and Caudal Fin Robot) is conceived to study BCF locomotion. The robot, owing to having a generic design, can adopt multiple motion patterns related to different kinds of natural organisms in the BCF category. Moreover, a bio-inspired control scheme is proposed to develop swimming gaits for the robot. The schematics incorporate central pattern generators in a deep reinforcement learning (DRL) framework which allows the framework to develop/optimize distinct motion patterns and switch between them seamlessly and instantly. | en_US |
| dcterms.abstract | The first part of the study focuses on developing three gaits (namely anguilliform, subcarangiform, and carangiform). After development of these gaits, a thorough comparison is made between them by testing the gaits through simulation and on the hardware. In depth testing and comparison reveal different benefits associated with different gaits. One gait is found to be the fastest, another to be the most energy economical and one to be multi-modal. Hence, the benefits of three benchmark motion patterns can be well realized with the developed robot and can be freely switched and optimized with the developed DRL mechanism. This should be the first attempt for achieving a multi-modal pattern optimization and switching within a single BCFbot and demonstrating a successful motion generation regime similar to real animals. Oscillators integrated into a learning paradigm provide a bioinspired framework to systematically develop a variety of swimming gaits. | en_US |
| dcterms.abstract | The second part of the study focuses on on-surface navigation of the robot along desired trajectories. To realize this, dynamic motion primitives, which can represent a large range of motion behaviors, are combined with a decoupled reinforcement learning framework. The proposed architecture optimizes the motion primitives first to develop a travelling wave undulation pattern in the tail and then to navigate the robot along different predefined paths. Through this framework, effective swimming gait emerges, and the robot is able to navigate well on the surface of water. This framework combines the optimization potential of deep reinforcement learning with stability and generalization properties of dynamic motion primitives. The method is trained and tested on a simulated model of the robot to demonstrate the effectiveness of the method and conduct experimental testing on the real robot to verify the results. | en_US |
| dcterms.abstract | In the third part, the design and schematics are modified to perform 3D underwater locomotion and navigation. In literature, for depth control, mostly pectoral fins are employed which do not fall into the BCF category. In this part, the same caudal fin used in previous parts for planar locomotion is used to perform underwater locomotion by adding extra pitch joints to the robot. The robot’s ability to perform underwater locomotion is tested thoroughly by a series of experiments performed via teleoperation. Testing results reveal that the modified design and the architecture can well realize three-dimensional underwater locomotion. Finally, 3D path-following along desired trajectories is also realized in simulation. | en_US |
| dcterms.extent | 173 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2023 | en_US |
| dcterms.educationalLevel | Ph.D. | en_US |
| dcterms.educationalLevel | All Doctorate | en_US |
| dcterms.accessRights | open access | en_US |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14302