Undulation Adaptability Theory And Control For The Biomimetic Undulating Fins

Biomimetic undulating fins, as the quintessential case of biomimetics in the field of underwater vehicles or robots, learn valuable inspirations of higher efficiency, greater maneuvering or stabilization and better state-keeping at low speeds, from the fish which generally swim by undulations of their long-based ribbon fins. Even-tually, these undulating fins are expected to effectively complement conventional underwater propellers in the above-mentioned performance aspects, and therefore they have a bright prospective in both military and civilian applications. Since the performance of the pre-existing undulating fins is hardly satisfactory, this the-sis originally proposes the biomimetic undulation adaptability method in terms of the closed-loop biomimetics methodology. The studies in this thesis concentrate on the kinematics modeling, the undulation adaptability analyzing, the undulation adaptability control by iterative learning, and the corresponding experiments of the biomimetic undulating fins, with the following achievements and progress:(1) The undulation adaptability theory was proposed to mimic the functions and capabilities of the undulatory fish in adapting themselves to the outer flow and improving their swimming performance; and furthermore, the connotation, definition, classification, and key techniques of undulation adaptability were elaborately discussed and clarified.(2) A uniform descriptive method and its theoretical model were brought forward for both nature fish and bio-fish systems, on the basis of Martin's self-propulsion mathematical model; and then, a ruled surface based kinematics model was es-tablished with observations and measurements on the fundamental features in morphology and locomotion of the amiiform fish. The kinematics model not only described the morphological characteristics, such as the curve base line, non-uniform heights and non-zero depths of fin rays, but also better preserved periodicity, mono-apex, bi-directional swimming and asymmetry of the propul-sive waves in the waveform sequences.(3) The computational fluid dynamics (CFD) platform was set up for the undulation adaptability, and meanwhile, several algorithms were designed and implemented for the coming dynamic analysis. These algorithms are the undulation amplitude profile detecting algorithm, the smoothness-keeping fin movement algorithm, as well as the undulating dynamic mesh algorithm, respectively. (4) Based on the undulation adaptability platform, we discerned the morphology and locomotion adaptability of the biomimetic undulating fins. It was under-stood that the fins can find an optimal distribution when they are dissembled to various carriers. Then, the frequency effect, the wavelength effect and the asymmetric waveform effect were suggested and analyzed by using dynamic mesh CFD analysis, one by one. Succeeding this, the undulation adaptability engineering principles were strongly proposed, through the interrelation of the mentioned results and the locomotion observation data.(5) The Preisach-based dynamic model was established to describe the hysteresis nonlinearity which is remarkable within the biomimetic undulations. The theo-retical proofs were completed to confirm that the dynamic systems with hystere-sis preserve continuity and repeatability and that iterative learning control is applicable to improve the tracking performance of the hysteretic systems. And accordingly, the undulation adaptability iterative learning control algorithm was designed and implemented in this thesis. Simulations verified that the tracking error can decay to less than 10% of the original mode with ten iterations under some typical undulatory parameters.(6) The pre-existing biomimetic undulating fin, RoboGnilos, and the related un-dulation testbeds were upgraded to possess learnability with iterations, so that the corresponding experiments were able to be carried out to test both the un-dulation adaptability theory and the undulation adaptability control algorithm. On one hand, the three undulation effects (in terms of the frequency effect, the wavelength effect and the asymmetric waveform effect) were, respectively, val-idated by comparisons on the propulsive performance with various undulatory parameters; on the other hand, it was testified that the biomimetic fin is able to compensate the hysteresis nonlinearity by iterative learning and that the it-erative learning control algorithm was effective to improve upon the propulsive performance. Typically, the propulsive velocity may be enhanced nearly 40% with ten iterations.All the above achievements effectively facilitate the breakthrough from "imita-tion in shape" (mimicking the fin locomotion of the amiiform fish) to "inspiration in spirit" (unveiling the inherent property of the biomimetic undulation adaptability). Inspiringly, it is an in-depth development of the biomimetic undulation propulsion, and it is also full of significance to provide novel schemes and techniques for practical applications of the biomimetic underwater propulsors.