Towards an Esox lucius inspired multimodal robotic fish

Abstract

In this paper, we further explore multimodal locomotion via an updated robotic fish model based on Esox lucius. Besides the improved actuation properties like higher torque servomotors and powerful electronics, the robotic fish has some innovative mechanical design to pursue diverse swimming modes and superior performance. Specifically, we introduced a ±50° yawing head joint that functions as the neck for enhancing turning ability. A pair of pectoral mechanisms with two DOFs per fin is constructed to achieve 3-D swimming and to enrich multiple pectoral motions. At the control level, an improved central pattern generator (CPG) model allowing for free adjustment of the phase relationship among outputs is employed to produce rhythmic signals of multimodal swimming. Extensive experiments were carried out to examine how characteristic parameters in CPGs including amplitude, frequency, and phase lag affect the swimming performance. As a result, the robotic fish successfully performed various locomotion actions such as forward swimming, backward swimming, turning, diving, surfacing, as well as three pectoral motions in the form of pitching, heaving, and heaving-pitching. We found that small phase lag between oscillating joints which means large propulsive body wave length and undulation width could lead to a faster swimming in body and/or caudal fin (BCF) locomotion.

摘要

创新点

给出了一种改进的仿生机器狗鱼设计方案, 进一步探索了机器人在水中的多模态运动控制。 为了实现丰富的运动模态和良好的运动性能, 针对机器狗鱼进行了多项机构改进。 首先, 增加了能够偏航±50°的颈部, 以扩大其转向范围; 其次, 设计了灵巧的四自由度胸鳍机构以增强其三维机动能力和实现多种胸鳍运动。 在控制层面, 基于改进的CPG模型来进一步探索CPG特征参量对机器鱼游动性能的影响。 实验结果表明该机器狗鱼具有良好的多模态运动能力, 实现了包括直游、 倒游、 转向及潜浮在内的多种运动模式, 并进一步揭示了较小的鱼体关节相位差能够产生较快的游动速度。