This paper proposed a discrete-time integral terminal sliding mode (DITSM)-based forward speed tracking control for a robotic fish (RF). Due to the difficulty of quantification for the hydrodynamic model of RF during swimming, the head-yawing effect is considered in the obstructive thrust and then the RF dynamic model is further optimized via a data-driven approach. Compared with the traditional robotic fish model, the established model can provide sufficient thrust for the swimming of robotic fish and effectively improve the model accuracy. The DITSM control based on the accurate RF model is proposed by considering the head-yawing effect in order to achieve high-precision and robust speed tracking performance. The real-time experimental results are demonstrated to verify that the proposed control with head-yawing can provide with high-accuracy speed tracking for the RF.

本文提出了一种基于离散时间积分终端滑模（DITSM）的机器鱼（RF）前向速度跟踪控制。由于游泳时 RF 的流体动力学模型难以量化，因此在阻碍推力中考虑了头部偏航效应，然后通过数据驱动的方法进一步优化了 RF 动力学模型。与传统机器鱼模型相比，所建立的模型可以为机器鱼的游动提供足够的推力，有效提高模型精度。为了实现高精度和鲁棒的速度跟踪性能，提出了基于精确RF模型的DITSM控制，考虑了头部偏航效应。