Optimal Gait Design for a Soft Quadruped Robot via Multi-fidelity Bayesian Optimization

Kaige Tan,Xuezhi Niu,Qinglei Ji,Lei Feng,Martin Törngren

2024-06-11

Abstract:This study focuses on the locomotion capability improvement in a tendon-driven soft quadruped robot through an online adaptive learning approach. Leveraging the inverse kinematics model of the soft quadruped robot, we employ a central pattern generator to design a parametric gait pattern, and use Bayesian optimization (BO) to find the optimal parameters. Further, to address the challenges of modeling discrepancies, we implement a multi-fidelity BO approach, combining data from both simulation and physical experiments throughout training and optimization. This strategy enables the adaptive refinement of the gait pattern and ensures a smooth transition from simulation to real-world deployment for the controller. Moreover, we integrate a computational task off-loading architecture by edge computing, which reduces the onboard computational and memory overhead, to improve real-time control performance and facilitate an effective online learning process. The proposed approach successfully achieves optimal walking gait design for physical deployment with high efficiency, effectively addressing challenges related to the reality gap in soft robotics.

Robotics,Systems and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to improve the walking ability of soft - legged quadruped robots, especially by optimizing their gait design through online adaptive learning methods. Specifically, the paper focuses on how to achieve efficient gait design in soft - legged quadruped robots to overcome the shortcomings in traditional methods, such as the reality gap between simulation and reality, the complexity of soft material modeling, and the challenges of soft actuator control precision.

To achieve this goal, the paper proposes the following key points:

1. **Gait Generator Design**: A parameterized gait generator is designed using a model based on the Central Pattern Generator (CPG), which can generate periodic gait patterns.

2. **Multi - fidelity Bayesian Optimization (MFBO)**: To address the gap between the simulation and the actual robot, the paper proposes a multi - fidelity Bayesian optimization method. This method combines data from simulation and physical experiments and gradually optimizes gait parameters through online learning, thus achieving a smooth transition from simulation to actual deployment.

3. **Edge Computing Architecture**: To improve real - time control performance, the paper introduces an edge computing architecture, which offloads complex computing tasks to edge servers, reducing the computational and memory burden on the robot itself and improving the effectiveness of the online learning process.

Through these methods, the paper aims to solve the key challenges in the gait design of soft robots and achieve efficient and highly adaptable walking ability.