Deep Reinforcement Learning-Enhanced Event-Triggered Data-Driven Predictive Control for a 3D Cable-Driven Soft Robotic Arm
2026-06-24 • Robotics
Robotics
AI summaryⓘ
The authors address the difficulty of controlling soft robots, which are complicated because their behavior changes and is hard to predict. They improve an existing method called DeePC, which plans robot movements but can be slow for real-time use, by using reinforcement learning to decide when to run DeePC only when needed. Their method, called RL-ET-DeePC, reduces how often the system needs to do heavy calculations by up to two-thirds without losing control accuracy. Tests in simulations and on a real soft robotic arm show it works well, saving computing effort while keeping the robot on track.
soft robotsdata-enabled predictive controlreinforcement learningevent-triggered controlnonlinear dynamicsreal-time controloptimizationrobotic armzero-shot transfer
Authors
Cheng Ouyang, Moeen Ul Islam, Kaixiang Zhang, Zhaojian Li, Xiaobo Tan, Dong Chen
Abstract
Soft robots are challenging to control due to their nonlinear and time-varying dynamics. Data-enabled predictive control (DeePC) offers a model-free alternative by directly leveraging measured input-output trajectories to construct a predictive controller. However, its receding-horizon formulation requires solving a constrained optimization problem at every sampling instant, which can be computationally demanding for real-time deployment on resource-limited robotic platforms.To address this limitation, we propose an adaptive reinforcement-learning-based event-triggered DeePC (RL-ET-DeePC) framework for soft robotic control. A model-free RL policy is trained to determine when to invoke the DeePC optimizer based on the current system state representation, thereby reducing unnecessary optimization calls while preserving closed-loop performance.Simulation results show that RL-ET-DeePC reduces optimization frequency by up to 66% compared to periodic DeePC, while maintaining comparable tracking accuracy. Hardware experiments on a three-dimensional cable-driven soft robotic arm demonstrate zero-shot transfer, achieving a 34% reduction in optimization frequency with tracking accuracy comparable to periodic DeePC and more consistent performance than a static threshold-based event-triggered baseline.