Open Access

ARTICLE

Obstacle Avoidance Path Planning for Delta Robots Based on Digital Twin and Deep Reinforcement Learning

Hongxiao Wang¹, Hongshen Liu¹, Dingsen Zhang^1,*, Ziye Zhang¹, Yonghui Yue¹, Jie Chen²

1 College of Information Science and Engineering, Northeastern University, Shenyang, 110819, China
2 Xufeng Electronics Co., Ltd., Shenyang, 110819, China

* Corresponding Author: Dingsen Zhang. Email:

Computers, Materials & Continua 2025, 83(2), 1987-2001. https://doi.org/10.32604/cmc.2025.060384

Received 30 October 2024; Accepted 05 February 2025; Issue published 16 April 2025

Abstract

Despite its immense potential, the application of digital twin technology in real industrial scenarios still faces numerous challenges. This study focuses on industrial assembly lines in sectors such as microelectronics, pharmaceuticals, and food packaging, where precision and speed are paramount, applying digital twin technology to the robotic assembly process. The innovation of this research lies in the development of a digital twin architecture and system for Delta robots that is suitable for real industrial environments. Based on this system, a deep reinforcement learning algorithm for obstacle avoidance path planning in Delta robots has been developed, significantly enhancing learning efficiency through an improved intermediate reward mechanism. Experiments on communication and interaction between the digital twin system and the physical robot validate the effectiveness of this method. The system not only enhances the integration of digital twin technology, deep reinforcement learning and robotics, offering an efficient solution for path planning and target grasping in Delta robots, but also underscores the transformative potential of digital twin technology in intelligent manufacturing, with extensive applicability across diverse industrial domains.

---

Keywords

---