837. Sequential Actuator Placement Optimization for Aircraft Fuselage Assembly via Reinforcement Learning
REMOVEDInvited abstract in session Removed abstracts, stream Removed, cancelled and rejected abstracts.
Authors (first author is the speaker)
| 1. | Juan Du
|
| 2. | Peng Ye
|
| The Hong Kong University of Science and Technology (Guangzhou) |
Abstract
Precise assembly of composite fuselages is critical for aircraft assembly to meet the ultra-high precision requirements. However, precise composite fuselage assembly is challenging due to compliant structures, complex material properties, and dimensional variabilities of incoming fuselages. In practice, actuators are required to adjust fuselage dimensions by applying forces to the fuselage edge through pulling or pushing force actions. The placements and force settings of these actuators significantly influence the efficiency of the shape adjustments. The current literature usually predetermines the fixed number of actuators, which is not optimal in terms of overall quality and corresponding actuator costs. This paper introduces a reinforcement learning (RL) framework that enables sequential decision-making for actuator placement selection and optimal force computation. Specifically, our methodology employs the Dueling Double Deep Q-Learning (D3QN) Algorithm to refine the decision-making capabilities of sequential actuator placements. We formulate the actuator selection problem as a submodular function optimization problem, which allows us to leverage the properties of sub-modularity to achieve near-optimal solutions efficiently. The proposed methodology has been comprehensively evaluated through numerical case studies and comparisons with benchmarks, demonstrating its effectiveness and outstanding performance in enhancing assembly precision.
Keywords
- Industrial Optimization
- Practice of OR
- Analytics and Data Science
Status: accepted
Back to the list of papers