Aircraft Fuselage Subassembly Positioning with Neural Network Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current aircraft fuselage assembly methods require manual iterative processes for aligning large-format subassembly components, leading to increased assembly times and potential deformations due to external mechanical forces, which complicates achieving the precise target geometry needed for joining.
Innovation Solution
A device equipped with neural networks and open-loop/closed-loop control systems allows for automatic and precise positioning of subassembly components in three-dimensional space, using measuring devices like laser trackers or photogrammetric systems to achieve a predetermined target geometry, enabling simultaneous movement of positioning devices and reducing the regulatory effort required for alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Force
If manual iterative positioning processes are used to align subassembly components, then the forces acting on components can be limited, but the assembly time increases significantly
Solution Approach 1:
The patent replaces manual mechanical positioning operations with an automated control system that uses measurement data to directly control positioning devices. The control system automatically calculates target positions and controls multiple positioning devices simultaneously, eliminating the need for manual iterative adjustments while reducing assembly time.
Solution Approach 2:
The system performs self-positioning by automatically measuring the actual positions of subassembly components, calculating target positions based on predetermined target geometry, and controlling positioning devices to achieve the target positions without manual intervention. The control system serves itself by integrating measurement, calculation, and control functions.
2Productivity
If multiple positioning devices are moved simultaneously to reduce assembly time, then productivity increases, but the risk of causing undefined deformations increases
Solution Approach 1:
The patent uses measurement devices to continuously measure the actual positions of subassembly components during the positioning process. The control system receives this measurement data and uses it to calculate target positions, creating a closed-loop feedback system that ensures accurate positioning even when multiple devices move simultaneously.
Solution Approach 2:
The control system calculates target positions for all positioning devices based on predetermined target geometry before the positioning operation begins. This preliminary calculation ensures that all devices move to their target positions in a coordinated manner, preventing deformations while maintaining high productivity.
3Manufacturing precision
If complex non-linear mathematical algorithms are used to describe the geometry of picked-up components, then positioning precision can be achieved, but the device complexity increases
Solution Approach 1:
The patent replaces complex manual calculations and manual positioning operations with an automated control system that uses measurement data to directly control positioning devices. The control system handles the complex mathematical relationships between measurement data and target positions automatically, reducing the operational complexity despite the mathematical sophistication required.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a device for spatially orienting at least two large-size subgroup components, in particular at least one side shell (7,8), at least one upper shell (12), at least one lower shell (11), and/or at least one floor framework, in relation to each other to integrate a component, in particular a fuselage section of an aircraft, comprising: a) at least two positioning apparatuses for accommodating one subgroup component each, in particular at least two side shell positioners (2,3), at least one upper shell positioner (5), and/or at least one lower shell positioner (4), b) at least one measuring apparatus for detecting a plurality of measurement data, in particular position data of the subgroup components and/or of the positioning apparatuses, c) at least one control and/or regulating apparatus (19), in particular at least one CNC controller, and d) at least one neural network (20). The neural network (20) integrated according to the invention into the device allows the subgroup components to be brought into a specified target geometry by means of simultaneous movement of the positioning apparatuses in the shortest amount of time and with high accuracy. The control and/or regulating apparatus (19) is used preferably to subsequently orient the subgroup components in relation to each other by means of simultaneously executed linear movements of the positioners. The invention further relates to a method for orienting subgroup components.