Research on multi-source optical measurement–driven online evaluation of flexible assembly deviation for aircraft panels (<italic>inner cover paper·invited</italic>)

ObjectiveFlexible assembly has become a development trend for compliant assembly technologies of weak-stiffness aircraft panels. Online acquisition of the actual state during the assembly process and evaluation of assembly deviation are necessary prerequisites for realizing feedback control in flexible assembly. However, during the assembly process, weak-stiffness panels are easily affected by positioning constraints, gravity, and clamping loads, resulting in time-varying and random deformations. Existing methods have difficulty in acquiring the global surface shape state online, which makes online evaluation of assembly deviation challenging. Therefore, it is urgent to develop an online evaluation method for flexible assembly deviation.MethodsA multi-source optical measurement–driven online evaluation method for assembly deviation of aircraft panels is developed. First, a laser–vision–fiber collaborative multi-source optical measurement system is constructed to acquire the local displacement and strain information of the panel and its assembly datum components in real time under a unified reference coordinate system. By combining physical models, high-reliability real-time acquisition of the global geometric state of the panel and datum components is achieved. Second, geometric information of the contact surfaces between the panel and the assembly datum components is extracted from the optical sensing results, and an assembly deviation representation and propagation model is established, enabling the mapping of deviations from assembly interfaces to assembly process targets. By incorporating assembly process requirements, online evaluation of flexible assembly deviations, such as assembly gap deviation and panel profile tolerance deviation, is realized. Finally, a scaled experimental platform for aircraft horizontal tail panel assembly is built, and experimental validation is conducted to verify the accuracy and effectiveness of the proposed method.Results and DiscussionsThe experimental results show that during the gradual application of assembly loads, an optical measurement–based sensing approach for acquiring the actual state during the part assembly process is capable of reconstructing the displacement field of the front spar component, with a relative error of less than 2.86% (Tab.1) and a computation time of 0.005 s, satisfying the accuracy and real-time requirements for online displacement measurement of panel assembly datum components. The panel surface shape can be accurately reconstructed (Fig.9). A locally larger relative error appears at the lower assembly hole region of the panel; this error is not caused by sensing inaccuracies, but results from interpolation during the visualization of the continuous surface cloud map. Since the data at the assembly hole region are not included in error statistics and subsequent deviation evaluation, this visualization artifact does not affect the accuracy calculation or the validity of the conclusions. Compared with the reference values, the panel surface shape sensing results obtained by the proposed method exhibit an average absolute error of 0.627 mm, a maximum relative error of 19.87%, an average relative error of 7.47%, and a computation time of 0.009 s, which satisfy the accuracy and real-time requirements for online panel surface shape measurement during the assembly process. For assembly gap deviation evaluation, the proposed assembly deviation prediction method achieves a maximum absolute error of 0.881 mm and an average absolute error of 0.393 mm (Tab.2), which is better than one third of the maximum assembly gap measured during the experiments, with a computation time of 0.295 s, meeting engineering requirements for assembly gap evaluation. For panel profile tolerance deviation evaluation, the absolute error is 0.132 mm and the relative error is 2.38% (Tab.3), with a computation time of 0.051 s.ConclusionsThe requirements of deviation evaluation in the flexible assembly process of aircraft panels are addressed through a multi-source optical measurement–driven online evaluation method. The deformation sensing accuracy of assembly datum components is better than 2.86%, and the surface shape sensing accuracy of the panel is better than 0.627 mm with a computation time of 0.009 s. The average absolute error of assembly gap deviation evaluation is 0.393 mm with a computation time of 0.295 s, and the absolute error of panel profile tolerance deviation evaluation is 0.132 mm with a computation time of 0.051 s. The proposed method enables online evaluation of flexible assembly deviation of aircraft panels and provides necessary data support for execution control of flexible assembly of aircraft panels.