Trained Network Force Measurement System for Aircraft/Store Interfaces

This proposal addresses the development of a new methodology to predict reaction forces in weapon store connections of naval aircraft. The proposed methodology utilizes a strain gage-based measurement technique in which a series of sensors are calibrated with a set of known loading configurations. The sensitivity matrix relating the measured strains to the loads forms the core of a system that can be thought of as a neural network. The efficacy of this network will improve as more strain gages are added and more calibration load cases are applied. System identification and singular value decomposition procedures will be used in the training of the network. In addition, optimization routines will be adopted for strain gage placement in order to improve the strain-to-load mapping. The feasibility of the proposed technique on a simple but representative test specimen will be demonstrated. It is anticipated that system nonlinearities will pose challenges to the proposed methodology; hence the test specimen will be designed to incorporate this effect to provide realistic performance estimates. Included in the final report for Phase I will be a description of a streamlined version - hardware and software - of the Phase II interface-load measurement system.