Data-Driven Flight Load Prediction using Modal Decomposition Techniques
This paper proposes a novel application of modal decomposition techniques for the prediction of flight section loads from strain gauge measurements. The number of used strain sensors is relatively small and simulates a typical Structure Health Monitoring (SHM) application in flying vehicles. This makes the usage of commonly used modal decomposition techniques challenging, and the Modal Frequency Decomposition and Reconstruction (MFDR) is used for the decomposition. The decomposition relies heavily on the availability of suitable mode shapes. The proposed process is data-driven and does not require any model of the structure or ground vibration test. Suitable mode shapes are derived by analysing a large set of flight tests and building a database of flight modes for the decomposition process. It is shown that modal decomposition can be used to derive an estimate for section loads.
