Support vector regression of a high fidelity helicopter flight model

The traditional technique for the dynamic modelling of helicopters and their systems involves the collection of flight data and aircraft specifications from which physics-based theoretical equations are generated and validated. It is a time consuming process that requires the availability of a significant amount of data. The data required is often proprietary or commercial in confidence. The suggestion of a black box approach using machine learning techniques may provide the answer for a more simplistic method of simulation. This would only require data that is readily available to the owner of the helicopter platform. The application of Support Vector Machines (SVMs) as a machine learning technique for helicopter simulation is chosen for this investigation.

A high fidelity model of a SH-2G(A) Super Seasprite helicopter is initially developed and validated in the FLIGHTLAB simulation environment. This includes a new servo flap component designed using Chopra and Shen’s quasisteady adaption of the Theodorsen model. This flight model provides a basis of noise free flight data from which a number of SVM models are produced to simulate the longitudinal pitch dynamics of a helicopter in hover. Subsequently, the best performing SVM configuration is trained using real flight data and compared to the simulation results.

The SVM results show significant promise in the ability to represent aspects of a helicopter’s dynamics at a high fidelity, provided that the following is established. Firstly, it is important to provide the machine with knowledge of past inputs that encompass the delay characteristics of the helicopter dynamic system. Secondly, the relationship, rather than the mechanics, between the significant variables that represent the dynamic system must be well understood.