No abstract

No abstract

Flight models are used to accurately predict the aircraft performance and response and are used for aircraft development, engineering analysis, and pilot training. It is common practice to use linearized small perturbation equations of motion for the identification of the stability and control derivatives that form the basis of flight models using system identification. Due in large part to advances in computer technology and optimization techniques, it is feasible to use nonlinear equations of motion for time-domain system identification. This report compares two full flight envelope aircraft models that were developed with identical data: one developed using linearized equations of motion in a state space form, and the other with nonlinear equations. The global flight models were developed for the NRC Bell 412 helicopter in forward flight across its range of speeds, altitudes, and configurations. Use of the nonlinear equations of motion produced a model with less parameter variance and its corresponding global model had improved trim characteristics.

This paper proposes the use of pseudospectral methods to solve the nonlinear trajectory planning problem for freefloating robots. Specifically, three different optimization tools are analyzed. Using each tool, simulations are performed, and it is shown that each solver is capable of finding a deployment trajectory that minimizes the final attitude of a free-floating robot. Each solution is then validated using Pontryagin's minimum principle and Bellman's principle of optimality, as well as by propagating the control torques using a numerical integrator and the dynamics model. Experimental validation is performed at Carleton University's Spacecraft Robotics and Control Laboratory to further investigate the solutions obtained from each tool. Ultimately, it was determined that all solutions resulted in a reduced attitude disturbance at the end of the robotic deployment maneuver.

The authors have developed a novel physics-based nonlinear autoregressive exogeneous neural network model architecture for flight modelling across the entire flight envelope, called FlyNet. When using traditional parameter estimation and output-error methods, aircraft models are captured about a single point in the flight envelope using a first-order Taylor series to approximate forces and moments. To enable analysis throughout the entire flight envelope, the traditional models can be extended across the entire flight envelope by interpolating or stitching between a number of these single-condition models. This paper completes the evolutionary next step in aircraft modelling to consider all second-order Taylor series terms instead of a subset of those and by exploiting the ability of neural networks to capture more complex and nonlinear behaviour for the efficient development of a continuous flight simulation model valid across the entire flight envelope. This method is valid for fixed- and rotary-wing aircraft. The behaviour of a conventional model is compared to FlyNet using flight test data collected from the National Research Council of Canada’s Bell 412HP in forward flight.