Neural dynamic trajectory design for reentry vehicles : preprint /

The next generation of reentry vehicles is envisioned to have onboard autonomous capability of real-time trajectory planning to provide capability of responsive launch and delivering payload anywhere with precise flight termination. This capability is also desired to overcome, if possible, in-flight vehicle damage or control effector failure resulting in degraded vehicle performance. An aerial vehicle is modeled as a nonlinear multi-input-multi-output (MIMO) system. An ideal optimal trajectory control design system generates a series of control commands to achieve a desired trajectory under various disturbances and vehicle model uncertainties including aerodynamic perturbations caused by geometric damage to the vehicle. Conventional approaches suffer from the nonlinearity of the MIMO system, and the high-dimensionality of the system state space. In this paper, we apply a Neural Dynamic Optimization (NDO) based approach to overcome these difficulties. The core of an NDO model is a multilayer perceptron (MLP) neural network, which generates the control parameters online. The advantage of the NDO system is that it is very fast and gives the trajectory almost instantaneously. The bulk of the time consuming computation is required only during off-line training. The inputs of the MLP are the time-variant states of the MIMO systems. The outputs of the MLP are the near optimal control parameters.Report prepared in part by Knowledge Based Systems, Inc., College Station, Texas.Photocopy."Conference Paper Preprint: June 15, 2005 -- July 23, 2007.""July 2007."Includes bibliographic references (pages 10-11).The next generation of reentry vehicles is envisioned to have onboard autonomous capability of real-time trajectory planning to provide capability of responsive launch and delivering payload anywhere with precise flight termination. This capability is also desired to overcome, if possible, in-flight vehicle damage or control effector failure resulting in degraded vehicle performance. An aerial vehicle is modeled as a nonlinear multi-input-multi-output (MIMO) system. An ideal optimal trajectory control design system generates a series of control commands to achieve a desired trajectory under various disturbances and vehicle model uncertainties including aerodynamic perturbations caused by geometric damage to the vehicle. Conventional approaches suffer from the nonlinearity of the MIMO system, and the high-dimensionality of the system state space. In this paper, we apply a Neural Dynamic Optimization (NDO) based approach to overcome these difficulties. The core of an NDO model is a multilayer perceptron (MLP) neural network, which generates the control parameters online. The advantage of the NDO system is that it is very fast and gives the trajectory almost instantaneously. The bulk of the time consuming computation is required only during off-line training. The inputs of the MLP are the time-variant states of the MIMO systems. The outputs of the MLP are the near optimal control parameters.Mode of access: Internet