Nobuhiro Yokoyama

Modified Genetic Algorithm for Constrained Trajectory Optimization

Although direct methods using sequential quadratic programming, such as direct collocation, have been widely applied to trajectory optimization problems, optimization results for these methods are sensitive to initial solutions in some cases. For the purpose of finding an appropriate initial solution to gradient-based direct trajectory optimization, a numerical trajectory optimization method using a real-coded genetic algorithm (GA) is considered. Because a typical GA solves unconstrained optimization problems, the penalty function has usually been used as the selection criterion of surviving individuals. However, the convergence properties of this conventional method are sensitive to the penalty parameter. Therefore, a new method that robustly achieves global optimization of the objective function and the feasibility search even with a large penalty parameter is proposed. The proposed method is applied to the constrained brachistochrone problem and a space plane reentry problem. Results indicate promising performance of the proposed method in providing appropriate initial solutions for gradient-based trajectory optimization.

Multidisciplinary Design Optimization of Space Plane Considering Rigid Body Characteristics

DOI: 10.2514/1.19969 Multidisciplinary designoptimization isanimportant approachfor the conceptualdesignof space planesbecause these planes are characterized by disciplines that interact with one another. A multidisciplinary design optimization problemof asingle-stage-to-orbit space plane isformulated andsolved inthis study.Themodeling andoptimization ofrigidbodycharacteristics,suchastrimandstability,arefocusedonbecauseasingle-stage-to-orbitspaceplanehas a tendency for considerable shift of both the aerodynamic center and the center of gravity. Moreover, the design of the air-breathing engines are integrated with the airframe and its effect on the rigid body characteristics are also modeled in the framework of multidisciplinary design optimization. Using the all-at-once-based multidisciplinary design optimization approach, which incorporates sparse nonlinear programming and metamodeling, the design of the vehicle and its flight trajectory are successfully optimized. Finally, the characteristics of the optimal solution are investigated, especially the relationships among the airframe-engine integration, rigid body characteristics, and payload transportation capability.

Path Planning Algorithms for Skid-to-Turn Unmanned Aerial Vehicles

This study describes two types of algorithms for skid-to-turn unmanned aerial vehicles to plan paths between two waypoints under constant wind conditions. The first type of algorithm is a rigorous optimization algorithm based on the Euler-Lagrange formulation with analytical integration of the path. The second type of algorithm is a fast algorithm describing the path by two circular arcs connected by a line segment or another circular arc in the air mass frame, which is similar to the Dubins path. The latter algorithm is developed for actual airborne application, whereas the former algorithm is developed to check the quasi-optimality of the path calculated by the latter algorithm. We present a convergence proof of the latter algorithm under certain assumptions and its quasi-optimality in comparison with the former algorithm. Furthermore, the computational efficiency and the convergence reliability of the latter algorithm are demonstrated through numerical examples.

Parameter Estimation of Aircraft Dynamics via Unscented Smoother with Expectation-Maximization Algorithm

This paper proposes a new method for parameter estimation of aircraft dynamics modeled in state space. The developed method employs the smoother, which estimates the state and unknown parameters, combined with expectation-maximization algorithm to estimate unknown statistics in the problem, i.e., the mean and covariance of an initial state, and the noise covariances. To approximate the expectation values in the expectation-maximization with a reasonable computational cost, an unscented transform based on the estimates obtained by the unscented Rauch–Tung–Striebel smoother is employed. Moreover, nonconvex numerical optimization algorithms are not necessaryinthemaximizationstepoftheexpectation-maximization,becausetheoptimumsoftheunknownstatistics are given in analytical forms. Thus, the developed method achieves low computational cost and high robustness. Its effectiveness is demonstrated through two problems of estimating aircraft aerodynamic parameters.

Path Generation Algorithm for Turbulence Avoidance Using Real-Time Optimization

This paper describes a novel algorithm for generating flight paths in real time that avoid turbulent regions detected by an onboard Doppler lidar. Adopting the distance traveled rather than the time as the independent variable in the state equations enables an exact expression to be obtained for the point-mass dynamics and reduces the number of constraints imposed by turbulence avoidance. Moreover, the algorithm estimates the global optimum by applying second-order cone programming relaxation to the original nonconvex problem. Based on this estimate, it then applies convex quadratic programming. This algorithm is well suited for real-time applications since it is guaranteed to converge in polynomial computation time and it is expected to generate practically useful paths. The high computational speed of the algorithm is demonstrated and the physical feasibilities of the calculated paths are analyzed by numerically simulating turbulence avoidance.

Optimal Path Planning for Skid-to-Turn Unmanned Aerial Vehicle

Skid-to-turn is an effective way of turn for small unmanned aerial vehicles (UAVs) because of stable GPS receptions, data link receptions/transmissions, and easiness in target observation by gimbaled camera. This paper describes two types of path planning algorithms for the skid-to-turn UAVs, both of which calculate paths between two waypoints under constant wind. The one is a rigorous optimization algorithm based on the EulerLagrange formulation with analytical integration of the path. The other is a fast algorithm describing the path by two circular arcs connected with a straight segment or another circular arc in the air frame, which is similar to the Dubins path. The latter algorithm is developed for actual airborne application, whereas the former algorithm is developed to check the quasi-optimality of the path calculated by the latter algorithm. We present a convergence proof of the latter algorithm under some assumptions and its quasi-optimality in comparison with the former algorithm. Furthermore, the computational efficiency and the convergence reliability of the latter algorithm are demonstrated through numerical examples.

A Study on Online Flight Trajectory Search and its Flight Simulator Testing

[Abstract] Real -time flight trajectory optimization algorithms are developed and evaluated through numerical simulation and flight simulator testing. The purpose of this study is to guide and control an aircraft in emergency landing. Though our final goal is to develop an automatic control system for tracking flight trajectories generated with the online optimization, this study verifies the validity of the generated optimal trajectories with manual flight control. This paper considers the real-time direct trajectory optimization method that can deal with constraints more strictly.

Inference of Aircraft Intent via Inverse Optimal Control Including Second-Order Optimality Condition

This paper presents a method to infer the intent of an aircraft by using inverse optimal control. Assuming that the observed or estimated trajectory of the target aircraft is the approximate local optimum, the current method calculates the weight of each term of the objective function that represents the flight’s intent. The intent may include direct travel to a waypoint; circling around a waypoint; holding airspeed/heading/turn rate; avoiding certain regions such as those with adverse weather or special-use airspaces; and resolving conflicts with other aircraft. By incorporating the second-order optimality condition (i.e., the positive definiteness of the projected Hessian of the Lagrangian), the method approximately satisfies the sufficiency of the local optimality of the given trajectory. Moreover, the method also guarantees the uniqueness of the inferred intent. The effectiveness of the current method in terms of the reasonableness of the inferred intent as well as real-time applicability is demonstra...