A comprehensive design procedure based on extremum seeking for minimum power demand formation ight is presented, the rst with performance guarantees. The procedure involves the design of a new wake robust formation hold autopilot and transformation of the closed-loop aircraft dynamics to a form in which a newly available rigorous design procedure for extremum seeking is applicable. The design procedure is applied to a formation of Lockheed C-5s, extending the use of maximum performance formation ight to large transports. By the use of available experimental wake data of the C-5, a model of the aircraft in the wake is developed that models aerodynamic interference as feedback nonlinearities. Thus, this work is also the rst to attain stable extremum seeking for a plant with nonlinear feedback. Optimal formation ight is attained by online minimization of an easily measurable objective, the pitch angle of the wingman.

No abstract

We present a comprehensive design procedure based on extremum seeking for minimum power demand formation flight, the first with performance guarantees. The procedure involves the design of a new wake robust formation hold autopilot, and transformation of the closed loop aircraft dynamics to a form in which a newly available rigorous design procedure for extremum seeking is applicable. We apply the design procedure on a formation of Lockheed C-5s, extending the use of maximum performance formation flight to large transports. Using available experimental wake data of the C-5, we develop a model of the aircraft in the wake that models aerodynamic interference as feedback nonlinearities. Thus, our work is also the first to attain stable extremum seeking for a plant with nonlinear feedback. Optimal formation flight is attained by online minimization of an easily measurable objective, the pitch angle of the wingman. Motivation and P r i o r WorkBy flying in formation, two aircraft can achieve a significant reduction in power demand (up to 20%) through aerodynamic interference, by the wingman riding upon the upwash field of the leader, like a glider in a thermal. Given the potential payoffs, and the availability of enabling avionics and control algorithms, the problem of attaining the configuration for minimum power demand through automatic control of the wingman has lately been a subject of intense interest [3, 5, 81. Given the high sensitivity of formation benefits to positioning error (an error of just 10% of the aircraft wingspan can reduce the benefits by half [2]), and the uncertainty in aerodynamic interference modeling, i.e., the need for accurate steady state performance in the presence of modeling uncertainty, there is a definite need for adaptive feedback control. The wingman control system is based on a formation-hold autopilot', which is fed an estimate of the optimal separation generated by an adaptive feedback control scheme. The nature of the problem has led to the use of extremum seeking algorithms [3, 81. In [SI, a simple discrete time extremum seeking algorithm to maximize aileron deflection was used to attain a power demand reduction of 20% in experimental flight tests of two Dornier aircraft in formation. In [3], simulation studies of a continuous time extremum seeking algorithm to maximize induced lift were presented. A systematic design procedure was absent in both.' This work was supported in part by grants from AFOSR, Corresponding author. Email: kariyur@mae.ucsd.edu. ONR, and NSF' An autopilot capable of tracking relative position (with respect to the leader) reference signals. [3,8], and supplies a generally applicable, com;prehensiue design procedure for minimum power demand fo:mation flight with performance guarantees, and an easily measurable objective for eztremum seeking. This goal is attained through the following steps: This work solves the problems left open by1. 2. 3. 4. 5.Modeling of aerodynamic interference as a multiple feedback nonlinearity in the aircraft dynamics.Linking po...

An autonomous control system of a rover for planetary exploration can be subdivided in two distinct parts, the first devoted to planning and scheduling and the second to the plan execution. In this paper the concept of a hybrid intelligent system composed by three layers will be presented, focusing on the plan execution part which is embodied by a Behavioral and a Reflexive layer. The plan generated by the Planning layer is composed by a sequence of logical actions that can not be directly executed by the rover hardware. The Behavioral layer translates these actions in executable actions through a Fuzzy subsystem, and prepares the input for the Reflexive layer. This layer is in charge of the low level control of the rover hardware which is performed by an Artificial Neural Network. In this paper, the whole Execution system will be presented, focusing on the path following problem and reporting the results of the first tests executed on a real prototype

In recent years, microsatellites have attracted great attention for their low cost, rapid development, and utilization capability. Demand for more requirements is increasing, and reliable attitude control is indispensable for secure operation. An Earth sensor is typically used to detect the direction of Earth's center relative to the spacecraft's attitude. Unlike Sun sensors, an Earth sensor can be functional even in the eclipse region in orbit, which enables a spacecraft to determine its attitude regardless of orbital position. The thermopile-based Earth sensor developed in this research is designed to have full sky coverage, i.e., it is operational independent of satellite attitude. Therefore, it can be used in safe-hold-mode operation. As a result of this research, an Earth sensor is realized by utilizing multiple thermopile sensors. The geocentric vector can be determined from the output voltages of thermopiles that are mounted on the entire satellite in a distributed manner and pointed in different directions. In order to establish a data processing algorithm, the output voltage and angle characteristics of the sensors were investigated analytically and experimentally. A numerical model of the sensor was developed, and its performance in orbit was evaluated in a software-based simulation and verification environment.