Design of a distributed holographic interferometric sensor for measuring the surface displacement of a large segmented reflector is proposed. The reflector's surface is illuminated by laser light of two wavelengths and volume holographic gratings are formed in photorefractive crystals of the wavefront returned from the surface. The sensor is based on holographic contouring with a multiple frequency source. It is shown that the most stringent requirement of temporal stability affects both the temporal resolution and the dynamic range. Principal factor which limit the sensor performance include the response time of photorefractive crystal, laser power required to write a hologram, and the size of photorefractive crystal.

This paper investigates the use of an adaptive system to stabilize a launch vehicle with sloshable fuel. The adaptive system utilizes a laser sensor to measure fluid displacement in the tank, a recursive-least-squares estimator to estimate the natural frequency of the sloshing fuel, and a linear filter which stabilizes the unstable fuel slosh using the frequency estimate. The method was successful for a single slosh-pendulum model and requires no tank baffles or partitions.

The measurement of launch day winds for expendable launch vehicle (ELV) missions is performed to reduce the aerodynamic loads on the vehicle during the atmospheric ascent phase. To accomplish this, wind measurements are performed prior to launch to determine a steering profile that minimizes vehicle loading while still assuring mission success and range safety. However, the current methods of Jimspheres and Rawinsondes to measure launch day winds are time consuming and prone to inaccuracies. The common problem is the time required for the balloons to rise. Typically, Jimsphere data requires sixty minutes to provide a profile to 20 km (approx. 60,000 ft) and Rawinsondes slightly longer to reach 24 km (approx. 80,000 ft). Consequently, the collected wind profile may not be an accurate representation of the real-time wind profile. This uncertainty degrades the estimation of wind persistency and identification of wind gusts. Moreover, the balloons are subject to wind field drift and may not measure the actual winds the vehicle will fly through. As a result, a lidar wind profiler is being examined as an alternate method. In order to assess the utility of a lidar wind profiling system to support day of launch ELV activity, studies have been performed to identify the operational requirements a lidar system must satisfy. Studies were directed towards determining the lidar range, resolution, and accuracy requirements. Operational requirements with respect to weather conditions and launch availability are also considered. Simulation studies have also been performed to assess the effects of having more current wind data.

An optical velocity sensor, based on the sheet-pair transit-time technology, was designed, built to flightworthy standards and test flown on an F-16B to 50,000 ft and Mach 1.2 as part of an opto-avionic air data system. Brief descriptions of the work leading to 24 flights in January and February 1990 are given, with examples of data and discussions of experiences. Compared with conventional pneumatic sensors the system offers advantages that include potential improvements in accuracy, latency, calibration, dynamic range of speed and attitude, robustness, and possibly life, cost, and range of application, without modification of the vehicle skin contour. Measurements corresponding well with the aircraft system were obtained under all conditions except heavy cloud, which demands small design changes for future systems. The importance of modeling for software and hardware design optimization is stressed and measurements are presented.

An aircraft-based multisensor system consisting of forward-looking and down-looking sensor suites has been developed. Sensors which have been evaluated in this system include CO2 and near-IR laser radars, long-wave passive IR extending to 12 micrometers , and an 85.5 GHz millimeter wave (MMW) radar. Sensor descriptions are presented along with example imagery. The goals of this continual effort are deeper understanding of the data characteristics, including realistic phenomenology, and the delivery of co-registered, multi- dimensional sensor data for the development of detection and classification algorithms.

On-line discrimination between arbitrary petroleum products is necessary for optimal control of petroleum refinery and pipeline operation and process control involving petroleum distillates. There are a number of techniques by which petroleum products can be distinguished from one another. Among these, optical measurements offer fast, non-intrusive, real-time characterization. The application examined here involves optically monitoring the interface between dissimilar batches of fluids in a gasoline pipeline. After examination of near- infrared and mid-infrared absorption spectroscopy and Raman spectroscopy, Fourier transform mid-infrared (FTIR) spectroscopy was chosen as the best candidate for implementation. On- line FTIR data is presented, verifying the applicability of the technique for batch interface detection.

Photodiode arrays used in laser diffraction particle sizing instruments must be calibrated to account for detector-to-detector variations in sensitivity. We have calibrated an Insitec EPCS-P ring detector (planar-diffused, p on n silicon photodiode array) by scanning a focussed laser beam across the detector surface. A deconvolution of the known intensity distribution of the laser beam from the measured signal resulted in detector response as a function of position. Detector response was approximately constant over the region of the ideal active detector and it decreased exponentially in the region beyond the ideal detector boundary. A diffusion length constant of 50 microns gave the best fit to the measured data. Theoretical predictions of calibration factors based on measured detector response agreed reasonably well with Malvem and Insitec calibration factors obtained from the traditional uniform light illumination method. This indicates that edge effects in different ring detectors are similar.

Experimental results for the detection of the navigable space surrounding a mobile vehicle using a compact optical 3-D range sensor are presented. A first prototype of the optical sensor, which is based on the use of a double aperture mask in place of the diaphragm of a standard camera lens to acquire panoramic images of a room, is shown. Preliminary experiments concerning the integration of multiple acquisitions and creation of a map of the immediate free space environment around the vehicle, when the sensor is used in an exploratory mode, are presented.

An approach to segmentation of laser radar range images is presented which is based on a terrain variation model. Topics discussed include the laser radar system with low and high drop-out probabilities, statistical methods for target/nontarget classification and terrain segmentation, and the resulting segmentation with respect to range and drop-out probabilities.

This paper describes research into the use of discrete piezoelectric sensors and actuators for active modal control of flexible two-dimensional structures such as might be used as components for spacecraft. A dynamic coupling term is defined between the sensor/actuator and the structure in terms of structural model shapes, location and piezoelectric behavior. The relative size of the coupling term determines sensor/actuator placement. Results are shown for a clamped square plate and for a large antenna. An experiment was performed on a thin foot-square plate clamped on all sides. Sizable vibration control was achieved for first, second/third (degenerate) and fourth modes.

Modal domain optical fiber sensors, or distributed effect sensors, for active vibration suppression in flexible structures are considered. Preliminary modeling results indicate that these sensors can be used to sense vibrations in a flexible beam and the signal can be used to damp vibrations in the beam. Weighted distributed-effect sensors can be used to implement high order compensators with low order functional observers.

Technology is being developed to process signals from distributed sensors using distributed computations. These distributed sensors provide a new feedback capability for vibration control that has not been exploited. Additionally, the sensors proposed are of an optical and distributed nature and could be employed with known techniques of distributed optical computation (Fourier optics, etc.) to accomplish the control system functions of filtering and regulation in a distributed computer. This paper reviews a procedure for the analytic design of control systems for this application. For illustration, the procedure is applied to the problem of suppressing the vibrations of a simply supported beam. A simulator has been developed to study the effects of sensor and processing errors. An extensive study of the effects of these errors on estimation and regulation performance is presented.

The exposure of multiple Bragg gratings into a single optical fiber has been demonstrated. As sensors of strain and temperature, these devices promise an architecturally efficient means to conduct quasi-distributed measurements with an array of discrete gages. Such a configuration can have great benefit when embedded within composite structures. This paper will present instrumentation concepts and results from initial laboratory work to develop methods for multiplexed strain and temperature measurements using Bragg grating sensors.