Michael J. Roan

An application of principal component analysis for lower body kinematics between loaded and unloaded walking.

Load carriage is a very common daily activity at home and in the workplace. Generally, the load is in the form of an external load carried by an individual, it could also be the excessive body mass carried by an overweight individual. To quantify the effects of carrying extra weight, whether in the form of an external load or excess body mass, motion capture data were generated for a diverse subject set. This consisted of twenty-three subjects generating one hundred fifteen trials for each loading condition. This study applied principal component analysis (PCA) to motion capture data in order to analyze the lower body gait patterns for four loading conditions: normal weight unloaded, normal weight loaded, overweight unloaded and overweight loaded. PCA has been shown to be a powerful tool for analyzing complex gait data. In this analysis, it is shown that in order to quantify the effects of external loads and/or for both normal weight and overweight subjects, the first principal component (PC1) is needed. For the work in this paper, PCs were generated from lower body joint angle data. The PC1 of the hip angle and PC1 of the ankle angle are shown to be an indicator of external load and BMI effects on temporal gait data.

The effect of evenly distributed load carrying on lower body gait dynamics for normal weight and overweight subjects

The carrying of extra weight can cause significant injuries. This extra weight can be in the form of an external load carried by an individual or excessive body weight carried by an overweight individual. This study attempts to define the differences in lower body gait patterns caused by either external load carriage, excessive body weight, or a combination of both. Twenty-three subjects generated 115 trials of motion capture data for each loading condition. Path lengths of the phase portrait and the ranges of joint motions (hip, knee and ankle) were used to quantify subgroup differences. The study found significant gait differences due to external load carriage and excessive body weight. Within each class of normal weight and overweight subjects, differences were found in the hip and ankle path lengths when a subject carried an evenly distributed external load. This implies that these joints may be more prone to injury due to external load carriage.

Cooperative localization of an acoustic source using towed hydrophone arrays

We describe field experiments in which a team of autonomous underwater vehicles cooperatively localize an acoustic source. The team implements a data fusion algorithm to enhance the localization performance of each individual vehicle and implements a decentralized motion control algorithm so that each vehicle maneuvers to minimize the joint localization error of the acoustic source. Each autonomous underwater vehicle is equipped with a custom-designed towed hydrophone array that measures the bearing angle between the array and the acoustic source. The noise statistics of the hydrophone arrays are state-dependent, and a generalized Kalman filter that accounts for the state-dependant measurement noise is utilized for localization.

Gait analysis to classify external load conditions using linear discriminant analysis

There are many instances where it is desirable to determine, at a distance, whether a subject is carrying a hidden load. Automated detection systems based on gait analysis have been proposed to detect subjects that carry hidden loads. However, very little baseline gait kinematic analysis has been performed to determine the load carriage effect while ambulating with evenly distributed (front to back) loads on human gait. The work in this paper establishes, via high resolution motion capture trials, the baseline separability of load carriage conditions into loaded and unloaded categories using several standard lower body kinematic parameters. A total of 23 participants (19 for training and 4 for testing) were studied. Satisfactory classification of participants into the correct loading condition was achieved by employing linear discriminant analysis (LDA). Six lower body kinematic parameters including ranges of motion and path lengths from the phase portraits were used to train the LDA to discriminate loaded and unloaded walking conditions. Baseline performance from 4 participants who were not included in training data sets show that the use of LDA provides a 92.5% correct classification over two loaded and unloaded walking conditions. The results suggest that there are gait pattern changes due to external loads, and LDA could be applied successfully to classify the gait patterns with an unknown load condition.

Adaptive near-field beamforming techniques for sound source imaging

Phased array signal processing techniques such as beamforming have a long history in applications such as sonar for detection and localization of far-field sound sources. Two sometimes competing challenges arise in any type of spatial processing; these are to minimize contributions from directions other than the look direction and minimize the width of the main lobe. To tackle this problem a large body of work has been devoted to the development of adaptive procedures that attempt to minimize side lobe contributions to the spatial processor output. In this paper, two adaptive beamforming procedures-minimum variance distorsionless response and weight optimization to minimize maximum side lobes--are modified for use in source visualization applications to estimate beamforming pressure and intensity using near-field pressure measurements. These adaptive techniques are compared to a fixed near-field focusing technique (both techniques use near-field beamforming weightings focusing at source locations estimated based on spherical wave array manifold vectors with spatial windows). Sound source resolution accuracies of near-field imaging procedures with different weighting strategies are compared using numerical simulations both in anechoic and reverberant environments with random measurement noise. Also, experimental results are given for near-field sound pressure measurements of an enclosed loudspeaker.

Performance Bounds for Multisource Parameter Estimation Using a Multiarray Network

Networked sensors are increasingly being used to perform tasks such as detection, source localization, and tracking. It is intuitive to expect a performance increase by completing these tasks with networked arrays. Quantifying the increase in performance requires a generalized analysis, which is typically done using the Cramer-Rao bounds (CRB). Previously, the CRB for multisource-single-array and multiarray-single-source models were derived, but not the general multiarray-multisource case. In this paper, the general case is modeled and is shown to reduce to previously published cases. The model and CRB derived in this paper serve as a benchmark to which current and future research in multiarray, multisource parameter estimation algorithms can be compared. It is proven that the CRB of a scalar parameter in a model containing K + 1 sources is always higher than the CRB of K sources. In addition, using numerical analysis, it is shown that adding sources to a multiarray model has unique effects that cannot be predicted from less general models. The number of sources and sensors, the geometry of the model (i.e., source and sensor locations), and the source and noise power levels all affect the CRB. The effect of changing model parameters is shown for several multiarray-multisource examples when the estimated parameter is source location.

A comparison of near-field beamforming and acoustical holography for sound source visualization

Abstract Acoustical holography procedures make high-resolution visualization possible via estimation of the sound intensity on surfaces closer to the sources than the near-field measurement surface. Another source localization technique, beamforming, has been used to estimate the direction of arrival of sound from sources that typically lie in the far-field. However, little work has been done using beamforming as a visualization technique based on near-field measurements. As a result, the performance of beamforming and acoustical holography in terms of source resolution capabilities has not been directly compared when using near-field measurements. In this work, point source beamforming was used to visualize sources based on near-field measurements. Acoustic intensity estimated from beamformed pressure measurements was compared with the absolute intensity estimated using acoustical holography techniques. In addition to noise-free, anechoic simulations, cases of measurement pressure with random noise were generated and used to compare source resolution accuracy of acoustical holography and beamforming techniques in the presence of measurement noise. It was found that intensity estimated using acoustical holography provided the clearest image of sources when the measurement surface was conformal with the source geometry. However, sources can be resolved more accurately using near-field beamforming than acoustical holography at high frequencies when the sources are not located perfectly on a surface conformal with the measurement geometry.

A graph theoretical approach toward a switched feedback controller for pursuit-evasion scenarios

This research introduces a novel method for constructing a switched feedback control system to be used for an autonomous agent. The state space is partitioned into sets of states where a specific control is applied. Each partition is represented by nodes of a digraph where the success of the control in traversing the partitions is represented by a connecting edge. Using the concept of capture sets in the field of differential games, it is shown that a set of states included in a particular partition is capable of reaching the target set if the eigenvalues of the adjacency matrix representing the digraph are all zero and none of the partitions are invariant. The advantage of this method is that it is possible to assign finite horizon controls to each partition that are easier to calculate than infinite horizon methods, but still maintain the infinite horizon guarantee of reaching the target. An example is given to illustrate the implementation of the proposed controller.

Deconvolution and signal extraction in geophysics and acoustics

Deconvolution and signal extraction are fundamental signal processing techniques in geophysics and acoustics. An introductory overview of the standard second‐order methods and minimum entropy deconvolution is presented. Limitations of the second‐order methods are discussed and the need for more general methods is established. The minimum entropy deconvolution (MED), as proposed by Wiggins in 1977, is a technique for the deconvolution of seismic signals that overcomes limitations of the second‐order method of deconvolution. The unifying conceptual framework MED, as presented in the Donoho’s classical paper (1981) is discussed. The basic assumption of MED is that input signals to the forward filter are independent, identically distributed non‐Gaussian random processes. A forward convolution filter ‘‘makes’’ the output of the forward filter more Gaussian which increases its entropy. The minimization of entropy restores the original non‐Gaussian input. We also give an overview of recent developments in blind ...

Wavelet transform techniques for time varying propagation and scattering characterization

It is shown that linear time varying (LTV) propagation and scattering channels are efficiently characterized by wideband spreading functions that can be estimated using continuous wavelet transforms (CWT). We show that the CWT domain characterization of LTV propagation channels leads naturally to a wavelet domain implementation of maximum likelihood detectors of multi-highlight objects whose echos are distributed in range and time-scale.