Page 67 - ASME DSCC 2015 Program

Technical Program

a New Vibration controller design using consensus Technique EEG Stochastic nonlinear oscillator Models for Alzheimer’s Disease
Contributed regular paper. DSCC2015-9795 Invited session paper. DSCC2015-9676
Ehsan omidi, nima Mahmoodi, The University of Alabama, Tuscaloosa, AL, Parham Ghorbanian, Center for Nonlinear Dynamics and Control
United States - Villanova Uniersity, Villanova, PA, United States, Subramanian
Ramakrishnan, University of Minnesota, Duluth, Duluth, MN, United States,
This paper discusses the concept of a new methodology for active vibration
hashem ashrafiuon, Villanova University, Villanova, PA, United States
control of flexible structures using consensus control of network systems.
In the new approach, collocated actuation/sensing patches communicate In this article, we derive unique stochastic nonlinear coupled oscillator
with one another through a network with certain directed topology. A virtual models of EEG signals from an Alzheimer’s Disease (AD) study. EEG signals
leader is assigned to enforce the vibration amplitude at the place of each recorded during resting eyes-open (EO) and eyes-closed (EC) conditions
agent to zero. Since the modal states of the system are not available for in a pilot study with AD patients and age-matched healthy control subjects
the vibration control task, individual optimal observers are designed for (CTL) are employed. An optimization scheme is then utilized to match the
each agent first. After describing the controller and examining the stability output of the stochastic Duffing - van der Pol double oscillator network with
of the system, controller performance is verified using a clamped-clamped EEG signals recorded during each condition for AD and CTL subjects by se-
thin aluminum beam. According to the obtained numerical results, the new lecting the model parameters and noise intensity. The selected signal char-
control approach successfully suppresses the vibration amplitudes, while acteristics are power spectral densities in major brain frequency bands and
the consensus design ensures that all agents are synchronized during the Shannon and sample entropies to match the signal information content and
performance. complexity. It is shown that statistically significant unique models represent
the EC and EO conditions for both CTL and AD subjects. It is further shown
InVITED SESSIon that the inclusion of sample entropy in the optimization process significantly
2-11-1 fA2 Biomed and neural Systems enhances the stochastic nonlinear oscillator model performance. The study
George Bellows B 10:00am–12:00pm suggests that EEG signals recorded under different brain states as well as
those belonging to a brain disorder such as Alzheimer’s disease can be
Session Organizer: hashem ashrafiuon, Villanova University uniquely represented by stochastic nonlinear oscillators paving the way for
Session Organizer: Jin-oh Hahn, University of Maryland identification of new discriminants.
Session Organizer: Parham Ghorbanian, Villanova University
optimizing ultrafiltration rate profiles for the estimation of Blood
Session Organizer: Ramin Bighamian, University of Maryland
Volume During Hemodialysis
Session Chair: Yossi Chait, University of Massachusetts
Invited session paper. DSCC2015-9725
Session Co-Chair: hashem ashrafiuon, Villanova University
CV Hollot, Joe Horowitz, Yossi Chait, University of Massachusetts,
An optimization-Based Approach for Prosthesis Dynamic Modeling Amherst, MA, United States, RP Shrestha, Octet Research Inc., Boston,
and parameter Identification MA, United States, Michael J. Germain, Western New England Renal &
Invited session paper. DSCC2015-9637 Transplant Associates, PC, Springfield, MA, United States
Ting Yang, Harbin Institute of Technology, Harbin, China, fen Wu, Managing blood volume levels in end-stage kidney disease (ESRD) patients
Ming liu, Helen (He) Huang, North Carolina State University, Raleigh, NC, undergoing dialysis is an essential component of the overall treatment. Re-
United States moving too much, too little, and/or too fast fluid have been associated with
increased patient morbidity and mortality. Thus, knowledge of a subject’s
In this paper, we propose an effective approach to model the prosthetic
blood volume at the start of hemodialysis is crucial for successful treatment.
leg dynamics for amputees wearing active-transfemoral prosthesis (ATP).
In this paper, we show by analysis how to estimate initial blood volume from
To accommodate unexpected effects of thigh on knee joints, the dynamic
measurements made during hemodialysis. We show that it is difficult, if not
prosthesis model has been derived using both the thigh-knee-shank and
impossible, to estimate blood volume under present clinical practice param-
the knee-shank configurations. Correlated with the amputee’s walking data,
eters. We propose new ultrafiltration profiles that allow for estimation of an
a nonlinear optimization problem is then formulated to identify the model
individual’s initial blood volume.
parameters and the gains of the control input torque for the ATP, while re-
ducing measurement errors of the data. Moreover, the identified models are
validated by comparing the predicted dynamics with experimental measure-
ments. The advantages of proposed method in terms of simplicity, flexibility,
and accuracy are demonstrated by the high correlation coefficients and the
low root-mean-square errors.

67