G. Ron Chen, Ph.D., Associate Professor, Electrical and Computer Engineering, UH
Figure 1. A WRS block diagram shows part of the water recovery and supply system equipped with the new fuzzy expert system recently implemented on the ALSS computer simulation program at the NASA-JSC.
Researchers seek to develop a fuzzy-logic-based expert system for computer-simulated fault management of water recovey and supply (WRS), which can relieve human operators and perform management functions semi-automatically under regular and irregular conditions. The project is closely related to the Advanced Life Support System (ALSS) currently under development at the NASA-Johnson Space Center.
Under support of an ISSO grant, the basic framework of the fuzzy expert system has been established, including knowledge acquisition, the fuzzy rule base for fault management, and computer interface software development using the G2 software on a SunStation. Figure 1 shows part of the water supply and recovery system equipped with the new fuzzy expert system that developed bu UH researchers, which has been implemented on the ALSS computer simulation program at the NASA-JSC.
Back to Table of Contents
Albert M. K. Cheng, Ph.D., Assistant Professor Real-Time Systems Lab, Computer Science, UH
Research is focused on sufficient conditions and scheduling strategies for preemptive scheduling of a set of real-time tasks on n processors with various types of constraints. We have shown a sufficient condition for preemptive scheduling of a set of independent, periodic tasks on n identical processors with task migration constraints so that each ti can complete its computation time Ci before its deadline Di. Any set of independent, periodic tasks satisfying this condition is guaranteed a feasible schedule at run time. This condition is optimal for a set of m periodic tasks running on an n-processor system which requires a task migration time of R(R > 1) cycles.
Let T = gcd( D1 , D2 , ... , Dm ).
If the utilization factor (U) of a set of periodic tasks is at most n*(TR+1)/T, a feasible schedule exists. If R <= 1, then the sufficient condition becomes U <= n. This computation solves the open problem stated as a conjecture in Dertouzos and Mok's work that the condition U <= n is both necessary and sufficient for feasible scheduling.
We next applied this result together with imprecise computation techniques to schedule tasks in real-time multimedia communication systems, making it possible to yield a balanced tradeoff between the quality of the multimedia information (such as image/video) transmitted and the available resources for transmission. Images and videos are among the most demanding multimedia information to process and transmit. We modified an existing progressive transmission algorithm to implement the imprecise computation approach. Both the milestone and sieve approaches have been applied to yield imprecise image/video transmission algorithms. The computation in the form of incremental refinement of the final results can be divided into a sequence of sieve functions. The milestone approach is used to save intermediate results after distinct stages of the computation.
Back to Table of Contents
Raj S. Chhikara, Ph.D., Professor and Lie June Shiau, Ph.D., School of Natural and Applied Sciences, UHCL
Investigation is focused on two issues: (1) the role of metabolic gases in the tissue bubble formation during hypobaric exposures, which was empirically evaluated using a NASA data set of 426 simulated extravehicular activities (EVA) observations, and (2) the sensitivity of the bubble growth model to different EVA related parameters using inputs varying over a possible range of values of these parameters.
In the first study, a logistic regression model was used to assess the probability of circulating microbubbles (CMB) occurrence expressed as a function of the tissue ratio which is a measure of decompression stress. Analysis consisted of modelfit and parameter estimation using the maximum likelihood method. Investigators determined a computation-of-tissue ratio across a range of 13 values including zero for the metabolic gas tension (Pmg). The halftime period consisted of 9 different values, which ranged from 240 minutes to 700 minutes.
The likelihood value proved maximum when the halftime period was 330 minutes and Pmg = 186 hPa. This combination provided the optimum tissue ratio that best described the outcome of circulating microbubbles (CMB). Analysis indicated a significant participation of metabolic gases in bubble formation.
In the second investigation, parametric equations based on physiological processes were carried out through a range of biological values for the parameters. The most suitable model reflects very consistently PbN2 for a bubble in a steady state and could be used in further derived equations. It also shows that a certain value seems to be critical for the pressure in a bubble; and so, larger than this critical value, PbN2 is kept constant. Variations in parameters Rb, t and D, in some models, do not express any significant changes. The only variable influencing quantitatively is Kl, halftime period of the tissue of interest. The model for the bubble growth as a function of time as considered by Van Liew and Hlastala using Ficks' law for diffusion of gases across the surface of a bubble located in a tissue, reflects with good accuracy the bubbled growth in real time throughout EVA duration. The model shows that large radii bubbles generated from tissues with halftimes up to 700 minutes could explain CMB detected toward the end of EVA and existence of persistent symptoms. Its parametric values are in a limited but physiological range.
Contents
ISSO -- Institute for Space Systems Operations
1994-1995 Annual Report
|
|
