Comparison and verification of the performance of a Stateflow model on a C166 microcontroller board

ScholarWorks Home
→
University of Alaska Fairbanks
→
UAF Graduate School
→
Older Theses Not Clearly Affiliated with a Current College
→
View Item

Comparison and verification of the performance of a Stateflow model on a C166 microcontroller board

Grandham, Jyothsna

Date: 2005-05

Type: Thesis

Table of Contents: Introduction -- Disease description -- State machine -- Introduction -- Description of a finite state machine -- Simulink model -- Stateflow model -- Data analysis -- Timing issues -- Model analysis with a microcontroller board -- Hardware in the loop -- Matlab/Simulink environment -- Target- infineon C166 microcontroller -- Bootstrap mode -- Method 1 : using the C166 microcontroller toolbox of Matlab -- Method 2 : using RTC167-target toolbox and tasking compiler -- Code generation using tasking compiler -- Method 3 : using RTC167-target toolbox and keil compiler -- Results and discussion -- Introduction -- Extracting and analyzing patient data -- Stateflow model analysis -- Analysis of the different implementation methods -- Speed comparison -- Conclusions and future work.

Abstract:

The IMPROVE data library is a collection of continuous and intermittent physiological observations of 59 patients in the Intensive Care Unit (ICU), recorded for one complete day at the Kuopio University Hospital. These data consisted of patient data stored as annotation and record files. The IMPROVE data library also provided the information regarding four disorders (Hypovolaemia, Cardiac Failure, Hypervolaemia and Oxygen-Content related problems) in terms of the patient data. An available model designed in Stateflow generated the risk levels for the patients. The goal of this thesis is to analyze the risk levels generated from the simulation and compare them with the physician's diagnosis. The main objective of this work is to evaluate the performance of the Stateflow model that interprets the risk level of the four disorders, for a particular patient at any instant of time and later implement it on an Infineon C 166 microcontroller board. The various limitations encountered during the implementation on the microcontroller board are discussed in the thesis. To demonstrate the rapid prototyping capabilities of using the microcontroller board a speed-comparison analysis of implementing the Stateflow model on various computers was performed.