Related Experiment Videos
Symbolic programs for structural identification of linear pharmacokinetic systems
Computer Programs in Biomedicine
|September 1, 1981
Summary
Structural identification determines if system parameters are uniquely identifiable from experimental data, even with noise. This study presents symbolic computation tools to solve this problem for linear systems, with pharmacokinetic examples.
Area of Science:
- Systems Biology
- Mathematical Modeling
- Control Theory
Background:
- Parameter estimation in dynamic systems often assumes unique identifiability.
- Real-world experimental data can be noisy, challenging unique parameter determination.
- Structural identification addresses the fundamental question of parameter uniqueness.
Purpose of the Study:
- To investigate the problem of structural identification for linear systems.
- To develop computational tools for assessing parameter identifiability.
- To demonstrate the application of these tools in pharmacokinetic modeling.
Main Methods:
- Utilized symbolic matrix calculus for theoretical analysis.
- Developed two computer programs to aid in structural identification.
- Applied the methods to linear system models, including pharmacokinetic examples.
Main Results:
- The presented programs facilitate the analysis of structural identifiability.
- Demonstrated that not all parameters are uniquely determinable from input-output data.
- Provided practical insights through pharmacokinetic case studies.
Conclusions:
- Structural identification is crucial for reliable parameter estimation in dynamic systems.
- Symbolic computation offers a powerful approach to solving identifiability problems.
- The developed tools enhance the analysis of linear systems and their parameters.