A comparison of traditional versus object-oriented models for simulation of plant development /
| Main Author: | |
|---|---|
| Other Authors: | , , |
| Format: | Thesis Book |
| Language: | English |
| Published: |
1990.
|
| Subjects: | |
| Online Access: | ProQuest, Abstract Link to OAKTrust copy |
| Abstract: | This dissertation accomplished three goals: a procedural model was adapted and validated for Texas, this procedural model was modified to include plant fruiting structure, and an object-oriented model was developed and its behavior verified. Data to characterize cotton plant development was obtained through intensive field studies on six cotton cultivars in College Station, TX during 1987-1989. Fiber quality properties and yield of individual fruits at different nodal positions were determined for a short-season cultivar, TAMCOT CD3H. The development of fruit cohorts was simulated in the procedural model using a modification of an algorithm that simulates the distribution of growth rates with a time distributed delay. The model predicted observed fruiting data from stressed and non-stressed plants reasonably well. The ability to simulate fruit location discretely on a plant resulted in a more accurate determination of harvest price than is possible with currently available cotton development models. A second model based on the representation of individual plant organs was developed. This object-oriented, behavior-based program permitted a more explicit representation of age and size structure than current procedural models. The object-based model provided additional information for management, especially identification of individual plant organs at discrete locations resulting in the ability to represent site-specific fiber properties and pest vulnerability. Variability in light interception and thus carbohydrate supply for leaves at specific positions led naturally to variability in position-dependent fiber yield. Thus, the behavior of individual organs depended upon the interaction with other organs in the plant. The procedural alternative is the pre-assignment of vectors or multidimensional arrays to represent age and size structure in a partial differential equation structure or stochastic cohort models which are much less explicit, less tractable and require variability as input. Other advantages of object-oriented simulation are summarized. |
|---|---|
| Item Description: | Typescript (photocopy). Vita. "Major subject: Bioengineering." |
| Physical Description: | xix, 227 leaves : illustrations ; 29 cm |
| Bibliography: | Includes bibliographical references. |