Development and Analysis of a Feedback Treatment Strategy for Parturient Paresis of Cows
This document has been relocated to http://scholarsmine.mst.edu/mec_aereng_facwork/3418
There were 12 downloads as of 28 Jun 2016.
Abstract
An intelligent on-line feedback treatment strategy based on nonlinear optimal control theory is presented for the parturient paresis of cows. A limitation in the development of an existing nonlinear mathematical model for the homogeneous system is addressed and further modified to incorporate a control input. A neural network based optimal feedback controller is synthesized for the treatment of the disease. Detailed studies are used to analyze the effectiveness of a feedback medication strategy and it is compared with the current "impulse" strategy. The results show that while the current practice may fail in some cases, especially if it is carried out before the condition of a patient deteriorates, the proposed continuous medication process may be initiated at any time. Moreover the proposed on-line continuous infusion strategy never leads to severe hypercalcemic problems, thereby avoiding an associated disastrous consequence of cardiac arrest. A comparison study with linear quadratic regulator theory brings out the advantages of the nonlinear control synthesis approach.
