Control of tumor growth by modern control methodologies

Abstract

The role of modern control engineering in physiological controls cannot be questioned. However, practitioners have to face with many challenges in the field. The imprecise information of the state variables of the system to be controlled, significant inter- and intra-patient variability, limitations regard to the applied sampling frequency are just a few of them. The current study investigates a possible solution for those issues related to control of tumor growth. In order to describe the parameter variabilities Linear Parameter Varying (LPV) method has been used and extended by applying Tensor Product (TP) model transformation. We formulated the goals of the control by using Linear Matrix Inequalities (LMI). Parallel Distributed Control can be used based on the state-feedback gains obtained through LMI optimization. The unmeasurable states can be estimated by using Extended Kalman Filtering. By using these techniques we were able to realize a control framework which enforces our original nonlinear system to behave as a given reference system within limitations. We have found that the developed control framework operates satisfactory by reaching all of the determined goals of the control.