Regularized error function-based extended Kalman filter for estimating the cancer chemotherapy dosage: impact of improved grey wolf optimization

Abstract

Abstract Objectives The main aim of this work is to introduce a robust controller for controlling the drug dosage. Methods The presented work establishes a novel robust controller that controls the drug dosage and it also carried out parameters estimation. Along with this, a Regularized Error Function-based EKF (REF-EKF) is introduced for estimating the tumor cells that could be adapted for different conditions. It also assists in solving the overfitting problems, which occur during the drug dosage estimation. Moreover, the performance of the adopted controller is compared over other conventional schemes, and the attained outcomes reveal the appropriate impact of drug dosage injection on immune, normal, and tumor cells. It is also ensured that the presented controller does a robust performance on the parameter uncertainties. Moreover, to enhance the performance of the proposed system and for fast convergence, it is aimed to fine-tune the initial state of EKF optimally using a new Improved Gray Wolf Optimization (GWO) termed as Adaptive GWO (AGWO). Finally, analysis is held to validate the betterment of the presented model. Results The outcomes, the proposed method has accomplished a minimal value of error with an increase in time, when evaluated over the compared models. Conclusions Thus, the improvement of the proposed REF-EKF-AGWO model is proved from the attained results.