Fault Detection and Isolation Using Sliding Mode Observer for Descriptor Systems

Abstract

In modern technological era, research in medical science has drastically increased. Cancer is one of the most common diseases in these days and cancer treatment is most expensive treatment which is difficult to afford for a common man. Diethylstibestrol is an artificial estrogen whose use was limited because of the fact that a small quantity of under infusion or over infusion can cause cardiovascular toxicity. According to a research a certain low dose of diethylstibestrol is very effective for treatment of prostate cancer. Diethylstibestrol is inexpensive and effective candidate when used with accurate dosage for CRPC (metastatic castration-resistant prostate cancer) treatment and ADT androgen deprivation therapy which are done before chemotherapy. In this research work a descriptor system model of DES transferring in human body is considered which is affected by some unknown exogenous input. Exogenous input is estimated and closed loop fault tolerant nonovershooting scheme is presented for infusion of drugs diethylstibestrol combined with aspirin. An unknown exogenous infusion input appearing in descriptor system model of DES transfer in human body is estimated through sliding mode observer. Nonovershooting fault tolerant tracking controller is used to remove fault from the system model without over injection which may be caused due to overshooting behavior.