Automated multi-drug anesthesia presents a significant control challenge due to complex drug interactions and the need for high-reliability, resource-efficient systems. This paper introduces a novel event-based fractional order control framework that, contrary to conventional expectations, achieves improved control performance while simultaneously reducing computational demands. The proposed strategy integrates Event-Based sampling with robust fractional order controllers and a steady-state decoupling matrix to manage a 4x4 MIMO anesthesia model. Comprehensive validation on a cohort of 24 virtual patient models reveals that the event-based framework reduces control computations by 59.6% compared to its equivalent time-triggered implementation (Ts=1 s). More importantly, this efficiency gain is accompanied by a statistically significant performance enhancement, with an average of 20.6% improved precision and 37.2% improved tracking accuracy across all physiological outputs ( p < 0 . 001 ). The system maintained safety compliance above 98% for both physiological and infusion rate constraints. This work presents the first simulation-based evidence that for complex clinical systems, intelligent, event-based sampling is not a performance trade-off but a direct pathway to achieving both superior control and the computational feasibility potentially suitable for clinical hardware deployment.
Hegedüs et al. (Sat,) studied this question.