A gradient-based time-delay optimization algorithm for evaluating control strategies in a fractional-order infectious...

Mathematical models have played a crucial role in developing strategies to control various diseases by identifying effective interventions. However, balancing different interventions to develop an optimal cost-effective strategy remains a significant challenge. In this study, the eight-compartmental SIDARTHE model is refined with realistic cost and control functions. Four control interventions are undertaken to provide insights into the model. We combined the Adams-Bashforth method and the composite trapezoidal rule to design an innovative gradient-based time-delay optimization algorithm. To the best of our knowledge, this is the first study to perform cost optimization for parameter changes over different time periods in the field of fractional-order infectious disease modeling. The implementation of control interventions results in a cost reduction of approximately 35.61% compared to the initial cost when no control is used. This cost optimization effectively demonstrates the financial benefits of implementing control interventions in the model. Furthermore, this study highlights the progress achieved by implementing the Caputo-Fabrizio fractional derivatives, which provides a closer match with real-case disease incidence data from Italy, especially for fractional order of 0.7. This clearly shows the influence of the memory properties of the fractional operator. The numerical results obtained provide valuable insight into the benefits of implementing control interventions. Future research and practice can build on this foundation to enhance treatment strategies using our time-delay optimization algorithm.