Evolutionary Dynamics-Driven Learning of Optimal Decisions for Nonlinear System: Extend-Policy Iterative Algorithm

An extend-policy iterative algorithm is proposed for solving the ecological evolving-lung cancer cells growth inhibition optimal drug delivery scheme. With the analysis of the cell proliferation-apoptosis process of lung cancer cells with primitive immune system and external drug interventions, such as chemotherapeutic drugs and immunological agents, a model of ecological containment of lung cancer cells mimicking injection labeling is constructed. The HJB equation for biological tissue damage has also been established by considering the concentration of lung cancer cells in the blood and the amount of drug administered. The final simulation experiment proved the effectiveness of the drug delivery scheme.