Solving the Individual Control Strategy Tasks Using the Optimal Complexity Models Built on the Class of Similar Objects

摘要

The conventional approach for calculating individual optimal strategies assumes that the best control actions are determined for the same object that has been studied by monitoring or conducting active trials. However, the class of objects for which is impossible to organize repeated tests is widespread. An example is patients with a particular disease, for each of which it is impossible to organize separate trials to study possible strategies for its cure. This paper proposes an approach to formulate the individual strategies optimization task that uses observational data obtained during the monitoring or active experiment on a sample of similar objects. It is proposed to obtain the state models of the optimal complexity object that are nonlinear in the parameters and initial conditions of the object and linear in control actions, to construct an effective calculation technology. As a modeling tool, algorithms of Group Method of Data Handling (GMDH) are used. The optimization task of individual strategies is formed after substituting the individual values of object parameters in the model of functional and models of constraints. The final calculation procedure takes the form of a linear programming problem. Limitations of the approach and an example of calculating the individual strategy are considered.