Skip to Main Content
In choosing the best parameters for compensation networks for systems including the human operator, the problem of nonconstant human operator parameters is encountered. In particular, some operator parameters vary in an adaptive way with plant dynamics. A satisfactory design must predict the operator's parameters for a particular system. A new technique regards the pilot as an optimal controller in one sense and the compensator as an optimal controller in another sense. The combination of man, machine, and compensator may then be represented by a dual optimization problem. To account for restricted numbers of variable parameters and unavailable states, the problem is modeled as a dual suboptimal (limited state feedback), control problem. Analytic expressions are obtained for the necessary conditions for (dual) optimality. An algorithm based on the method of successive approximations is described. This provides for rapid solution to the optimization problem. The method by which any considered design configuration can be analyzed to rapidly establish its potential with respect to 1) designers cost and 2) pilots work load is illustrated, and results are included.