A generalized reduced gradient method for the optimal control of multiscale dynamical systems

This paper considers the problem of computing optimal state and control trajectories for a multiscale dynamical system comprised of many interacting dynamical systems, or agents. A generalized reduced gradient (GRG) approach is presented for distributed optimal control (DOC) problems in which the agent dynamics are described by a small system of stochastic differential equations (SDEs). A new set of optimality conditions is derived using calculus of variations, and used to compute the optimal macroscopic state and microscopic control laws. An indirect GRG approach is used to solve the optimality conditions numerically for large systems of agents. By assuming a parametric control law obtained from the superposition of linear basis functions, the agent control laws can be determined via set-point regulation, such that the macroscopic behavior of the agents is optimized over time, based on multiple, interactive navigation objectives.