ISBN:  Unavailable

Category: Unavailable

Page count: 16

The backflow cell-network model in state and adjoint variables is employed to obtain estimates of the optimal control and response trajectories for a nonlinear parabolic distributed system. The process under study is that of a heat exchange chemical reactor with axial mixing. The use of feedback control to obtain the priming control trajectory is shown to be a most useful technique. Resetting gradient methods for maximizing the Hamiltonian are shown to be stable and give fast initial convergence. Estimates of the optimal conditions for the continuous system may be obtained by extrapolating the results for an increasing number of cells. (Author).