Computing optimal control With a hyperbolic partial differential equation

From MaRDI portal





The authors consider the optimal control problem with distributed parameters of the form \[ \int_\Omega L(x,t,z,u)d\Omega\to \underset{u_{\min}\leq u(x,t)\leq u_{\max}}{\text{maximum}} \] \[ \text{subject to}\quad {\partial^2 z\over\partial x \partial t}+ a(x,t,u(x, t))\cdot{\partial z\over\partial x}+ b(x,t,u(x,t))\cdot {\partial z\over\partial t}= f(x,t,u(x,t)) \] with the boundary conditions \(z(x,0)= p(x)\), \(z(0,t)= q(t)\). This problem is discretized and reduced to an ordinary optimization problem. Fast automatic differentiation is applied to calculate the exact gradient of the discretized problem so that existing optimization algorithms may be applied. A practical problem arising from the process of gas absorption is solved with the given method.











This page was built for publication: Computing optimal control With a hyperbolic partial differential equation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4237806)