OpenDDPT

OpenDiscreteDynamicProgrammingTemplate : founds optimal constrainted parameters of a discrete controls with second order optimization template replacing Hessian with directional derivatives and backpropagation for digital filter(as neural network) Th

Advertisements:

OpenDiscreteDynamicProgrammingTemplate : founds optimal constrainted parameters of a discrete controls with second order optimization template replacing Hessian with directional derivatives and backpropagation for digital filter(as neural network) The physic of the real world can be formulated into mathematical form using differential equations. Differential equation and systems of differential equations are the natural language trough we can describe any real devices like electronic circuit, automatic controls, elettro-magnetic wave emission or any other engineering problems. The purpose of OpenDDPT project is to build a software that can aids people to build automatically these control devices. The strategy used by this software is based on the concept that differential equations generally are parameters dependent so we can found optimal control parameters using an optimization template. The optimization tool-box used by OpenDDPT is an innovative non-convex second-order method that replaces the hessian structure (the hessian matrix has heavy memory storage and an expensive computational-cost) with directional-derivatives , reducing the second order information computational cost of one order without approximation.Control's local informations (i.e. gradient and it's directional derivate) are found trough an extended form of the back-propagation rule used for neural-network and that new extended form can perform not only weighted-sum and sigmoidal saturations but also multiplication ,addition, time-delay and any non-linearity and it can be used as a generic expression for realizing any non linear discrete time control device.