We present a novel method to approximate optimal feedback laws for nonlinear optimal control based on low‐rank tensor train (TT) decompositions. The approach is based on the Dirac–Frenkel variational principle with the modification that the optimization uses an empirical risk. Compared to current state‐of‐the‐art TT methods, our approach exhibits a greatly reduced computational burden while achieving comparable results. A rigorous description of the numerical scheme and demonstrations of its performance are provided.
The operation of industrial supply technology is a broad field for optimization. Industrial cooling plants are often (a) composed of several components, (b) linked using network technology, (c) physically interconnected, and (d) complex regarding the effect of set-points and operating points in every entity. This leads to the possibility of overall optimization. An example containing a cooling tower, water circulations, and chillers entails a non-linear optimization problem with five dimensions. The decomposition of such a system allows the modeling of separate subsystems which can be structured according to the physical topology. An established method for energy performance indicators (EnPI) helps to formulate an optimization problem in a coherent way. The novel optimization algorithm OptTopo strives for efficient set-points by traversing a graph representation of the overall system. The advantages are (a) the ability to combine models of several types (e.g., neural networks and polynomials) and (b) an constant runtime independent from the number of operation points requested because new optimization needs just to be performed in case of plant model changes. An experimental implementation of the algorithm is validated using a simscape simulation. For a batch of five requests, OptTopo needs 61 min while the solvers Cobyla, SDPEN, and COUENNE need 0.3 min, 1.4 min, and 3.1 min, respectively. OptTopo achieves an efficiency improvement similar to that of established solvers. This paper demonstrates the general feasibility of the concept and fortifies further improvements to reduce computing time.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.