T. Kmet, M. Kmetova: Radial Basis Function Networks Simulation of Age-Structure Population. In Theory and Practice of Natural Computing (Lecture Notes in Computer Science 11324). Cham, CH : Springer, pp. 414-425, 2018. ISSN 0302-9743 link

Abstract: Based on radial basis function networks (RBFN) we propose a new method to solve optimal control problems for systems governed by hyperbolic partial differential equations. RBFN are able to approximate continuous function with given precision [1]. This property of RBFN was used to estimate optimal control, optimal state trajectory and optimal co-state trajectory by adaptive critic design. A new algorithm was verified and compared with indirect methods on the harvesting model of age-dependent population.