Hill Tunneling Optimization: Applications To Optimal Channel Networks
Adams, Taylor 1
1 CU Boulder
Hill Tunneling Optimization is a member of the Reactive Search Optimization class of algorithms that attempts to reproduce the global-minimum-finding capabilities of Simulated Annealing algorithms while increasing computational efficiency in certain classes of applications. The determination of the Optimal Channel Network for a specified grid is chosen as a test problem to evaluate the performance of Hill Tunneling Optimization as compared to Simulated Annealing, and the results of computational experiments comparing the two are presented. Both methods of optimization are then utilized to find globally optimal network configurations for networks whose energy input (precipitation) is correlated to a randomized initial network configuration, and these networks are compared to the global optimums found for the same grid with uniform energy input.
Colaiori, F., et al, 1997, Analytical and numerical study of optimal channel networks: Physical Review E, v. 55-2, p. 1298-1310.