Flood Algorithm (FLA): A Nature-Inspired Optimization Technique
The Flood Algorithm (FLA) is a new meta-heuristic optimization method inspired by how water behaves during floods in river basins. It mimics natural processes such as water flowing downhill, varying flow rates, soil permeability, and changes in water levels caused by rain and evaporation.
FLA models these natural events using mathematical formulas. It then uses them to guide a population of candidate solutions toward better results. This process connects real-world flood behavior with computational optimization in a meaningful way.
FLA runs in two main phases:
-
Regular Movement Phase – The algorithm moves the population toward current best solutions.
-
Flooding Phase – It introduces random shifts to explore new possibilities and maintain diversity.
New solutions enter the population regularly. At the same time, the algorithm removes weaker candidates—similar to how rising and falling water levels occur in nature. This cycle helps prevent stagnation and promotes innovation.
Tests on benchmark and real-world problems show strong results:
-
On CEC2005 functions, FLA outperformed or matched 16 other algorithms in basic and enhanced forms.
-
On CEC2014 benchmarks, it handled dimensions of 30, 50, and 100, competing with 20 optimization methods.
-
In 12 real-world engineering problems, FLA achieved competitive and reliable results.
These tests confirm that the Flood Algorithm offers a robust, scalable, and effective approach for both theoretical and practical optimization challenges.
Downloadable files
