Genetic Programming With Niching for Uncertain Capacitated Arc Routing Problem

The uncertain capacitated arc routing problem is an important optimization with many real-world applications. Genetic programming considered a promising hyper-heuristic technique to automatically evolve policies that can make effective real-time decisions in environment. Most existing research on genetic for the only focused test performance aspect. As result, evolved by are usually too large and complex, hard comprehend. To effective, smaller, simpler policies, this article proposes novel approach, which simplifies during evolutionary process using niching technique. simplified stored external archive. We also developed new elitism, parent selection, breeding schemes generating offspring from original population experimental results show newly proposed approach achieve significantly better than current state-of-the-art algorithms problem. thus potentially more interpretable.