Analysis of the Chemical Reaction Optimization Algorithm

Author First name, Last name, Institution

Sinan Salman, Zayed University

Document Type

Conference Proceeding

Source of Publication

Proceedings of the Institute for Industrial and Systems Engineers (IISE) Annual Conference & Expo 2023, New Orleans, LA, USA

Publication Date



This paper presents an analysis of the Chemical Reaction Optimization (CRO) algorithm using the W-Model, a discrete and tunable optimization benchmarking objective function. It compares three different literature-based variants of the CRO algorithm including the conical, elitist, and adaptive variants. In addition, a Probabilistic Random Search Optimization (PRSO) algorithm is included in the comparison to investigate the significance of energy and population management schemes embedded in the CRO algorithm. All algorithms are set to use the same solution space search operators to ensure objective design comparisons. The results show that the conical CRO algorithm struggles to converge in the benchmarking problem, but its performance significantly improves when elitism is incorporated into the algorithm. In addition, the adaptive and elitist variant of the algorithm significantly outperforms the conical and elitist variants. Comparing the elitist and adaptive CRO variants to the PRSO algorithm shows that the energy and population management schemes embedded in CRO provide an advantage over a simple probabilistic random search, but only when elitism and adaptiveness are utilized. The only algorithm that achieved optimality in all optimization runs was the elitist and adaptive CRO variant, and it accomplished this with fewer objective function evaluations. Inspection of population compositions during optimization runs reveals that the adaptive variants use differentiated search operator deployment to achieve higher performance. The results also highlight population collapses as a weakness in the conical algorithm. These findings can be valuable for researchers investigating the use of CRO variants in discrete optimization applications.


Computer Sciences


Discrete Optimization, Optimization Heuristics, Algorithm Design, Nature-inspired Algorithms

Indexed in Scopus


Open Access


Open Access Type

Bronze: This publication is openly available on the publisher’s website but without an open license