Chemical Reaction Optimization for Combinatorial Problem Solving

Chemical Reaction Optimization (CRO) is a nature-inspired metaheuristic that mimics the interaction and transformation of molecules in chemical processes to address combinatorial optimisation challenges. In CRO, candidate solutions are encoded as molecules with associated potential energy representing objective costs. A set of elementary reactions—on-wall ineffective collision, inter-molecular ineffective collision, decomposition and synthesis—allows the algorithm to explore and exploit the solution space by adjusting energy levels and molecular structures. Through adaptive energy buffers and reaction probabilities, CRO balances diversification and intensification, making it well suited to discrete problems such as the quadratic assignment, travelling salesman and shortest common supersequence problems. Recent advances have refined operator design, introduced hybrid schemes and tailored energy management strategies to improve convergence speed, solution quality and robustness. Practical applications span logistics, network design and bioinformatics, where CRO has demonstrated competitive performance against established metaheuristics while offering a flexible framework for customisation and parallel implementation.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Foundational work has been consolidated in a comprehensive tutorial that formalises CRO mechanisms and provides guidelines for implementation. This tutorial demonstrates the method’s versatility across both discrete and continuous domains, with successful applications to benchmark combinatorial problems including assignment and routing scenarios, and outlines potential future research directions in adaptive operator control and parallel architectures.

Building on this foundation, a recent study introduced an opposition-based learning enhancement to CRO for the shortest common supersequence problem. By generating an opposite population alongside a random initial set, the enhanced algorithm selects the fittest molecules to form a diverse initial pool. Iterative application of standard reaction operators then yields solutions with significantly reduced runtime—more than 50 percent faster than earlier variants—while preserving or improving sequence coverage on DNA and protein datasets.

Chemical Reaction Optimization for Combinatorial Problem Solving publication trend

The graph below shows the total number of articles in chemical reaction optimization for combinatorial problem solving across all publications each year (not limited to Nature Index journals).

Technical terms

Metaheuristic: A high-level strategy guiding subordinate heuristics to efficiently explore large solution spaces.

Molecule: A data structure encoding a candidate solution, analogous to a reactant in a chemical system.

Potential Energy: A scalar value representing the fitness or cost of a solution, which the algorithm seeks to minimise.

Elementary Reaction: An operator that transforms molecules, including decomposition, synthesis, on-wall ineffective and inter-molecular ineffective reactions.

Combinatorial Optimisation Problem: A challenge of selecting an optimal arrangement or subset from a finite, discrete set of candidates.

References

1. Chemical Reaction Optimization: a tutorial. Memetic Computing (2012).
2. An Opposition-Based Learning CRO Algorithm for Solving the Shortest Common Supersequence Problem. Entropy (2022).