Decentralized Supply Chain Optimization via Swarm Intelligence

When optimised, supply chains can bring tremendous benefits to all its participants. Supply chains therefore can be framed as a networked optimization problem to which swarm intelligence techniques can be applied. Given recent trends of globalization and e-commerce, we propose a supply chain that uses an open e-commerce business model, where all participants have equal access to the market and are free to trade with each other based on mutually agreed prices and quantities. Based on this model, we improve upon the Particle Swarm Optimization algorithm with constriction coefficient (CPSO), and we demonstrate the use of a new random jump algorithm for consistent and efficient handling of constraint violations. We also develop a new metric called the `improvement multiplier' for comparing the performance of an algorithm when applied to a problem with different configurations.