Agent swarm optimisation,

Agent swarm optimisation (ASO) is a new paradigm based on particle swarm optimisation that exploits distributed or swarm intelligence and borrows some ideas from multi-agent based systems. It is aimed at supporting decision-making processes by solving either single or multi-objective optimisation problems. Classical methods of optimisation have been shown to be poorly suited for many real-world problems since they are unable to deal with highly-dimensional, multimodal, non-linear problems; and process inaccurate, noisy, discrete and complex data. Robust methods of optimisation are often required to generate suitable results. ASO offers robustness through a common framework where a plurality of population-based algorithms co-exist, thereby offering superior performance by dynamically combining the strengths of multiple metaheuristics. In this work the ASO framework is used to solve a complex problem in water management, namely the optimal design of water supply systems (including sizing of components, reliability, renewal, and rehabilitation strategies) using a multi-objective approach. Conditions for the correct development of the Pareto front are described. In addition, during the solution process, the users, working in parallel with computational algorithms, can force the recruitment of new agents/swarms to the environment and even contribute to the solution process with expert-based personal proposals that are later ‘learned’ by the algorithms.