Alejandro Tapia Córdoba, , Pablo Millán Gata
Nature Inspired Computing for Data Science, pp 169-200
The use of Micro-Hydro Power Plants (MHPP) has established itself as a fundamental tool to address the problem of energy poverty in rural isolated areas, having become the most used renewable energy source not just in this field but also in big scale power generation. Although the technology used has made important advances in the last few decades, it has been generally applied to big scale hydro-power systems. This fact has relegated the use of isolated MHPPs to the background. In this context, there is still a vast area of improvement in the development of optimization strategies for these projects, which in practice remains limited to the use of thumb rules. It results in a sub-optimal use of the available resources. This work proposes the use of a Genetic Algorithm (GA) to assist the design of MHPP, finding the most suitable location of the different elements of a MHPP to achieve the most efficient use of the resources. For this, a detailed model of the plant is first developed, followed by an optimization problem for the optimal design, which is formulated by considering the real terrain topographic data. The problem is presented in both single (to minimize the cost) and multi-objective (to minimize cost while maximizing the generated power) mode, providing a deep analysis of the potentiality of using GAs for designing MHPP in rural isolated areas. To validate the proposed approach, it is applied to a set of topographic data from a real scenario in Honduras. The achieved results are compared with a baseline integer-variable algorithm and other meta-heuristic algorithms, demonstrating a noticeable improvement in the solution in terms of cost.