Computational Intelligence Methods for Agricultural Engineering Design Optimisation Under Multi-Objective Constraints in Nigerian Contexts

This study develops and applies computational intelligence methods for solving multi-objective agricultural engineering design optimisation problems characteristic of Nigerian agricultural development contexts, contributing original algorithmic advances and applied engineering design outputs. Many critical agricultural engineering design problems in Nigeria involve multiple competing objectives, complex non-linear constraint structures, and mixed continuous-discrete decision variables that make them intractable using conventional optimisation methods. Computational intelligence approaches including genetic algorithms, particle swarm optimisation, multi-objective evolutionary algorithms, and neural network surrogate models offer powerful alternatives for these problem classes. This study develops a computational intelligence optimisation framework incorporating adaptive multi-objective evolutionary algorithms hybridised with machine learning surrogate models to reduce computational cost. The framework is applied to four priority agricultural engineering design problems in Nigeria: multi-objective irrigation canal network design, solar-biomass hybrid energy system sizing, post-harvest cold chain network layout optimisation, and agricultural machine component design for local manufacturing. For each problem, mathematical formulations are developed, computational experiments are conducted, and Pareto-optimal solution sets are generated and analysed. Field implementation of selected optimal designs validates computational predictions. Findings demonstrate that the proposed framework generates solution quality 22 to 41 percent superior to single-objective formulations across problem types, revealing important trade-off structures between cost, performance, and equity objectives that conventional design practice misses. The study contributes an original Nigerian agricultural engineering optimisation toolkit and recommends its adoption in postgraduate engineering design courses.

Keywords: computational intelligence, multi-objective optimisation, agricultural engineering, genetic algorithms, Nigeria.