A multi-objective Giant Pacific Octopus optimizer for scheduling resource-constrained multi-project systems with transfer constraints

This study introduces a novel bio-inspired algorithm, the Multi-Objective Giant Pacific Octopus Optimizer (MOGPOO), designed to solve the Resource-Constrained Multi-Project Scheduling Problem with resource transfer (RCMPSP with resource transfer). MOGPOO integrates multi-objective optimization principles with the adaptive search behavior of the Giant Pacific Octopus to effectively balance exploration and exploitation in complex, high-dimensional search spaces. The algorithm explicitly considers both time and cost in multi-mode resource transfers across projects – an aspect often neglected in existing scheduling methods. Computational experiments demonstrate that MOGPOO significantly outperforms benchmark algorithms, including Multi-objective Slime Mold Algorithm (MOSMA), Non-dominated Sorting Genetic Algorithm II (NSGA-II), Multi-objective Elk Herd Optimization (MOEHO) and Multi-Objective Crested Porcupines Optimization (MOCPO), across several performance metrics such as makespan, total cost, average project delay (APD) and portfolio percentage delay (PDEL). Further assessment using standard quality indicators – Hypervolume (HV), Inverted Generational Distance Plus (IGD+), and Spacing (SP) – confirms its superior convergence and solution diversity. The algorithm’s effectiveness is reinforced by a theoretical convergence analysis using a Markov chain framework. These findings highlight MOGPOO as a robust and efficient solution for complex multi-project scheduling scenarios with dynamic inter-project resource coordination.