Task Decomposition for Cloud Manufacturing under Industrial Internet Using an Improved Genetic Algorithm

Abstract

In the industrial internet environment, task decomposition and reconfiguration are key links connecting user demands with manufacturing resources in a cloud manufacturing platform. To address the issues of traditional methods relying on process logic while neglecting enterprise capability differences, a capability matching‑based multi‑dimensional task decomposition method is proposed. A multi‑dimensional task correlation model covering technological, capability, geographical, and cost correlations is constructed, followed by a multi‑objective reconfiguration optimization model targeting efficiency, cost, and quality. For efficient solution, an improved genetic algorithm is designed (integer encoding, order‑preserving crossover, heuristic mutation, and adaptive parameter adjustment). A case study on aero‑engine impeller manufacturing demonstrates that, while maintaining the optimal makespan, the optimized solution reduces cost by 5.69% and improves quality by 0.33%, verifying the effectiveness of the proposed method.