Selective disassembly sequence optimization based on the improved immune algorithm

Abstract

Purpose

The purpose of this paper is to improve the automation of selective disassembly sequence planning (SDSP) and generate the optimal or near-optimal disassembly sequences.

Design/methodology/approach

The disassembly constraints is automatically extracted from the computer-aided design (CAD) model of products and represented as disassembly constraint matrices for DSP. A new disassembly planning model is built for computing the optimal disassembly sequences. The immune algorithm (IA) is improved for finding the optimal or near-optimal disassembly sequences.

Findings

The workload for recognizing disassembly constraints is avoided for DSP. The disassembly constraints are useful for generating feasible and optimal solutions. The improved IA has the better performance than the genetic algorithm, IA and particle swarm optimization for DSP.

Research limitations/implications

All parts must have rigid bodies, flexible and soft parts are not considered. After the global coordinate system is given, every part is disassembled along one of the six disassembly directions –X, +X, –Y, +Y, –Z and +Z. All connections between the parts can be removed, and all parts can be disassembled.

Originality/value

The disassembly constraints are extracted from CAD model of products, which improves the automation of DSP. The disassembly model is useful for reducing the computation of generating the feasible and optimal disassembly sequences. The improved IA converges to the optimal disassembly sequence quickly.

Citation

Ji, J. and Wang, Y. (2023), "Selective disassembly sequence optimization based on the improved immune algorithm", Robotic Intelligence and Automation, Vol. 43 No. 2, pp. 96-108. https://doi.org/10.1108/RIA-06-2022-0156