An improved evaluation model for supplier selection based on particle swarm optimisation‐back propagation neural network. Issue 4 (29th September 2022)

• Record Type: Journal Article
• Authors: Yu, Jun; Li, Daming; Wang, Aihui; Liu, Ping; Song, Jingwen; Han, Xiaobo
• Is Part Of: IET collaborative intelligent manufacturing. Volume 4:Issue 4(2022)
• Page Start: 316
• Page End: 325
• Publication Date: 2022-09-29
• Subjects: enterprise evaluation -- neural network -- particle swarm optimisation -- supplier selection; Production management -- Periodicals; Production engineering -- Periodicals; Production management; Production engineering; Electronic journals; Periodicals; 658.5
• DOI: 10.1049/cim2.12067
• Languages: English
• ISSNs: 2516-8398

Abstract

Abstract: With the trend of supply chain globalisation, competition among enterprises is becoming more intense. Enterprises urgently need to improve their core competitiveness, and the enhancement of the competencies can depend on technologies services and the quality of suppliers. Since external factors are less controllable, this study starts with the quality of suppliers and proposes a supplier evaluation method that combines particle swarm optimisation with neural network algorithm to maximise the interests of enterprises. The particle swarm algorithm to lock the approximate location of the global optimum is first employed. Based on this, we establish an evaluation model of suppliers to train for the minimum errors between the desired and predicted values by constructing a back propagation (BP) neural network. Finally, the output results of the proposed method is compared with the BP neural network without the particle swarms optimisation. The proposed model is less empirically sensitive to the initialisation and can quickly converge to the local optimums, which overcomes the shortage of traditional neural networks and is more applicable to supplier evaluation. Abstract : A supplier evaluation method combining particle swarm with neural network algorithms is proposed to maximize the interests of enterprises. This method greatly overcomes the shortage of traditional neural network which is sensitive to initialization and converges to local optimum easily. The accuracy of the model established in this paper is verified, indicating that the model in this paper is more applicable to supplier evaluation.