Process Knowledge Recommendation System for Mechanical Product Design

Zhenyong Wu,Haotian Liu,Lu Li

doi:10.1109/access.2020.3002922

Abstract

To solve the problems of high time cost caused by the low efficiency of knowledge acquisition, this paper proposes a knowledge service framework based on case set. Three knowledge retrieval methods are designed based on parts keywords, customer orders and manufacturing processes. Additionally, a VSM-based (vetor space model) knowledge recommender method is proposed by using similarity matching algorithm to improve the efficiency of knowledge acquisition and transfer. Finally, application scenarios and evaluation of the proposed approach in a mechanical product design project are given to illustrate the application effect of knowledge service system.

Highlights

With the ever-growing in the generation technology trends, the manufacturing industry is realizing the importance of knowledge in having a competitive advantage over its competitors
PERSPECTIVE This paper proposes a vector space model-based method to enhance the ability of knowledge acquirement during product development
The demonstration of the knowledge recommendation system show the benefits of the implementation of the proposed method

Summary

INTRODUCTION

With the ever-growing in the generation technology trends, the manufacturing industry is realizing the importance of knowledge in having a competitive advantage over its competitors. Process knowledge here mainly refines from process data, tools and documents of product development, product design cases. Since it would be taken a long term for an engineer to be an expert. From the annual Global Most Admired Knowledge Enterprises (MAKE) [3], the organizations have been recognized as leaders in effectively transforming enterprise knowledge into wealth-creating ideas, products, and solutions Following these examples, more enterprises start to learn the knowledge acquisition and reuse methods to support product development. A new knowledge service method should be proposed for product development and manufacturing to boost the efficiency of product design, reduce the time-cost.

RELATED WORK

SUPPORTING TECHNOLOGIES FOR PROPOSED FRAMEWORK

DESIGN OF SIMILARITY MODEL FOR KNOWLEDGE RECOMMENDATION

Findings

CONCLUSION AND FUTURE PERSPECTIVE