Research on the Method of Material Scheme Matching Based on Deep Learning

Abstract

Based on the research of the deep learning network and the material scheme matching method, a material scheme matching method based on the combination of materials, solid ID data and multi-layer perceptrons is proposed. Based on the relationship among engineering design standards, general equipment selection and material procurement standards, a material and solidified ID data information system is formed. Then, collect, sort, and model electrical primary and secondary equipment and line information to form a model structure of plans and materials; finally, integrate and analyze historical data to form a typical plan material matching library. The training of the perceptron network obtains the material plan matching network. Experimental results show that the matching method of material schemes using materials, solidified ID data and multilayer perceptron network can achieve 96% matching accuracy, which solves the problem of information barriers in design standards, general equipment requirements and material procurement standards. The development of the material plan provides new ideas.