Steel structure building has become an important direction of green and industrialized transformation of the construction industry due to its advantages of high strength, light weight and high recyclability. As an important technical link in the construction process of steel structure buildings, the level of process planning will directly affect the safety and service life of steel structure buildings. As a highly knowledge-intensive work, the design quality of welding process planning largely depends on the support of process knowledge. In the process of engineering practice, steel structure construction enterprises have accumulated a large amount of process data, but the low degree of structuring of these process data and the lack of knowledge reuse mechanism make it difficult for process planners to effectively obtain relevant knowledge and apply it to the process planning process.

Knowledge engineering provides new ideas for solving welding process planning problems through structured representation, systematic storage and reasoning application of domain knowledge. Therefore, based on the knowledge engineering technology, this paper investigates the method of acquiring, representing, storing and applying reasoning to the domain knowledge of welding process of steel structural components, and designs a welding process planning system to assist the process planners in decision making, so as to improve the efficiency and effect of process design. The main research contents of this paper are as follows:

(1) The framework of welding process planning method for steel structure components. In order to support the application of knowledge engineering in welding process planning for steel structure components, we analyzed the welding process planning process, sorted out the related theories of knowledge engineering, and proposed a knowledge engineering-based welding process planning method framework for steel structure components based on the knowledge engineering, including “welding knowledge representation”, “welding knowledge storage”, “welding knowledge reasoning” and “welding knowledge-driven application”.

(2) Welding process knowledge representation and knowledge base modeling for steel structure components. The characteristics of welding process knowledge are clarified, and according to the knowledge sources and application scenarios, welding process knowledge is categorized into factual, case-based and rule-based knowledge. Aiming at the multi-source heterogeneous characteristics of process knowledge, a hybrid knowledge representation method is adopted to realize structured expression. Specifically, the welding process knowledge framework is constructed based on ontology modeling technology, the hierarchical representation and attribute description of the case knowledge is realized by frame representation, the logic of the process rules is described by generative representation, and the knowledge is combined with the database through the mapping rules, which lays the foundation for the integrated storage and efficient retrieval of the knowledge.

(3) Design of welding process planning method based on knowledge reasoning. Based on the structured representation of the welding process knowledge, a hybrid reasoning framework of case-based reasoning and fuzzy reasoning is established to realize the intelligent planning of the welding process through the hybrid reasoning mechanism of “case matching-parameter correction”. First of all, through the K-nearest neighbor algorithm and hierarchical analysis weighting method, a multi-dimensional feature similarity calculation model is constructed to retrieve and match historical process cases according to the target demand. On this basis, the fuzzy inference model is designed for the correction of process parameters, and the fuzzy affiliation function and Mamdani inference model are used to deal with the uncertain parameters in order to realize the accurate output of process parameters. And the feasibility of the method is proved by example verification.

(4) Design of welding process planning system for steel structural components. In order to better integrate the entire knowledge modeling and reasoning process with real-world engineering scenarios, based on the previous research on the knowledge representation and reasoning mechanism of welding process, a welding process planning system for steel structure components is further designed to meet the practical requirements. Based on the analysis of system requirements and overall architecture, functional modules such as welding process knowledge management, knowledge retrieval and process planning are designed to improve the fragmentation of knowledge management, low design efficiency and lack of knowledge reuse mechanism faced by welding process planning.

This paper combines theory and practice, on the one hand, based on the knowledge engineering related technology to study the knowledge management and planning method of welding process of steel structural components, which enriches the methodology system of knowledge-driven decision-making in the construction field. On the other hand, it designs and builds a welding process planning system to realize the digital precipitation of industry knowledge, which provides methodological support for the high quality and sustainable development of steel structure construction

Key words: Prefabricated building, Construction Supply chain, Blockchain, Quality information disclosure, Stackelberg Game