Research On Fast Matching Method Of Process Information Of Key Parts Of Marine Diesel Engine Based On Association Rule Mining

The new generation of artificial intelligence is booming globally,injecting new momentum into economic and social development,and profoundly changing people’s production and lifestyle.Manufacturing is an important application field of artificial intelligence technology,and the deep integration of artificial intelligence and manufacturing is leading to a profound industrial revolution.Against this backdrop,the field of process design also faces a new round of technical challenges: the existing process design system has accumulated a large number of process instances,but lacks effective use and mining,and it takes a lot of time to search and query repetitive process design cases,reducing the efficiency of process design.Therefore,this thesis proposes a method for rapid matching of key component process information of marine diesel engines based on association rule mining,and studies the ontology-based process knowledge expression model,the process knowledge generation method based on association rule mining,and the process instance retrieval and knowledge reuse method based on the nearest neighbor algorithm.A process information rapid matching function module was developed and verified.The module includes three modules: process knowledge ontology creation,process knowledge mining,and process instance retrieval and knowledge reuse.The main research contents of this thesis are as follows:(1)A method for creating a process knowledge expression model based on ontologyIn order to solve the problem that existing machining process knowledge lacks a unified representation,a process knowledge expression model based on ontology was designed.First,the composition of machining process knowledge is analyzed,and key concepts and attributes are refined.Secondly,the ontology representation method is used to formalize the description of key concepts and attributes,and the process ontology knowledge model is constructed.Then,a process knowledge expression method based on the ontology model is designed to convert process knowledge into the ontology model,realizing effective organization and management of process knowledge.Finally,the part process ontology knowledge base is created for orderly storage and arrangement of process knowledge,laying the foundation for subsequent knowledge mining.(2)Created a process knowledge generation method based on association rule miningFirstly,historical process information and data are collected,and the data is discretized by the equal-width method.Secondly,data preprocessing is carried out by means of data cleaning and normalization to ensure the quality and accuracy of the data.Then,the Apriori algorithm is used to analyze and integrate the processed data,mine the association between process information,and form corresponding process knowledge.Finally,taking the process information of the key part piston of the diesel engine as an example,the process knowledge is generated through the method of association rule mining.(3)Proposed a process instance retrieval and knowledge reuse method based on the nearest neighbor algorithmFirst,a knowledge reasoning mechanism that integrates Case-Based Reasoning(CBR)and Rule-Based Reasoning(RBR)is established.CBR is the main reasoning method,extracting the machining feature type of parts and searching for corresponding process instances to obtain the process parameter information of the current feature type.RBR is used as a post-supplement to verify the accuracy of CBR.Secondly,a process instance retrieval and knowledge reuse algorithm process is established to realize efficient matching of to-beprocessed part features and historical cases,and combined with rule reasoning,it provides precise parameter suggestions for process design.Finally,using the nearest neighbor algorithm,the feature type of the to-be-processed part and the process instance are matched to realize the reuse of process knowledge.(4)Developed a process information rapid matching function moduleIn order to realize the rapid matching of process information of key components of marine diesel engines,the diesel engine cylinder head and connecting rod are used as test objects,and the MPD-Processer software for 3D machining process design system is used as the platform.On this basis,three module components of process knowledge ontology creation,process knowledge mining,and process instance retrieval and knowledge reuse are developed.Combining the application effect of each functional module,the effectiveness and practicality of the method proposed in this thesis are verified.