Research On Key Technology Of Automobile Harness Intelligent CAPP System

Before the production of automobile wiring harness,it is necessary to carry out process design of each link,including wire cutting process design,inline process planning,wiring harness pre-assembly planning,etc.Finally,various process cards are used to guide the production of wiring harness.This paper takes the process of wiring harness process design as the object,studies three key technologies of wiring harness product information identification and input,inline process planning and pre assembly process planning,establishes the corresponding mathematical model and uses the applicable intelligent optimization algorithm to solve the problem,which improves the intelligent degree of the existing CAPP system.The main research contents include the following three aspects:(1)The process and method of connector wiring information recognition based on deep neural network are proposed.Firstly,the position of connector in wiring harness design drawing is identified by using YOLO deep neural network,and the connector image containing wiring label information is extracted.Then,the connector is detected twice by using YOLO network,and the area of wiring information corresponding to each hole of connector is segmented,Finally,Fast R-CNN is used to recognize the number,color and other character information of the circuit connected to the hole.(2)Based on the idea of loop splitting,the process model of wire in-line lapping branch is transformed into a combinatorial optimization model of wire head in-line point allocation;Then,the framework of differential evolution algorithm is used to solve the lowest cost in-line programming scheme with the material cost as the objective.(3)An assembly efficiency driven intelligent process planning method for wire harness pre-assembly is proposed.The unassembled connector is selected as the main part,and the sub-assembly is extracted by using the connection constraints of circuit and connector;Then,according to the deletion conditions,the circuits and connectors to be deleted are deleted to obtain the simplest subassemblies.After sorting the connectors,select the main components from the non pre-installed connectors in this order for pre-installation.Different sorting of connectors will produce different pre installation planning schemes.According to this model,the problem of wiring harness pre assembly planning is abstracted into the problem of connector permutation and combination optimization.According to the production experience,the number of subassemblies,the number of wires to be connected and the coincidence degree of circuits are selected as the optimization objectives.The analytic hierarchy process is used to score the three factors,and the judgment matrix is constructed to calculate the weight coefficient.Finally,the genetic algorithm is used to solve the preassembly scheme with the highest assembly efficiency.