Method and system for determining maintenance policy of complex forming device

Patent metadata

Field	Value
Publication number	US-10318931-B2
Application number	US-201816137558-A
Country	US
Kind code	B2
Filing date	Sep 21, 2018
Priority date	Sep 22, 2017
Publication date	Jun 11, 2019
Grant date	Jun 11, 2019

Abstract

The present invention discloses a method and system for determining a maintenance policy of a complex forming device. The method and system include: establishing a performance judgement standard; obtaining actual operation data; and determining a performance stage of the complex forming device, and determining the maintenance policy of the complex forming device. The establishing a performance judgement standard includes allocating data located at different performance stages to corresponding working condition units; setting data located at a normal performance stage to normal reference data, and calculating, by using a shortest path algorithm, distances between operation data of different working condition units in different functional modules and the normal reference data, to obtain performance statuses of the different functional modules. In view of this, by using the method and system provided in the present invention, a suitable maintenance policy can be selected for the complex forming device, thereby reducing maintenance costs.

First claim

What is claimed is:

1. A method executed by a computer for determining a maintenance policy of a complex forming device, wherein the maintenance policy determining method comprises:
obtaining a plurality of sample data sets, wherein the sample data sets are sample data sets with maintenance policy labels, and the maintenance policy labels comprise a normal performance label, a caring label, a maintenance label, a repair label, and a replacement label;
obtaining functional modules of the complex forming device and working condition units comprised in the functional modules;
establishing a working condition hierarchical identification model according to the functional modules and the working condition units comprised in the functional modules; and
classifying data items in each sample data set respectively according to the working condition hierarchical identification model, to obtain a plurality of subsets with normal performance labels, a plurality of subsets with caring labels, a plurality of subsets with maintenance labels, a plurality of subsets with repair labels, and a plurality of subsets with replacement labels, wherein the data items in each subset is operation data describing each working condition unit; specifically, classifying the data items in each sample data set respectively according to a description requirement of each functional module in the working condition hierarchical identification model, to obtain a plurality of feature subsets with normal performance labels, a plurality of feature subsets with caring labels, a plurality of feature subsets with maintenance labels, a plurality of feature subsets with repair labels, and a plurality of feature subsets with replacement labels, wherein the data item in the feature subset is operation data describing the functional module; obtaining operating status sample data of each working condition unit; using a neural network allocation algorithm according to the operating status sample data of each working condition unit to obtain a trained neural network model; inputting each feature subset with a normal performance label, each feature subset with a caring label, each feature subset with a maintenance label, each feature subset with a repair label, and each feature subset with a replacement label into the trained neural network model, to obtain a plurality of subsets with normal performance labels, a plurality of subsets with caring labels, a plurality of subsets with maintenance labels, a plurality of subsets with repair labels, and a plurality of subsets with replacement labels;
processing each subset with a normal performance label to obtain a plurality of reference subsets, wherein the data item in the reference subset is operation data describing each working condition unit in a normal performance status of the complex forming device; specifically, each subset with a normal performance label comprises a plurality of numeric data items and a plurality of non-numeric data items; for the numeric data items, calculating an average value of the plurality of numeric data items; for the non-numeric data items, obtaining a standard value of the plurality of non-numeric data items by using a mode method; obtaining the reference subset of each working condition unit according to the average value and the standard value;
separately calculating distances between each reference subset and each subset with a normal performance label, each subset with a caring label, each subset with a maintenance label, each subset with a repair label, and each subset with a replacement label by using a shortest path algorithm, and determining a performance set with a normal performance label, a performance set with a caring label, a performance set with a maintenance label, a performance set with a repair label, and a performance set with a replacement label of the complex forming device, wherein elements in the performance sets are the distances between the reference subset and the subset with a normal performance label, the subset with a caring label, the subset with a maintenance label, the subset with a repair label, and the subset with a replacement label respectively; specifically, separately calculating the distances between each reference subset and each subset by using the shortest path algorithm; obtaining a weight of each distance according to a cross validation algorithm; determining the performance set with a normal performance label, the performance set with a caring label, the performance set with a maintenance label, the performance set with a repair label, and the performance set with a replacement label of the complex forming device according to the distances between the reference subset and each subset and the weights of the distances;
obtaining actual operation data of a working cycle in the complex forming device;
determining an actual operation performance set of the complex forming device according to the actual operation data, the working condition hierarchical identification model, and the shortest path algorithm; specifically, classifying the actual operation data according to the working condition hierarchical identification model, to obtain a plurality of subsets of the actual operation data; calculating, by using the shortest path algorithm, distances between the reference subset and each subset of the actual operation data, and determining the actual operation performance set of the complex forming device;
separately calculating Euclidean distances between the actual operation performance set and the performance set with a normal performance label, the performance set with a caring label, the performance set with a maintenance label, the performance set with a repair label, and the performance set with a replacement label, and selecting Euclidean distances within a specified threshold range;
determining a winning Euclidean distance according to the selected Euclidean distances by using a voting method; and
according to the winning Euclidean distance, determining a label, with a maintenance policy, of the sample data set corresponding to the winning Euclidean distance, and determining the maintenance policy denoted in the label with the maintenance policy as the maintenance policy of the complex forming device.

2. A system for determining a maintenance policy of a complex forming device, wherein the maintenance policy determining system comprises:
an obtaining part and a processing part; wherein the obtaining part is communicatively connected to the processing part; and
the processing part comprises one or more non-volatile memories, and a processor, wherein the processor comprises a sample data set obtaining module configured to obtain a plurality of sample data sets by using the obtaining part; and the sample data sets are sample data sets with maintenance policy labels, and the maintenance policy labels comprise a normal performance label, a caring label, a maintenance label, a repair label, and a replacement label;
a functional module and working condition unit obtaining module configured to obtain functional modules of the complex forming device and working condition units comprised in the functional modules by using the obtaining part;
a working condition hierarchical identification model establishment module configured to establish a working condition hierarchical identification model according to the functional modules and the working condition units comprised in the functional modules;
a subset obtaining module configured to classify data items in each sample data set respectively according to the working condition hierarchical identification model, to obtain a plurality of subsets with normal performance labels, a plurality of subsets with caring labels, a plurality of subsets with maintenance labels, a plurality of subsets with repair labels, and a plurality of subsets with replacement labels, wherein the dat

Assignees

• Univ Hefei Technology

Inventors

• Zhang Qiang
• Wang Wanying
• Peng Zhanglin
• Lu Xiaonong
• Pei Jun
• Fan Yu
• Yan Jianwen

Classifications

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06Q10/20 Primary

  Administration of product repair or maintenance · CPC title

• G06N3/08

  Learning methods · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/09

  Supervised learning · CPC title