# US 12,007,377 B2 · System and method for building detection model based on standard value to confirm soldering status > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,007,377 B2 - Document ID: US12007377B2 - Status: Granted - Kind: B2 - Application number: 17/013,663 - Filed: 2020-09-07 - Published: 2021-03-11 - Granted: 2024-06-11 - Earliest claimed: 2019-09-09 - Assignees: Inventec (Pudong) Technology Corporation; Inventec Corporation - Applicants: Inventec (Pudong) Technology Corporation; Inventec Corporation - Inventors: Li Yu - Web page: https://patentgrep.com/patent/US12007377B2 ## CPC - G01N 33/207: Welded or soldered joints; Solderability - G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q) ## Abstract The present disclosure provides a method and a system for building a detection model based on a standard value to confirm a soldering status. The present disclosure generates the solder joint classification based on the standard value data, and uses the detection model to analyze whether the second solder joint data generated by detecting the PCB has poor soldering conditions after building the detection model based on the first solder joint data corresponding to the solder joint included in the solder joint classification, thereby achieving the technical effect of reducing the number of solder joints that are misjudged as poor soldered and shortening the time required for manual re-judgment. ## Claims ### Claim 1 A method for building a detection model based on a standard value to confirm a soldering status, comprising at least: collecting a plurality of standard value data and a plurality of first solder joint data, wherein each of the plurality of first solder joint data corresponds to one of solder joints of an existing PCB layout, and each of the plurality of standard value data corresponds to at least one of the solder joints; generating one or more solder joint classifications according to the plurality of standard value data, wherein each of the solder joint classifications includes at least one of the solder joints; building a detection model based on the first solder joint data corresponding to each of the solder joints included in each of the solder joint classifications; collecting at least one second solder joint data of a to-be-tested PCB with the existing PCB layout, each of the at least one second solder joint data corresponds to one solder joint of the to-be-tested PCB; and sequentially analyzing whether each of the at least one second solder joint data includes poor soldering by using the detection model; wherein the collecting of the plurality of standard value data comprises: collecting upper and lower limits of a volume, area, height, horizontal offset, and vertical offset of each of the solder joints; or collecting a width, height, and area of a plurality of pads including at least one of the solder joints. ### Claim 2 (depends on 1) The method for building a detection model based on a standard value to confirm a soldering status according to claim 1, wherein after sequentially analyzing whether each of the at least one second solder joint data includes poor soldering by using the detection model, the method further comprises: adjusting the one or more solder joint classifications according to a ratio of poor soldering in the plurality of second solder joint data. ### Claim 3 (depends on 1) The method for building a detection model based on a standard value to confirm a soldering status according to claim 1, wherein the building of the detection model based the solder joint data corresponding to each of the solder joints included in each of the solder joint classifications further comprises: training the detection model by using a part of the plurality of first solder joint data, and predicting accuracy by using the other part of the plurality of first solder joint data until the accuracy reaches a predetermined value. ### Claim 4 (depends on 1) The method for building a detection model based on a standard value to confirm a soldering status according to claim 1, wherein the collecting of the at least one second solder joint data comprises: collecting solder joint data corresponding to a solder joint with poor soldering detected by a Solder Paste Inspection (SPI) device. ### Claim 5 A system for building a detection model based on a standard value to confirm a soldering status, comprising at least: a data collection module to collect a plurality of standard value data and a plurality of first solder joint data, and to collect at least one second solder joint data, wherein each of the plurality of first solder joint data corresponds to one of solder joints of an existing PCB layout, and each of the plurality of standard value data corresponds to at least one of the solder joints, wherein each of the at least one second solder joint data corresponds to one solder joint of a to-be-tested PCB with the existing PCB layout; a solder joint classification module to generate one or more solder joint classifications according to the plurality of standard value data, wherein each of the solder joint classifications includes at least one of the solder joints; a model building module to build a detection model based on solder joint data corresponding to each of the solder joints included in each of the solder joint classifications; and a data analysis module to sequentially analyze whether each of the at least one second solder joint data includes poor soldering by using the detection model; wherein the plurality of standard value data comprises: upper and lower limits of a volume, area, height, horizontal offset, and vertical offset of each of the solder joints; or a width, height, and area of a plurality of pads including at least one of the solder joints. ### Claim 6 (depends on 5) The system for building a detection model based on a standard value to confirm a soldering status according to claim 5, wherein the solder joint classification module further adjusts the one or more solder joint classifications according to a ratio of poor soldering in the plurality of second solder joint data. ### Claim 7 (depends on 5) The system for building a detection model based on a standard value to confirm a soldering status according to claim 5, wherein the model building module trains the detection model by using a part of the plurality of first solder joint data, and predicts accuracy by using the other part of the plurality of first solder joint data until the accuracy reaches a predetermined value. ### Claim 8 (depends on 5) The system for building a detection model based on a standard value to confirm a soldering status according to claim 5, wherein the at least one second solder joint data comprises solder joint data corresponding to a solder joint with poor soldering detected by an SPI device.