CN115616386A - Flying probe testing method based on artificial intelligence - Google Patents
Flying probe testing method based on artificial intelligence Download PDFInfo
- Publication number
- CN115616386A CN115616386A CN202211444643.6A CN202211444643A CN115616386A CN 115616386 A CN115616386 A CN 115616386A CN 202211444643 A CN202211444643 A CN 202211444643A CN 115616386 A CN115616386 A CN 115616386A
- Authority
- CN
- China
- Prior art keywords
- pcb
- data
- component
- target
- analysis module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/2801—Testing of printed circuits, backplanes, motherboards, hybrid circuits or carriers for multichip packages [MCP]
- G01R31/281—Specific types of tests or tests for a specific type of fault, e.g. thermal mapping, shorts testing
- G01R31/2817—Environmental-, stress-, or burn-in tests
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/2801—Testing of printed circuits, backplanes, motherboards, hybrid circuits or carriers for multichip packages [MCP]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Microelectronics & Electronic Packaging (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Tests Of Electronic Circuits (AREA)
Abstract
The invention relates to a flying probe testing method based on artificial intelligence, which comprises the following steps: the first analysis module constructs a high-quality domain of the PCB according to the element bit sequence data of the PCB and the specification working data of all electronic elements of the PCB. And the second analysis module inputs the element bit sequence data of the PCB and the element position data and the element reading data of all electronic elements of the PCB into the circuit board analysis model so as to output the circuit board state characteristics of the PCB. And the second analysis module acquires the predicted state characteristic of the PCB according to the state characteristic of the circuit board and the element bit sequence data and acquires the circuit board aging direction of the PCB according to the state characteristic of the circuit board and the predicted state characteristic. The third analysis module acquires real-time state points of the PCB according to the state characteristics of the circuit board, continuously ages the real-time state points of the PCB in the high-quality domain based on the aging direction of the circuit board to obtain critical points of the high-quality domain, and then calculates the distance between the real-time state points of the PCB and the critical points of the high-quality domain.
Description
Technical Field
The invention relates to the field of artificial intelligence and flying probe testing, in particular to a flying probe testing method based on artificial intelligence.
Background
In the production of PCB boards, the quality of PCB boards does not meet the design requirements for various reasons. Therefore, the PCB needs to be tested before the PCB is shipped from a factory to ensure the quality of the PCB.
The PCB is tested, and various data information is involved. Patent publication No. (CN 110736911B) "flying needle test method, flying needle test apparatus, flying needle test device, and storage medium" discloses dividing a PCB board into a plurality of auxiliary areas, selecting an auxiliary point for each auxiliary area, and then accurately positioning the positioning point of a flying needle test probe according to the auxiliary point, thereby improving the accuracy of a PCB open point.
Therefore, in the prior art, a needle bed is replaced by a test probe of a flying needle test device, and a plurality of electric probes which are driven by a motor and can move rapidly are used for contacting pins of a device and performing electric measurement to determine whether an open point exists on a PCB so as to judge whether the PCB can have a fault. However, some PCBs have poor component quality or poor component contact, which may cause the PCBs to have no open point at the current moment and thus can be used normally, but the PCBs have short service lives and cannot be used for a long time under various working conditions. Therefore, detecting a faulty PCB by only the flying probe test may cause a user to lose the quality and life of the PCB.
Along with the use of the PCB board is more and more extensive, how to design a flying probe test result through the PCB board at the present moment, through the reliability of artificial intelligence technique prediction PCB board long-term service under all kinds of operating modes, promptly, select the PCB board that the quality is poor short-lived according to the flying probe test result at the present moment through artificial intelligence technique is the problem that technical staff in the field await solution urgently.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a flying probe testing method based on artificial intelligence, which comprises the following steps: the test terminal sends circuit principle data of the PCB to the artificial intelligence platform; a first analysis module of the artificial intelligence platform performs component extraction and bit sequence analysis on the circuit principle data of the PCB to obtain component bit sequence data of the PCB, and obtains all electronic components of the PCB according to the component bit sequence data;
the first analysis module acquires the standard working data of all electronic elements of the PCB from the database, and constructs a high-quality domain of the PCB according to the element bit sequence data of the PCB and the standard working data of all electronic elements of the PCB;
a test probe of the flying probe test equipment acquires the PCB serial number of the PCB, the element position data and the element reading data of all electronic elements of the PCB and sends the element position data and the element reading data to the artificial intelligence platform;
the second analysis module generates first training data and second training data from a database according to circuit principle data of the PCB, performs model training according to the first training data and the second training data to obtain a circuit board analysis model, and inputs element bit sequence data of the PCB, element position data of all electronic elements of the PCB and element reading data into the circuit board analysis model to output circuit board state characteristics of the PCB;
the second analysis module acquires the predicted state characteristic of the PCB according to the circuit board state characteristic and the element bit sequence data of the PCB, and acquires the circuit board aging direction of the PCB according to the circuit board state characteristic and the predicted state characteristic of the PCB;
the third analysis module acquires real-time state points of the PCB according to the circuit board state characteristics of the PCB, continuously ages the real-time state points of the PCB in a high-quality domain based on the circuit board aging direction to obtain critical points of the high-quality domain, and then calculates the distance between the real-time state points of the PCB and the critical points of the high-quality domain;
the third analysis module normalizes the distance between the real-time state point of the PCB and the critical point of the high-quality domain to obtain the quality value of the PCB, and compares the quality value with a quality threshold value; and sending the PCB number and the quality value of the PCB to a test terminal when the quality value is smaller than the quality threshold.
According to a preferred embodiment, the circuit principle data is a circuit diagram of the PCB board and a model number of each electronic component in the circuit diagram; the element bit sequence data is used for representing the element distribution condition and the element connection relation of all electronic elements of the PCB.
According to a preferred embodiment, the component position data is a position parameter of a test point positioned by a test probe of the flying probe test equipment; the element reading data is voltage reading, current reading and measured reading of the electronic element collected by a test probe of the flying probe test equipment.
According to a preferred embodiment, the test terminal is an intelligent device with communication function and data transmission function used by a tester, and comprises: smart phones, laptops, tablets, and desktops.
According to a preferred embodiment, the specification operating data of the electronic component comprises a nominal value of the electronic component, a maximum current and a maximum voltage of the electronic component during normal operation, a minimum current and a minimum voltage of the electronic component during normal operation.
According to a preferred embodiment, the constructing a high quality domain of the PCB board according to the component bit sequence data of the PCB board and the specification working data of all electronic components of the PCB board by the first analysis module comprises:
the first analysis module randomly selects one electronic element on the PCB as a target element, acquires the electronic element with circuit connection with the target element according to element bit sequence data of the PCB, and then takes the electronic element with circuit connection with the target element as an adjacent element of the target element;
the first analysis module constructs a working constraint function of the target element based on the standard working data of the target element, and acquires a plurality of element aging directions of the target element according to the standard working data of the target element and the standard working data of each adjacent element of the target element;
the first analysis module continuously ages the target element according to the working constraint function of the target element and the aging directions of the plurality of elements of the target element to obtain a standard working domain of the target element;
the first analysis module traverses all electronic elements of the PCB to obtain a standard working domain of each electronic element of the PCB;
the first analysis module generates a high-quality domain of the PCB according to the standard working domain of each electronic element of the PCB and the element bit sequence data of the PCB.
According to a preferred embodiment, the first analysis module obtaining a plurality of component aging directions of the target component from the normative working data of the target component and the normative working data of each of the neighboring components of the target component comprises:
the first analysis module extracts the characteristics of the standard working data of the target element and maps the characteristics of the standard working data of the target element to a multidimensional vector space to obtain a standard characteristic vector of the target element;
the first analysis module randomly selects one adjacent element of the target element as an object element, extracts the characteristics of the standard working data of the object element, and then maps the characteristics of the standard working data of the object element to a multi-dimensional vector space to obtain the standard characteristic vector of the object element;
the first analysis module calculates a characteristic included angle between the target element and the object element according to the standard eigenvector of the target element and the standard eigenvector of the object element, and determines an element aging direction of the target element on the object element connecting line according to the characteristic included angle;
the first analysis module traverses all adjacent elements of the target element to obtain an element aging direction of the target element on each adjacent element connection line.
According to a preferred embodiment, the step of obtaining the standard working domain of the target element by the first analysis module according to the aging direction of the target element comprises the following steps:
the method comprises the steps that a first analysis module randomly selects one element aging direction of a target element as a target aging direction, continuously ages the target element in the target aging direction, and then judges whether to stop aging the target element according to a working constraint function of the target element;
when the target element is stopped being aged, the first analysis module obtains element aging characteristics of the target element and maps the element aging characteristics to a multi-dimensional characteristic space to obtain an aging point of the target element in a target aging direction;
the first analysis module traverses all element aging directions of the target element to obtain an aging point of the target element in each element aging direction;
the first analysis module connects the aging points of the target elements in each element aging direction to obtain the standard working domain of the target elements.
According to a preferred embodiment, the inputting, by the second analysis module, the component bit sequence data of the PCB and the component position data and the component reading data of all electronic components of the PCB into the circuit board analysis model to output the circuit board status characteristic of the PCB comprises:
extracting the component position characteristic and the component reading characteristic of each electronic component from the component position data and the component reading data of each electronic component by utilizing two convolution kernels, and carrying out convolution operation on the component position characteristic and the component reading characteristic of each electronic component to capture the relevance between the component position and the component reading of each electronic component so as to obtain the component relevance characteristic of each electronic component;
extracting element bit sequence characteristics of element bit sequence data by using a convolution kernel in a convolution layer of the circuit board analysis model;
pooling the element association characteristics of each electronic element by a pooling layer of the circuit board analysis model to compress and remove redundancy of the element association characteristics of each electronic element, and performing characteristic fusion on the element association characteristics of all the electronic elements according to element bit sequence characteristics to obtain first state characteristics of the PCB;
the deconvolution layer of the circuit board analysis model performs transposition convolution on the first state characteristic of the PCB to obtain a second state characteristic of the PCB;
and the anti-pooling layer of the circuit board analysis model performs anti-pooling on the second state characteristic to obtain a circuit board state characteristic of the PCB.
According to a preferred embodiment, the third analysis module normalizing the distance between the real-time status point of the PCB and the critical point of the high-quality domain to obtain the quality value of the PCB comprises:
wherein R is the mass coefficient, S is the mass value of the PCB,
the distance between the real-time state point of the PCB and the critical point of the high-quality domain,
the radius of the high quality domain.
The invention has the following beneficial effects: the PCB is tested by the test probe of the flying probe test equipment, so that the element position data and the element reading data of each electronic element on the PCB can be flexibly and quickly acquired.
In addition, the flying probe test result of the PCB at the current moment is obtained through the flying probe test equipment, and the flying probe test result at the current moment is analyzed through the artificial intelligence technology to obtain the long-term service reliability of the PCB under various working conditions, so that defective products without faults but with poor quality are screened out, and the condition that the loss is brought to users due to the use of the defective PCB is reduced.
On the other hand, this application combines flying probe test technique and artificial intelligence technique, thereby not only realize the flying probe test result of quick nimble collection PCB board but also can carry out the analysis to flying probe test result through artificial intelligence technique and elect the wastrel of PCB board.
Drawings
FIG. 1 is a flow chart of a method for artificial intelligence based flying probe testing according to an exemplary embodiment.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It is to be understood that although the terms first, second, third, etc. may be used herein to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. The word "if" as used herein may be interpreted as "at" \8230; "or" when 8230; \8230; "or" in response to a determination ", depending on the context.
Referring to FIG. 1, in one embodiment, an artificial intelligence based flying probe testing method may include:
s1, a test terminal sends circuit principle data of a PCB to an artificial intelligence platform; a first analysis module of the artificial intelligence platform performs component extraction and bit sequence analysis on the circuit principle data of the PCB to obtain component bit sequence data of the PCB, and obtains all electronic components of the PCB according to the component bit sequence data.
Optionally, the circuit principle data is a circuit diagram of the PCB and a model number of each electronic component in the circuit diagram; the element bit sequence data is used for representing the element distribution condition and the element connection relation of all electronic elements of the PCB.
The component distribution condition of the electronic component is relative position information and absolute position information of the electronic component distributed on the PCB, the relative position information is a relative position parameter of the electronic component and the surrounding electronic component, and the absolute position information is an absolute position parameter of the electronic component on the PCB.
The element connection relationship is used for describing the connection relationship between the electronic element and the electronic element on the PCB.
Electronic components include rectifier diodes, switching diodes, resistors, capacitors, photocouplers, transformers, operational amplifiers, etc., rectifier diodes:
common models of electronic components, rectifier diodes: 1N4001 1N4001; a switching diode: 1N4148 1N4150 1N4448; photoelectric coupler: 4N35 4N36 N37; and a transistor PNP:8050 9015A 92; and a transistor NPN:9012 9013 9014 9015 9018.
Optionally, the test terminal is an intelligent device with a communication function and a data transmission function, and the intelligent device is used by a tester and includes: smart phones, laptops, tablets, and desktops.
S2, the first analysis module acquires the standard working data of all electronic elements of the PCB from the database, and constructs a high-quality domain of the PCB according to the element bit sequence data of the PCB and the standard working data of all electronic elements of the PCB.
Optionally, the specification operating data of the electronic component includes a nominal value of the electronic component, a maximum current and a maximum voltage of the electronic component during normal operation, and a minimum current and a minimum voltage of the electronic component during normal operation.
Alternatively, the nominal value of the electronic component is a value specified according to the actual situation of the electronic component as a standard value of the product, namely the nominal value of the electronic component.
In a specific embodiment, the constructing, by the first analysis module, the high quality domain of the PCB according to the component bit sequence data of the PCB and the specification working data of all electronic components of the PCB comprises:
the first analysis module randomly selects one electronic element on the PCB as a target element, acquires the electronic element with circuit connection with the target element according to element bit sequence data of the PCB, and then takes the electronic element with circuit connection with the target element as an adjacent element of the target element;
the first analysis module constructs a working constraint function of the target element based on the standard working data of the target element, and acquires a plurality of element aging directions of the target element according to the standard working data of the target element and the standard working data of each adjacent element of the target element;
the first analysis module continuously ages the target element according to the working constraint function of the target element and the aging directions of the plurality of elements of the target element to obtain a standard working domain of the target element;
the first analysis module traverses all electronic elements of the PCB to obtain a standard working domain of each electronic element of the PCB;
the first analysis module generates a high-quality domain of the PCB according to the standard working domain of each electronic element of the PCB and the element bit sequence data of the PCB.
Optionally, the operating constraint function of the target element is used to represent a constraint relationship between the voltage reading, the current reading and the element reading when the target element is not damaged.
In a specific embodiment, the obtaining, by the first analysis module, a plurality of component aging directions of the target component according to the specification working data of the target component and the specification working data of each adjacent component of the target component includes:
the first analysis module extracts the characteristics of the standard working data of the target element and maps the characteristics of the standard working data of the target element to a multidimensional vector space to obtain a standard characteristic vector of the target element;
the first analysis module randomly selects one adjacent element of the target element as an object element, extracts the characteristics of the standard working data of the object element, and then maps the characteristics of the standard working data of the object element to a multi-dimensional vector space to obtain the standard characteristic vector of the object element;
the first analysis module calculates a characteristic included angle between the target element and the object element according to the standard eigenvector of the target element and the standard eigenvector of the object element, and determines an element aging direction of the target element on the object element connecting line according to the characteristic included angle;
the first analysis module traverses all adjacent elements of the target element to obtain an element aging direction of the target element on each adjacent element connection line.
In a specific embodiment, the obtaining, by the first analysis module, the standard working domain of the target element by continuously aging the target element according to a plurality of element aging directions of the target element includes:
the method comprises the steps that a first analysis module randomly selects one element aging direction of a target element as a target aging direction, continuously ages the target element in the target aging direction, and then judges whether to stop aging the target element according to a working constraint function of the target element;
when the target element is stopped being aged, the first analysis module obtains element aging characteristics of the target element and maps the element aging characteristics to a multi-dimensional characteristic space to obtain an aging point of the target element in a target aging direction;
the first analysis module traverses all element aging directions of the target element to obtain an aging point of the target element in each element aging direction;
the first analysis module connects the aging points of the target elements in each element aging direction to obtain the standard working domain of the target elements.
Optionally, the aging point is used to characterize the state of the target element when it fails to function properly and is damaged.
And S3, acquiring the PCB serial number of the PCB and the element position data and the element reading data of all electronic elements of the PCB by a test probe of the flying probe test equipment, and sending the element position data and the element reading data to the artificial intelligence platform.
Optionally, the component position data is a position parameter of a test point of the test electronic component located by a test probe of the flying probe test equipment; the element reading data is voltage reading, current reading and measured reading of the electronic element collected by a test probe of the flying probe test equipment.
Optionally, the component position data of the electronic component is an absolute position parameter of a test point on the PCB board when the test probe tests the electronic component.
Optionally, in one embodiment, the test probe obtains a voltage reading, a current reading, and actual values of the resistance and the resistance across the resistor; the test probe obtains a voltage reading, a current reading, and an actual value of the capacitance across the capacitance.
Optionally, the PCB board number is used to uniquely identify the PCB board.
The PCB is tested by the test probe of the flying probe test equipment, so that the element position data and the element reading data of each electronic element on the PCB can be flexibly and quickly acquired. In addition, the present application prevents reading interference by electrically shielding other electronic components while the test probes test the component reading data of the electronic components.
And S4, the second analysis module generates first training data and second training data from a database according to the circuit principle data of the PCB, performs model training according to the first training data and the second training data to obtain a circuit board analysis model, and inputs element bit sequence data of the PCB and element position data and element reading data of all electronic elements of the PCB into the circuit board analysis model to output circuit board state characteristics of the PCB.
Optionally, the circuit board analysis model comprises: convolutional layers, pooling layers, anti-convolutional layers, and anti-pooling layers.
In a specific embodiment, the inputting, by the second analysis module, the component bit sequence data of the PCB and the component position data and the component reading data of all electronic components of the PCB into the circuit board analysis model to output the circuit board status characteristic of the PCB comprises:
extracting the component position characteristic and the component reading characteristic of each electronic component from the component position data and the component reading data of each electronic component by using two convolution kernels by the convolution layer of the circuit board analysis model, and carrying out convolution operation on the component position characteristic and the component reading characteristic of each electronic component to capture the relevance between the component position and the component reading of each electronic component so as to obtain the component relevance characteristic of each electronic component;
extracting element bit sequence characteristics of element bit sequence data by using convolution kernels from convolution layers of the circuit board analysis model;
pooling the element association characteristics of each electronic element by a pooling layer of the circuit board analysis model to compress and remove redundancy of the element association characteristics of each electronic element, and performing characteristic fusion on the element association characteristics of all the electronic elements according to element bit sequence characteristics to obtain first state characteristics of the PCB;
the deconvolution layer of the circuit board analysis model performs transposition convolution on the first state characteristic of the PCB to obtain a second state characteristic of the PCB;
and the anti-pooling layer of the circuit board analysis model performs anti-pooling on the second state characteristic to obtain the circuit board state characteristic of the PCB.
And S5, the second analysis module acquires the predicted state characteristic of the PCB according to the circuit board state characteristic of the PCB and the element bit sequence data, and acquires the circuit board aging direction of the PCB according to the circuit board state characteristic of the PCB and the predicted state characteristic.
Optionally, the predicted state feature of the PCB is a state feature of the PCB at the next time; the circuit board state characteristic of the PCB is the state characteristic of the PCB at the current moment.
S6, the third analysis module acquires real-time state points of the PCB according to the circuit board state characteristics of the PCB, continuously ages the real-time state points of the PCB in the high-quality domain based on the circuit board aging direction to obtain critical points of the high-quality domain, and then calculates the distance between the real-time state points of the PCB and the critical points of the high-quality domain.
Optionally, the critical point of the high-quality domain is a state point of the PCB in a state where the PCB cannot work normally, and the state point of the PCB is used for representing a state characteristic of the PCB. The real-time state point of the PCB is used for representing the state characteristic of the PCB at the current moment.
Optionally, the distance between the real-time status point of the PCB and the critical point of the high-quality domain is used to represent the characteristic distance between the status feature of the current status of the PCB and the status feature of the PCB when the PCB cannot work normally.
S7, the third analysis module normalizes the distance between the real-time state point of the PCB and the critical point of the high-quality domain to obtain the quality value of the PCB, and compares the quality value with a quality threshold value; and transmitting the PCB number and the quality value of the PCB to a test terminal when the quality value is smaller than the quality threshold.
In a specific embodiment, the third analysis module normalizes the distance between the real-time status point of the PCB and the critical point of the high-quality domain to obtain the quality value of the PCB includes:
wherein R is the mass coefficient, S is the mass value of the PCB,
the distance between the real-time state point of the PCB and the critical point of the high-quality domain,
is the radius of the high quality domain.
Optionally, the mass coefficient is preset according to actual conditions, and the radius of the high-mass domain is the distance from the center of the high-mass domain to the critical point of the high-mass domain.
Optionally, the quality value of the PCB is used to indicate the quality and the service life of the PCB and the reliability of the PCB in long-term service under various working conditions.
Optionally, the quality threshold is preset according to actual conditions, for example, a PCB that a certain factory wishes to leave a factory can last for at least 5 years under various working conditions, then the quality threshold is calculated according to a 5-year lifetime, a PCB with a quality value smaller than the quality threshold is likely not to last for 5 years under various working conditions, and at this time, a PCB with a quality value smaller than the quality threshold is selected as a defective product.
Thereby this application tests the PCB board through flying probe test equipment's test probe and realizes nimble quick acquisition PCB board on every electronic component's component position data and component reading data.
In addition, the flying probe test result of the PCB at the current moment is obtained through the flying probe test equipment, and the flying probe test result at the current moment is analyzed through the artificial intelligence technology to obtain the long-term service reliability of the PCB under various working conditions, so that defective products without faults but with poor quality are screened out, and the condition that the user is damaged due to the fact that the defective PCB is used is reduced.
On the other hand, this application combines flying probe test technique and artificial intelligence technique, thereby not only realize the flying probe test result of quick nimble collection PCB board but also can carry out the analysis to flying probe test result through artificial intelligence technique and elect the wastrel of PCB board.
In addition, functional units in the embodiments herein may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solutions in the present invention substantially or partially contribute to the prior art, or all or part of the technical solutions may be embodied in the form of a software product stored in a storage medium, which includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the methods described in the embodiments of the present invention, and the foregoing storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, and various media capable of storing program codes.
The principles and embodiments of this document are explained herein using specific examples, which are presented only to aid in understanding the methods and their core concepts; meanwhile, for the general technical personnel in the field, according to the idea of this document, there may be changes in the concrete implementation and the application scope, in summary, this description should not be understood as the limitation of this document.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A flying probe testing method based on artificial intelligence is characterized in that a testing terminal sends circuit principle data of a PCB to an artificial intelligence platform; a first analysis module of the artificial intelligence platform performs component extraction and bit sequence analysis on the circuit principle data of the PCB to obtain component bit sequence data of the PCB, and obtains all electronic components of the PCB according to the component bit sequence data;
the first analysis module acquires the standard working data of all electronic elements of the PCB from the database, and constructs a high-quality domain of the PCB according to the element bit sequence data of the PCB and the standard working data of all electronic elements of the PCB;
a test probe of the flying probe test equipment acquires the PCB serial number of the PCB, and element position data and element reading data of all electronic elements of the PCB and sends the element position data and the element reading data to the artificial intelligence platform;
the second analysis module generates first training data and second training data from a database according to circuit principle data of the PCB, performs model training according to the first training data and the second training data to obtain a circuit board analysis model, and inputs element bit sequence data of the PCB, element position data of all electronic elements of the PCB and element reading data into the circuit board analysis model to output circuit board state characteristics of the PCB;
the second analysis module acquires the predicted state characteristic of the PCB according to the circuit board state characteristic and the element bit sequence data of the PCB, and acquires the circuit board aging direction of the PCB according to the circuit board state characteristic and the predicted state characteristic of the PCB;
the third analysis module acquires real-time state points of the PCB according to the circuit board state characteristics of the PCB, continuously ages the real-time state points of the PCB in a high-quality domain based on the circuit board aging direction to obtain critical points of the high-quality domain, and then calculates the distance between the real-time state points of the PCB and the critical points of the high-quality domain;
the third analysis module normalizes the distance between the real-time state point of the PCB and the critical point of the high-quality domain to obtain the quality value of the PCB, and compares the quality value with a quality threshold; and transmitting the PCB number and the quality value of the PCB to a test terminal when the quality value is smaller than the quality threshold.
2. The method of claim 1, wherein the circuit principle data is a circuit diagram of the PCB board and a model number of each electronic component in the circuit diagram; the element bit sequence data is used for representing the element distribution condition and the element connection relation of all electronic elements of the PCB.
3. The method of claim 2, wherein the component position data is a position parameter of a test point located by a test probe of the flying probe test equipment; the element reading data is voltage reading, current reading and measured reading of the electronic element collected by a test probe of the flying probe test equipment.
4. The method according to claim 3, wherein the test terminal is a smart device with communication function and data transmission function used by a tester, and comprises: smart phones, laptops, tablets, and desktops.
5. The method of claim 4, wherein the specification operating data of the electronic component comprises a nominal value of the electronic component, a maximum current and a maximum voltage of the electronic component during normal operation, and a minimum current and a minimum voltage of the electronic component during normal operation.
6. The method of claim 5, wherein the first analysis module constructing the high quality domain of the PCB board according to the component bit sequence data of the PCB board and the specification working data of all electronic components of the PCB board comprises:
the first analysis module randomly selects one electronic element on the PCB as a target element, acquires the electronic element with circuit connection with the target element according to element bit sequence data of the PCB, and then takes the electronic element with circuit connection with the target element as an adjacent element of the target element;
the first analysis module constructs a working constraint function of the target element based on the standard working data of the target element, and acquires a plurality of element aging directions of the target element according to the standard working data of the target element and the standard working data of each adjacent element of the target element;
the first analysis module continuously ages the target element according to the working constraint function of the target element and the aging directions of the plurality of elements of the target element to obtain a standard working domain of the target element;
the first analysis module traverses all electronic elements of the PCB to obtain a standard working domain of each electronic element of the PCB;
the first analysis module generates a high-quality domain of the PCB according to the standard work domain of each electronic element of the PCB and the element bit sequence data of the PCB.
7. The method of claim 6, wherein the first analysis module obtaining a number of component aging directions for the target component based on the canonical operating data for the target component and the canonical operating data for each neighboring component of the target component comprises:
the first analysis module extracts the characteristics of the standard working data of the target element and maps the characteristics of the standard working data of the target element to a multidimensional vector space to obtain a standard characteristic vector of the target element;
the first analysis module randomly selects one adjacent element of the target element as an object element, extracts the characteristics of the standard working data of the object element, and then maps the characteristics of the standard working data of the object element to a multi-dimensional vector space to obtain the standard characteristic vector of the object element;
the first analysis module calculates a characteristic included angle between the target element and the object element according to the standard eigenvector of the target element and the standard eigenvector of the object element, and determines an element aging direction of the target element on the object element connecting line according to the characteristic included angle;
the first analysis module traverses all adjacent elements of the target element to obtain an element aging direction of the target element on each adjacent element connection line.
8. The method of claim 7, wherein the step of continuously aging the target component by the first analysis module according to the plurality of component aging directions of the target component to obtain the standard working domain of the target component comprises:
the method comprises the steps that a first analysis module randomly selects one element aging direction of a target element as a target aging direction, continuously ages the target element in the target aging direction, and then judges whether to stop aging the target element according to a working constraint function of the target element;
when the target element is stopped being aged, the first analysis module obtains element aging characteristics of the target element and maps the element aging characteristics to a multi-dimensional characteristic space to obtain an aging point of the target element in a target aging direction;
the first analysis module traverses all element aging directions of the target element to obtain an aging point of the target element in each element aging direction;
the first analysis module connects the aging points of the target elements in each element aging direction to obtain the standard working domain of the target elements.
9. The method of claim 8, wherein the second analysis module inputting the component bit sequence data of the PCB and the component position data and the component reading data of all electronic components of the PCB into the circuit board analysis model to output the circuit board status characteristics of the PCB comprises:
extracting the component position characteristic and the component reading characteristic of each electronic component from the component position data and the component reading data of each electronic component by utilizing two convolution kernels, and carrying out convolution operation on the component position characteristic and the component reading characteristic of each electronic component to capture the relevance between the component position and the component reading of each electronic component so as to obtain the component relevance characteristic of each electronic component;
extracting element bit sequence characteristics of element bit sequence data by using a convolution kernel in a convolution layer of the circuit board analysis model;
pooling the element association characteristics of each electronic element by a pooling layer of the circuit board analysis model to compress and remove redundancy of the element association characteristics of each electronic element, and performing characteristic fusion on the element association characteristics of all the electronic elements according to element bit sequence characteristics to obtain first state characteristics of the PCB;
the deconvolution layer of the circuit board analysis model performs transposition convolution on the first state characteristic of the PCB to obtain a second state characteristic of the PCB;
and the anti-pooling layer of the circuit board analysis model performs anti-pooling on the second state characteristic to obtain the circuit board state characteristic of the PCB.
10. The method of claim 9, wherein the third analysis module normalizing the distance between the real-time status point of the PCB and the critical point of the high-quality domain to obtain the quality value of the PCB comprises:
wherein R is the mass coefficient, S is the mass value of the PCB,
the distance between the real-time state point of the PCB and the critical point of the high-quality domain,
is the radius of the high quality domain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211444643.6A CN115616386B (en) | 2022-11-18 | 2022-11-18 | Flying probe testing method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211444643.6A CN115616386B (en) | 2022-11-18 | 2022-11-18 | Flying probe testing method based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115616386A true CN115616386A (en) | 2023-01-17 |
| CN115616386B CN115616386B (en) | 2023-03-14 |
Family
ID=84878478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211444643.6A Active CN115616386B (en) | 2022-11-18 | 2022-11-18 | Flying probe testing method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115616386B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| BRMU8702892U2 (en) * | 2007-10-08 | 2010-05-18 | Solectron Ind Com Serv E Expor | device for automated inspection of component placement faults on printed circuit boards by statistical data training |
| CN103165203A (en) * | 2011-12-14 | 2013-06-19 | 中国广东核电集团有限公司 | Detection method for nuclear power station circuit board component |
| CN106815425A (en) * | 2017-01-12 | 2017-06-09 | 侯海亭 | Multi-layer PCB board fault detection method and system |
| CN108181572A (en) * | 2017-12-29 | 2018-06-19 | 大族激光科技产业集团股份有限公司 | Flying probe tester test method, device, computer equipment and storage medium |
| CN112053318A (en) * | 2020-07-20 | 2020-12-08 | 清华大学 | Two-dimensional PCB defect real-time automatic detection and classification device based on deep learning |
-
2022
- 2022-11-18 CN CN202211444643.6A patent/CN115616386B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| BRMU8702892U2 (en) * | 2007-10-08 | 2010-05-18 | Solectron Ind Com Serv E Expor | device for automated inspection of component placement faults on printed circuit boards by statistical data training |
| CN103165203A (en) * | 2011-12-14 | 2013-06-19 | 中国广东核电集团有限公司 | Detection method for nuclear power station circuit board component |
| CN106815425A (en) * | 2017-01-12 | 2017-06-09 | 侯海亭 | Multi-layer PCB board fault detection method and system |
| CN108181572A (en) * | 2017-12-29 | 2018-06-19 | 大族激光科技产业集团股份有限公司 | Flying probe tester test method, device, computer equipment and storage medium |
| CN112053318A (en) * | 2020-07-20 | 2020-12-08 | 清华大学 | Two-dimensional PCB defect real-time automatic detection and classification device based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 孙扩等: "基于SVM专家系统的模拟电路PCB故障诊断研究", 《国外电子测量技术》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115616386B (en) | 2023-03-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110514924B (en) | Power transformer winding fault positioning method based on deep convolutional neural network fusion visual identification | |
| US7437271B2 (en) | Methods and apparatus for data analysis | |
| US6792373B2 (en) | Methods and apparatus for semiconductor testing | |
| CN108196143A (en) | Power Transformer Faults depth diagnostic method and terminal device | |
| CN115729761B (en) | Hard disk fault prediction method, system, equipment and medium | |
| CN118091520B (en) | Intelligent regulation and control method and system for 10us square wave surge testing equipment | |
| CN115864759A (en) | Control method and system for automatic motor test work station | |
| CN115616386B (en) | Flying probe testing method based on artificial intelligence | |
| CN108828437B (en) | Analog circuit fault feature extraction method based on cloud correlation coefficient matrix | |
| DE10323230A1 (en) | Methods and apparatus for characterizing board test coverage | |
| CN105159826B (en) | A kind of method and apparatus of wrong sentence in positioning target program | |
| JP2000121705A (en) | Board model correcting method and equipment | |
| US5432460A (en) | Apparatus and method for opens and shorts testing of a circuit board | |
| van Schaaijk et al. | Automatic generation of in-circuit tests for board assembly defects | |
| CN117031223A (en) | Diagnostic method for partial discharge defect of switch cabinet | |
| CN111579910A (en) | Electrical detection method for component faults | |
| CN112865860B (en) | Calibration method and device for trillion passive optical network BOB equipment | |
| CN110967615B (en) | Circuit board fault diagnosis device and method | |
| CN114325455A (en) | Power supply early warning method and device and electronic equipment | |
| US6792385B2 (en) | Methods and apparatus for characterizing board test coverage | |
| CN106992873B (en) | Protection group processing method and device | |
| JP3696009B2 (en) | Semiconductor test apparatus, semiconductor test method, and recording medium | |
| JP4745820B2 (en) | Circuit board inspection apparatus and circuit board inspection method | |
| CN112988792B (en) | Searching method and device for wafer yield problem database | |
| CN117871986A (en) | Method, device, equipment and medium for detecting resonant inductor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |