TWI779649B - Testing system, judging device, testing method for testing insulation of a circuit board, and computer-readable recording medium thereof - Google Patents

Abstract

An embodiment of a detection system is used to detect the insulation of a circuit board. The detection system includes: a current measuring device, an analog-to-digital conversion device, a recording device, and a judging device. The power supply device is used to provide power to the detection system. The current measuring device is used to measure a current signal of a circuit to be tested on the circuit board, and output a measurement result. The analog-to-digital conversion device is used to convert the measurement result into a digital signal. The recording device records the digital signal. The judging device judges the insulation of the circuit board according to the changing waveform of the digital signal to output a judging result, wherein the judging result is output based on at least one previous measurement waveform of the previous measurement.

Description

A detection system, a judging device, and a detection system for detecting the insulation of a circuit board Test method and its computer-readable recording medium

The present invention relates to a circuit board testing technology, more particularly to a testing system, a judging device, a testing method and a computer-readable recording medium for testing the insulation of a circuit board.

In the conventional circuit board inspection technology, the insulation state between the circuit patterns of the printed substrate is detected by applying an inspection voltage between the circuit patterns of the printed substrate through the inspection probe, and according to the inspection voltage and the current value flowing in the inspection probe Find the insulation resistance value, and check whether the insulation condition is good or not based on the insulation resistance value. In addition, there is also a conventional circuit board inspection technology that checks the rising slope of the voltage of the probe after applying the inspection voltage to determine whether there is a risk of sparks.

Generally speaking, the conventional detection technology is to combine the signals of multiple test points of multiple circuit patterns on the printed substrate to the test equipment for insulation state detection. Under the framework of converging the test points to the test equipment, the conventional detection technology can only detect whether the overall insulation state of the printed substrate is good, but fails to point out which test point has the risk of sparks. Furthermore, the method of judging whether the insulation state is good or not and the parameters of the judging criteria are determined by human experience, which is very time-consuming and depends on the ability of the individual, and there is also the possibility of misjudgment.

Therefore, the detection technology of the insulation state of the printed substrate still needs to be improved.

One object of the present invention is to provide a detection technique for detecting the insulation of a circuit board. As far as individual test points are concerned, this technology uses a judging device based on at least one previous measurement waveform of the previous measurement, and according to the change waveform of the digital signal corresponding to the measurement result returned by the circuit board under test, to judge the The insulation of the circuit board is used to output the judgment result.

In order to achieve at least the above-mentioned purpose, according to a first aspect of the present invention, an embodiment of a detection system is proposed for detecting the insulation of a circuit board. The detection system includes: a current measuring device, an analog-to-digital conversion device, A recording device and a judging device. The power supply device is used to provide power to the detection system. The current measuring device is used to measure a current signal of a circuit to be tested on the circuit board, and output a measurement result. The analog-to-digital conversion device is used to convert the measurement result into a digital signal. The recording device records the digital signal. The judging device judges the insulation of the circuit board according to the changing waveform of the digital signal to output a judging result, wherein the judging result is output based on at least one previous measurement waveform of the previous measurement.

In an embodiment of the detection system, the scan frequency range provided by the power supply device is about 20 megahertz to about 80 megahertz.

In an embodiment of the detection system, the detection system further includes: a waveform database storing the previously measured waveforms.

In an embodiment of the detection system, the judgment device includes an artificial intelligence engine trained by the previous measurement waveform, and the judgment result is output by the artificial intelligence engine.

In an embodiment of the detection system, the artificial intelligence engine uses an autoencoder model.

In an embodiment of the detection system, the judging device includes a comparator, which compares the changing waveform with the previous measurement waveform to output the judging result.

In order to achieve at least the above purpose, according to a second aspect of the present invention, an embodiment of a judging device is proposed, which is used in a detection system to detect the insulation of a circuit board. A current measuring device for outputting a measurement result of a current signal of a circuit to be tested, and a recording device for recording a changing waveform of the digital signal of the measurement result, the judging device includes: a waveform database and a AI engine. The waveform database stores at least one previously measured waveform. An artificial intelligence engine trained by the previous measurement waveform is used for judging the insulation of the circuit board according to the changing waveform.

In an embodiment of the judging device, the detection system further includes a power supply device for providing power to the detection system, and the provided scanning frequency is about 20 megahertz to about 80 megahertz.

In an embodiment of the judgment device, the detection system further includes an analog-to-digital conversion device for converting the measurement result into the digital signal.

In an embodiment of the judging device, the detection system further includes a recording device for recording the digital signal.

In an embodiment of the judging device, the artificial intelligence engine adopts an autoencoder model.

In order to achieve at least the above-mentioned purpose, according to a third aspect of the present invention, an embodiment of a detection method is proposed, which is used to detect the insulation of a circuit board, including: measuring a current signal of a circuit to be tested on the circuit board, And output a measurement result; convert the measurement result into a digital signal; record the digital signal; and judge the insulation of the circuit board according to the change waveform of the digital signal to output a judgment result, wherein the judgment result is based on At least one previously measured waveform of the previous measurement is output.

In an embodiment of the judging method, the judging step includes: outputting the judging result by an artificial intelligence engine, wherein the artificial intelligence engine is trained by the previously measured waveform.

In an embodiment of the judging method, the detecting method further includes: inputting the previously measured waveform to the artificial intelligence engine; and training the artificial intelligence engine with a preset result of the previously measured waveform.

In an embodiment of the judging method, the judging step includes: using a comparator to compare the changing waveform with the previous measurement waveform to output the judging result.

In order to achieve at least the above-mentioned purpose, according to a fourth aspect of the present invention, an embodiment of a testing method is proposed for testing the insulation of a circuit board, including: recording a digital signal, the digital signal is converted from a measurement result , the measurement result is output by measuring a current signal of a circuit to be tested on the circuit board; and judging the insulation of the circuit board according to the change waveform of the digital signal to output a judgment result, wherein the judgment result is based on At least one previously measured waveform of the previous measurement is output.

In an embodiment of the judging method, the step of judging the insulation of the circuit board according to the changing waveform of the digital signal includes: outputting the judging result by an artificial intelligence engine, wherein the artificial intelligence engine is based on the previously measured waveform trained.

In an embodiment of the judging method, the detecting method further includes: inputting the previously measured waveform to the artificial intelligence engine; and training the artificial intelligence engine with a preset result of the previously measured waveform.

In an embodiment of the judging method, the step of judging the insulation of the circuit board according to the changing waveform of the digital signal includes: comparing the changing waveform with the previous measured waveform by a comparator to output the judging result.

In order to achieve at least the above-mentioned purpose, according to a fifth aspect of the present invention, a non-transitory computer-readable recording medium is proposed, which stores a plurality of instructions, and after an electronic device reads the instructions, it can execute the above-mentioned The judging method of the fourth aspect or one of its embodiments or a combination thereof.

A number of embodiments of various implementations of a detection technology for detecting the insulation of a circuit board are presented above. As far as individual test points are concerned, this technology uses a judging device based on at least one previous measurement waveform of the previous measurement, and according to the change waveform of the digital signal corresponding to the measurement result returned by the circuit board under test, to judge the The insulation of the circuit board is used to output the judgment result. The judging device can be further implemented with artificial intelligence technology, such as machine learning, so as to reduce the misjudgment rate.

1A, 1B, 1C: detection system

3: Training system

10: Power supply device

12: Current measuring device

14: Analog-to-digital conversion device

16: Recording device

18: Judgment device

20:Waveform database

21: Artificial intelligence engine

22: Comparator

30: Electric measuring equipment

32: computing device

34: Database

36: Artificial intelligence engine

40: Autoencoder Models

41: Encoder

42: latent code

43: Decoder

410: Input layer

420: hidden layer

440: hidden layer

450: output layer

CB, CBX: circuit board

PB1, PB2: test probe

S110~S160: steps

WR: changing waveform data

IN: input data

OUT: output data

FIG. 1A is a schematic block diagram of an embodiment of a testing system for testing the insulation of a circuit board.

FIG. 1B is a schematic block diagram of another embodiment of a testing system for testing the insulation of a circuit board.

FIG. 1C is a schematic block diagram of yet another embodiment of a testing system for testing the insulation of a circuit board.

Fig. 2 is a schematic block diagram of an embodiment of the judging device of the detection system.

Fig. 3 is a schematic block diagram of an embodiment of the judging device of the detection system.

FIG. 4 is a flowchart of an embodiment of a testing method for testing the insulation of a circuit board.

FIG. 5 is a flowchart of an embodiment of a method for training an artificial intelligence engine.

FIG. 6 is a schematic diagram of an embodiment of a training system for training an artificial intelligence engine.

FIG. 7 is a schematic diagram of an embodiment of using an autoencoder to build an artificial intelligence engine.

Fig. 8 is a schematic diagram of an embodiment of an autoencoder.

FIG. 9 is a schematic diagram of the reconstruction error of the verification data of the autoencoder in FIG. 8 .

FIG. 10 is a schematic diagram of a confusion matrix used to determine the performance of the autoencoder in FIG. 8 .

In order to fully understand the purpose, characteristics and effects of the present invention, the present invention will be described in detail by the following specific embodiments and the accompanying drawings, as follows: Please refer to Figure 1A, which is used for testing A schematic block diagram of an embodiment of a circuit board insulation detection system.

As shown in FIG. 1A , a testing system 1A is used for testing the insulation of a circuit board. The detection system 1A includes a power supply device 10 , a current measurement device 12 , an analog-to-digital conversion device 14 , a recording device 16 , and a judging device 18 .

The power supply device 10 is used to provide power to the detection system 1A. For example, in FIG. 1A, the detection system 1A applies a higher voltage provided by the power supply device 10 between the two circuit patterns of the circuit board CB within a certain period of time through the test probes PB1 and PB2 to detect these circuits. Whether the insulation state between the patterns is good, and whether there is any defect caused by electric sparks. In FIG. 1A, the power supply device 10 is electrically connected to the test probe PB1, the test probe PB1 is used to electrically connect to one circuit pattern of the circuit board CB, and the test probe PB2 is grounded and used to electrically connect to another circuit pattern. circuit pattern.

The current measuring device 12 is used to measure the current signal of the circuit to be tested (such as the above two circuit patterns) on the circuit board CB, and output a measurement result. As shown in FIG. 1A , the current measuring device 12 is electrically connected to the test probe PB1 , and is used to measure a current signal when the test probes PB1 and PB2 are electrically contacted with two circuit patterns. If the insulation of the two circuit patterns is poor, it will be reflected in the changing waveform of the current signal, and a waveform with relatively large fluctuations will be generated. The current measurement device 12 represents the measurement result by, for example, a voltage signal.

The analog-to-digital conversion device 14 is used to convert the measurement result into a digital signal.

The recording device 16 records the digital signal. The judging device 18 judges the insulation of the circuit board according to the changing waveform of the digital signal to output a judging result, wherein the judging result is output based on at least one previous measurement waveform of the previous measurement. For example, the recording device 16 and the judging device 18 are respectively realized by using a computing device, such as a memory unit and a processing unit in a computer. In another example, the recording device 16 is a memory, such as a volatile or non-volatile memory, or a storage device such as hardware, solid state hardware, etc.; the judging device 18 can also be a microprocessor, a microcontroller or a digital signal processor, or can be based on a circuit using a field programmable gate array (FPGA), or an application specific integrated circuit (ASIC) or a complex programmable logic device (CPLD). It can be realized by one or more circuits, and can also be realized by using a dedicated circuit or module.

Please refer to FIG. 1B , which is a schematic block diagram of another embodiment of a testing system for testing the insulation of a circuit board. The difference between the detection system 1B of FIG. 1B and the detection system 1A of FIG. 1A is that the current measuring device 12 of the detection system 1B of FIG. 1B is electrically connected to the test probe PB2. When the test probes PB1 and PB2 electrically contact two circuit patterns, if the insulation of the two circuit patterns is poor, it will be reflected in the changing waveform of the current signal measured by the current measuring device 12 of the detection system 1B , a waveform with large fluctuations is produced.

Please refer to FIG. 1C , which is a schematic block diagram of another embodiment of a testing system for testing the insulation of a circuit board. The difference between the detection system 1C of FIG. 1C and the detection system 1A of FIG. 1A and the detection system 1B of FIG. 1B is that the detection system 1C of FIG. 12C is electrically connected to test probes PB1 and PB2. When the test probes PB1 and PB2 are in electrical contact with the two circuit patterns, the current measurement module 12C uses the two current measurement devices 12 to measure the circuit to be tested on the circuit board CB (such as the aforementioned two circuit patterns) The current signal, and the output measurement Result; the analog-to-digital conversion module 14C converts the measurement result into two digital signals through the two analog-to-digital conversion devices 14 . The two digital signals can be processed, such as subtracted, to achieve the effect of anti-interference. The recording device 16 of the detection system 1C in FIG. 1C records the two digital signals. The judging device 18 judges the insulation of the circuit board according to the changing waveforms of the digital signals to output a judging result, wherein the judging result is output based on at least one previous measurement waveform of the previous measurement.

In some embodiments of the detection system 1A, 1B or 1C, the scan frequency range provided by the power supply device 10 is about 20 megahertz to about 80 megahertz.

Please refer to FIG. 2 , which is a schematic block diagram of an embodiment of the judging device 18 of the detection system 1A, 1B or 1C. As shown in FIG. 2 , the judgment device 18 may include an artificial intelligence engine 21 trained by the previous measurement waveform, and the judgment result is output by the artificial intelligence engine 21 . The previously measured waveforms are stored in the waveform database 20, for example. The artificial intelligence engine 21 shown in FIG. 2 is a trained model. In one embodiment, the artificial intelligence engine 21 adopts an autoencoder model. An example description will be given later on the auto-encoding model. The judging device 18 may further include a waveform database 20 , for example, storing measured waveforms judged to represent generation of electric sparks.

Please refer to FIG. 3 , which is a schematic block diagram of an embodiment of the judging device 18 of the detection system 1A. As shown in FIG. 3 , the judging device 18 may include a comparator 22 , which compares the changing waveform with the previous measured waveform to output the judging result. For example, the previously measured waveform is a surge waveform representing an electric spark, and the comparator 22 uses image processing techniques, such as scanning window (sliding window) or related techniques, to search for whether there is any connection with the waveform in the changing waveform. burst waveform. If there is the surge waveform in the changing waveform, output the judgment result as the risk of electric spark; if there is no surge waveform in the changing waveform, output the judgment result as normal.

Please refer to FIG. 4 , which is a flowchart of an embodiment of a detection method for detecting the insulation of a circuit board. The detection method can be realized by using the above-mentioned detection system 1A, 1B or 1C. As shown in FIG. 4 , a detection method for detecting the insulation of a circuit board includes the following steps S110-S160.

As shown in step S110, measure a current signal of a circuit to be tested on the circuit board, and output a measurement result. Such as using the current measurement device 12 of the detection system 1A, 1B or 1C to realize.

As shown in step S120, the measurement result is converted into a digital signal. Such as using the analog-to-digital conversion device 14 of the detection system 1A, 1B or 1C to realize.

As shown in step S150, the digital signal is recorded. This is done with the recording device 16 of the detection system 1A, 1B or 1C.

As shown in step S160, the insulation of the circuit board is judged according to the changing waveform of the digital signal to output a judgment result, wherein the judgment result is output based on at least one previous measurement waveform of the previous measurement. Such as using the judgment device 18 of the detection system 1A, 1B or 1C to realize.

In an embodiment of the judging method, the step S160 (judging step) includes: outputting the judging result by an artificial intelligence engine 21, wherein the artificial intelligence engine 21 is trained by the previous measurement waveform. The trained one shown in FIG. 2 is realized by using the artificial intelligence engine 21 .

In an embodiment of the judging method, the step S160 (judging step) includes: a comparator 22 compares the changing waveform with the previous measurement waveform to output the judging result. As shown in FIG. 3, the comparator 22 is used for realization.

Please refer to FIG. 5 , which is a flowchart of an embodiment of a method for training an artificial intelligence engine. As shown in FIG. 5 , the method for training an artificial intelligence engine includes steps S130 and S140: as shown in step S130, input the previously measured waveform to the artificial intelligence engine.

As shown in step S140, the artificial intelligence engine is trained with a preset result of the previous measurement waveform.

In one embodiment, the detection method in FIG. 4 may further include steps S130 and S140 in FIG. 5, for example, before steps S150 and S160, use steps S130 and S140 to train the untrained artificial intelligence engine 21 to obtain the Trained artificial intelligence engine 21 . When executing steps S150 and S160, the trained artificial intelligence engine 21 is used to make a judgment.

In addition, in some embodiments, steps S130 and S140 in FIG. 5 can be implemented on a first computing device (such as a local computer or a cloud), and steps S150 and S160 in FIG. 4 can be realized on a second computing device (such as local computer or cloud), wherein the first computing device and the second computing device can be the same or different computing devices.

In addition, according to an aspect of the present invention, steps S150 and S160 in FIG. 4 can also be used as a detection method for detecting the insulation of a circuit board, and the detection method is executed on a computing device (such as a local computer or a cloud). superior. In addition, in some embodiments of the detection method, steps S130 and S140 in FIG. 5 are further included.

The following example illustrates the implementation of the artificial intelligence engine in FIG. 2 .

Please refer to FIG. 6 , which is a schematic diagram of an implementation manner of a training system for training an artificial intelligence engine. As shown in FIG. 6 , the training system 3 includes an electric measuring device 30 , a computing device 32 , a database 34 , and an artificial intelligence engine 36 .

The electrical measuring equipment 30 includes, for example, the power supply device 10, the current measurement device 12, and the analog-to-digital conversion device 14 in the detection system 1A, 1B, or 1C, and the power supply device 10, the current measurement device 12, and the analog-to-digital conversion device 14 utilize Electrically connect and configure as shown in FIG. 1A, 1B or 1C. The electrical measuring equipment 30 is used to operate through the power supply device 10 , the current measuring device 12 , and the analog-to-digital conversion device 14 It is used to detect the insulation of one or more circuit patterns in one or more circuit boards CBX, and thus obtain various digital signals representing changing waveforms.

The computing device 32 is, for example, a computer, and is used to record the digital signal output by the electric measuring device 30 as various detection waveform data. The changing waveform data WR can be collected and stored in the database 34 and used as a data set for training the artificial intelligence engine 36 . A technician can record various changing waveforms WR to the database 34 on the computing device 32 . The artificial intelligence engine 36 is a model of the artificial intelligence engine to be trained.

According to the architecture of the training system 3 in FIG. 6, in one embodiment, the database 34 and the artificial intelligence engine 36 can be implemented on a first computing device (such as a local computer or cloud), and use the training system shown in FIG. The method of the artificial intelligence engine is used to train the artificial intelligence engine 36 . In another embodiment, the first computing device can also be realized by the computing device 32 .

As shown in FIG. 6, the training system 3 can be used to prepare training data, such as changing waveform data WR. For example, the changing waveform data WR may include waveforms W1-W4, wherein the waveform W1 represents a normal waveform, the waveform W2 represents a waveform that may cause sparks, and the waveforms W3 and W4 represent waveforms without sparks. The waveform W2 has the characteristics of a local surge not in the waveform W1, for example, the peak and trough of the local surge have a voltage difference of at least, for example, 20V or more, and the peak and trough of the local surge change within a relatively short period of time. Occurs within intervals, so the changes of the local peaks and troughs, which are different from the waveforms W3 and W4, occur within a longer time interval. The above-mentioned expert field knowledge belonging to the judgment of the waveform that may cause sparks can be used as a data set for selecting useful waveforms. In addition, a data augmentation technique can be used to add more samples to the training data set based on the waveforms W1 - W4 . In addition, the above-mentioned graphics corresponding to the local surges of the sparks and the graphics corresponding to the local peaks and valleys of the non-sparks are established in the artificial intelligence engine 36, such as When implementing the artificial intelligence engine 36 with a deep learning model, it can be further used as features.

For example, in one embodiment, the artificial intelligence engine 36 may include a supervised learning model, which is a model trained to detect a waveform that may cause a spark. In another embodiment, the artificial intelligence engine 36 may include an unsupervised learning model. The unsupervised learning model performs clustering or classification operations on the input data set, and uses it to discover or detect possible The waveform that causes the spark.

Please refer to FIG. 7 , which is a schematic diagram of an embodiment of using an autoencoder to build an artificial intelligence engine 36 . Autoencoders are neural networks that belong to unsupervised learning. As shown in FIG. 7 , the autoencoder model 40 includes an encoder (encoder) 41 , a latent code (latent code) 42 , and a decoder (decoder) 43 . The encoder 41 includes an input layer 410 and at least one hidden layer 420 . The decoder 43 includes at least one hidden layer 440 and an output layer 450 . As shown in the lower part of Figure 7, the input layer 410 of the encoder 41 is used to convert the input data IN, such as a digital signal corresponding to the measured waveform, into a waveform pattern and use it as the input data IN, which is processed by the encoder 41 to obtain a latent code. 42. The latent code 42 is input to the decoder 43 and converted into a waveform pattern as output data OUT.

Each layer in the autoencoder model 40 may use fully-connected layers, or partially connected layers such as convolution layers or pooling layers.

In one embodiment, the encoder 41 includes at least one convolutional layer. For example, the convolutional layer uses the aforementioned graphs corresponding to local spikes of sparks and graphs corresponding to local peaks and troughs of non-sparks as features to build feature maps, which are extracted from the waveform pattern of the input data IN Take out features. The encoder 41 further includes at least one pooling layer, which performs a dimensionality reduction operation on the output of the convolutional layer, and finally the encoder 41 generates a latent code 42 .

In one embodiment, by designing the convolutional layer and the pooling layer in the encoder 41, the latent code 42 can represent whether the waveform is a normal waveform (such as waveform W1), a waveform that may cause sparks (such as waveform W2), or no waveform. The codes classified by the waveforms of sparks (such as waveforms W3 and W4 ) are achieved and represented by one-hot encoding.

In an embodiment, the decoder 43 may be implemented symmetrically with the encoder 41 .

In one embodiment, the decoder 43 can be realized as variational with the encoder 41 according to the requirement, so as to generate the output data OUT with variation. For example, there is a color change (such as red) at the spikes of the waveform that may cause sparks to generate the output data OUT, such as achieved by heat map techniques. For another example, there are special graphics or boxes as indications at the surges of the waveform that may cause sparks, so as to generate the output data OUT.

As shown in FIG. 8 , it is a schematic diagram of an embodiment of an autoencoder model 40 . In this embodiment, the input layer 410, the hidden layer 420, the latent code (or can be regarded as a hidden layer) 42, the hidden layer 440, and the output layer 450 have 128, 64, 32, 64, and 128 neurons, respectively. For example, the optimizer (optimizer) to speed up the convergence speed of the autoencoder model 40 is an Adam optimizer, and each neuron uses a rectified linear unit (Rectified Linear Unit, ReLU) as an activation function, and the mean square error (Mean square error, MSE) as the loss function.

Please refer to FIG. 9 , which is a schematic diagram of reconstruction error of the validation data of the autoencoder in FIG. 8 . As shown in Figure 9, the reconstruction error of the verification data. The verification data in Figure 9 is 50 normal signals plus 10 abnormal signals (representing the possibility of sparks). It can be seen that the reconstruction error of the last 10 data is an outlier, which is detected according to the "abnormal score threshold". .

Please refer to FIG. 10 , which is a schematic diagram of a confusion matrix used to determine the performance of the autoencoder in FIG. 8 . There are 1054 verification data, of which 1000 are normal data and 54 are abnormal data with sparks. The data identification detection is carried out through the automatic encoder model 40, and the detection result shows that the data identification accuracy rate is 95.6%. Among them, 46 were misjudged, and the source was normal but was identified as abnormal (with sparks).

After training to reduce the misjudgment rate to a suitable or desirable value, a final autoencoder model 40 is obtained, which can be used to implement the judging device 18 of the detection system 1A, 1B or 1C.

In some application scenarios, the above-mentioned judging device 18 can also be regarded as a part of the electric measuring equipment, or implemented in the electric measuring equipment. Inspection equipment such as electrical testing equipment with multiple inspection probes and automated machinery to receive circuit boards for testing. In this way, the detection technology is helpful to point out the judgment results individually for multiple test points of the circuit board, so that technicians can efficiently find out the position of the circuit board that needs further detection.

According to another aspect of the present invention, an embodiment of a non-transitory computer-readable recording medium is proposed, which stores a plurality of instructions, and after an electronic device (or computing device) reads these instructions, it can execute the The judging method in steps S150 and 160 or one of its embodiments or a combination thereof. Examples of computer-readable recording media include but are not limited to: optical information storage media, magnetic information storage media, hard disks, solid state disks, or memories such as memory cards, firmware, or ROM or RAM. For example, the computing device includes a communication unit, at least one processing unit and at least one storage device, wherein the processing unit is electrically coupled to the communication unit and the memory unit. The storage device includes, for example, a primary storage device and/or an auxiliary storage device, such as any example of the aforementioned storage media. The at least one processing unit is used to communicate with a communication network through the communication unit in a wireless or wired manner, so as to communicate with other computing devices such as terminal devices. The processing unit may include one or more processors, and the computing device may also include other devices such as a graphics processor for computing. In some embodiments, the computing device can execute an operating system, and can further One-step implementation using one or more of various technologies such as web services, script engines, web applications, or web application programming interface (API) servers to provide application services for browsers on the user end, applications, etc.

As described above, several examples of various implementations of a detection technique for detecting the insulation of a circuit board are proposed. As far as individual test points are concerned, this technology uses a judging device based on at least one previous measurement waveform of the previous measurement, and according to the change waveform of the digital signal corresponding to the measurement result returned by the circuit board under test, to judge the The insulation of the circuit board is used to output the judgment result. The judging device can be further implemented with artificial intelligence technology, such as machine learning, so as to reduce the misjudgment rate.

The present invention has been disclosed above with preferred embodiments, but those skilled in the art should understand that the embodiments are only used to describe the present invention, and should not be construed as limiting the scope of the present invention. It should be noted that all changes and substitutions equivalent to these embodiments should be included within the scope of the present invention. Therefore, the scope of protection of the present invention should be defined by the scope of the patent application.

1A: Detection system

10: Power supply device

12: Current measuring device

14: Analog-to-digital conversion device

16: Recording device

18: Judgment device

CB: circuit board

PB1, PB2: test probe

Claims (17)

A detection system for detecting the insulation of a circuit board, comprising: a power supply device for providing power to the detection system; a current measuring device for measuring a circuit to be tested on the circuit board a current signal, and output a measurement result; an analog-to-digital conversion device, used to convert the measurement result into a digital signal; a recording device, recording the digital signal; and a judging device, according to the digital signal changing the waveform to judge the insulation of the circuit board to output a judgment result, wherein the judgment result is output based on at least one previous measurement waveform of the previous measurement, wherein the judgment device includes an artificial intelligence engine, which is based on the previous measurement The waveform is trained, and the judgment result is output by the artificial intelligence engine. The detection system as claimed in claim 1, wherein the scanning frequency range provided by the power supply device is about 20 megahertz to about 80 megahertz. The inspection system as described in claim 1, wherein the inspection system further includes: a waveform database storing the previously measured waveforms. The detection system as described in claim 1, wherein the artificial intelligence engine adopts an autoencoder model. The detection system as claimed in claim 1, wherein the judging device includes a comparator, which compares the changing waveform with the previous measurement waveform to output the judging result. A judging device used in a detection system to detect the insulation of a circuit board, the detection system includes a current measurement for measuring a current signal of a circuit to be tested on the circuit board to output a measurement result device, and a recording device for recording a changing waveform of the digital signal of the measurement result, the judging device includes: A waveform database storing at least one previous measurement waveform; and an artificial intelligence engine trained by the previous measurement waveform for judging the insulation of the circuit board according to the changing waveform. The judging device as claimed in claim 7, wherein the detection system further includes a power supply device for providing power to the detection system, and the provided scanning frequency is about 20 megahertz to about 80 megahertz. The judging device as described in Claim 7, wherein the detection system further includes an analog-to-digital conversion device for converting the measurement result into the digital signal. The judging device as described in Claim 9, wherein the detection system further includes a recording device for recording the digital signal. The judging device according to claim 7, wherein the artificial intelligence engine adopts an autoencoder model. A detection method for detecting the insulation of a circuit board, comprising: measuring a current signal of a circuit to be tested on the circuit board, and outputting a measurement result; converting the measurement result into a digital signal; recording the digital signal; and judging the insulation of the circuit board according to the changing waveform of the digital signal to output a judging result, wherein the judging result is output based on at least one previous measurement waveform of the previous measurement, and is output by an artificial intelligence engine Outputting the judgment result, wherein the artificial intelligence engine is trained by the previous measurement waveform. The detection method as described in claim 11, wherein the detection method further comprises: inputting the previously measured waveform to the artificial intelligence engine; and training the artificial intelligence engine with a preset result of the previously measured waveform. The detection method according to claim 11, wherein the judging step includes: a comparator compares the changing waveform with the previous measurement waveform to output the judging result. A detection method for detecting the insulation of a circuit board, comprising: recording a digital signal, the digital signal is converted from a measurement result of measuring a current of a circuit to be tested on the circuit board and output a judgment result by judging the insulation of the circuit board according to the changing waveform of the digital signal, wherein the judgment result is output based on at least one previous measurement waveform of the previous measurement, and is output by an artificial intelligence engine Outputting the judgment result, wherein the artificial intelligence engine is trained by the previous measurement waveform. The detection method as claimed in claim 14, wherein the detection method further comprises: inputting the previously measured waveform to the artificial intelligence engine; and training the artificial intelligence engine with a preset result of the previously measured waveform. The detection method according to claim 14, wherein the step of judging the insulation of the circuit board according to the changing waveform of the digital signal includes: using a comparator to compare the changing waveform with the previous measurement waveform to output the judgment result. A non-transitory computer-readable recording medium, which stores a plurality of instructions, and an electronic device can execute the detection method according to any one of claims 14-16 after reading the instructions.

Images