CN117310591A - Small-size equipment for testing equipment calibration accuracy detection - Google Patents

Abstract

The invention discloses a small-sized device for detecting calibration precision of test equipment, which comprises: the test device comprises an upper computer, a standard instrument and a test fixture, wherein the test fixture comprises a simulation board and a test bottom board; the user sends a test control instruction through the upper computer to enable a signal source of the standard instrument to output standard signals, the standard signals are connected to a signal input interface of the test fixture through a cable, and the standard signals pass through the simulation board and then enter the test bottom board through the ejector pins on the test fixture. Therefore, the test efficiency can be ensured, and the accuracy and reliability of the test result can also be ensured.

Description

Small-size equipment for testing equipment calibration accuracy detection

Technical Field

The invention relates to the technical field of intelligent precision detection, in particular to a small-sized device for detecting calibration precision of test equipment.

Background

In the development stage of a product, a tester needs to make many tests for each functional point of the product to verify the reliability and integrity thereof. For example, different values are input at the front end, and different results are obtained at the terminal, thereby verifying different situations. If manual, a test item may take tens of minutes or more, and for thousands of test items, the effort is enormous. If errors occur in the test process, repeated debugging is needed, and the test period is longer. Prior to mass production, pilot-scale tests are also required to be performed on the product, and verification tests are performed on the product to ensure the quality of mass production. In this process, the factors of yield, efficiency, quality and the like need to be considered. The tester needs to ensure the test efficiency as much as possible and ensure the accuracy and reliability of the test result.

For more complex test items, there is often difficulty in using manual means, and automated testing becomes a solution to this situation. By automating large-scale complex testing, manual operation is reduced, and efficient matching between human and machine is realized, so that the testing efficiency is improved, and even the time of links such as product research and development, manufacturing and the like is reduced.

For some measurement circuits, some shifts in performance can occur as critical devices age and ambient temperature change. This requires some testers to check the automated test equipment at regular intervals and recalibrate the automated test equipment for severe performance drift so that the test results are accurate and reliable. However, the manual inspection and calibration is inefficient and may be subject to false positives.

Thus, an optimized calibration accuracy detection scheme for test equipment is desired.

Disclosure of Invention

The embodiment of the invention provides a small-sized device for detecting the calibration precision of test equipment, which comprises the following components: the test device comprises an upper computer, a standard instrument and a test fixture, wherein the test fixture comprises a simulation board and a test bottom board; the user sends a test control instruction through the upper computer to enable a signal source of the standard instrument to output standard signals, the standard signals are connected to a signal input interface of the test fixture through a cable, and the standard signals pass through the simulation board and then enter the test bottom board through the ejector pins on the test fixture. Therefore, the test efficiency can be ensured, and the accuracy and reliability of the test result can also be ensured.

The embodiment of the invention also provides a small-sized device for detecting the calibration precision of the test device, which comprises: the test device comprises an upper computer, a standard instrument and a test fixture, wherein the test fixture comprises a simulation board and a test bottom board; the user sends a test control instruction through the upper computer to enable a signal source of the standard instrument to output standard signals, the standard signals are connected to a signal input interface of the test fixture through a cable, and the standard signals pass through the simulation board and then enter the test bottom board through the ejector pins on the test fixture.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

In the drawings: fig. 1 is a simplified block diagram of a station automated test equipment calibration principle.

Fig. 2 is a block diagram of a small-sized apparatus for testing calibration accuracy detection of an apparatus according to an embodiment of the present invention.

FIG. 3 is a schematic simplified block diagram of a workstation automated test equipment calibration system provided in an embodiment of the present invention.

Fig. 4 is a schematic and simplified block diagram of a calibration device audio Kit provided in an embodiment of the present invention.

FIG. 5 is a schematic illustration of the basic mechanical dimensions-130 mmX130mmX32mm provided in an embodiment of the present invention.

FIG. 6 is a flow chart of a method for calibration accuracy detection of a test device in a compact form, in accordance with an embodiment of the present invention.

Fig. 7 is a schematic diagram of a system architecture of a method for detecting calibration accuracy of a test device in a small-sized embodiment of the present invention.

Fig. 8 is an application scenario diagram of a small-sized device for testing calibration accuracy detection of a device according to an embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings. The exemplary embodiments of the present invention and their descriptions herein are for the purpose of explaining the present invention, but are not to be construed as limiting the invention.

Unless defined otherwise, all technical and scientific terms used in the examples of this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present application.

In the description of the embodiments of the present application, unless otherwise indicated and defined, the term "connected" should be construed broadly, and for example, may be an electrical connection, may be a communication between two elements, may be a direct connection, or may be an indirect connection via an intermediary, and it will be understood by those skilled in the art that the specific meaning of the term may be understood according to the specific circumstances.

It should be noted that, the term "first\second\third" in the embodiments of the present application is merely to distinguish similar objects, and does not represent a specific order for the objects, it is to be understood that "first\second\third" may interchange a specific order or sequence where allowed. It is to be understood that the "first\second\third" distinguishing objects may be interchanged where appropriate such that the embodiments of the present application described herein may be implemented in sequences other than those illustrated or described herein.

With the overall development of automated testing, automated testing is widely used. In the development stage of a product, a tester needs to make many tests for each functional point of the product to verify the reliability and integrity thereof. For example, different values are input at the front end, and different results are obtained at the terminal, thereby verifying different situations. If manual, a test item may take tens of minutes or more, and for thousands of test items, the effort is enormous. If errors occur in the test process, repeated debugging is needed, and the test period is longer. Prior to mass production, pilot-scale tests are also required to be performed on the product, and verification tests are performed on the product to ensure the quality of mass production. In this process, the factors of yield, efficiency, quality and the like need to be considered. The tester needs to ensure the test efficiency as much as possible and ensure the accuracy and reliability of the test result. For some measurement circuits, some shifts in performance can occur as critical devices age and ambient temperature change. This requires some testers to check the automated test equipment at regular intervals and recalibrate the automated test equipment for severe performance drift so that the test results are accurate and reliable.

For more complex test items, there is often difficulty in using manual means, and automated testing becomes a solution to this situation. By automating large-scale complex testing, manual operation is reduced, and efficient matching between human and machine is realized, so that the testing efficiency is improved, and even the time of links such as product research and development, manufacturing and the like is reduced. And in combination with the automatic test of the customer demands, different solutions are provided for different test conditions. For a scenario of repeated testing with a large number of test items, an automated test scheme based on a script can be provided; for testers without programming capability, a code-free automated test scheme can be provided; for the scenario where automated testing is to be integrated into a test platform, an automated integration test scheme may be provided; for customers without relevant background, custom automated test protocols may be provided.

In the face of the situation that the scale and complexity of an automatic test system are increasingly improved, the requirements on automatic test equipment are also higher and higher, and the automatic test equipment also faces new challenges, wherein one of the challenges is the calibration of the automatic test equipment.

The manner in which the automated test calibration of the present station is mainly as follows, as shown in fig. 1.

The standard instrument is controlled by a USB bus, a user sends a command to enable a signal source of the standard instrument to output standard signals through an upper computer, the standard signals are connected to a signal input interface of the tool through a cable, and the standard signals enter a test bottom plate through a thimble on the tool after passing through a simulation board. The test base plate is controlled by adopting a network port, a user sends a command through the upper computer to enable a control board on the test base plate to switch a signal path on the test base plate, a standard signal is connected to a corresponding functional module, and the control board reads the standard signal acquired by the functional module through a control bus and uploads the standard signal to the upper computer through the network port. The accuracy of the acquired signals is calculated to see if it is within the nominal accuracy of the functional module. If yes, the functional module is not required to be calibrated; otherwise, calculating a calibration coefficient, and writing the calibration coefficient into a memory EEPROM on the functional module through the network port control panel. And in the same way, similar signal calibration is performed, and the same calibration process is adopted, so that the accuracy of data acquisition is ensured.

The test base plate is controlled by adopting a network port, and a user sends a command through the upper computer to enable a control board on the test base plate to switch a signal path on the test base plate, so that signals output by related functional modules on the test base plate are connected to the ejector pins. The signals enter the interface connector through the simulation board. And then accessed to standard instruments through cables. The standard instrument is controlled by a USB bus, a user sends a command through the upper computer to enable the universal meter of the standard instrument to collect input signals, and a measurement result is uploaded to the upper computer through the USB bus. The accuracy of the acquired signals is calculated to see if it is within the nominal accuracy of the functional module. If yes, the functional module is not required to be calibrated; otherwise, calculating a calibration coefficient, and writing the calibration coefficient into a memory EEPROM on the functional module through the network port control panel. And in the same way, similar signal calibration is performed, and the accuracy of a data result is ensured by adopting the same calibration process.

The existing automatic test and calibration method for the work station has the problems that the calibration is difficult, the calibration instrument is expensive, the standard instrument occupies large space and is heavy, the power consumption is large, the operation of professional technicians who are familiar with the instrument is needed, the wiring is complex, the error is easy to occur, the calibration error is caused, and the like. The standard instrument is used for automatic testing and calibration of the workstation, and a proper standard instrument is needed to be selected first. Secondly, the selection of cables, and the wiring of standard instruments all need professional personnel to operate. A workstation often requires several standard instruments for calibration, each of which is relatively bulky and takes up a large amount of field space to stack together. And the power consumption of each device is larger, and the power consumption of the devices which are overlapped together is larger. The price of each standard instrument is tens of thousands, even hundreds of thousands, the high price increases the burden of enterprises, the working principle of the standard instrument is mastered by knowing the calibration of the work station, the functions required in the calibration of the work station are selected for optimization, and an effective small-sized device for testing the calibration precision detection of the device is customized, so that the effect of reducing the power consumption can be effectively achieved. The cost generated in the production process can be effectively reduced, and the benefit is increased.

The invention aims to design a small-sized device for detecting the calibration precision of test equipment, which is used for solving the problems of difficult calibration, expensive calibration instrument, large and heavy occupation space of standard instrument, large power consumption, complicated wiring, error-prone calibration and the like in a site work station.

The calibrating equipment for calibrating the field station equipment is portable, light and easy to operate, and is convenient for field staff to maintain the machine.

In one embodiment of the present invention, fig. 2 is a block diagram of a small-sized apparatus for testing calibration accuracy detection of an apparatus according to an embodiment of the present invention. As shown in fig. 2, a small-sized apparatus 100 for testing calibration accuracy detection of an apparatus according to an embodiment of the present invention includes: the device comprises an upper computer 1, a standard instrument 2 and a test jig 3, wherein the test jig 3 comprises a simulation board 31 and a test bottom board 32; the user sends a test control command to enable the signal source of the standard instrument 2 to output a standard signal through the upper computer 1, the standard signal is connected to the signal input interface of the test fixture 3 through the cable 4, and the standard signal passes through the simulation board 31 and then enters the test bottom board 32 through the thimble 33 on the test fixture 3.

The test base plate is controlled by adopting a network port, a user sends a command through the upper computer to enable a control board on the test base plate to switch a signal path on the test base plate, a standard signal is connected to a corresponding functional module, and the control board reads the standard signal acquired by the functional module through a control bus and uploads the standard signal to the upper computer through the network port. The accuracy of the acquired signals is calculated to see if it is within the nominal accuracy of the functional module. If yes, the functional module is not required to be calibrated; otherwise, calculating a calibration coefficient, and writing the calibration coefficient into a memory EEPROM on the functional module through the network port control panel. And in the same way, similar signal calibration is performed, and the same calibration process is adopted, so that the accuracy of data acquisition is ensured.

In particular, it is contemplated that user-diversified test requirements need to be considered in an automated test procedure. That is, different solutions need to be provided for different test cases. For example, for a tester without programming capability, it is desirable to provide a code-free automated test scheme. Based on the above, in the technical solution of the present application, the small-sized device for testing calibration accuracy detection of a device further includes a code-free automatic testing module.

Here, in order to realize the function of the code-free automatic test module, the technical concept of the application is as follows: and automatically generating test codes according to test requirement text description input by a user by using a natural language processing technology based on deep learning, and running the test codes on the upper computer to generate the test control instruction so as to complete the function of automatic test.

In one embodiment of the present application, the apparatus further comprises: a code-free automatic test module; wherein, the code-free automated test module comprises: the demand text testing unit is used for acquiring test demand text description from the upper computer; the data preprocessing unit is used for preprocessing the data of the test requirement text description to obtain a sequence of test requirement description sentences; the semantic coding unit is used for carrying out semantic coding on the sequence of the test requirement description sentence to obtain a test requirement semantic coding feature vector; and the control instruction generating unit is used for generating the test control instruction based on the test requirement semantic coding feature vector.

Specifically, in the technical scheme of the application, first, a test requirement text description is obtained from the upper computer. And then, carrying out data preprocessing on the test requirement text description to obtain a sequence of test requirement description sentences.

In a specific example of the present application, the encoding process for performing data preprocessing on the test requirement text description to obtain a sequence of test requirement descriptions includes: firstly, the test requirement text description is passed through an AIGC model-based requirement description perfect expressive machine to obtain perfect test requirement text description; and then carrying out sentence dividing processing on the complete test requirement text description to obtain a sequence of test requirement description sentences.

And then, carrying out semantic coding on the sequence of the test requirement description sentence to obtain a test requirement semantic coding feature vector. That is, semantic feature information containing user intent contained in the sequence of test requirement description sentences is captured.

In a specific example of the present application, the semantic coding unit includes: a sentence understanding subunit, configured to pass the sequence of the test requirement description sentence through a Bert model-based sentence understanding device to obtain a sequence of test requirement description sentence feature vectors; and a context coding subunit, configured to pass the sequence of test requirement description sentence feature vectors through a context coder based on a converter to obtain the test requirement semantic coding feature vector.

Wherein the context encoding subunit is configured to: one-dimensional arrangement is carried out on the sequence of the test requirement description sentence feature vectors to obtain global test requirement description sentence feature vectors; calculating the product between the global test requirement description sentence feature vector and the transpose vector of each test requirement description sentence feature vector in the sequence of the test requirement description sentence feature vectors to obtain a plurality of self-attention correlation matrices; respectively carrying out standardization processing on each self-attention correlation matrix in the plurality of self-attention correlation matrices to obtain a plurality of standardized self-attention correlation matrices; obtaining a plurality of probability values by using a Softmax classification function through each normalized self-attention correlation matrix in the normalized self-attention correlation matrices; and weighting each test requirement description sentence feature vector in the sequence of test requirement description sentence feature vectors by taking each probability value in the plurality of probability values as a weight to obtain the test requirement semantic coding feature vector.

In one embodiment of the present application, the control instruction generating unit is configured to: automatically generating the test demand semantic coding feature vector through a test code based on an AIGC model to obtain a test code; and running the test code on the upper computer to generate the test control instruction.

Further, the test requirement semantic coding feature vector is automatically generated through a test code based on an AIGC model to obtain a test code; and running the test code on the upper computer to generate the test control instruction.

In one embodiment of the present application, the small-sized apparatus for testing apparatus calibration accuracy detection further includes a training module for training the aicc model-based demand description perfect expressive, the Bert model-based sentence comprehener, the converter-based context encoder, and the aicc model-based test code automatic generator; wherein, training module includes: the training data acquisition unit is used for acquiring training data, wherein the training data comprises training test requirement text description and a true value of a test code; the training requirement description perfect expression unit is used for enabling the training test requirement text description to pass through the AIGC model-based requirement description perfect expression unit so as to obtain training perfect test requirement text description; the training clause processing unit is used for carrying out clause processing on the training perfect test requirement text description so as to obtain a sequence of training test requirement description sentences; the training sentence understanding unit is used for passing the sequence of the training test requirement description sentence through the Bert model-based sentence understanding device to obtain a sequence of training test requirement description sentence feature vectors; the training context coding unit is used for enabling the sequence of the training test requirement description sentence characteristic vectors to pass through the context coder based on the converter so as to obtain training test requirement semantic coding characteristic vectors; the training optimization unit is used for optimizing the position-by-position characteristic values of the training test requirement semantic coding characteristic vectors to obtain optimized test requirement semantic coding characteristic vectors; the training test code automatic generation unit is used for enabling the semantic coding feature vector of the optimization test requirement to pass through the AIGC model-based test code automatic generator to obtain a training test code; and a training unit for calculating a cross entropy function value between the training test code and a true value of the test code as a loss function value to train the AIGC model-based demand description perfect expressive device, the Bert model-based sentence comprehener, the converter-based context encoder, and the AIGC model-based test code automatic generator.

Here, each training test requirement description sentence feature vector in the sequence of training test requirement description sentence feature vectors expresses text semantic features of the training perfect test requirement text description under a local semantic space, so that after the sequence of training test requirement description sentence feature vectors passes through a context encoder based on a converter, text semantic feature context associated features among local semantic spaces under a global semantic space can be further extracted, and the obtained training test requirement semantic coding feature vectors have a local semantic space-diversified text semantic feature representation under the global semantic space. In this way, when the training test requirement semantic coding feature vector passes through the AIGC model-based test code automatic generator, since the AIGC model-based test code automatic generator converts the training test requirement semantic coding feature vector from the semantic space feature domain to the semantic probability density domain to obtain the training test code based on the semantic probability density distribution, the problem that the diversified text semantic feature representation of the training test requirement semantic coding feature vector in the semantic space brings local feature distribution sparsification to the overall feature representation thereof, namely, the sub-manifold is thinned out of the distribution relative to the overall high-dimensional feature manifold, is considered, and therefore, when the training test requirement semantic coding feature vector passes through the AIGC model-based test code automatic generator, the convergence of the training test requirement semantic coding feature vector to the predetermined regression probability representation in the semantic probability density domain is poor, and the code quality of the obtained training test code is affected.

Therefore, preferably, the training test requirement semantic coding feature vector is optimized by position feature value, specifically: optimizing the position-by-position characteristic values of the training test requirement semantic coding characteristic vectors by using the following optimization formula to obtain optimized test requirement semantic coding characteristic vectors; wherein, the optimization formula is:wherein (1)>Is the training test requirement semantically encoded feature vector, < >>Is the training test requirement semantic coding feature vector +.>Is>Characteristic value of individual position->Is the +.f. of the optimization test requirement semantically encoded feature vector>Characteristic values of the individual positions.

That is, sparse distribution within a high-dimensional feature space is processed by re-probability based regularization to activate the training test requirement semantic codingFeature vectorNatural distribution transfer of geometric manifold into probability density space in high-dimensional feature space, thereby semantically encoding feature vector by demand for said training test>The distribution sparse sub-manifold of the high-dimensional feature manifold is subjected to a smooth regularization mode based on the re-probability, so that the convergence of the complex high-dimensional feature manifold with high space sparsity under the preset regression probability distribution is improved, and the semantic coding feature vector required by the training test is improved >Code quality of training test codes obtained by an AIGC model-based test code automatic generator.

In summary, a small-sized device 100 for testing device calibration accuracy detection according to an embodiment of the present invention is illustrated, which utilizes a deep learning-based natural language processing technology to automatically generate a test code according to a text description of a test requirement input by a user, and runs the test code on the host computer to generate the test control command, thereby completing an automatic test function.

In one embodiment of the present application, the system of the present invention is composed of a host computer, a switch, calibration equipment, and a simulation board placed in a jig. As shown in fig. 3.

In one embodiment of the present application, a DC voltage source calibration is provided. The user sends a control instruction to a control board-2 on the jig through the network port of the upper computer. The control board-2 receives the instruction, controls the corresponding module-1, outputs direct-current voltage, and transmits the direct-current voltage to the simulation board through the corresponding thimble. The direct current voltage is transmitted to the input end of the calibration equipment through the cable between the simulation board and the calibration equipment, and enters the direct current voltage measurement function of the direct current measurement unit through the circuit switching unit.

The user sends a control instruction to the control panel-1 of the calibration equipment through the network port of the upper computer. The control board-1 receives the instruction, controls the direct current voltage measuring function of the corresponding direct current measuring unit, reads the direct current voltage measuring result of the direct current measuring unit through the bus, and uploads the measuring result to the upper computer through the network port. Meanwhile, the reference value of the direct-current voltage in the EEPROM is read and uploaded to the upper computer together. The reference value in the memory EEPROM is the measurement result of the direct current voltage measuring function on the calibration equipment by using a standard instrument signal source to output direct current voltage.

The upper computer receives the measurement result of the direct current voltage, compares the measurement result with a corresponding reference value in the EEPROM of the memory, judges whether the precision and uncertainty of the direct current voltage are in a design range, if so, indicates that the module-1 to be measured on the jig is normal, otherwise, judges that the module-1 to be measured is abnormal, the measurement result is deviated, and recalibration is needed.

In an embodiment of the present application, a DC current source calibration is provided. The user sends a control instruction to the control panel-2 of the jig through the network port of the upper computer. The control board-2 receives the instruction and controls the corresponding module-2 to output direct current, and the direct current is transmitted to the simulation board through the corresponding thimble. The direct current flows through the cable between the simulation board and the calibration equipment, is transmitted to the input end of the calibration equipment, and enters the direct current measuring function of the direct current measuring unit through the circuit switching unit.

The user sends a control instruction to a control board-1 on the calibration equipment through the network port of the upper computer. The control board-1 receives the instruction, controls the direct current measuring function of the corresponding direct current measuring unit, reads the measuring result of the direct current measuring function through the bus, and uploads the measuring result to the upper computer through the network port. Meanwhile, the reference value of direct current in the EEPROM is read and uploaded to the upper computer together. The reference value in the memory EEPROM is the result of direct current output by a standard instrument signal source and measurement by a direct current measurement unit.

The upper computer receives the measurement result of the direct current, compares the measurement result with a corresponding reference value in the EEPROM of the memory, judges whether the accuracy and uncertainty of the direct current are in a design range, if so, indicates that the module-2 to be measured on the jig is normal, otherwise, judges that the module-2 to be measured is abnormal, the measurement result is deviated, and recalibration is needed.

In an embodiment of the present application, an ac voltage source calibration is provided. The user sends a control instruction to the control panel-2 of the jig through the network port of the upper computer. The control board-2 receives the instruction and controls the corresponding module-3 to output alternating voltage, and the alternating voltage is transmitted to the simulation board through the corresponding thimble. The alternating voltage is transmitted to the input end of the calibration equipment through a cable between the simulation board and the calibration equipment, and enters the alternating voltage measuring function of the alternating voltage measuring unit through the circuit switching unit.

The user sends a control instruction to the control panel-1 of the calibration equipment through the network port of the upper computer. The control board-1 receives the instruction, controls the alternating voltage measuring function of the corresponding alternating current measuring unit, reads the alternating voltage measuring result of the alternating current measuring unit through the bus, and uploads the measuring result to the upper computer through the network port. And simultaneously, reading the reference value of the alternating voltage in the EEPROM, and uploading the reference value to the upper computer. The reference value in the memory EEPROM is the result of the AC voltage output by the standard instrument signal source and measured by the AC measuring unit.

The upper computer receives the measurement result of the alternating voltage, compares the measurement result with a corresponding reference value in the EEPROM of the memory, judges whether the precision and uncertainty of the alternating voltage are in a design range, if so, indicates that the module-3 to be measured on the jig is normal, otherwise, judges that the module-3 to be measured is abnormal, and the measurement result is deviated and needs to be recalibrated.

In an embodiment of the present application, a direct voltage measurement calibration is provided. The user sends a control instruction to the control panel-1 of the calibration equipment through the network port of the upper computer. The control board-1 receives the instruction and controls the corresponding direct current source unit to output direct current voltage, and the direct current voltage is transmitted to the output end of the calibration equipment through the circuit switching unit. The direct voltage is transmitted to the simulation board via a cable between the calibration device and the simulation board. And then the workpiece is transmitted to the input end of the jig through the corresponding thimble.

The user sends a control instruction to the control panel-2 of the jig through the network port of the upper computer. The control board-2 receives the instruction, controls the direct-current voltage measuring function of the corresponding module-4, reads the direct-current voltage measuring result through the bus, and uploads the measuring result to the upper computer through the network port. Meanwhile, the reference value of the direct-current voltage in the EEPROM is read and uploaded to the upper computer together. The reference value in the memory EEPROM is the direct current voltage output by the direct current source unit and is measured by a standard instrument multimeter.

The upper computer receives the measurement result of the direct current voltage, compares the measurement result with a corresponding reference value in the EEPROM of the memory, judges whether the precision and uncertainty of the direct current voltage are in a design range, if so, indicates that the module-4 to be measured on the jig is normal, otherwise, judges that the module-4 to be measured is abnormal, the measurement result is deviated, and recalibration is needed.

In an embodiment of the present application, a direct current measurement calibration is provided. The user sends a control instruction to the control panel-1 of the calibration equipment through the network port of the upper computer. The control board-1 receives the instruction and controls the corresponding direct current source unit to output direct current, and the direct current is transmitted to the output end of the calibration equipment through the circuit switching unit. The direct current flows through the cable between the calibration device and the emulation board and is transmitted to the emulation board. And then the workpiece is transmitted to the input end of the jig through the corresponding thimble.

The user sends a control instruction to the control panel-2 of the jig through the network port of the upper computer. The control board-2 receives the instruction, controls the direct current measuring function of the corresponding module-5, reads the direct current measuring result through the bus, and uploads the measuring result to the upper computer through the network port. Meanwhile, the reference value of direct current in the EEPROM is read and uploaded to the upper computer together. The reference value in the memory EEPROM is the direct current output by the direct current source unit, and the direct current result is measured by a standard instrument multimeter.

The upper computer receives the measurement result of the direct current, compares the measurement result with a corresponding reference value in the EEPROM of the memory, judges whether the accuracy and uncertainty of the direct current are in a design range, if so, the module-5 to be measured of the jig is normal, otherwise, the module-5 to be measured is abnormal, the measurement result is deviated, and recalibration is needed.

In an embodiment of the present application, an ac voltage measurement calibration is provided. The user sends a control instruction to the control panel-1 of the calibration equipment through the network port of the upper computer. The control board-1 receives the instruction and controls the corresponding alternating current source unit to output alternating current voltage, and the alternating current voltage is transmitted to the output end of the calibration equipment through the circuit switching unit. The ac voltage is transmitted to the simulation board via a cable between the calibration device and the simulation board. And then the workpiece is transmitted to the input end of the jig through the corresponding thimble.

The user sends a control instruction to the control panel-2 of the jig through the network port of the upper computer. The control board-2 receives the instruction, controls the alternating voltage measuring function of the corresponding module-6, reads the alternating voltage measuring result through the bus, and uploads the measuring result to the upper computer through the network port. And simultaneously, reading the reference value of the alternating voltage in the EEPROM, and uploading the reference value to the upper computer. The reference value in the memory EEPROM is the result of the AC voltage output by the AC source unit and measured by a standard instrument multimeter.

The upper computer receives the measurement result of the alternating voltage, compares the measurement result with a corresponding reference value in the EEPROM of the memory, judges whether the precision and uncertainty of the alternating voltage are in a design range, if so, indicates that the module-6 to be measured on the jig is normal, otherwise, judges that the module-6 to be measured is abnormal, and the measurement result is deviated and needs to be recalibrated.

In an embodiment of the present application, an LCR measurement calibration is provided. The user sends a control instruction to the control panel-1 of the calibration equipment through the network port of the upper computer. The control board-1 receives the instruction and controls the corresponding standard element unit to be switched to the standard inductance L, the standard capacitance C or the standard resistance R, and the standard element is transmitted to the output end of the calibration equipment through the circuit switching unit. The standard element is transmitted to the simulation board via a cable between the calibration device and the simulation board. And then the workpiece is transmitted to the input end of the jig through the corresponding thimble.

The user sends a control instruction to the control panel-2 of the jig through the network port of the upper computer. The control board-2 receives the instruction, controls the standard element LCR measuring function of the corresponding module-7, reads the standard element measuring result through the bus, and uploads the measuring result to the upper computer through the network port. And simultaneously, reading the reference values of the standard inductance L, the standard capacitance C or the standard resistance R in the EEPROM, and uploading the reference values to an upper computer. The reference value in the memory EEPROM is the result of the standard element unit switching to the standard inductance L, the standard capacitance C or the standard resistance R, which are measured with the standard instrument LCR meter.

The upper computer receives the measurement results of the standard inductance L, the standard capacitance C or the standard resistance R, compares the measurement results with the corresponding reference values in the EEPROM, judges whether the accuracy and uncertainty of the LCR are in the design range, if so, the module-7 to be measured on the jig is normal, otherwise, the module-7 to be measured is abnormal, the measurement results deviate, and recalibration is needed.

Further, in one embodiment of the present application, the calibration device audio Kit is mainly composed of a power supply portion DC/DC Unit, a Control Board, a memory Unit EEPROM, a direct current measurement Unit DC Measurement Unit, an alternating current measurement Unit AC Measurement Unit, an alternating current Source Unit AC Source Unit, a direct current Source Unit DC Source Unit, and a standard component Standard Components. The functional block diagram is shown in fig. 4 below.

And in the Power supply part, a Power output port of the external Power adapter is connected to a Power input port Power Header of the calibration equipment Audio Kit, so that Power is supplied to the calibration equipment audio Kit. And then different voltages, such as + -7V, 5.3V, 5V and 3.3V, are output through the power conversion unit, and power is supplied to different devices respectively.

The Control Board is a core Board using All Programmable System-on-Chip (SoC) as a main controller. The Soc with the core integrates the dual-core ARM Cotex-A9 MPU and the Artix-7 FPGA, so that the Soc has abundant hardware peripheral interfaces, mature system support and strong parallel data processing capability. The PS (ARM) end of the core board is provided with 1GB DDR3 SDRAM, 4GB eMMC Flash and 128M-bit Quad-SPI Flash, which provide support for the operation of an operating system. The PS end, namely the ARM end, is also provided with a 10M/100M/1000M Ethernet interface, a USB 2.0 OTG high-speed interface and 11 digital IOs (6 of 1.8V and 5 of 3.3V), and the digital IOs of the PS end CAN be multiplexed and configured into a group of SDIO, two groups of IIC, two groups of CAN or a group of UART and the like besides being used as general IOs. The PS end has two groups of UART, one group of UART is connected with a USB-to-UART chip on the module, and the UART is only used for outputting the system log. The other group can be multiplexed by external IO of PS. Besides PS end interface, PL end has 124 digital IO, these digital IO are divided into three banks: bank34, bank35, bank13, wherein there are 50 Bank34 IOs, 49 Bank35 IOs, 25 Bank13 IOs, each of which can be independently configured as 3.3V, 2.5V, or 1.8V. The IO of these PLs may be configured as functions such as SPI, IIC, UART by the IP core of the software. The Soc on board has two 12-bit 1MSPS ADCs inside, one ADC is used for measuring the external input analog voltage, and the other ADC is used for measuring the temperature.

The control board receives and uploads data through the network port, controls each control circuit, controls each functional module unit, reads the data and reads and writes the EEPROM data of the storage unit.

The memory cell EEPROM is a charged erasable programmable read only memory. The memory chip is free from data loss after power failure. The EEPROM may be reprogrammed by erasing the existing information on a computer or on a dedicated device.

The storage unit EEPROM stores the calibration coefficients of each functional module unit, and ensures the accuracy in the measuring process and the source output process. And meanwhile, the standard instrument is used for calibration, and measurement results of all functional unit modules on the calibration equipment, namely reference values of various parameters stored in the EEPROM are saved. The reference value is uploaded to the upper computer along with the return instruction when the calibration equipment receives the upper computer instruction.

The direct voltage to be measured is transmitted to a circuit switching unit formed by K4 and K3 through an input Connector of the calibration equipment. K4 and K3 are in a normally closed state, and the direct current voltage to be measured is transmitted to a circuit switching unit formed by K2 and K2. K2 and K1 are in a normally closed state, and the direct current voltage to be measured is transmitted to the input end of the direct current measuring unit.

Filter composed of devices R56, R14 and C4-C6 for direct current voltage to be measuredAnd filtering the common mode interference signal and the differential mode interference signal. The devices OPA2-OPA4 and R7-R13 form a separated three-operational-amplifier instrument amplifier circuit, and the direct-current voltage to be measured is amplified by the separated three-operational-amplifier instrument amplifier circuit to obtain proper gain. And then the DC voltage to be measured is raised upwards by a bias circuit formed by OPA1, R5 and R6, so that the measurement is facilitated. The direct current voltage to be tested is filtered out common mode interference signals and differential mode interference signals through a filter formed by devices R1, R2 and C1-C3. Wherein the gain of the instrument amplifier circuit of the three separated operational amplifiers is related to the resistance value of R7-R13, and the resistance is selected from R11=R12, R9=R10 and R7=R8, so that the output voltage V of the instrument operational amplifier circuit _OUT =(V _P - V _N )*[1+2*(R11/R13)]* (R7/R9). To obtain a high performance differential amplifier, it is critical to obtain a high common mode rejection ratio CMRR, and therefore a good resistance ratio and relative drift matching is required for the resistance of the instrumentation amplifier. For true differential amplification, the resistance matching ratio is very important, i.e. r7/r9=r8/R10. Manufacturing tolerances of the resistor, and parametric behavior of different times and temperatures affect the resistance matching ratio. Assuming a perfect unity gain difference with infinite common mode rejection, a tolerance of 1% resistance match results in a common mode rejection of only 34 dB. Therefore, it is recommended to use a resistance of at least 0.01% or better. The thermal noise of resistor R13 is amplified by the instrumentation op-amp and a value must be chosen low enough to reduce the thermal noise at the output while still providing an accurate measurement. The operational amplifiers OPA3 and OPA4 have high gain, and thus high-precision, low-bias voltage and low-noise amplifiers are selected. The operational amplifier OPA2 operates at a much smaller gain, even a gain of 1, whose type selection requirements are relatively different, and the input voltage noise, the input bias voltage of the operational amplifier are the main considerations.

The direct current voltage to be measured is subjected to analog-to-digital conversion through the ADC device U1, so that a measurement result is obtained. The control board controls the ADC device through the SPI bus, reads the measurement result and uploads the measurement result to the upper computer through the network port.

The measurement of the direct current to be measured is similar to the measurement of the direct voltage, but the direct current to be measured at the input end flows through the circuit switching unit formed by K4 and K3K4 and K3 are normally closed. And then, the K2 and the K1 are in a normally open state through a circuit switching unit formed by the K2 and the K1. Direct current to be measured flows through sampling resistor R _SNS Thereby converting the direct current to be measured into direct voltage, and the subsequent measurement process is consistent with the measurement of the direct current to be measured, thereby obtaining the measurement result of the direct current to be measured.

The alternating voltage to be measured is transmitted to a circuit switching unit formed by K4 and K3 through an input Connector of the calibration device. K4 and K3 are in a normally open state, and the alternating voltage to be measured is transmitted to the input end of the alternating current measuring unit.

The alternating voltage to be measured passes through a coupling capacitor C12, an RC filter formed by devices R25 and C11, and then a low-pass filter formed by OPA6, C10 and R22-R25, so as to filter interference signals outside the audio frequency band of 20Hz-20 kHz. The devices OPA5, OPA7, C7-C9 and R15-R21 form an ADC driving circuit of professional audio. Wherein V on ADC _COM The reference voltage output provides a suitable input common mode reference, V _COM The voltage is buffered by a voltage follower formed by the operational amplifiers OPA7 and R17, and the common-mode voltage is driven to V of the operational amplifier OPA5 _C Pins, which bias the average output voltage of OPA5 to V _COM Voltage value. The signal gain of the circuit is typically set to about 0.25, compatible with commonly used audio line levels. The gain can be adjusted, if necessary, by varying the values of R20 and R21, and the feedback resistors R18 and R19, to obtain optimum noise performance. R15, R16 and C7 provide input filters for the ADC, debug their parameter values to perform satisfactorily, and adjustment of some values may help optimize different performances of the ADC. It is important to maintain an accurate resistance match across resistors R18-R21 to achieve good differential signal balance, and it is recommended that better performance be achieved with a 1% resistance. Capacitors C7-C9 should be carefully selected to achieve good distortion performance, polystyrene, polypropylene, NPO ceramic and mica type capacitors are selected, and polyester and high-k ceramic type capacitors are not selected.

The alternating voltage to be measured is subjected to analog-to-digital conversion through the audio ADC device U2, so that a measurement result is obtained. The control board controls the audio ADC device through the I2S bus, reads the measurement result and uploads the measurement result to the upper computer through the network port.

The Control Board receives the command from the upper computer via the network port and needs to output AC voltage. The control board controls the integrated chip U3 to output the set alternating voltage through the I2S bus.

The alternating voltage forms a bipolar low-pass filter circuit with differential input through devices OPA8, OPA10, R26-R33 and C13-C16 to obtain proper gain amplification and static voltage, and interference signals outside the audio frequency band 20Hz-20kHz are filtered. The alternating voltage forms a differential-to-single-ended signal circuit through devices OPA9 and R34-R37, and is connected to the output end of the alternating current source unit through a coupling capacitor C17 and an output resistor R38. Wherein V on an Audio DAC _Q The pins output a static voltage that biases the differential signal to a suitable voltage. The values of feedback resistors R32 and R33 are gain adjusted, if necessary, by changing R26 and R27. Low pass filters are commonly used in signal processing applications to reduce noise and prevent aliasing. The phase and amplitude response of the DAC depends on external analog circuitry. The gain of the differential-to-single-ended signal circuit is related to the resistance value of R34-R37, and the resistance is R37/R34=R36/R35, so that the differential-to-single-ended signal circuit outputs the voltage V _OUT =(V _P - V _N ) (R37/R34). The common mode rejection ratio of the differential pair operational amplifier is required to be higher, the precision of the peripheral resistor is not matched, and the error is large, so that the common mode rejection ratio of the operational amplifier is influenced, and the precision of the resistor is selected to be higher as much as possible so as to meet the measurement.

The alternating voltage forms a circuit switching unit via K5 and K6. K5 and K6 are in a normally closed state, and an alternating voltage is transmitted to the output end of the calibration device Connector through the circuit switching unit.

The Control Board receives the command from the upper computer via the network port and needs to output DC voltage. The control board outputs the set direct-current voltage through the circuit switching unit.

The voltage source PP7V converts the output reference voltage REF5V via the reference source chip. The reference voltage REF5V is buffered by a voltage follower formed by the devices OPA14 and R46, and the reference voltage enters a resistor voltage dividing unit formed by R47 to R50, and the set dc voltage is obtained by different circuit states of a circuit switching unit formed by K16 and K17. The devices OPA15, OPA16, C19 and R51-R53 constitute a remote feedback circuit of the dc voltage source. Because the transmission line, the connector, the line and the like of the power supply have impedance and contact resistance, the power supply is equivalent to a resistor connected between a direct-current voltage source and a load in series, and the voltage at two ends of the load can be reduced, namely the wire loss according to kirchhoff's law. The voltage at the two ends of the load can be increased to the normal voltage through the adjustment of the far-end feedback circuit. The resistors R52 and R53 can normally output the set direct current voltage when the far end is not connected with a feedback circuit, but the size of the line loss can influence the accuracy of outputting the direct current voltage at the two ends of the load. The set DC voltage is transmitted to the output end of the DC source unit through the normal start end of the circuit switching unit K18.

The set direct current voltage forms a circuit switching unit through K5-K7. K5-K7 are in a normally open state, and the set direct-current voltage is transmitted to the output end of the calibration equipment Connector through the circuit switching unit.

The control board receives the instruction sent by the upper computer through the network port and needs to output direct current. The control board outputs the set direct current through the circuit switching unit.

The voltage source PP7V converts the output reference voltage REF5V via the reference source chip. The reference voltage REF5V is buffered by a voltage follower constituted by the devices OPA11 and R39, and the reference voltage enters a constant current source circuit constituted by the devices OPA12, OPA13, and R40 to R44. The magnitude of the constant current depends on the magnitude of the resistors R40-R44. Resistor selection r40=r41, r42=r43, constant current source size I _OUT =(V _P -V _N )/(G*R _O )= (V _P -V _N ) /(R40/R42 x R44). If r40=r41= K, R42 =r43=r44=1k, then I _OUT = (5-0)/(5K/1K) =1 mA. The set direct current flows through the normally closed end of the circuit switching unit K18 and is transmitted to the output end of the direct current source unit.

The set direct current constitutes a circuit switching unit via K5 and K6. K5 and K6 are in a normally closed state, and the set direct current is transmitted to the output end of the calibration device Connector through the circuit switching unit.

Standard components consist of devices R54, R55, C20 and L1. The standard element selects a device with high precision and high stability. The devices K8-K15 form a circuit switching unit for four-wire measurement of LCR. According to different combination states of K8-K15, different standard elements are connected to an input end of external measuring equipment for measurement through a four-wire measuring method.

In one embodiment of the present application, the mechanical dimensions may be adjusted according to the actual situation, and the basic mechanical dimensions are shown in fig. 5 below.

The technical scheme of the invention has the beneficial effects that: the calibration equipment for the calibration precision detection of the field station equipment has the advantages of automatic calibration, low cost, low power consumption, small volume, light weight, portability, simple wiring, convenience in installation and debugging and simplicity in operation, and is convenient for field staff to maintain the machine.

In one embodiment of the present application, an alternative is provided. The alternative technical scheme one is similar to the detailed technical scheme. The calibration device is divided into a main control part, a connecting cable and various functional accessories. For example, some workstation devices only measure the parameters dc voltage and dc current. In order to adapt to the use of the actual situation on site, the main control part and the direct current measurement unit function module can be connected to the calibration system, and then the automatic calibration system is started to detect the calibration precision of the test equipment. For other work stations, the required functional module units can be selected according to actual conditions, and the power module units are accessed to a calibration system to finish calibration work. This alternative solution can completely replace the effects of the above-mentioned solutions.

Alternative technical scheme II is similar to the technical scheme in detail. The difference is that an adapter plate is added between the calibration equipment and the simulation board, and the adapter plate mainly comprises a circuit switching unit. The control bus of the adapter plate is connected to the calibration device, which provides the control bus. The calibration equipment transmits a control signal to switch the states of the circuit switching units of the adapter plate through the control bus, and the different circuit states of the circuit switching units correspond to the measurement of different functional units, so that the number of the middle connecting cables can be greatly reduced by carrying out time-sharing multiplexing through the same connecting cable, and the function of the technical scheme can be completely replaced.

Alternative solution three is similar to the detailed solution described above. The communication mode with the upper computer is changed into a USB communication interface from a network port. This alternative solution can completely replace the effects of the above-mentioned solutions.

FIG. 6 is a flow chart of a method for calibration accuracy detection of a test device in a compact form, in accordance with an embodiment of the present invention. Fig. 7 is a schematic diagram of a system architecture of a method for detecting calibration accuracy of a test device in a small-sized embodiment of the present invention. As shown in fig. 6 and 7, a compact method for testing calibration accuracy detection of a device includes: 210, a user sends a test control instruction through an upper computer to enable a signal source of a standard instrument to output a standard signal, the standard signal is connected to a signal input interface of a test fixture through a cable, and the standard signal passes through a simulation board and then enters a test bottom board through a thimble on the test fixture; the method for sending the test control instruction by the user through the upper computer comprises the following steps: 211, acquiring test requirement text description from the upper computer; 212, performing data preprocessing on the test requirement text description to obtain a sequence of test requirement description sentences; 213, performing semantic coding on the sequence of the test requirement description sentence to obtain a test requirement semantic coding feature vector; and, 214, generating the test control instruction based on the test requirement semantically encoded feature vector.

It will be appreciated by those skilled in the art that the specific operation of each step in the above-described method for testing device calibration accuracy detection of a small size has been described in detail in the above description of the small size device for testing device calibration accuracy detection with reference to fig. 1 to 5, and thus, repetitive description thereof will be omitted.

Fig. 8 is an application scenario diagram of a small-sized device for testing calibration accuracy detection of a device according to an embodiment of the present invention. As shown in fig. 8, in this application scenario, first, a test requirement text description (e.g., C as illustrated in fig. 8) is acquired at the upper computer; the obtained test requirement text description is then input into a server (e.g., S as illustrated in fig. 8) deployed with a small-scale test device calibration accuracy detection algorithm, wherein the server is capable of processing the test requirement text description based on the small-scale test device calibration accuracy detection algorithm to generate the test control instructions.

The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.

Claims (8)

1. A compact device for testing calibration accuracy detection of a device, comprising: the test device comprises an upper computer, a standard instrument and a test fixture, wherein the test fixture comprises a simulation board and a test bottom board; the user sends a test control instruction through the upper computer to enable a signal source of the standard instrument to output standard signals, the standard signals are connected to a signal input interface of the test fixture through a cable, and the standard signals pass through the simulation board and then enter the test bottom board through the ejector pins on the test fixture.

2. The miniature apparatus for testing calibration accuracy detection of an apparatus according to claim 1, further comprising: a code-free automatic test module; wherein, the code-free automated test module comprises: the demand text testing unit is used for acquiring test demand text description from the upper computer; the data preprocessing unit is used for preprocessing the data of the test requirement text description to obtain a sequence of test requirement description sentences; the semantic coding unit is used for carrying out semantic coding on the sequence of the test requirement description sentence to obtain a test requirement semantic coding feature vector; and the control instruction generation unit is used for generating the test control instruction based on the test requirement semantic coding feature vector.

3. The small-sized apparatus for testing equipment calibration accuracy detection according to claim 2, wherein the data preprocessing unit is configured to: the test requirement text description is passed through an AIGC model-based requirement description perfect expressive machine to obtain perfect test requirement text description; and carrying out sentence segmentation on the perfect test requirement text description to obtain a sequence of the test requirement description sentences.

4. A compact device for testing device calibration accuracy detection according to claim 3, wherein said semantic coding unit comprises: a sentence understanding subunit, configured to pass the sequence of the test requirement description sentence through a Bert model-based sentence understanding device to obtain a sequence of test requirement description sentence feature vectors; and a context coding subunit, configured to pass the sequence of test requirement description sentence feature vectors through a context encoder based on a converter to obtain the test requirement semantic coding feature vector.

5. The small-form factor test device calibration accuracy detection apparatus of claim 4, wherein the context encoding subunit is configured to: one-dimensional arrangement is carried out on the sequence of the test requirement description sentence feature vectors to obtain global test requirement description sentence feature vectors; calculating the product between the global test requirement description sentence feature vector and the transpose vector of each test requirement description sentence feature vector in the sequence of the test requirement description sentence feature vectors to obtain a plurality of self-attention correlation matrices; respectively carrying out standardization processing on each self-attention correlation matrix in the plurality of self-attention correlation matrices to obtain a plurality of standardized self-attention correlation matrices; obtaining a plurality of probability values by using a Softmax classification function through each normalized self-attention correlation matrix in the normalized self-attention correlation matrices; and weighting each test requirement description sentence feature vector in the sequence of test requirement description sentence feature vectors by taking each probability value in the plurality of probability values as a weight to obtain the test requirement semantic coding feature vector.

6. The apparatus for small-sized test apparatus calibration accuracy detection according to claim 5, wherein the control instruction generating unit is configured to: automatically generating the test demand semantic coding feature vector through a test code based on an AIGC model to obtain a test code; and running the test code on the upper computer to generate the test control instruction.

7. The small-form factor test device calibration accuracy detection apparatus of claim 6, further comprising a training module for training the AIGC model-based demand description perfect expressive, the Bert model-based sentence comprehender, the converter-based context encoder, and the AIGC model-based test code automatic generator; wherein, training module includes: the training data acquisition unit is used for acquiring training data, wherein the training data comprises training test requirement text description and a true value of a test code; the training requirement description perfect expression unit is used for enabling the training test requirement text description to pass through the AIGC model-based requirement description perfect expression unit so as to obtain training perfect test requirement text description; the training clause processing unit is used for carrying out clause processing on the training perfect test requirement text description so as to obtain a sequence of training test requirement description sentences; the training sentence understanding unit is used for passing the sequence of the training test requirement description sentence through the Bert model-based sentence understanding device to obtain a sequence of training test requirement description sentence feature vectors; the training context coding unit is used for enabling the sequence of the training test requirement description sentence characteristic vectors to pass through the context coder based on the converter so as to obtain training test requirement semantic coding characteristic vectors; the training optimization unit is used for optimizing the position-by-position characteristic values of the training test requirement semantic coding characteristic vectors to obtain optimized test requirement semantic coding characteristic vectors; the training test code automatic generation unit is used for enabling the semantic coding feature vector of the optimization test requirement to pass through the AIGC model-based test code automatic generator to obtain a training test code; and a training unit for calculating a cross entropy function value between the training test code and a true value of the test code as a loss function value to train the AIGC model-based demand description perfect expressive machine, the Bert model-based sentence comprehener, the converter-based context encoder, and the AIGC model-based test code automatic generator.

8. The small-sized apparatus for testing the calibration accuracy detection of the apparatus according to claim 7, wherein the training optimizing unit is configured to: optimizing the position-by-position characteristic values of the training test requirement semantic coding characteristic vectors by using the following optimization formula to obtain optimized test requirement semantic coding characteristic vectors; wherein, the optimization formula is:wherein (1)>Is the training test requirement semantically encoded feature vector, < >>Is the training test requirement semantic coding feature vector +.>Is>Characteristic value of individual position->Is the +.f. of the optimization test requirement semantically encoded feature vector>Characteristic values of the individual positions.