Automatic test equipment error calibration method and device, terminal equipment and storage medium

Abstract

The invention relates to the technical field of automatic test equipment, in particular to an automatic test equipment error calibration method and device, terminal equipment and a storage medium, and the method comprises the steps: collecting working data in the operation process of the automatic test equipment in real time; performing error analysis on the working data, and determining an error category; establishing an error calibration model according to the error category and the working data, and training the error calibration model; and when the automatic test equipment works, integrating the error calibration model into the automatic test equipment, and carrying out real-time calibration on the automatic test equipment. And more comprehensive calibration operation can be carried out on the automatic test equipment.

Description

Technical Field

[0001] This application relates to the technical field of automatic test equipment, and particularly to a method, device, terminal device and storage medium for calibrating errors of automatic test equipment. Background Art

[0002] Automatic Test Equipment (ATE) is a type of equipment used for automatically testing the functions and performance of electronic components or systems. With the rapid development of electronic technology, ATE plays an indispensable role in semiconductor, electronic component production and communication equipment manufacturing. ATE equipment can quickly detect unqualified products during the production process to ensure the quality and performance of products. However, during the operation of ATE equipment, due to reasons such as equipment aging, changes in environmental conditions, hardware failures or system deviations, the actual output value often has a certain deviation from the preset input value.

[0003] Most current ATE equipment calibration methods rely on static calibration boards to recalibrate the equipment regularly. This calibration method can reduce errors to a certain extent, but it cannot achieve real-time correction. Especially during long-term operation, when the equipment state changes, errors will gradually accumulate, resulting in inaccurate test results. In addition, the existing calibration methods have poor effects in dealing with errors and cannot handle complex equipment output fluctuations. Summary of the Invention

[0004] In view of this, embodiments of this application provide a method for calibrating errors of automatic test equipment, which can effectively solve the problem of inability to handle complex equipment output fluctuations, etc.

[0005] In a first aspect, embodiments of this application provide a method for calibrating errors of automatic test equipment, including:

[0006] Real-time collect the working data during the operation of the automatic test equipment;

[0007] Perform error analysis on the working data to determine the error category;

[0008] According to the error category and the working data, establish an error calibration model and train the error calibration model;

[0009] When the automatic test equipment is working, integrate the error calibration model into the automatic test equipment to perform real-time calibration on the automatic test equipment.

[0010] In one embodiment, the working data includes a preset input value and an actual output value;

[0011] The performing error analysis on the working data to determine the error category includes:

[0012] Fit the preset input value and the actual output value to obtain a fitting result;

[0013] If the fitting result is linear, the error category is a linear error; if the fitting result is non-linear, the fitting result is a non-linear error.

[0014] In one embodiment, establishing an error calibration model based on the error category and the working data, and training the error calibration model includes:

[0015] Construct a corresponding error calibration model according to the error category;

[0016] Substitute the working data into a preset loss function for calculation, and adjust the parameters in the error calibration model according to the result of the loss function until the output accuracy rate of the error calibration model reaches a preset value.

[0017] In one embodiment, when the automatic test equipment is working, integrating the error calibration model into the automatic test equipment for real-time calibration of the automatic test equipment includes:

[0018] Integrate the error calibration model into the automatic test equipment, and collect the input preset value and the actual value output by the automatic test equipment in real time;

[0019] The error calibration model outputs a calibration value according to the preset value and the actual value, and feeds back the calibration value to the system of the automatic test equipment.

[0020] In one embodiment, constructing a corresponding error calibration model according to the error category includes:

[0021] When the error category is a linear error, construct a linear regression model as the error calibration model;

[0022] When the error category is a non-linear error, construct a polynomial regression model, a support vector machine or a neural network model as the error calibration model.

[0023] In one embodiment, substituting the working data into the error calibration model for calculation, and adjusting the parameters in the error calibration model until the output accuracy rate of the error calibration model reaches a preset value includes:

[0024] When the error calibration model is a linear regression model, perform the least squares method on the working data to determine the slope and intercept in the linear regression model;

[0025] And / or, when the error calibration model is a polynomial regression model, perform the least squares method on the working data to fit the regression coefficients in the polynomial regression model;

[0026] And / or, when the error calibration model is a support vector machine, determine the kernel function, and perform SVM training based on the kernel function and the working data to obtain the weight vector and kernel intercept;

[0027] And / or, when the error calibration model is a neural network model, input the working data into the designed neural network model for training, construct a loss function, calculate the loss value according to the loss function, and when the loss value is within a preset interval, determine that the error calibration model is successfully trained.

[0028] In one embodiment, the method further includes:

[0029] Monitor the working conditions of the automatic test equipment in real time. When the working conditions change, adjust the parameters of the error correction model in real time according to the new working conditions;

[0030] Obtain the actual value output by the automatic test equipment in real time, calculate the error value according to the sensor, and feedback and adjust the parameters of the error correction model according to the error value.

[0031] In a second aspect, the present application further provides an automatic test equipment error calibration device, including:

[0032] An acquisition module, configured to acquire the working data during the operation of the automatic test equipment in real time;

[0033] A classification module, configured to perform error analysis on the working data to determine the error category;

[0034] A training module, configured to establish an error calibration model according to the error category and the working data, and train the error calibration model;

[0035] A calibration module, configured to integrate the error calibration model into the automatic test equipment when the automatic test equipment is working, and perform real-time calibration on the automatic test equipment.

[0036] In a third aspect, the present application further provides a terminal device, where the terminal device includes a processor and a memory, the memory stores a computer program, and the processor is configured to execute the computer program to implement the automatic test equipment error calibration method described above.

[0037] In a fourth aspect, the present application further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed on a processor, the automatic test equipment error calibration method described above is implemented.

[0038] The embodiments of the present application have the following beneficial effects:

[0039] By determining the error type and establishing a targeted error calibration model according to the error type, when calibrating the errors of the automatic test equipment subsequently, a more appropriate model can be used for calibration, improving the calibration accuracy. And because it supports various different types of error calibration models, it can handle more complex equipment fluctuations, enabling a wider calibration range and increasing the calibration efficiency and calibration effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0041] Figure 1 FIG. shows a schematic flowchart of a method for calibrating errors of an automatic test equipment according to an embodiment of the present application;

[0042] Figure 2 FIG. shows a schematic structural diagram of a calibration plate and an automatic test equipment according to an embodiment of the present application;

[0043] Figure 3 FIG. shows a schematic structural diagram of an error calibration device for an automatic test equipment according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments.

[0045] The components of the embodiments of the present application described and illustrated in the drawings here are usually arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present application provided in the drawings below is not intended to limit the scope of the present application claimed, but only represents the selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application.

[0046] In the following, the terms "comprising", "having" and their cognates that can be used in various embodiments of the present application are only intended to represent specific features, numbers, steps, operations, elements, components or combinations of the foregoing items, and should not be construed as first excluding the existence of one or more other features, numbers, steps, operations, elements, components or combinations of the foregoing items or precluding the possibility of adding one or more features, numbers, steps, operations, elements, components or combinations of the foregoing items. In addition, the terms "first", "second", "third", etc. are only used for differential description and cannot be construed as indicating or implying relative importance.

[0047] Unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which various embodiments of the present application pertain. The terms (such as those defined in a commonly used dictionary) will be interpreted as having the same meaning as their contextual meaning in the relevant technical field and will not be interpreted as having an idealized meaning or an overly formal meaning unless clearly defined in various embodiments of the present application.

[0048] The following will, with reference to the accompanying drawings, elaborate on some embodiments of the present application. Without conflict, the following embodiments and the features in the embodiments may be combined with each other.

[0049] The calibration version of the current automatic test equipment cannot provide error calibration for complex relationships. The present application provides an automatic test equipment error calibration method. By determining the error categories and establishing different error calibration models for different error categories, the calibration board can, based on this method, correct a variety of complex errors, minimize the errors, and at the same time improve the effect of error correction.

[0050] The following will, with reference to some specific embodiments, illustrate the automatic test equipment error calibration method.

[0051] Figure 1 A flowchart of an automatic test equipment error calibration method according to an embodiment of the present application is shown. Exemplarily, the automatic test equipment error calibration method includes:

[0052] Step S100, collect the working data during the operation of the automatic test equipment in real time.

[0053] An automatic test equipment is a type of equipment used for automatically testing the functions and performances of electronic components or systems. Due to reasons such as equipment aging, changes in environmental conditions, hardware failures or system deviations, the actual output value often has a certain deviation from the preset input value. The error calibration in this embodiment is used to calibrate such a deviation.

[0054] Therefore, the working data mentioned in the steps refers to the preset input values and actual output values of the automatic test equipment. The values to be obtained are mainly used for subsequent error model modeling, so multiple working data will be continuously collected.

[0055] Among them, the preset input values include signals such as current and voltage, which simulate the input data of the equipment under different working conditions and can be continuous values distributed between [0, 100].

[0056] Actual output value: The actual output of the corresponding equipment. There are errors in this actual output. The actual output value contains various linear and nonlinear errors, and these linear and nonlinear errors need to be corrected.

[0057] In addition, it also includes some environmental variables. It can be understood that external conditions such as temperature, humidity, and load during the operation of the equipment will affect the output performance of the equipment.

[0058] After obtaining the data, data cleaning operations will be performed. By removing outliers and noise, the quality of the data set is ensured. Extreme values can be identified and removed through statistical methods.

[0059] Then, the data will be standardized to eliminate the influence of different dimensions. The commonly used standardization formula is:

[0060]

[0061] Among them, x is the original data, μ is the mean of the data, σ is the standard deviation of the data, and x norm is the standardized data.

[0062] All kinds of working data and environmental data collected can be processed in the above way to obtain standardized data.

[0063] Step S200, perform error analysis on the working data to determine the error category.

[0064] Perform error analysis on the collected data. In this embodiment, the errors are divided into linear errors and nonlinear errors. For the linear error part, a linear regression model is used for correction; for the nonlinear error part, advanced algorithms such as polynomial regression, support vector machines, and neural networks can be used for error analysis and correction. By classifying the errors, corresponding methods can be selected for model training to achieve the effect of targeted processing of different errors.

[0065] Among them, linear error: There is a fixed proportional relationship between the output value and the preset value of the equipment, and such errors can be processed through a linear regression model.

[0066] Nonlinear error: There is a complex nonlinear relationship between the output value and the preset value of the device, which may be affected by environmental variables. This type of error needs to be corrected through models such as polynomial regression, support vector machine (SVM), and neural network.

[0067] According to the different mathematical relationships between the above linear error and nonlinear error, through the method of linear fitting, it can be determined whether there is a fixed proportional relationship between the output value and the preset value of the device. If so, it can be directly determined as a linear error, otherwise it can be determined as a nonlinear relationship.

[0068] Step S300, based on the error category and the working data, establish an error calibration model and train the error calibration model.

[0069] As Figure 2 shown, the establishment of the error calibration model in this step S300 includes:

[0070] Step S310, according to the error category, construct a corresponding error calibration model.

[0071] The error category has been judged in the previous steps. Therefore, for different error categories, the corresponding error calibration models can be established.

[0072] For linear errors, use a linear regression model as the error calibration model.

[0073] Assume that there is a linear relationship between the actual output value y and the preset input value x, and the formula is:

[0074] y = a * x + b;

[0075] Among them, a and b are the regression coefficients of the model, representing the slope and intercept of the correction. Determine a and b through the least squares method, and its goal is to minimize the sum of the squares of the errors between the predicted value and the actual output value:

[0076]

[0077] In the formula, m is the number of working data.

[0078] By taking the partial derivative of J(a, b) and solving, the optimal a and b can be obtained, thus completing the correction of the linear error.

[0079] For nonlinear errors, a polynomial regression model can be used as the error calibration model.

[0080] The polynomial regression model is used to handle simple nonlinear errors. When the error shows quadratic or cubic curve characteristics, it can be corrected through polynomial regression. The basic formula of polynomial regression is:

[0081] y = a0 +a 1 x + a 2 x 2 +…+a n x n

[0082] where a 0 , a 1 , …, a n ; are the coefficients of the polynomial regression, and these coefficients are fitted by the least squares method. Polynomial regression can capture relatively simple non-linear features in the device output, such as the trend of the device output value accelerating or decelerating as the preset input value changes.

[0083] The preset input value x can be polynomially extended first, such as generating eigenvectors such as x, x 2 and so on. This can transform the original linear data into a high-degree polynomial form. Then, the extended polynomial features are used to fit the coefficients by the least squares method. Finally, the corrected output value is predicted through the polynomial model.

[0084] For non-linear relationships, a support vector machine (SVM) model can also be used in this embodiment.

[0085] Support vector machine regression (SVR) is used to handle more complex non-linear errors. The core idea of SVM is to map the data to a high-dimensional space by selecting an appropriate kernel function and find an optimal hyperplane in the high-dimensional space to fit the input-output relationship.

[0086] The goal of support vector machine regression is to find a function f(x) such that the error of the data points from this function does not exceed a certain threshold ∈. Its basic form is:

[0087] f(x) = w·φ(x) + b

[0088] where w is the weight vector and φ(x) is the kernel function that maps the data from the original space to a high-dimensional space.

[0089] In this embodiment, the steps for constructing the SVM model include:

[0090] Select the kernel function: Select the kernel function according to the complexity of the data. Commonly used kernel functions include the radial basis function (RBF), and its form is:

[0091] K(x, x′) = exp(-γ∥x - x′∥ 2 )

[0092] where γ controls the width of the kernel function.

[0093] Model training: Use the SVM algorithm to train the model, solve for the optimal w and b, and find the high-dimensional hyperplane that can minimize the error.

[0094] Perform prediction through the trained model, correct the output, and thus perform prediction calibration output.

[0095] In this embodiment, a neural network model can also be used to construct an error correction model to correct the error relationship of the non-linear relationship.

[0096] The neural network model is suitable for error correction problems with high complexity and strong non-linearity. Its basic structure includes an input layer, a hidden layer, and an output layer. Through the connection of multiple neurons and non-linear activation functions, it can learn the complex relationship of the device output.

[0097] For the convenience of explanation, this embodiment takes the commonly used neural network model multi-layer perceptron (MLP) as an example. This neural network model includes:

[0098] Input layer: Input the pre-input settings of the device.

[0099] Hidden layer: Contains several neurons, and each neuron receives the output of the previous layer and performs a non-linear transformation.

[0100] Output layer: Finally output the corrected device result.

[0101] The neural network trains the weights and biases through the error backpropagation algorithm to minimize the loss function. Commonly used activation functions are ReLU or Sigmoid functions. In addition, other functions can also be used.

[0102] In this embodiment, the steps for constructing the error correction model based on the neural network model include:

[0103] Input the working data into the designed neural network model for training, construct a loss function, calculate the loss value according to the loss function, and when the loss value is within the preset interval, determine that the error calibration model training is successful.

[0104] The network structure of the neural network model needs to be designed in advance, mainly involving the number of layers of the network and the number of neurons in each layer. Simple errors can be processed by a single-layer network with a small number of neurons, while complex non-linear problems require a multi-layer neural network.

[0105] During model training, the collected working data is divided into a test set and a training set. Through a large amount of data training, the gradient descent method and the error backpropagation algorithm are used. The data in the training set is used for the training process, and the data in the test set is used for testing after the training is completed to adjust the weights and biases in the network.

[0106] Specifically, the training set can account for 80% of the total data volume and is used to fit the model to help the machine learning algorithm find the optimal parameter combination. The training of the model is based on a large amount of device output and error data under different scenarios to ensure that the model can handle various device conditions.

[0107] The test set can account for 20% and is used to adjust the hyperparameters of the model (such as regularization coefficient, learning rate, etc.) to ensure that the model does not have the problem of overfitting. The validation set is used to select the most suitable model and parameter settings.

[0108] Step S320, bring the working data into a preset loss function for calculation, and adjust the parameters in the error calibration model according to the result of the loss function until the output accuracy rate of the error calibration model reaches a preset value.

[0109] Among them, for each of the above error correction models (linear regression, polynomial regression, SVM, neural network), train on the training set and adjust the model parameters to minimize the loss function. The loss function usually adopts the mean squared error (MSE), and the formula is:

[0110]

[0111] In the formula, y i is the actual output value, is the model prediction value, and n is the number of samples. The mean squared error (MSE) is used to evaluate the difference between the prediction value and the actual output value. The smaller the MSE, the better the fitting effect of the model.

[0112] After the model training is completed, evaluate the model on the validation set. Use the validation set to test the generalization ability of the model and prevent the model from overfitting. In addition to MSE, the commonly used evaluation metric can also use R 2 (coefficient of determination) to measure the explanatory power of the model:

[0113]

[0114] Among them, is the average value of the actual output values. R 2 close to 1 indicates that the model can well explain the changes in the data and has strong generalization ability.

[0115] In the validation stage, if it is found that the prediction effect of the model is not good, the model can be optimized by adjusting the hyperparameters. For example:

[0116] For the linear regression model, overfitting can be avoided through regularization (such as L1 or L2 regularization).

[0117] For the polynomial regression model, the order of the polynomial can be adjusted to control the model complexity.

[0118] For a support vector machine, the kernel function parameters (such as the γ value in the radial basis function) can be adjusted to optimize the model.

[0119] For a neural network model, hyperparameters such as the number of hidden layers, the number of neurons in each layer, and the learning rate can be adjusted to optimize the network.

[0120] Specifically, how to adjust needs to be set according to the actual equipment and working environment.

[0121] Step S400, when the automatic test equipment is working, integrate the error calibration model into the automatic test equipment to perform real-time calibration on the automatic test equipment.

[0122] According to the operations of the foregoing steps, the model training is completed. At this time, the calibration model can be integrated into the control system of the automatic test equipment, and the system can perform error correction by collecting the output data of the equipment in real time.

[0123] For example, during the operation of the equipment, the preset input value and the actual output value are collected in real time, and the trained calibration model is used to adjust the output of the equipment. The model automatically predicts the corrected output value according to the input preset value, feeds it back to the control system, and adjusts the actual output of the equipment to make it closer to the preset value. In this way, the purpose of real-time correction can be achieved.

[0124] At the same time, this embodiment can also provide a method for dynamically adjusting the model. The system continuously monitors the output of the equipment and dynamically adjusts the parameters of the calibration model according to the magnitude of the current error. Through the feedback control mechanism, when the environmental conditions (such as temperature, humidity, load, etc.) of the equipment change, the equipment can perform self-adaptive correction to ensure the accuracy and stability of the output value. Moreover, as the equipment runs for a long time, the model may need to be retrained and optimized. By continuously collecting and analyzing historical data, the system regularly updates the parameters of the model to ensure that the calibration effect of the equipment remains in the best state for a long time.

[0125] Among them, the feedback control can be implemented by a PID controller. The formula of the PID controller is:

[0126]

[0127] In the formula, e(t) is the error between the real-time output value and the preset value, K p is the proportional gain, K i is the integral gain, K d is the derivative gain. The PID controller can dynamically adjust the parameters of the equipment to make the output error gradually converge.

[0128] The system calculates the current error value e(t) based on the actual output and preset value of the device. The error is measured in real time by a sensor and fed back to the control system. The system adjusts the operating parameters of the device through a PID controller to reduce the error.

[0129] For example, if the output current of the device is too high, the PID controller will adjust the preset value of the current according to the magnitude and change trend of the current error, so that the output current gradually returns to the expected range.

[0130] When the operating conditions of the device (such as temperature, load, etc.) change drastically, the system can adjust the operating parameters of the device through the above feedback control, and can also dynamically adjust the parameters of the error correction model (such as the kernel function parameters of SVM, the learning rate of neural networks, etc.) according to real-time data, so that the model can quickly adapt to the error situation in the new environment.

[0131] The system continuously monitors the error change between the output of the device and the preset value. As time goes by, the feedback system will gradually reduce the error until the device output completely matches the preset value. During this process, the system will optimize the gain coefficient of the PID controller according to the change law of historical errors, further improving the control effect, so as to achieve the technical effect of real-time monitoring and feedback optimization.

[0132] The method for calibrating the error of the automatic test equipment in this embodiment comprehensively corrects the equipment error by combining multiple regression models (linear regression, polynomial regression, support vector machine, neural network), so that the error can be corrected by a suitable regression model, significantly improving the accuracy of the equipment output. And in the actual calibration work, it combines machine learning technology and feedback control mechanism, enabling the system to have self-adaptive ability. When the device environment or working conditions change, the system can automatically adjust the parameters of the correction model to maintain the stability of the device output. Compared with traditional manual calibration, the present invention realizes true automatic error correction and can respond to external changes during the operation of the device in real time. Thus, it reduces the operation process of frequent shutdown and manual adjustment by staff, fully automates the calibration and adaptive update operations, reduces the requirements for manual operation, improves the operation efficiency of the device, and further reduces the maintenance cost. Especially in large-scale production lines, reducing equipment downtime and manual maintenance plays an important role in improving the overall production efficiency.

[0133] Figure 3 FIG. shows a schematic structural diagram of an error calibration device for an automatic test equipment according to an embodiment of the present application. Exemplarily, the device includes:

[0134] An acquisition module 10 for real-time acquisition of working data during the operation of the automatic test equipment;

[0135] A classification module 20, configured to perform error analysis on the working data to determine an error category;

[0136] A training module 30, configured to establish an error calibration model according to the error category and the working data, and train the error calibration model;

[0137] A calibration module 40, configured to integrate the error calibration model into the automatic test device to perform real-time calibration on the automatic test device when the automatic test device is working.

[0138] It can be understood that the device in this embodiment corresponds to the method in the above embodiment, and the optional items in the above embodiment are also applicable to this embodiment, so they will not be described repeatedly here.

[0139] This application further provides a terminal device, where the terminal device includes a processor and a memory, the memory stores a computer program, and the processor is configured to execute the computer program to implement the automatic test device error calibration method described above.

[0140] This application further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed on a processor, the automatic test device error calibration method described above is implemented.

[0141] Among them, the processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including at least one of a central processing unit (CPU), a graphics processing unit (GPU), a network processor (NP), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc., and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application.

[0142] The memory can be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electric Erasable Programmable Read-Only Memory (EEPROM), etc. Among them, the memory is used to store computer programs, and after receiving the execution instruction, the processor can execute the computer program accordingly.

[0143] The computer-readable storage medium is used to store the computer program used in the above terminal device. For example, the computer-readable storage medium may include, but is not limited to: various media that can store program codes such as USB flash drives, mobile hard disks, Read-Only Memory (ROM), Random Access Memory (RAM), magnetic disks, or optical discs.

[0144] In several embodiments provided in this application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely illustrative. For example, the flowcharts and structure diagrams in the drawings show the possible architectures, functions, and operations of the devices, methods, and computer program products according to multiple embodiments of this application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in the alternative implementation, the functions marked in the blocks can occur in a different order from that marked in the drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and sometimes they can also be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the structure diagram and / or flowchart, as well as the combination of blocks in the structure diagram and / or flowchart, can be implemented by a dedicated hardware-based system that executes the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0145] In addition, in each embodiment of this application, each functional module or unit can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part.

[0146] When the above-mentioned functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a terminal device (which can be a smart phone, a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application.

[0147] As described above, the above are only specific implementation manners of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application.

Claims

1. A method for calibrating an automatic test equipment error, characterized in that: include: Real-time collection of working data during the operation of automatic test equipment; Performing error analysis on the working data to determine error categories; Establishing an error calibration model according to the error category and the working data, and training the error calibration model; When the automatic test equipment is working, the error calibration model is integrated into the automatic test equipment to perform real-time calibration on the automatic test equipment.

2. The automatic test equipment error calibration method according to claim 1, characterized in that: The working data includes preset input values ​​and actual output values; The performing error analysis on the working data to determine the error category includes: Fitting the preset input value and the actual output value to obtain a fitting result; If the fitting result is linear, the error category is a linear error; if the fitting result is nonlinear, the fitting result is a nonlinear error.

3. The automatic test equipment error calibration method according to claim 1, characterized in that: The step of establishing an error calibration model according to the error category and the working data, and training the error calibration model, comprises: According to the error category, construct a corresponding error calibration model; The working data is brought into a preset loss function for calculation, and according to the result of the loss function, the parameters in the error calibration model are adjusted until the output accuracy of the error calibration model reaches a preset value.

4. The automatic test equipment error calibration method according to claim 1, characterized in that: When the automatic test equipment is working, the error calibration model is integrated into the automatic test equipment to perform real-time calibration on the automatic test equipment, including: Integrate the error calibration model into an automatic test device to collect input preset values ​​and actual values ​​output by the automatic test device in real time; The error calibration model outputs a calibration value according to the preset value and the actual value, and feeds back the calibration value to the system of the automatic test equipment.

5. The automatic test equipment error calibration method according to claim 3, characterized in that: The step of constructing a corresponding error calibration model according to the error category includes: When the error category is a linear error, a linear regression model is constructed as an error calibration model; When the error category is a nonlinear error, a polynomial regression model, a support vector machine or a neural network model is constructed as an error calibration model.

6. The automatic test equipment error calibration method according to claim 3, characterized in that: The step of bringing the working data into the error calibration model for calculation, and adjusting parameters in the error calibration model until the output accuracy of the error calibration model reaches a preset value, includes: When the error calibration model is a linear regression model, performing least square method on the working data to determine the slope and intercept in the linear regression model; and or, when the error calibration model is a polynomial regression model, performing least square method on the working data to fit the regression coefficients in the polynomial regression model; and or, when the error calibration model is a support vector machine, determining a kernel function, performing SVM training according to the kernel function and the working data, and obtaining a weight vector kernel intercept; And or, when the error calibration model is a neural network model, the working data is input into a designed neural network model for training, a loss function is constructed, and a loss value is calculated according to the loss function. When the loss value is within a preset interval, it is determined that the error calibration model training is successful.

7. The automatic test equipment error calibration method according to claim 1, characterized in that: Also includes: monitoring the working conditions of the automatic test equipment in real time, and when the working conditions change, adjusting the parameters of the error correction model in real time according to the new working conditions; The actual value output by the automatic test equipment is acquired in real time, an error value is calculated according to a sensor, and the parameters of the error correction model are adjusted by feedback according to the error value.

8. An automatic test equipment error calibration device, characterized in that: include: The acquisition module is used to collect the working data of the automatic test equipment in real time during operation; A classification module, used for performing error analysis on the working data and determining error categories; A training module, used to establish an error calibration model according to the error category and the working data, and train the error calibration model; The calibration module is used to integrate the error calibration model into the automatic test equipment when the automatic test equipment is working, so as to perform real-time calibration on the automatic test equipment.

9. A terminal device, characterized in that: The terminal device comprises a processor and a memory, wherein the memory stores a computer program, and the processor is configured to execute the computer program to implement the automatic test equipment error calibration method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The device stores a computer program, which, when executed on a processor, implements the automatic test equipment error calibration method according to any one of claims 1 to 7.

Images