A test bench data inference method and system based on intelligent data analysis

Abstract

The present invention relates to the field of electrical digital data processing technology, and specifically to a test bench data inference method and system based on intelligent data analysis. The present invention realizes the intelligent training of test bench data inference models for different vibration tests. When using the present invention, the user only needs to input the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test, and the corresponding test bench data inference model can be called. The experimental data automatically collected by the data acquisition system of the vibration test bench is input into the test bench data inference model, and the test results of the vibration test can be obtained through data intelligent analysis, without the need for manual review and evaluation of each vibration test, and solves the technical problem that the vibration test data may have errors under human eye observation, that is, the vibration data analysis may have errors and low reliability, and greatly improves the efficiency, reliability and intelligence of the vibration test.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical digital data processing, and in particular to a test bench data inference method and system based on intelligent data analysis. Background Art

[0002] The main purpose of vibration testing is to simulate various vibration environments that a product may encounter during transportation, installation and use, in order to test whether it can withstand the impact of these environments. Specifically, the purpose of vibration testing is to:

[0003] Confirm the reliability of the product: Through vibration testing, the vibration resistance of the product in the expected use environment can be evaluated to ensure that it can work normally in a vibration environment.

[0004] ‌Discover and solve design problems in advance‌: Vibration testing can help discover weak links in product design and make improvements in advance to avoid product failures in actual use.

[0005] ‌Evaluate product lifespan‌: Through vibration testing, the product lifespan in long-term use can be estimated to ensure product durability.

[0006] ‌Screening out defective products‌: Vibration testing can screen out defective products before they leave the factory, thus improving the overall quality of the products.

[0007] However, in the prior art, since the product names, product quantities, and various setting parameters of the vibration test benches are often different for different vibration tests, and even the vibration tests on the same vibration machine may often change due to the timing of the vibration tests, the test results of each vibration test are often reviewed and evaluated manually, but the vibration test data may have errors under human observation, that is, the vibration data analysis may have errors and low reliability, and each vibration test is manually reviewed and evaluated, which wastes a lot of time and manpower, and has low detection efficiency.

[0008] Therefore, the prior art needs to be further developed. Summary of the invention

[0009] The purpose of the present invention is to overcome the above technical deficiencies and provide a test bench data inference method and system based on intelligent data analysis to solve the problems existing in the prior art.

[0010] To achieve the above technical objectives, according to a first aspect of the present invention, the present invention provides a test bench data inference method based on intelligent data analysis, comprising:

[0011] S100, using the MES manufacturing system to obtain the product name, product quantity, various setting parameters of the vibration test bench, vibration test results and experimental data of any vibration test input by the user as a process data set of a vibration test, obtain a first preset number of process data sets of vibration tests, and group each process data set according to the product name, product quantity and various setting parameters of the vibration test bench of each process data set, that is, the process data sets with the same product name, product quantity and various setting parameters of the vibration test bench are grouped into the same group, and each group of process data sets is divided into a training data set and a verification data set according to a first preset ratio;

[0012] S200, using the training data set and the verification data set corresponding to each group of process data sets, respectively training the test bench data inference model corresponding to each group of process data sets, using the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets;

[0013] S300. After the training of the test bench data inference model corresponding to each group of process data sets is completed, the product name, product quantity, various setting parameters of the vibration test bench and experimental data of the vibration test bench for the next vibration test input by the user are obtained, and the corresponding test bench data inference model is called according to the product name, product quantity and various setting parameters of the vibration test bench for the next vibration test input by the user, and the experimental data of the vibration test bench is input into the test bench data inference model, so as to determine and obtain the test result of the vibration test.

[0014] Specifically, the experimental data of the vibration test bench include:

[0015] Amplitude, frequency, temperature, acceleration, vibration time, vibration direction, spectral density, ambient temperature, ambient humidity.

[0016] Specifically, the various setting parameters of the vibration test bench include:

[0017] Vibration type, frequency range, acceleration amplitude, vibration direction, orbit scanning mode, cumulative damage parameters, displacement limit, ambient temperature, ambient humidity.

[0018] Specifically, the method uses the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, and uses the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets, including:

[0019] The training data set is input into the preset network layer in batches for training, and the preset network layer includes a Transformer network layer. The Transformer network layer is used to predict the experimental data of the vibration test bench for the next vibration test with the same product name, product quantity, and various setting parameters of the vibration test bench according to the experimental data of the current vibration test bench through forward propagation, so as to obtain the predicted loss value; the loss value of the preset network layer is calculated and input into the optimizer for optimization to determine the direction in which the parameter gradient of the test bench data inference model decreases fastest; the test bench data inference model performs back propagation according to the loss value and the parameter gradient of the model to optimize the parameters of the test bench data inference model.

[0020] Specifically, the method uses the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, uses the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimizes the model parameters, and completes the training of the test bench data inference model corresponding to each group of process data sets, and also includes:

[0021] After each training, the verification data set is input into the preset network layer of the previous training in batches for model parameter verification, and cyclic training is performed, with the total number of training rounds set to the first preset round; the loss value of the preset network layer is recorded, and it is determined whether the loss value meets the first preset condition, and whether to end the training and output the model parameters based on the judgment result.

[0022] Specifically, the determining whether the loss value satisfies the first preset condition, and determining whether to end the training and output the model parameters according to the determination result, includes:

[0023] If the loss value meets the first preset condition, the training is terminated and the parameters of the current test bench data inference model are output; if the loss value does not meet the first preset condition, the training continues.

[0024] Specifically, the first preset condition is: after the second preset round of training, the loss values ​​obtained in the next round of training are greater than or equal to the loss values ​​that occurred during the second preset round of training.

[0025] Specifically, the test result of the vibration test includes a normal test result and an abnormal test result, and the experimental data of the vibration test bench is input into the test bench data inference model to determine the test result of the vibration test, including:

[0026] If the test result of the vibration test is normal, a prompt signal indicating that the test result of the vibration test is normal is output.

[0027] Specifically, the step of inputting the experimental data of the vibration test bench into the test bench data inference model, and then determining and obtaining the test result of the vibration test, further includes:

[0028] If the test result of the vibration test is abnormal, a prompt signal indicating that the test result of the vibration test is abnormal is output.

[0029] According to a second aspect of the present invention, there is provided a test bench data inference system based on intelligent data analysis, comprising:

[0030] An acquisition module is used to use the MES manufacturing system to acquire the product name, product quantity, various setting parameters of the vibration test bench, vibration test results and experimental data of any vibration test input by the user as a process data set of a vibration test, and acquire the process data sets of the first preset number of vibration tests; and to acquire the product name, product quantity, various setting parameters of the vibration test bench and experimental data of the vibration test bench for the next vibration test input by the user after the test bench data inference model training corresponding to each group of process data sets is completed;

[0031] A control module is used to group each process data set according to the product name, product quantity, and various setting parameters of the vibration test bench of each process data set, that is, the process data sets grouped into the same group are the process data sets with the same product name, product quantity, and various setting parameters of the vibration test bench, and each group of process data sets is divided into a training data set and a verification data set according to a first preset ratio; used to use the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, use the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets; used to call the corresponding test bench data inference model according to the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test input by the user, input the experimental data of the vibration test bench into the test bench data inference model, and then determine the test result of the vibration test.

[0032] Beneficial effects:

[0033] The present invention realizes the intelligent training of test bench data inference models for different vibration tests. When using the present invention, the user only needs to input the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test, and the corresponding test bench data inference model can be called. The experimental data automatically collected by the data acquisition system of the vibration test bench is input into the test bench data inference model, and the test result of the vibration test can be obtained through data intelligent analysis. There is no need for manual review and evaluation of each vibration test, and the technical problem that the vibration test data may have errors under human eye observation, that is, the vibration data analysis may have errors and low reliability is solved. The efficiency, reliability and intelligence of the vibration test are greatly improved, and the labor and time costs of the vibration test are greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flow chart of a test bench data inference method based on intelligent data analysis provided in a specific embodiment of the present invention;

[0035] Figure 2 Schematic diagram of the system composition of a test bench data inference system based on intelligent data analysis provided in a specific embodiment of the present invention;

[0036] Figure 3 It is a structural schematic diagram of the test bench data inference model provided in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention is clearly and completely described below in conjunction with the accompanying drawings of the present invention. Based on the embodiments in this application, other similar embodiments obtained by ordinary technicians in this field without making creative work should all fall within the scope of protection of this application. In addition, the directional words mentioned in the following embodiments, such as "up", "down", "left", "right", etc., are only reference to the directions of the accompanying drawings. Therefore, the directional words used are used to illustrate rather than limit the invention.

[0038] The present invention will be further described below in conjunction with the accompanying drawings and preferred embodiments.

[0039] See also Figure 1-2 The present invention provides a test bench data inference method based on intelligent data analysis, comprising:

[0040] S100. Utilize the MES manufacturing system to obtain the product name, product quantity, various setting parameters of the vibration test bench, vibration test results and experimental data of any vibration test input by the user, and record them as a process data set of a vibration test; obtain a first preset number of process data sets of vibration tests; group each process data set according to the product name, product quantity and various setting parameters of the vibration test bench of each process data set, that is, the process data sets grouped into the same group are the process data sets with the same product name, product quantity and various setting parameters of the vibration test bench; and divide each group of process data sets into a training data set and a verification data set according to a first preset ratio.

[0041] Specifically, the experimental data of the vibration test bench include:

[0042] Amplitude, frequency, temperature, acceleration, vibration time, vibration direction, spectral density, ambient temperature, ambient humidity.

[0043] It should be noted here that vibration testing is a test method to evaluate the product's ability to resist vibration in a real service environment, and usually focuses on the following key parameters:

[0044] 1. Vibration frequency: This is the core parameter of the vibration test, which indicates the number of vibrations per second (Hz). It determines the vibration mode of the test, such as sinusoidal vibration, random vibration, etc.

[0045] 2. Acceleration: The unit is usually g (multiples of the earth's gravity acceleration). Acceleration is an important indicator for measuring vibration intensity, which directly affects the stress state of the internal components of the product.

[0046] 3. Vibration amplitude: usually expressed as peak value or effective value, it is a parameter to measure the magnitude of vibration displacement and is related to acceleration and frequency.

[0047] 4. Vibration direction: Vibration tests may need to be conducted on three orthogonal axes (X, Y, and Z axes, corresponding to up and down, left and right, and front and back directions) to simulate multi-axis vibration in actual environments.

[0048] 5. Vibration time: The duration of the test, which may be a specific period or until the product reaches a preset failure condition.

[0049] 6. Spectral density: In random vibration tests, this parameter describes the distribution of vibration intensity at different frequencies.

[0050] 7. Environmental conditions: Environmental factors such as temperature and humidity may need to be controlled simultaneously during vibration testing to simulate a more realistic usage environment.

[0051] Specifically, the various setting parameters of the vibration test bench include:

[0052] Vibration type, frequency range, acceleration amplitude, vibration direction, orbit scanning mode, cumulative damage parameters, displacement limit, ambient temperature, ambient humidity.

[0053] It is understandable that before the vibration experiment, the following parameters generally need to be set:

[0054] 1. Vibration type: Before the vibration test, it is necessary to determine whether to conduct a sinusoidal vibration test or a random vibration test, or other types of vibration simulation such as impact, drop, etc. Each vibration type corresponds to different physical effects and application environments.

[0055] 2. Frequency range: Sets the lower and upper limits of the vibration frequency, which is a key parameter that determines the impact of vibration on the product. The frequency is usually determined by the product's usage environment and the expected vibration source.

[0056] 3. Acceleration amplitude: Determine the peak value or root mean square value of the acceleration of the vibration test according to relevant standards or engineering requirements. This parameter will affect the vibration intensity that the product is subjected to.

[0057] 4. Vibration direction: Determine the direction of vibration, including longitudinal, lateral and vertical directions, or set it according to the three-dimensional spatial coordinates of the product during actual operation.

[0058] 5. Vibration Time and Cycle: Set the total duration of the vibration test and the test time at each frequency point. This ensures sufficient exposure time to fully evaluate the vibration resistance of the product.

[0059] 6. Orbital sweep mode: In a sinusoidal vibration test, you need to decide whether to use linear sweep, logarithmic sweep, or other sweep mode, which affects the speed of frequency change.

[0060] 7. Cumulative damage parameters (RMS isochrones): For random vibration, cumulative damage parameters such as RMS isochrones need to be set to ensure that the cumulative effect of vibration at different frequencies is the same.

[0061] 8. Displacement Limits: For some sensitive products, it may be necessary to set displacement limits to prevent excessive deformation or damage.

[0062] 9. Temperature environment: If the actual environment needs to be simulated, the vibration test may need to be carried out at a specific temperature, so the corresponding temperature parameters need to be set.

[0063] It should be noted that the steps before step S100 include:

[0064] A first preset number, a first preset ratio, a first preset round, and a second preset round are preset in the control module.

[0065] It can be understood that the first preset number, the first preset ratio, the first preset round, and the second preset round can be specifically set according to the actual needs of the user of the present invention. The present invention does not limit the specific values ​​of the above parameters, and there is no regularity in setting the specific values ​​of the above parameters, as long as it is applicable to the test bench data inference method based on intelligent data analysis proposed in the present invention.

[0066] Preferably, the present invention sets the first preset number to 100 groups, the present invention sets the first preset ratio to 8:2, the present invention sets the first preset round to 20,000 times, and the present invention sets the second preset round to 300. The above settings can further improve the performance of the model and the efficiency of model training, and greatly improve the accuracy and reliability of the prediction of the test bench data inference model of the present invention.

[0067] It can be understood that, assuming that the test results of the vibration test are normal, in vibration experiments with the same product name, product quantity, and various setting parameters of the vibration test bench, there is a certain coupling law between the experimental data of the vibration test bench of different vibration tests, that is, there is a certain dynamic balance relationship. If the test result of a certain vibration test is abnormal, the experimental data of the vibration test bench of this vibration test will break such dynamic balance relationship. The present invention designs a test bench data inference model based on this.

[0068] It can be understood that the experimental data of the vibration test bench is automatically collected by the data acquisition system of the vibration test bench and transmitted to the control module. The data acquisition system of the vibration test bench is a prior art and the present invention will not be elaborated in detail here.

[0069] S200. Use the training data sets and verification data sets corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, and use the verification data sets to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets.

[0070] Specifically, the method uses the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, and uses the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets, including:

[0071] The training data set is input into the preset network layer in batches for training, and the preset network layer includes a Transformer network layer. The Transformer network layer is used to predict the experimental data of the vibration test bench for the next vibration test with the same product name, product quantity, and various setting parameters of the vibration test bench according to the experimental data of the current vibration test bench through forward propagation, so as to obtain the predicted loss value; the loss value of the preset network layer is calculated and input into the optimizer for optimization to determine the direction in which the parameter gradient of the test bench data inference model decreases fastest; the test bench data inference model performs back propagation according to the loss value and the parameter gradient of the model to optimize the parameters of the test bench data inference model.

[0072] It can be understood that Transformer is a deep learning model based on the self-attention mechanism, which consists of a multi-head attention mechanism and a feedforward neural network. It can process sequence data and capture long-distance dependencies. The Transformer model does not contain traditional convolutional layers, but processes each element in the input sequence through an attention layer. The typical structure of Transformer includes an encoder and a decoder, wherein the encoder is composed of multiple identical layers stacked together, and the decoder adds a self-attention layer on the basis of the encoder to process sequence generation tasks. The present invention uses Transformer to supplement global context information, enhances the model's capabilities in feature extraction and global understanding, greatly improves the performance of the model, and greatly reduces the computational cost.

[0073] Specifically, the preset network layer includes a ResNet network layer, and the ResNet network layer is used to classify and predict the output results according to the collected experimental data of the vibration test bench, and the classification results include normal test results and abnormal test results.

[0074] It should be noted here that ResNet is a deep convolutional neural network that solves the gradient vanishing problem in deep network training by introducing residual connections. The core component of ResNet is the residual block, which usually contains two or three convolutional layers and a jump connection, which allows the gradient to bypass these convolutional layers directly. The network structure of ResNet can be very deep. These networks extract the features of the experimental data of the vibration test bench by stacking multiple residual blocks, and classify and predict the output results through global average pooling and fully connected layers. The classification results include normal test results and abnormal test results.

[0075] See also Figure 3The present invention combines the ResNet network layer and the Transformer network layer. ResNet is good at extracting local features of the experimental data of the vibration test bench in the training data set, while Transformer can supplement the global context information. This combination can enhance the model's ability in feature extraction and global understanding, greatly improve the performance of the model, and greatly reduce the computational cost.

[0076] It can be understood that the ResNet network layer is used to classify and predict the output results through the experimental data of the vibration test bench collected in the training data set and the verification data set. The specific steps are:

[0077] (1) First, put the training data set into the Backbones network layer of the model. Backbone refers to a series of convolutional layers that constitute the skeleton of the neural network. The main function of these layers is to extract the features of the input data. The Backbone network is usually composed of multiple convolutional layers, pooling layers, and activation functions. It can extract meaningful feature representations from the original data. Backbones is composed of multiple Backbones, and the core network layer in Backbone is the Resnet network layer. After the data set is input into the model, the training data set will be input into Backbones for training;

[0078] (2) When constructing Backbone, the test task number and Resnet network layer are added. After the training data set is input, Resnet starts training layer by layer to extract the features of the experimental data of the vibration test bench, and assist Transformer in predicting the test results of the vibration test. The extracted features obtain the number data corresponding to the test task with the help of the test task number, which facilitates relevant staff to quickly locate the vibration test with abnormal test results, further improving the intelligence and usability of the present invention, and further improving the efficiency of the vibration test.

[0079] Specifically, the method uses the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, uses the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimizes the model parameters, and completes the training of the test bench data inference model corresponding to each group of process data sets, and also includes:

[0080] After each training, the verification data set is input into the preset network layer of the previous training in batches for model parameter verification, and cyclic training is performed, with the total number of training rounds set to the first preset round; the loss value of the preset network layer is recorded, and it is determined whether the loss value meets the first preset condition, and whether to end the training and output the model parameters based on the judgment result.

[0081] Specifically, the determining whether the loss value satisfies the first preset condition, and determining whether to end the training and output the model parameters according to the determination result, includes:

[0082] If the loss value meets the first preset condition, the training is terminated and the parameters of the current test bench data inference model are output; if the loss value does not meet the first preset condition, the training continues.

[0083] Specifically, the first preset condition is: after the second preset round of training, the loss values ​​obtained in the next round of training are greater than or equal to the loss values ​​that occurred during the second preset round of training.

[0084] Specifically, after each training, the verification data set is input into the preset network layer of the previous training in batches to verify the model parameters, and the cyclic training is performed, including the following methods:

[0085] When the number of training times is not enough and the model needs to be optimized, the training can only be performed from the beginning. Therefore, the present invention develops a method for continuing the following training based on the previously trained model parameter file;

[0086] (1) Provide custom parameters, which can be set according to user needs to continue training based on any training parameters;

[0087] (2) Load the parameters of the custom model file;

[0088] (3) Put the verification data set into the custom parameter network layer for verification, obtain the loss value and record the model parameters of the model corresponding to the minimum loss value during the training process;

[0089] (4) Based on this set of model parameters, forward propagation training and back propagation are performed, and the optimizer optimizes the parameters to obtain a new set of data;

[0090] (5) Put the verification data into the new data network layer for inference to obtain the loss value;

[0091] (6) Perform cyclic training to obtain the best model and complete model optimization.

[0092] It is understandable that the training method adopted by the present invention is batch training, which means that each time the model parameters are updated, only a part of the samples in the validation data set is used, which is called a batch. The advantages of batch training are that it can reduce memory consumption, speed up training, increase randomness, and facilitate the generalization of the model.

[0093] It should be noted here that, in the present invention, batch_size=2, and the batch_size is the batch size. For example, when batch_size=2, it means that two process data sets are selected from the verification data set each time and put into the model for verification.

[0094] Specifically, the determining whether the loss value satisfies the first preset condition, and determining whether to end the training and output the model parameters according to the determination result, includes:

[0095] If the loss value meets the first preset condition, the training is terminated and the parameters of the current test bench data inference model are output; if the loss value does not meet the first preset condition, the training is continued;

[0096] The first preset condition is that after the second preset round of training, the loss values ​​obtained in the next round of training are greater than or equal to the loss values ​​that occurred during the second preset round of training.

[0097] It can be understood that the present invention preferably sets the maximum round of cyclic training to 20,000 times, which can effectively ensure the accuracy of the model prediction of the present invention and record the loss value of the model training. The present invention sets the second preset round to 300, that is, when the loss value of the training no longer decreases within 300 rounds of training, the round with the smallest loss value in the 300 rounds of training is saved as the best verification round parameter to generate the best model file. The above setting makes the training rounds no longer rely on manual judgment, but uses the deep learning model for automatic training, which improves the accuracy of the model prediction, and automatically ends the training when the accuracy meets the requirements, which greatly saves the training time, and can also effectively prevent overfitting, which greatly improves the intelligence level of the present invention and the efficiency of model training.

[0098] Furthermore, the specific process of the test bench data inference model of the present invention is as follows:

[0099] (1) Obtain the latent vector mean mu and logarithmic variance logvar during the forward propagation of the test bench data inference model;

[0100] (2) Calculate a reasonable loss value representing the training effect:

[0101] Calculate the divergence: k1=-0.5*(1+logvar-mu^2-(e^logvar));

[0102] Calculate the loss between the experimental data and the predicted values ​​of the vibration test bench in the validation data set:

[0103] loss=L1+k1*k1_weight;

[0104] Among them, loss is the loss value, k1 is the divergence, k1_weight is the proportion of k1, the value is 10, L1 is the absolute difference between the predicted value and the true value, and the calculation formula of L1 is as follows:

[0105] ;

[0106] Among them, A is the L1 loss in pytorch, which calculates the average or sum of the absolute differences between two tensors; A has a shape of N*C*H*W, which indicates that it is a four-dimensional tensor, N represents the number of samples (Batch Size), indicating that there are N data points, C represents the number of channels (Channels), H and W represent the height and width respectively, which are usually the two-dimensional spatial dimensions of the input features; P represents the padding value, which is a Boolean type, and the shape of P is c*h*w*1. The dimensions of P are the same as those of A in terms of channel, height, and width, but there is no sample number N, which means that P is a mask or padding information for a feature map of each sample, and the Boolean value of P may be used to mark whether the elements in A are filled (True) or valid (False).

[0107] (3) The loss value when using the validation data set to validate the current model network;

[0108] (4) If the validation loss of this round is less than that of the previous round, the best model parameters are replaced with the training model of this round;

[0109] (5) Put the rounds and loss values ​​into the dictionary one by one;

[0110] (6) Compare the loss value of the current round with the loss values ​​of the 300 rounds before this round (parameters are provided and can be set according to actual needs, the default is 300), that is, traverse the loss values ​​from epoch-300 to epoch-1 (epoch is the training round, and comparison is performed when epoch is greater than 300). If the loss values ​​of the current round of epoch are all greater than the loss values ​​from (epoch-300) to (epoch-1), then stop training.

[0111] (7) Save the best model parameters in the cyclic verification as a ckpt model file.

[0112] S300. After the training of the test bench data inference model corresponding to each group of process data sets is completed, the product name, product quantity, various setting parameters of the vibration test bench and experimental data of the vibration test bench for the next vibration test input by the user are obtained, and the corresponding test bench data inference model is called according to the product name, product quantity and various setting parameters of the vibration test bench for the next vibration test input by the user, and the experimental data of the vibration test bench is input into the test bench data inference model, so as to determine and obtain the test result of the vibration test.

[0113] Specifically, the test result of the vibration test includes a normal test result and an abnormal test result, and the experimental data of the vibration test bench is input into the test bench data inference model to determine the test result of the vibration test, including:

[0114] If the test result of the vibration test is normal, a prompt signal indicating that the test result of the vibration test is normal is output.

[0115] Specifically, the step of inputting the experimental data of the vibration test bench into the test bench data inference model, and then determining and obtaining the test result of the vibration test, further includes:

[0116] If the test result of the vibration test is abnormal, a prompt signal indicating that the test result of the vibration test is abnormal is output.

[0117] It can be understood that the present invention realizes the intelligent training of test bench data inference models for different vibration tests. When using the present invention, the user only needs to input the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test, and the corresponding test bench data inference model can be called. The experimental data automatically collected by the data acquisition system of the vibration test bench is input into the test bench data inference model, and the test result of the vibration test can be obtained through intelligent data analysis. There is no need for manual review and evaluation of each vibration test, and the technical problem that the vibration test data may have errors under human eye observation, that is, the vibration data analysis may have errors and low reliability is solved. The efficiency, reliability and intelligence of the vibration test are greatly improved, and the labor and time costs of the vibration test are greatly reduced.

[0118] See also Figure 3 The present invention provides another embodiment, which provides a test bench data inference system based on intelligent data analysis, and the test bench data inference system based on intelligent data analysis includes:

[0119] The acquisition module 100 is used to use the MES manufacturing system to obtain the product name, product quantity, various setting parameters of the vibration test bench, vibration test results and experimental data of any vibration test input by the user as a process data set of a vibration test, and obtain the process data sets of the first preset number of vibration tests; after the test bench data inference model training corresponding to each group of process data sets is completed, the product name, product quantity, various setting parameters of the vibration test bench and experimental data of the vibration test bench for the next vibration test input by the user are obtained;

[0120] The control module 200 is used to group each process data set according to the product name, product quantity, and various setting parameters of the vibration test bench of each process data set, that is, the process data sets grouped into the same group are the process data sets with the same product name, product quantity, and various setting parameters of the vibration test bench, and each group of process data sets is divided into a training data set and a verification data set according to a first preset ratio; it is used to use the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, and use the verification data set to optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets; it is used to call the corresponding test bench data inference model according to the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test input by the user, input the experimental data of the vibration test bench into the test bench data inference model, and then determine the test result of the vibration test.

[0121] It should be noted here that the present invention realizes the intelligent training of test bench data inference models for different vibration tests. When using the present invention, users only need to input the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test, and the corresponding test bench data inference model can be called. The experimental data automatically collected by the data acquisition system of the vibration test bench is input into the test bench data inference model, and the test results of the vibration test can be obtained through intelligent data analysis. There is no need for manual review and evaluation of each vibration test, and the technical problem that the vibration test data may have errors under human eye observation, that is, the vibration data analysis may have errors and low reliability is solved. The efficiency, reliability and intelligence of the vibration test are greatly improved, and the labor and time costs of the vibration test are greatly reduced.

[0122] In a preferred embodiment, the present application further provides an electronic device, the electronic device comprising:

[0123] A memory; and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the test bench data inference method based on intelligent data analysis is implemented. The computer device can be broadly a server, a terminal, or any other electronic device with necessary computing and / or processing capabilities. In one embodiment, the computer device may include a processor, a memory, a network interface, a communication interface, etc. connected via a system bus. The processor of the computer device can be used to provide necessary computing, processing and / or control capabilities. The memory of the computer device may include a non-volatile storage medium and an internal memory. An operating system, a computer program, etc. may be stored in or on the non-volatile storage medium. The internal memory may provide an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface and the communication interface of the computer device can be used to connect and communicate with external devices through a network. When the computer program is executed by the processor, the steps of the method of the present invention are performed.

[0124] The present invention may be implemented as a computer-readable storage medium having a computer program stored thereon, which causes the steps of the method of an embodiment of the present invention to be executed when executed by a processor. In one embodiment, the computer program is distributed on a plurality of computer devices or processors coupled to a network so that the computer program is stored, accessed and executed in a distributed manner by one or more computer devices or processors. A single method step / operation, or two or more method steps / operations, may be performed by a single computer device or processor or by two or more computer devices or processors. One or more method steps / operations may be performed by one or more computer devices or processors, and one or more other method steps / operations may be performed by one or more other computer devices or processors. One or more computer devices or processors may perform a single method step / operation, or perform two or more method steps / operations.

[0125] It will be appreciated by a person skilled in the art that the method steps of the present invention may be performed by instructing related hardware such as a computer device or a processor through a computer program, and the computer program may be stored in a non-temporary computer-readable storage medium, which causes the steps of the present invention to be performed when the computer program is executed. Depending on the circumstances, any reference to memory, storage, database, or other media herein may include non-volatile and / or volatile memory. Examples of non-volatile memory include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid state disk, etc. Examples of volatile memory include random access memory (RAM), external cache memory, etc.

[0126] It can be understood that the present invention realizes the intelligent training of test bench data inference models for different vibration tests. When using the present invention, the user only needs to input the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test, and the corresponding test bench data inference model can be called. The experimental data automatically collected by the data acquisition system of the vibration test bench is input into the test bench data inference model, and the test result of the vibration test can be obtained through intelligent data analysis. There is no need for manual review and evaluation of each vibration test, and the technical problem that the vibration test data may have errors under human eye observation, that is, the vibration data analysis may have errors and low reliability is solved. The efficiency, reliability and intelligence of the vibration test are greatly improved, and the labor and time costs of the vibration test are greatly reduced.

[0127] The various technical features described above can be combined arbitrarily. Although all possible combinations of these technical features are not described, any combination of these technical features should be considered to be covered by this specification as long as there is no contradiction in such combination.

[0128] The specific implementation of the present invention described above does not constitute a limitation on the protection scope of the present invention. Any other corresponding changes and modifications made based on the technical concept of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A test bench data inference method based on intelligent data analysis, characterized in that: The method comprises: S100, using the MES manufacturing system to obtain the product name, product quantity, various setting parameters of the vibration test bench, vibration test results and experimental data of any vibration test input by the user as a process data set of a vibration test, obtain a first preset number of process data sets of vibration tests, and group each process data set according to the product name, product quantity and various setting parameters of the vibration test bench of each process data set, that is, the process data sets with the same product name, product quantity and various setting parameters of the vibration test bench are grouped into the same group, and each group of process data sets is divided into a training data set and a verification data set according to a first preset ratio; S200, using the training data set and the verification data set corresponding to each group of process data sets, respectively training the test bench data inference model corresponding to each group of process data sets, using the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets; S300, after the training of the test bench data inference model corresponding to each group of process data sets is completed, the product name, product quantity, various setting parameters of the vibration test bench and experimental data of the vibration test bench for the next vibration test input by the user are obtained, and the corresponding test bench data inference model is called according to the product name, product quantity and various setting parameters of the vibration test bench for the next vibration test input by the user, and the experimental data of the vibration test bench is input into the test bench data inference model, so as to determine and obtain the test result of the vibration test; The experimental data of the vibration test bench include: Amplitude, frequency, temperature, acceleration, vibration time, vibration direction, spectral density, ambient temperature, ambient humidity; The various setting parameters of the vibration test bench include: Vibration type, frequency range, acceleration amplitude, vibration direction, orbit scanning mode, cumulative damage parameters, displacement limit, ambient temperature, ambient humidity; The method of using the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, and using the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets, includes: The training data set is input into a preset network layer in batches for training. The preset network layer includes a ResNet network layer. The ResNet network layer is used to classify and predict the output results through the experimental data of the vibration test bench collected in the training data set and the verification data set. The classification results include normal test results and abnormal test results.

2. The test bench data inference method based on intelligent data analysis according to claim 1 is characterized in that: The method of using the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, and using the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets, includes: The preset network layer includes a Transformer network layer, and the Transformer network layer is used to predict the experimental data of the vibration test bench for the next vibration test with the same product name, product quantity, and various setting parameters of the vibration test bench according to the experimental data of the current vibration test bench through forward propagation, and then obtain the predicted loss value; calculate the loss value of the preset network layer and input it into the optimizer for optimization, and determine the direction in which the parameter gradient of the test bench data inference model decreases fastest; the test bench data inference model performs back propagation according to the loss value and the parameter gradient of the model to optimize the parameters of the test bench data inference model.

3. The test bench data inference method based on intelligent data analysis according to claim 1, characterized in that: The method further comprises: using the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, using the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimizing the model parameters, and completing the test bench data inference model training corresponding to each group of process data sets; After each training, the verification data set is input into the preset network layer of the previous training in batches for model parameter verification, and cyclic training is performed, with the total number of training rounds set to the first preset round; the loss value of the preset network layer is recorded, and it is determined whether the loss value meets the first preset condition, and whether to end the training and output the model parameters based on the judgment result.

4. The test bench data inference method based on intelligent data analysis according to claim 3 is characterized in that: The determining whether the loss value satisfies the first preset condition, and determining whether to end the training and output the model parameters according to the determination result, includes: If the loss value meets the first preset condition, the training is terminated and the parameters of the current test bench data inference model are output; if the loss value does not meet the first preset condition, the training continues.

5. The test bench data inference method based on intelligent data analysis according to claim 3 is characterized in that: The first preset condition is that after the second preset round of training, the loss values ​​obtained in the next round of training are greater than or equal to the loss values ​​that occurred during the second preset round of training.

6. The test bench data inference method based on intelligent data analysis according to claim 5, characterized in that: The test result of the vibration test includes a normal test result and an abnormal test result. The test data of the vibration test bench is input into the test bench data inference model to determine the test result of the vibration test, including: If the test result of the vibration test is normal, a prompt signal indicating that the test result of the vibration test is normal is output.

7. The test bench data inference method based on intelligent data analysis according to claim 6, characterized in that: The step of inputting the experimental data of the vibration test bench into the test bench data inference model to determine and obtain the test result of the vibration test further includes: If the test result of the vibration test is abnormal, a prompt signal indicating that the test result of the vibration test is abnormal is output.

8. A test bench data inference system based on intelligent data analysis, characterized in that: The test bench data inference method based on intelligent data analysis according to any one of claims 1 to 6 comprises: An acquisition module is used to use the MES manufacturing system to acquire the product name, product quantity, various setting parameters of the vibration test bench, vibration test results and experimental data of any vibration test input by the user as a process data set of a vibration test, and acquire the process data sets of the first preset number of vibration tests; and to acquire the product name, product quantity, various setting parameters of the vibration test bench and experimental data of the vibration test bench for the next vibration test input by the user after the test bench data inference model training corresponding to each group of process data sets is completed; A control module is used to group each process data set according to the product name, product quantity, and various setting parameters of the vibration test bench of each process data set, that is, the process data sets grouped into the same group are the process data sets with the same product name, product quantity, and various setting parameters of the vibration test bench, and each group of process data sets is divided into a training data set and a verification data set according to a first preset ratio; used to use the training data set and the verification data set corresponding to each group of process data sets to respectively train the test bench data inference model corresponding to each group of process data sets, use the verification data set to respectively optimize the test bench data inference model corresponding to each group of process data sets, optimize the model parameters, and complete the training of the test bench data inference model corresponding to each group of process data sets; used to call the corresponding test bench data inference model according to the product name, product quantity, and various setting parameters of the vibration test bench for the next vibration test input by the user, input the experimental data of the vibration test bench into the test bench data inference model, and then determine the test result of the vibration test.

Images