A product detection method, a terminal and a computer readable storage medium

Abstract

The application discloses a product detection method, a terminal and a computer readable storage medium. The motor of each product to be detected in a batch to be detected is controlled to perform variable-speed rotating motion; a vibration acceleration signal of each product to be detected in the batch to be detected is acquired; the vibration acceleration signal is collected by an acceleration sensor of the product to be detected during the variable-speed rotating motion of the motor; a sweep frequency characteristic curve of the product to be detected is generated according to the vibration acceleration signal; the sweep frequency characteristic curve is used to represent the vibration condition of the product to be detected during the variable-speed rotating motion; all sweep frequency characteristic curves of the batch to be detected are compared to determine outliers of the batch to be detected; whether there is an abnormal product in the products to be detected is determined according to the outliers of the batch to be detected, so that the uniformity detection of the batch to be detected is completed. Through the above scheme, accurate and effective uniformity detection of products can be realized automatically, and the product delivery efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of testing technology, and in particular to a product testing method, terminal, and computer-readable storage medium. Background Technology

[0002] During product manufacturing, defects may occur, such as: imbalance of mechanical moving parts, improper fit of transmission components, poor gear engagement, and excessive clearance in journal bearings. Therefore, before products leave the factory, they usually need to undergo uniformity testing to check the stability of product quality, quantify the deviation of each product in the same batch, thereby screening out abnormal products and ensuring the quality of products leaving the factory.

[0003] In real-world scenarios, product defects may occur inside the product or be minor. Manual identification alone is often insufficient for accurate and effective uniformity testing. Furthermore, uniformity testing is inefficient and requires significant manpower and resources, impacting product delivery efficiency.

[0004] Therefore, providing an automated product uniformity testing solution has become an urgent technical problem to be solved. Summary of the Invention

[0005] The main objective of this invention is to provide a product testing method, terminal, and computer-readable storage medium, aiming to solve the problems in the prior art where accurate and effective product uniformity testing is impossible, and where uniformity testing is inefficient and costly, thus affecting product delivery efficiency.

[0006] To achieve the above objectives, embodiments of the present invention provide a product testing method, the method comprising:

[0007] Control the motor of each product to be tested in the batch to rotate at a variable speed.

[0008] Acquire the vibration acceleration signal of each product to be tested in the batch to be tested;

[0009] Each product to be tested includes several motors and acceleration sensors; the vibration acceleration signal is collected by the acceleration sensors during the motor's variable speed rotation.

[0010] Based on the vibration acceleration signal, a sweep frequency characteristic curve of the product under test is generated;

[0011] Compare all the sweep frequency characteristic curves of the batch to be tested to determine the outliers of the batch to be tested;

[0012] Based on the outliers in the batch to be tested, determine whether there are any abnormal products in the products to be tested, so as to complete the uniformity test of the batch to be tested.

[0013] Optionally, based on the vibration acceleration signal, a swept frequency characteristic curve of the product under test is generated, specifically including:

[0014] Peak-valley identification is performed on the vibration acceleration signal to obtain the corresponding envelope curve; the envelope curve includes: peak envelope curve and valley envelope curve;

[0015] The peak envelope curve and the valley envelope curve are spliced ​​together to obtain the sweep frequency characteristic curve of the product under test.

[0016] Optionally, the peak envelope curve and the valley envelope curve are spliced ​​together to obtain the sweep frequency characteristic curve of the product under test, specifically including:

[0017] The peak envelope curve and the valley envelope curve are filtered by a preset multi-channel moving average filter to remove high-frequency noise from the peak envelope curve and the valley envelope curve.

[0018] The filtered peak envelope curve and valley envelope curve are resampled respectively to obtain the resampled peak envelope curve and the resampled valley envelope curve.

[0019] The resampled peak envelope curve and the resampled valley envelope curve are spliced ​​together to obtain the sweep frequency characteristic curve of the product under test.

[0020] Optionally, all sweep frequency characteristic curves of the batch to be tested are compared to determine outliers of the batch to be tested, specifically including:

[0021] From all the sweep frequency characteristic curves of the batch to be tested, discrete points at the same sampling time are obtained to form a corresponding data set, so as to obtain several data sets of the batch to be tested; wherein, the sampling time of each data set is different;

[0022] Calculate the geometric center point of the dataset based on the location information of each discrete point in each dataset;

[0023] The distance between each feature point and its corresponding geometric center point in each dataset is used as the outlier value;

[0024] If the outlier value is greater than the first preset threshold, the discrete point corresponding to the outlier value is determined as the outlier point in the dataset.

[0025] Outliers in each dataset are considered as outliers in the batch to be tested.

[0026] Optionally, before obtaining feature points at the same sampling time from all frequency sweep characteristic curves to form the corresponding dataset, the method further includes:

[0027] Using a preset dimensionality reduction method, the dimensionality of each sweep frequency characteristic curve is reduced to two dimensions.

[0028] Optionally, based on outliers in the batch to be tested, determine whether there are any abnormal products in the products to be tested, specifically including:

[0029] Identify the product to be tested that corresponds to the outlier in the batch to be tested;

[0030] Count the number of outliers corresponding to the outliers of the product to be tested;

[0031] If the number of outliers in the product to be tested exceeds the second preset threshold, the product to be tested is determined to be an abnormal product.

[0032] Optionally, before generating the sweep frequency characteristic curve of the product under test based on the vibration acceleration signal, the method further includes:

[0033] The vibration acceleration signal is filtered by a preset low-pass filter to remove the high-frequency components of the vibration acceleration signal.

[0034] Optionally, the variable speed rotational motion is:

[0035] Uniform rotational motion or non-uniform rotational motion;

[0036] Uniformly variable rotational motion includes: uniformly accelerated rotational motion and uniformly decelerated rotational motion.

[0037] To achieve the above objectives, embodiments of the present invention also provide a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement the steps in the product testing method described above.

[0038] To achieve the above objectives, embodiments of the present invention also provide a terminal, characterized in that it includes: a processor and a memory; the memory stores a computer-readable program that can be executed by the processor; when the processor executes the computer-readable program, it implements the steps of the product testing method as described above.

[0039] This invention, in its embodiments, controls the motor of each product in a batch to undergo variable-speed rotation. An accelerometer of each product collects its vibration acceleration signal during this rotation, thus obtaining the vibration acceleration signal for each product in the batch. A sweep frequency characteristic curve is then generated based on these signals. These curves are compared across the batch to identify outliers, thereby determining abnormal products and completing the uniformity testing of the batch. This method automates the uniformity testing of products before shipment, saving significant manpower and resources and improving production efficiency. Furthermore, by utilizing the product's own motor and accelerometer, the uniformity testing process fully leverages the product itself, eliminating the need for external devices and further conserving testing resources and improving efficiency. Attached Figure Description

[0040] Figure 1 A flowchart of a product testing method provided in an embodiment of the present invention;

[0041] Figure 2 A flowchart of step S104 provided in an embodiment of the present invention;

[0042] Figure 3 A flowchart of step S105 provided in an embodiment of the present invention;

[0043] Figure 4 A flowchart of step S106 provided in an embodiment of the present invention;

[0044] Figure 5 This is a schematic diagram of the terminal structure provided in an embodiment of the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0046] This invention provides a product testing method, such as... Figure 1 As shown, a product testing method may include at least the following steps:

[0047] S101 controls the motor in each product to be tested in the batch to rotate at a variable speed.

[0048] The variable-speed rotational motion can be either uniformly variable-speed rotational motion or non-uniformly variable-speed rotational motion. Furthermore, uniformly variable-speed rotational motion includes uniformly accelerating rotational motion and uniformly decelerating rotational motion. Additionally, the variable-speed rotational motion is identical for each product under test.

[0049] In this embodiment of the invention, the product to be tested is a product including a motor and an acceleration sensor. The product to be tested may contain multiple motors or a single motor; similarly, the acceleration sensor may be one or more, and no specific limitation is made in this embodiment. Furthermore, it is understood that other devices capable of collecting vibration acceleration signals (such as an inertial measurement unit) may also be present in the product to be tested, and these can also serve as the acceleration sensor in this embodiment.

[0050] Specifically, the controller controls the motor of each product under test to perform variable speed rotation. When a product under test has multiple motors, the variable speed rotation of each motor can be the same or different, and no specific limitation is made in this embodiment of the invention.

[0051] For example, product A to be tested has motor 1 and motor 2. The maximum rotational speed of motor 1 is set to ω1 and the maximum rotational speed of motor 2 is set to ω2. Within time T, the controller controls motor 1 to accelerate from 0 to the maximum rotational speed ω1 and motor 2 to accelerate from 0 to the maximum rotational speed ω2.

[0052] It is understandable that the batch to be tested can refer to the same type of product produced in the same batch of feed, on the same production line, and in the same shift.

[0053] S102, acquire the vibration acceleration signal of each product to be tested in the batch to be tested.

[0054] Specifically, the vibration acceleration signal of the product under test is collected by the acceleration sensor of the product under test during the variable speed rotation of the motor.

[0055] For example, product A to be tested has motor 1 and motor 2. The maximum rotational speed of motor 1 is set to ω1, and the maximum rotational speed of motor 2 is set to ω2. Within time T, the controller controls motor 1 to accelerate from 0 to its maximum rotational speed ω1, and motor 2 to accelerate from 0 to its maximum rotational speed ω2. Then, within time T, the accelerometer uses a sampling frequency F... s Sampling is performed to obtain the vibration acceleration signal of the product under test. This vibration acceleration signal is a time series signal Y: Y = {y i =y(i·dt)i=0,1,2,K,L-1}. Where, F s = 1 / dt, T = L·dt, where dt is the sampling interval, L is the number of samples, and T is the measurement time. Compared to last year, the time stamp corresponding to the vibration acceleration signal Y can be denoted as X = {xi =i·dti=0,1,2,K,L-1}.

[0056] S103 filters the vibration acceleration signal using a preset low-pass filter to remove high-frequency components of the vibration acceleration signal.

[0057] In this embodiment of the invention, the vibration acceleration signal of each product to be tested can be input into a preset low-pass filter. The vibration acceleration signal is filtered by the preset low-pass filter to remove high-frequency components from the vibration acceleration signal, that is, to remove high-frequency noise. This makes the product testing based on the vibration acceleration signal with high-frequency components removed more accurate and improves the accuracy of product uniformity testing.

[0058] S104 generates the sweep frequency characteristic curve of the product under test based on the vibration acceleration signal.

[0059] Among them, the sweep frequency characteristic curve is used to characterize the amplitude of the product under test at different frequencies.

[0060] In embodiments of the present invention, such as Figure 2 As shown, step S104 above can be achieved through at least the following steps:

[0061] S201 identifies the peak-valley values ​​of the vibration acceleration signal to obtain the corresponding envelope curve.

[0062] The envelope curves include: peak envelope curve and valley envelope curve.

[0063] Specifically, peak-valley identification can be performed using the following formula:

[0064]

[0065] Using the above formula, the upper and lower envelope information in the vibration acceleration signal can be extracted to obtain the peak envelope curve Y. p (i.e., the upper envelope curve of the vibration acceleration signal Y), the valley envelope curve Y v (i.e., the lower envelope curve of the vibration acceleration signal Y).

[0066] Y p ={(x p,k ,y p,k k = 0, 1, 2, K, L p -1},

[0067] Y v ={(x v,k ,y v,k k = 0, 1, 2, K, L v -1};

[0068] Among them, yp,k x represents the sequence of peak points in the vibration acceleration signal Y. p,k This corresponds to the x-axis of time, and the number of peaks is L. p y v,k This represents the sequence of valley points in the vibration acceleration signal Y, x v,k This corresponds to the x-axis of time, and the number of valleys is L. v .

[0069] S202 filters the peak envelope curve and valley envelope curve respectively through a preset multi-channel moving average filter to remove high-frequency noise from the peak envelope curve and valley envelope curve.

[0070] In this embodiment of the invention, after obtaining the envelope curve through step S201, the peak envelope curve and the valley envelope curve can be input into a preset multi-channel moving average filter to filter the peak envelope curve and the valley envelope curve respectively, so as to remove high-frequency noise and further improve the accuracy of product detection.

[0071] S203, resample the filtered peak envelope curve and valley envelope curve respectively to obtain the resampled peak envelope curve and resampled valley envelope curve.

[0072] Specifically, resampling can be performed using interpolation methods to make the data more uniform. This interpolation method can be linear interpolation, which can be performed using the following formula:

[0073]

[0074]

[0075] The resampling peak envelope curve can be obtained using the method described above. Resampling valley envelope curve in,

[0076]

[0077]

[0078] S204. The resampled peak envelope curve and the resampled valley envelope curve are spliced ​​together to obtain the sweep frequency characteristic curve of the product under test.

[0079] Will and The curves are stitched together to obtain the sweep frequency characteristic curve S(Y) of the product under test, where:

[0080]

[0081] S105. Compare all sweep frequency characteristic curves of the batch to be tested to determine the outliers of the batch to be tested.

[0082] like Figure 3 As shown, step S105 in this embodiment of the invention can be implemented through the following steps:

[0083] S301 uses a preset dimensionality reduction method to reduce the dimensionality of each sweep frequency characteristic curve to a preset dimension.

[0084] In this embodiment of the invention, the preset dimensionality reduction method can be the T-SNE algorithm, and the preset dimension can be two-dimensional. First, the T-SNE algorithm is used to reduce the dimensionality of the frequency sweep characteristic curve, thereby simplifying subsequent calculations based on the frequency sweep characteristic curve and saving computational resources.

[0085] In this embodiment of the invention, the distance function value between products to be detected can be calculated based on the sweep frequency characteristic curve of the products to be detected, and the sweep frequency characteristic curve can be reduced in dimensionality using the T-SNE algorithm based on the distance function value.

[0086] The method for calculating the distance function value is as follows:

[0087]

[0088] Wherein, F(S(Y1),S(Y2)) are the distance function values ​​of products 1 and 2 to be tested, S(Y1) is the sweep frequency characteristic curve of product 1 to be tested, and S(Y2) is the sweep frequency characteristic curve of product 2 to be tested.

[0089] The T-SNE algorithm is a commonly used dimensionality reduction method in the field of artificial intelligence. It belongs to unsupervised learning and can be implemented using the Python open-source library "sklearn.manifold.TSNE". This method is a classic method in the field, and its mathematical principles will not be elaborated here.

[0090] In this embodiment of the invention, the sweep frequency characteristic curve of each product to be tested in the batch to be tested can be used to form a dataset D.

[0091] Where, D={S(Y k Let |k = 0, 1, 2, K, N-1}, where N represents the number of products in the batch to be tested.

[0092] The T-SNE algorithm can be used to reduce the dimensionality of dataset D to two dimensions, resulting in dataset D2:

[0093] D2={S2(Y k )|(m k ,n k(k = 0, 1, 2, K, N-1)

[0094] Among them, (m k ,n k ) represents the two-dimensional coordinates after dimensionality reduction.

[0095] It should be noted that, in addition to the T-SNE algorithm mentioned above, other existing preset dimensionality reduction methods can also be used, and no specific limitation is made in this embodiment of the invention. Furthermore, the preset dimension can be two-dimensional or three-dimensional, depending on the actual situation, and no specific limitation is made in this embodiment of the invention.

[0096] S302, obtain discrete points at the same sampling time from all sweep frequency characteristic curves of the batch to be tested, and form a corresponding data set from the discrete points at the same sampling time to obtain several data sets of the batch to be tested.

[0097] The sampling times are different for each dataset.

[0098] It is understandable that the sweep frequency characteristic curve of step S302 is the same as the sweep frequency characteristic curve after dimensionality reduction in step S301.

[0099] As described above, each product in the batch to be tested has a corresponding discrete point at the same sampling time. Combining the discrete points at the same sampling time forms the data set corresponding to that sampling time. The sampling time can be obtained according to a preset sampling interval.

[0100] S303, calculate the geometric center point of the dataset based on the location information of each discrete point in the dataset.

[0101] In this embodiment of the invention, the discrete points in the dataset can be labeled on a coordinate system using a visualization tool based on their location information for visualization purposes. Based on the coordinates of each discrete point in the dataset, the outermost scattered outliers are identified, which are the abnormal products in this batch.

[0102] Specifically, the set center point can be calculated based on the coordinates (i.e., location information) of each discrete point in the dataset.

[0103] As can be seen from the above, when reducing the sweep frequency characteristic curve to two dimensions, the geometric center point (m) of D2 is calculated. * ,n * )for:

[0104]

[0105] S304 calculates the distance between each discrete point in the dataset and its corresponding geometric center point, and uses this distance as the outlier value.

[0106] When the frequency sweep characteristic curve is reduced to two dimensions as described above, the geometric center point (m) of D2 is calculated. * ,n * For example, in D2, the k-th sample (m) k ,n k Outlier Dist(m) k ,n k ) is defined as:

[0107] Dist(m k ,n k )=(m k -m * ) 2 -(n k -n * ) 2 ;k = 0, 1, 2, K, N-1.

[0108] S305, if the outlier value is greater than the first preset threshold, determine the discrete point corresponding to the outlier value as the outlier point of the data set.

[0109] The first preset threshold can be a fixed value or an adjustable value. For example, the outliers in each dataset can be sorted in ascending order, and discrete points with large outliers can be selected as outliers according to a preset ratio (e.g., 1% or 5%).

[0110] In this embodiment of the invention, outliers are determined by identifying whether a discrete point is an outlier based on its outlier value, thereby identifying outliers in each dataset. It is understood that a dataset may contain no outliers, one outlier, or multiple outliers; this is not limited in this embodiment of the invention.

[0111] S306, outliers in each dataset are considered as outliers in the batch to be tested.

[0112] By using step S305 above, the outliers of each dataset can be obtained. By combining the outliers of all datasets together, the outliers of the batch to be tested can be obtained.

[0113] S106, Based on the outliers of the batch to be tested, determine whether there are any abnormal products in the products to be tested, so as to complete the uniformity test of the batch to be tested.

[0114] like Figure 4 As shown, step S106 in this embodiment of the invention can be implemented by at least the following steps:

[0115] S401, Identify the product to be tested corresponding to the outlier in the batch to be tested.

[0116] Outliers in each dataset can be identified through steps S303-S304 described above. As mentioned above, each discrete point corresponds to a product to be tested, thus identifying the product to be tested corresponding to each outlier. The outliers in all datasets are then identified by statistically analyzing the products to be tested corresponding to each outlier.

[0117] S402, count the number of outliers corresponding to the outliers of the product to be tested.

[0118] Among them, the number of outliers is greater than or equal to 0.

[0119] In other words, the product to be tested may or may not have outliers; that is, not all products in the batch to be tested have outliers.

[0120] S403, if the number of outliers in the product to be tested is greater than the second preset threshold, the product to be tested shall be regarded as an abnormal product.

[0121] It is understandable that if the number of outliers in the product to be tested is less than the second preset threshold, the product to be tested is determined to be a normal product.

[0122] By following the steps S401-S403 above, the uniformity test of the batch to be tested can be completed, and the corresponding abnormal products can be identified.

[0123] The product testing method provided in this invention involves controlling the motor of each product in the batch to rotate at variable speed. An accelerometer of each product collects vibration acceleration signals during this rotation, obtaining the vibration acceleration signal for each product in the batch. A sweep frequency characteristic curve is then generated based on these signals. The sweep frequency characteristic curves of the entire batch are compared to identify outliers, thus determining abnormal products and completing the uniformity testing of the batch. This method automates the uniformity testing of products before shipment, saving significant manpower and resources and improving production efficiency. Furthermore, by utilizing the product's own motor and accelerometer, the uniformity testing can fully leverage the product itself, eliminating the need for external devices and further saving testing resources and improving efficiency.

[0124] Based on the above product testing method, this embodiment of the invention also provides a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement the steps in the product testing method described in the above embodiments.

[0125] Based on the above product testing method, the present invention also provides a terminal, such as... Figure 5 As shown, it includes at least one processor 50; a display screen 51; and a memory 52, and may also include a communications interface 53 and a bus 54. The processor 50, display screen 51, memory 52, and communications interface 53 can communicate with each other via the bus 54. The display screen 51 is configured to display a preset user guide interface in the initial setup mode. The communications interface 53 can transmit information. The processor 50 can invoke logical instructions in the memory 52 to execute the methods described in the above embodiments.

[0126] Furthermore, the logic instructions in the aforementioned memory 52 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0127] The memory 52, as a computer-readable storage medium, can be configured to store software programs, computer-executable programs, such as program instructions or modules corresponding to the methods in the embodiments of this disclosure. The processor 50 executes functional applications and data processing by running the software programs, instructions, or modules stored in the memory 52, thereby implementing the methods in the above embodiments.

[0128] The memory 52 may include a program storage area and a data storage area. The program storage area may store the operating system and application programs required for at least one function; the data storage area may store data created based on the use of the terminal. Furthermore, the memory 52 may include high-speed random access memory (RAM) and non-volatile memory. Examples include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks; these can also be transient storage media.

[0129] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the terminal and medium embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the description of the method embodiments.

[0130] The terminals and media provided in this application are one-to-one with the methods. Therefore, the terminals and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the terminals and media will not be repeated here.

[0131] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0132] Of course, those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be a memory, magnetic disk, optical disk, etc.

[0133] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A product testing method, characterized in that the method... include: Control the motor of each product to be tested in the batch to rotate at a variable speed. The variable speed rotational motion is: Uniform rotational motion or non-uniform rotational motion; Uniformly variable rotational motion includes: uniformly accelerated rotational motion and uniformly decelerated rotational motion; Acquire the vibration acceleration signal of each product to be tested in the batch to be tested; Each product to be tested includes several motors and acceleration sensors; the vibration acceleration signal is collected by the acceleration sensors during the motor's variable speed rotation. Based on the vibration acceleration signal, a swept frequency characteristic curve of the product under test is generated, specifically including: Peak-valley identification is performed on the vibration acceleration signal to obtain the corresponding envelope curve; the envelope curve includes: peak envelope curve and valley envelope curve; The peak envelope curve and the valley envelope curve are spliced ​​together to obtain the sweep frequency characteristic curve of the product under test; Compare all the sweep frequency characteristic curves of the batch to be tested to determine the outliers of the batch to be tested; Based on the outliers in the batch to be tested, determine whether there are any abnormal products in the products to be tested, so as to complete the uniformity test of the batch to be tested.

2. The method according to claim 1, characterized in that, The peak envelope curve and the valley envelope curve are spliced ​​together to obtain the frequency sweep characteristic curve of the product under test, which specifically includes: The peak envelope curve and the valley envelope curve are filtered by a preset multi-channel moving average filter to remove high-frequency noise from the peak envelope curve and the valley envelope curve. The filtered peak envelope curve and valley envelope curve are resampled respectively to obtain the resampled peak envelope curve and the resampled valley envelope curve. The resampled peak envelope curve and the resampled valley envelope curve are spliced ​​together to obtain the sweep frequency characteristic curve of the product under test.

3. The method according to claim 1, characterized in that, The sweep frequency characteristic curves of all batches to be tested are compared to identify outliers in the batch, specifically including: From all the sweep frequency characteristic curves of the batch to be tested, discrete points at the same sampling time are obtained to form a corresponding data set, so as to obtain several data sets of the batch to be tested; wherein, the sampling time of each data set is different; Calculate the geometric center point of the dataset based on the location information of each discrete point in each dataset; The distance between each feature point and its corresponding geometric center point in each dataset is used as the outlier value; If the outlier value is greater than the first preset threshold, the discrete point corresponding to the outlier value is determined as the outlier point in the dataset. Outliers in each dataset are considered as outliers in the batch to be tested.

4. The method according to claim 3, characterized in that, Before extracting feature points at the same sampling time from all frequency sweep characteristic curves to form the corresponding dataset, the method also includes: Using a preset dimensionality reduction method, the dimensionality of each sweep frequency characteristic curve is reduced to two dimensions.

5. The method according to claim 1, characterized in that, Based on outliers in the batch to be tested, determine whether there are any abnormal products in the product to be tested, specifically including: Identify the product to be tested that corresponds to the outlier in the batch to be tested; Count the number of outliers corresponding to the outliers of the product to be tested; If the number of outliers in the product to be tested exceeds the second preset threshold, the product to be tested is determined to be an abnormal product.

6. The method according to claim 1, characterized in that, Before generating the sweep frequency characteristic curve of the product under test based on the vibration acceleration signal, the method also includes: The vibration acceleration signal is filtered by a preset low-pass filter to remove the high-frequency components of the vibration acceleration signal.

7. A computer-readable storage medium, characterized in that, A computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the product testing method as claimed in any one of claims 1-6.

8. A terminal, characterized in that, include: Processor and memory; The memory stores a computer-readable program that can be executed by a processor; when the processor executes the computer-readable program, it implements the steps of the product testing method as claimed in any one of claims 1-6.

Images