Abnormal sound detection device

Abstract

The utility model relates to an abnormal sound detection device, which is arranged in a test box and comprises a tool clamp, a detection device and a control device, the tool clamp is installed on the bottom wall of the test box and used for clamping a to-be-tested product. An internal fixture; the internal fixing piece is mounted on the top wall of the test box and is arranged above the tool clamp; a pressing device; the pressing device is arranged on the internal fixing piece and is used for pressing a key of the product to be tested and enabling the product to be tested to generate air medium noise; the microphone is arranged on the internal fixing piece and is used for collecting audio data of the air medium noise and sending the audio data to an acoustic analyzer; the acoustic analyzer is electrically connected with the microphone; and the acoustic analyzer detects whether the to-be-detected product generates abnormal sound or not according to the audio data. According to the invention, manual detection is replaced by a mechanical device, the sound is accurately analyzed, and the efficiency and accuracy of abnormal sound detection are improved.

Description

Technical Field

[0001] This utility model relates to the field of abnormal noise detection technology, and in particular to an abnormal noise detection device. Background Technology

[0002] In recent years, with the increasing demands for the quality of keyboards and mice, the requirements for their manufacturing have become increasingly stringent. In addition to some parameters in the manufacturing process affecting the quality of the keyboard and mouse, whether the keyboard and mouse are assembled properly also affects their overall quality. Besides improper assembly causing abnormal noises, abnormal products will also produce sounds different from normal products when vibrating. Therefore, abnormal noise detection is a common detection method in the quality inspection of keyboards and mice.

[0003] Currently, most keyboard and mouse noise detection relies on manual inspection by experienced workers. This involves people vibrating the product to generate sound, listening to it, and then making a judgment. This method is not only labor-intensive and costly, but also inefficient, inaccurate, and prone to misjudgment. Utility Model Content

[0004] Therefore, the technical problem to be solved by this utility model is to overcome the problems of high cost, low efficiency and low accuracy of product abnormal noise detection relying on manual inspection in the prior art; in order to solve the above technical problems, this utility model provides an abnormal noise detection device to replace manual inspection and ensure detection efficiency and accuracy.

[0005] To achieve the above objectives, the following components are provided: a tooling fixture mounted on the bottom wall of the test chamber for holding the product under test; an internal fastener mounted on the top wall of the test chamber and positioned above the tooling fixture; a pressing device mounted on the internal fastener for pressing a button on the product under test to generate airborne noise; a microphone mounted on the internal fastener for collecting audio data of the airborne noise and sending the collected audio data to an acoustic analyzer; the acoustic analyzer being electrically connected to the microphone; and the acoustic analyzer detecting whether the product under test generates abnormal noise based on the received audio data.

[0006] In one embodiment of this utility model, the product under test includes a keyboard or a mouse.

[0007] In one embodiment of this utility model, the tooling fixture is provided with a groove for placing the product to be tested.

[0008] In one embodiment of the present invention, the internal fixing component includes a first fixing plate and a second fixing plate; one end of the first fixing plate is installed on the inner wall of the top of the test box, and the other end is connected to the second fixing plate; the second fixing plate is disposed above the tooling fixture.

[0009] In one embodiment of this utility model, the microphone is disposed on the second fixing plate.

[0010] In one embodiment of this utility model, the pressing device includes at least two pressing components; the pressing components include: a pusher; the pusher is installed on the outer wall of the test chamber; a guide cable; one end of the guide cable is disposed on the second fixed plate, and the other end is disposed on the pusher; a metal striker; the metal striker is disposed at one end of the guide cable on the second fixed plate, and is exactly above the button of the product under test.

[0011] In one embodiment of this utility model, the metal firing pin is provided with a rubber protective head.

[0012] In one embodiment of this utility model, the pushing member includes: an external fixing member; the external fixing member is disposed on the outer wall of the test box; a cylinder; the cylinder is disposed inside the external fixing member; and a pressure plate is disposed on the cylinder.

[0013] In one embodiment of the present invention, the external fixing member includes a third fixing plate, a fourth fixing plate, and a fifth fixing plate; the fifth fixing plate is disposed on the outer wall of the test box, the third fixing plate and the fourth fixing plate are mounted on the fifth fixing plate, and a cavity is formed between the third fixing plate, the fourth fixing plate, and the fifth fixing plate.

[0014] In one embodiment of the present invention, the other end of the guide cable passes through the third fixing plate and is located in the cavity; the other end of the guide cable is provided with a pressure block.

[0015] The above-mentioned technical solution of this utility model has the following advantages compared with the prior art:

[0016] The abnormal noise detection device described in this utility model uses a pressing device to replace manual operation to press the buttons on the product under test, causing the product to produce sound, thereby automating the detection of abnormal noise, saving costs and improving efficiency; and analyzes the collected sound using an acoustic analyzer to determine whether the sound is abnormal, thus judging the product's condition and improving the accuracy of abnormal noise detection; this application replaces manual inspection with mechanical inspection, improving the efficiency and quality of abnormal noise detection for keyboards and mice. Attached Figure Description

[0017] To make the content of this utility model easier to understand, the present utility model will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0018] Figure 1 This is a schematic diagram of the overall structure of the abnormal noise detection device according to a preferred embodiment of the present invention.

[0019] Explanation of reference numerals in the accompanying drawings: 101, First fixing plate; 102, Second fixing plate; 2, Guide cable; 3, Microphone; 4, Acoustic analyzer; 5, Metal striking pin; 6, Rubber protective head; 7, Product to be tested; 8, Tooling fixture; 9, Pressure plate; 10, Cylinder; 11, Pressure block; 121, Third fixing plate; 122, Fourth fixing plate; 123, Fifth fixing plate; 13, Test box. Detailed Implementation

[0020] To better explain and facilitate understanding of this invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0021] Reference Figure 1 As shown, in a preferred embodiment of this utility model, an abnormal noise detection device is provided, which is disposed inside a test chamber 13 and includes: a fixture 8, an internal fixing component, a pressing device, a microphone 3, and an acoustic analyzer 4; the fixture 8 is installed on the inner wall of the bottom of the test chamber 13 for clamping the product under test 7; the internal fixing component is installed on the inner wall of the top of the test chamber 13 and is disposed above the fixture 8; the pressing device is disposed on the internal fixing component for pressing the button of the product under test 7 and causing the product under test 7 to generate air medium noise; the microphone 3 is disposed on the internal fixing component for collecting audio data of air medium noise and sending the collected audio data to the acoustic analyzer 4; the acoustic analyzer 4 is electrically connected to the microphone 3; the acoustic analyzer 4 detects whether the product under test 7 generates abnormal noise based on the received audio data.

[0022] Specifically, to prevent the device from collecting other noises and causing misjudgment, this application sets the entire testing device inside the test chamber 13. The test chamber 13 is made of metal flat sound-absorbing material to block external sound, making the entire testing device in a relatively quiet environment. Inside the test chamber 13, a tooling fixture 8 is set at the bottom, and a groove is set in the middle of the tooling fixture 8 for installing the product under test 7. An internal fixing component is set above the tooling fixture 8 and is installed on the top of the test chamber 13 for installing a pressing device and a microphone 3. The pressing device is used to press the button on the product under test 7, so that the product under test 7 generates air medium noise. The microphone 3, located above the product under test 7, collects the audio data of the source air medium noise and sends the collected audio data to the acoustic analyzer 4 for analysis. The acoustic analyzer 4 analyzes whether the audio data is abnormal, thereby determining whether the product is abnormal.

[0023] In this embodiment, the acoustic analyzer 4 is located outside the test chamber 13 and electrically connected to the microphone 3; part of the pressing device is inside the test chamber 13 and the other part is outside the test chamber 13; the product under test 7 includes a keyboard or mouse.

[0024] The internal fixing components include a first fixing plate 101 and a second fixing plate 102; one end of the first fixing plate 101 is installed on the inner wall of the top of the test chamber 13, and the other end is connected to the second fixing plate 102; the second fixing plate 102 is positioned above the tooling fixture 8. The microphone 3 is mounted on the second fixing plate 102.

[0025] Specifically, the internal fixing components include a first fixing plate 101 and a second fixing plate 102; wherein, the second fixing plate 102 is in the horizontal direction and is used to install the part of the pressing device inside the test box 13 and the microphone 3. The second fixing plate 102 is set above the product under test 7, so that in the initial state, the pressing device is exactly above the button of the product under test 7, and the microphone 3 can collect the sound generated by the product under test 7 to the maximum extent; the first fixing plate 101 is in the vertical direction and is set on the inner wall of the top of the test box 13, and is used to connect the second fixing plate 102 and the test box 13.

[0026] The pressing device includes at least two pressing components.

[0027] Specifically, the pressing device includes multiple pressing components, each pressing component pressing one button, and the number of pressing components is appropriately set according to the number of buttons on the product under test 7.

[0028] In this embodiment, the fixture 8 holds the mouse to be tested. The mouse has two buttons, so two pressing components are provided. If the product to be tested 7 is a keyboard, the area of ​​the second fixing plate 102 can be made larger, and the pressing components can be set according to the number of buttons. Alternatively, after setting multiple pressing components, the position of the product to be tested 7 can be moved in the fixture 8 so that the pressing components can be pressed multiple times to complete the detection of the keyboard buttons.

[0029] The push-button assembly includes a pusher, a guide cable 2, and a metal striker 5. The pusher is mounted on the outer wall of the test chamber 13. One end of the guide cable 2 is mounted on the second fixed plate 102, and the other end is mounted on the pusher. The metal striker 5 is positioned at one end of the guide cable 2 on the second fixed plate 102, and is positioned directly above the button on the product under test 7. A rubber protective head 6 is provided on the metal striker 5. A pressure block 11 is provided at the other end of the guide cable 2.

[0030] Specifically, the push assembly includes a pusher, a guide cable 2, and a metal striker 5. The pusher is mounted on the outer wall of the test chamber 13 to prevent the sound generated by the push from affecting the test results. The guide cable 2 is L-shaped, with the longer end mounted on the pusher and the shorter end mounted on the second fixed plate 102. A pressure block 11 is mounted on one end of the pusher, and a metal striker 5 is mounted on one end of the second fixed plate 102, with the metal striker 5 positioned directly above the button on the product under test 7. When the pusher pushes the pressure block 11, the pressure block 11 moves the guide cable 2 toward the product under test 7, and the metal striker 5 on the guide cable 2 presses down to press the button on the product under test 7. When the pusher retracts, the pressure block 11 automatically rebounds, causing the guide cable 2 to retract, and the guide cable 2 causes the metal striker 5 to release the button. During the button press and rebound, the microphone 3 above the product under test 7 collects audio data and sends it to the acoustic analyzer 4. To protect the product under test 7, the bottom of the metal striker 5 is equipped with a protective rubber head 6. In this embodiment, the guide cable 2 is a steel guide cable.

[0031] The pushing component includes an external fixing member and a cylinder 10; the external fixing member is disposed on the outer wall of the test chamber 13; the cylinder 10 is disposed inside the external fixing member; a pressure plate 9 is disposed on the cylinder 10. The external fixing member includes a third fixing plate 121, a fourth fixing plate 122, and a fifth fixing plate 123; the fifth fixing plate 123 is disposed on the outer wall of the test chamber 13, the third fixing plate 121 and the fourth fixing plate 122 are mounted on the fifth fixing plate 123, and a cavity is formed between the third fixing plate 121, the fourth fixing plate 122, and the fifth fixing plate 123. The other end of the guide cable 2 passes through the third fixing plate 121 and is located within the cavity;

[0032] Specifically, the pushing component includes an external fixing component and a cylinder 10. The external fixing component is disposed on the outer wall of the test chamber 13, and a cavity is formed inside the external fixing component. The cylinder 10 is disposed inside the external fixing component. A pressure plate 9 is disposed on the cylinder 10, and the pressure plate 9 is used to push the pressure block 11 on the guide cable 2. The external fixing component includes a third fixing plate 121, a fourth fixing plate 122, and a fifth fixing plate 123. The fifth fixing plate 123 is disposed on the outer wall of the test chamber 13. The third fixing plate 121 and the fourth fixing plate 122 are sequentially installed on the fifth fixing plate 123, and a cavity is formed between the third fixing plate 121, the fourth fixing plate 122, and the fifth fixing plate 123. In this embodiment, the pressure plate 9 is a metal pressure plate.

[0033] In this embodiment, for ease of use, all pressing components share a common external fixing component, that is, all cylinders 10 are installed in the same external fixing component.

[0034] The guide cable 2 has a pressure block 11 at one end, which passes through the third fixing plate 121, so that the pressure block 11 is in the cavity. This design allows the guide cable 2 to be in a stable state when it is not pushed by the cylinder 10, and it will not shake, causing the metal firing pin 5 to accidentally hit the button. When the guide cable 2 is pushed by the cylinder 10, the pressure block 11 prevents the guide cable 2 from sliding out from the third fixing plate 121.

[0035] The entire workflow of the abnormal noise detection device described in this application is as follows:

[0036] Inside the test chamber 13, a fixture 8 is provided at the bottom to fix the product under test 7. Above the fixture 8, a steel guide cable with a metal striker 5 and a microphone 3 for collecting audio data of airborne noise are installed via a second fixing plate 102. The other end of the steel guide cable leads to the outside of the test chamber 13 and is fixed by a third fixing plate 121 in the external fixing components. A cylinder 10 is installed on a fourth fixing plate 122. The cylinder 10 extends and pushes a metal pressure plate, which presses the pressure block 11 on the guide cable 2, driving the steel guide cable to the metal striker 5, causing the metal striker 5 to extend and press down on the keyboard or mouse button of the product under test 7. When the cylinder 10 retracts, the pressure block 11 on the guide cable 2 automatically rebounds, driving the steel guide cable to retract the metal striker 5 and release the sample button. During the button pressing and rebound, the microphone 3 collects audio data and sends the collected audio data to the acoustic analyzer 4 for analysis. To protect the sample, a protective rubber head 6 is installed at the bottom of the metal striker 5. By repeating this process, repeated testing and analysis can be performed.

[0037] After the acoustic analyzer 4 collects audio data, an abnormal noise detection method is implemented using a type of support vector machine. This method includes:

[0038] Data collection and preprocessing: Based on human hearing judgment, the sound data collected by the device is labeled. In abnormal noise detection, there is a large amount of data for normal sounds and a small amount for abnormal sounds. Therefore, during training, we only use sound data without abnormal sounds, while also collecting a portion of sound data with abnormal sounds as validation set data.

[0039] Feature extraction: Features are extracted from the collected data above. Here we extract the features of the Mel spectrum.

[0040] Model Training: In One-Class Support Vector Machines (OCSVM), a suitable kernel function, such as a linear kernel or an RBF (Radial Basis Function) kernel, is selected based on the characteristics of the data. Appropriate parameters are also chosen, such as the (nu) parameter in OCSVM, which defines the upper limit of the proportion of data considered outliers in the model. During training, the OCSVM model is trained using a training set containing only normal data. Simultaneously, some samples without outliers and samples with outliers are separated as a validation set to evaluate the trained model. Precision, recall, and other metrics can be used to assess model performance. Based on the results, it is determined whether further model tuning is needed. After tuning, the model is saved.

[0041] Anomaly detection: The trained model is deployed to the software of the acoustic analyzer 4 and applied to the real-time acquired audio data from the microphone 3 to determine whether they are normal or abnormal (abnormal noise).

[0042] In feature extraction, Mel spectral features are a commonly used feature representation in speech processing and audio analysis, particularly in speech recognition and music information retrieval. Based on the characteristics of the human auditory system, it converts frequencies to a Mel scale, thus more closely resembling human perception of sound. The steps for extracting Mel spectral features include:

[0043] Preprocessing: This typically involves normalizing the audio signal and cutting the audio into shorter frames.

[0044] Fast Fourier Transform (FFT): Applying FFT to each audio frame transforms the signal from the time domain to the frequency domain to obtain the spectrum.

[0045] Mapping to Mel Scale: Map the obtained spectrum to the Mel scale. The relationship between Mel frequency and Hz frequency is as follows:

[0046]

[0047] Applying the Mel filter bank: Each filter is applied to the FFT result, so that each filter captures the energy within the frequency range it covers.

[0048] Take the logarithm: Take the logarithm of the energy output of each filter. This step is based on the fact that humans perceive sound in logarithmic ways.

[0049] One-Class Support Vector Machine (OCSVM) is a variant of Support Vector Machine (SVM) specifically designed for anomaly detection (also known as outlier detection). The principle of OCSVM differs from traditional SVM. In traditional SVM, the goal is to find a hyperplane to separate two distinct classes. In OCSVM, the goal is to find a decision boundary such that most data points (normal points) fall on one side of this boundary, while maximizing the distance between the boundary and the data. This boundary is designed to be as far away from the center of the data as possible, so that newly emerging points (outliers) not belonging to the training data distribution fall on the other side of the boundary. The collected audio data, after preprocessing and feature extraction, yields samples {(x1,y1),(x2,y2),…,(x...}. n ,y n )}, where x is the sample feature and y is the sample label (no abnormality, no unusual sound). Assuming the sample features have L dimensions, then the sample in R L If the space can be represented as a coordinate point, then all samples without abnormal sounds exist within a range O in the high-dimensional control. x If O x For a hypersphere with radius R and center a, the decision function is f(x) = [x(xa)·xa)]-R 2 Sample x i The error can be defined as:

[0050]

[0051] For the decision boundary O x To maximize the coverage of samples without unusual noises while minimizing the "size" of the decision boundary (reducing the volume of coverage as much as possible), solve the constrained optimization problem:

[0052]

[0053] st<(xa)·(xa)>≤R 2 +ξ

[0054] ξ>0

[0055] Introducing Lagrange multipliers {a1, a2, ..., a n Using the KT conditions, the optimization problem is transformed into:

[0056]

[0057]

[0058] Solving the above optimization problem, the Lagrange multipliers {a1, a2, …, a n} can be obtained, where 0 < a m <c is the support vector of the sample x m lying on the spherical surface. To make the decision boundary more compact, the kernel function k(x i , x j ) is introduced. The use of the kernel function enables the original data (in the R L space) to be mapped to a higher-dimensional feature space (in the R S space, where S > L). In this higher-dimensional space, the distribution of the data may form a more compact sphere, which means that the normal data points are closer to each other in this space. The decision function is obtained as follows:

[0059]

[0060] In the formula, Those falling within the decision function region belong to one class (no abnormal sound), and those outside the region belong to another class (abnormal sound). The unknown sound data can be predicted and labeled using f(x).

[0061] In this embodiment, the size of the tooling fixture 8 is 300mm * 300 * 15mm, the length of the guide cable 2 is 1000mm, the length of the metal striker 5 is 50mm and the diameter is 3mm, and the size of the metal pressing plate is 100 * 50 * 10mm; the overall material is aluminum; the surface of the device is treated by sandblasting and then anodic oxidation. During blanking production and assembly, the components are processed with precision machining equipment according to the design drawings, and assembled after rechecking and confirming that the processing errors of each component meet the design requirements.

[0062] This application replaces manual inspection with an automated device and conducts precise analysis on the sound, thereby realizing the mechanized abnormal sound detection of keyboards and mice; this device can effectively solve the dilemma that the abnormal sound inside the keyboard and mouse can only be detected by manual shaking and listening, fill the blank of the abnormal sound detection inside the keyboard and mouse, and also provide noise and vibration data support for the development and design of new products of keyboards and mice.

[0063] In the description of this utility model, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this utility model, "a plurality of" means two or more, unless otherwise explicitly specified.

[0064] In this utility model, unless otherwise explicitly specified and limited, the terms "installation," "connection," "joining," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in this utility model can be understood according to the specific circumstances.

[0065] In this utility model, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "beneath" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0066] In the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0067] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make modifications, alterations, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. An abnormal noise detection device, installed inside a test chamber, characterized in that, The abnormal noise detection device includes: Tooling fixture; the tooling fixture is installed on the bottom wall of the test chamber and is used to hold the product to be tested; Internal fasteners; the internal fasteners are installed on the top wall of the test chamber and positioned above the tooling fixture; A pressing device; the pressing device is disposed on the internal fixing component and is used to press the button of the product under test and cause the product under test to generate air medium noise; A microphone, mounted on the internal fixing component, is used to collect audio data of airborne noise and send the collected audio data to an acoustic analyzer; the acoustic analyzer is electrically connected to the microphone; the acoustic analyzer detects whether the product under test produces abnormal noise based on the received audio data.

2. The abnormal noise detection device according to claim 1, characterized in that: The product under test includes a keyboard or mouse.

3. The abnormal noise detection device according to claim 1, characterized in that: The tooling fixture is provided with a groove for placing the product to be tested.

4. The abnormal noise detection device according to claim 1, characterized in that: The internal fasteners include a first fixing plate and a second fixing plate; One end of the first fixing plate is installed on the inner wall of the top of the test box, and the other end is connected to the second fixing plate; the second fixing plate is disposed above the tooling fixture.

5. The abnormal noise detection device according to claim 4, characterized in that: The microphone is mounted on the second mounting plate.

6. The abnormal noise detection device according to claim 5, characterized in that: The pressing device includes at least two pressing components; The pressing component includes: A pushing component; the pushing component is mounted on the outer wall of the test chamber; Guide cable; one end of the guide cable is disposed on the second fixing plate, and the other end is disposed on the pusher; Metal striker; the metal striker is disposed at one end of the guide cable located on the second fixing plate, and is positioned exactly above the button of the product under test.

7. The abnormal noise detection device according to claim 6, characterized in that: The metal firing pin is equipped with a rubber protective head.

8. The abnormal noise detection device according to claim 6, characterized in that: The pushing component includes: External fasteners; the external fasteners are disposed on the outer wall of the test chamber; A cylinder; the cylinder is disposed within the external fixing component; a pressure plate is disposed on the cylinder.

9. The abnormal noise detection device according to claim 8, characterized in that: The external fasteners include a third fastening plate, a fourth fastening plate, and a fifth fastening plate; The fifth fixing plate is disposed on the outer wall of the test box, and the third fixing plate and the fourth fixing plate are mounted on the fifth fixing plate, with a cavity formed between the third fixing plate, the fourth fixing plate and the fifth fixing plate.

10. The abnormal noise detection device according to claim 9, characterized in that: The other end of the guide cable passes through the third fixing plate and is located in the cavity; the other end of the guide cable is provided with a pressure block.

Images