Voiceprint sensor based on fiber network structure, preparation method and application
By using a fiber network-based acoustic fingerprint sensor, a three-dimensional composite fiber network is formed by silk fabric and MXene powder, combined with a polydopamine interface layer. This solves the problems of noise interference and insufficient generalization ability of transformer acoustic fingerprint sensors, and achieves high signal-to-noise ratio and stable acoustic fingerprint feature extraction, which is suitable for applications in multiple fields.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-07
AI Technical Summary
Existing transformer acoustic fingerprint sensors face problems in practical applications, such as significant environmental noise interference, insufficient generalization ability of fault feature extraction, lack of standardized datasets and lightweight diagnostic models, resulting in insufficient reliability.
A fiber network-based acoustic sensor was developed. A three-dimensional composite fiber network was formed using silk fabric and MXene powder, combined with a polydopamine interface layer. The fiber network structure sensing layer was prepared by ultrasonic treatment and centrifugation, and then encapsulated with copper foil and PE tape to construct a stable acoustic sensor.
It achieves high signal-to-noise ratio and stable voiceprint feature extraction, possesses intrinsic noise suppression capability, improves the generalization capability of fault feature extraction, provides clear electromechanical coupling relationship, lays the foundation for lightweight diagnostic models, and is suitable for applications in multiple fields.
Smart Images
Figure CN121815183A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to an acoustic fingerprint sensor based on a fiber network structure, its fabrication method, and its application, belonging to the field of dynamic pressure sensing technology. Background Technology
[0002] Acoustic sensors designed based on transformer vibration characteristics can monitor and analyze acoustic signals during equipment operation in real time, accurately identify mechanical faults and abnormal states, and provide key information for intelligent operation and maintenance of power equipment. Currently, research on transformer acoustic sensors mainly focuses on three directions: first, improving the accuracy of acoustic signal acquisition; second, optimizing fault feature extraction methods in noisy environments; and third, developing intelligent diagnostic algorithms based on deep learning. The ultimate goal is to achieve non-intrusive real-time monitoring of transformer mechanical conditions and early fault warning.
[0003] However, existing technologies still face significant bottlenecks in practical applications: environmental noise significantly interferes with signals, the generalization ability of fault feature extraction is insufficient, there is a lack of standardized voiceprint datasets in the industry, and there is a lack of lightweight diagnostic models with strong interpretability. These problems collectively restrict the reliability of sensors in engineering applications. Therefore, there is an urgent need to develop a voiceprint sensor with a simple fabrication process, high signal-to-noise ratio, and stable performance. Summary of the Invention
[0004] To address the problems existing in the prior art, this invention provides an acoustic fingerprint sensor based on a fiber network structure, its fabrication method, and its application, which features a simple fabrication process, high signal-to-noise ratio, and stable performance.
[0005] To achieve the above objectives, the present invention employs a fiber network structure-based acoustic fingerprint sensor, comprising a fiber network structure sensing layer and two copper foils.
[0006] The fiber network structure sensing layer includes silk fabric, MXene powder and dopamine hydrochloride. The silk fabric, which has undergone support and degumming treatment, is immersed in a mixed solution of MXene powder and dopamine hydrochloride. After ultrasonic treatment, it is centrifuged to obtain a fiber network structure sensing layer with a three-dimensional composite fiber network.
[0007] Two copper foils are respectively attached to the upper and lower surfaces of the fiber network structure sensing layer, and each copper foil is connected to a wire, thus obtaining a voiceprint sensor based on the fiber network structure.
[0008] As an improvement, two PE tapes are also included. The two PE tapes cover the outer sides of the copper foil on the upper and lower surfaces of the fiber network structure sensing layer, respectively, and are fixed to each other by the edges of the two PE tapes, thereby encapsulating the fiber network structure sensing layer and the two copper foils.
[0009] A second aspect of the present invention also provides a method for fabricating the aforementioned acoustic signature sensor based on a fiber network structure, comprising the following steps:
[0010] S1. Fabrication of fiber network structure sensing layer
[0011] S1-1. Place the silkworm on a flat plate and spin for a preset time to obtain flat silk. Cut the flat silk to a preset size and fix it in a clamp. Put it in hot water to boil to remove sericin. Then let it air dry at room temperature to obtain degummed silk fabric SFSC.
[0012] S1-2. Mix MXene powder, dopamine hydrochloride and deionized water to form a mixed solution. Immerse SFSC in the mixed solution and sonicate it. Then remove unreacted monomers by centrifugation to obtain a fiber network structure sensing layer.
[0013] S2. Two copper foils are respectively attached to the upper and lower surfaces of the fiber network structure sensing layer, and each copper foil is connected to a wire. The fiber network structure sensing layer and the copper foils are then bonded together with PE tape to obtain the voiceprint sensor based on the fiber network structure.
[0014] As an improvement, in step S1-1, the hot water temperature is 80-100℃ and the boiling time is 80-100 min.
[0015] As an improvement, in steps S1-2, the ratio of MXene powder, dopamine hydrochloride, and deionized water is 50 mg: (50-150) mg: (190-210) mL.
[0016] As an improvement, in steps S1-2, the ultrasonic treatment temperature is 25-35 ℃, the time is 5-7 h, the centrifugation speed is 400-600 rpm, and the time is 8-12 min.
[0017] As an improvement, step S1 specifically includes the following steps:
[0018] S1-1. Place 60 silkworms in a 20×20 cm 20×20 cm 2 On a flat plate, spinning was carried out for 60 hours to obtain flat yarn; the flat yarn was then cut into 3×3 cm pieces. 2 The silk fabric (SFSC) was fixed in the holder and boiled in 90℃ hot water for 90 minutes to remove the sericin. It was then air-dried at room temperature to obtain the degummed silk fabric.
[0019] S1-2. Mix 50 mg of MXene powder, 100 mg of dopamine hydrochloride and 200 mL of deionized water to form a mixed solution. Immerse SFSC in the mixed solution and sonicate at 30°C for 6 h. Then centrifuge at 500 rpm for 10 min to remove unreacted monomers and obtain the fiber network structure sensing layer.
[0020] A third aspect of the present invention also provides the aforementioned fiber network-based acoustic fingerprint sensor, or the fiber network-based acoustic fingerprint sensor prepared by the aforementioned method, in at least one of the following applications:
[0021] (1) Application in power equipment monitoring equipment;
[0022] (2) Application in medical equipment;
[0023] (3) Application in biological monitoring equipment;
[0024] (4) Applications in smart home devices;
[0025] (5) Application in security equipment.
[0026] The reaction mechanism of this invention:
[0027] In-situ polymerization and interfacial bridging of polydopamine (PDA) are the core of the formation of the three-dimensional composite fiber network: dopamine hydrochloride dissolves in water and undergoes oxidative self-polymerization under ultrasonic conditions to generate PDA; PDA, with its strong adhesion, rapidly and uniformly coats the surface of the support-degummed flat filament (SFSC); the functional groups in the PDA molecule firmly capture and anchor MXene nanosheets in the solution through intermolecular forces such as coordination bonds and hydrogen bonds; finally, a stable MXene conductive layer is formed on the SFSC fiber network, constructing a three-in-one core-shell conductive network of SFSC-PDA-MXene. Among them, SFSC, as a flexible three-dimensional porous support substrate, provides a loose and porous structural basis for the network; dopamine hydrochloride (converted to PDA) acts as a molecular glue to achieve a tight connection between the substrate and the active material; MXene, as the core conductive medium, ensures the electrical signal response.
[0028] Compared with the prior art, the acoustic fingerprint sensor based on the fiber network structure of the present invention has the following advantages:
[0029] 1. Intrinsic Noise Suppression, Overcoming Signal Interference Challenges. This invention abandons traditional dense or artificially designed sensitive structures, using a natural three-dimensional fiber network (SFSC) formed by the support and degumming process of silkworm silk as a substrate. Its spontaneously spun, randomly multi-scale porous and fluffy structure functions as both an "acoustic sponge" and a physical filter. This structure can resonate directionally with target acoustic signatures (such as specific frequencies of equipment malfunctions or physiological acoustic signatures) and amplify energy, while simultaneously scattering and attenuating environmental interference such as high-frequency random noise. Combined with the multi-element conductive path formed by uniformly attached MXene, it further focuses the target signal and reduces interference transmission, endowing the sensor with intrinsic noise suppression capabilities from a structural perspective, ensuring clear extraction of acoustic signature features even in noisy environments.
[0030] 2. High stability and strong consistency, overcoming generalization and drift bottlenecks. Addressing the issues of insufficient generalization ability in fault feature extraction and signal drift caused by the easy detachment of nanomaterials in existing technologies, this invention introduces a polydopamine (PDA) interface layer and employs a one-step ultrasonic in-situ modification process. PDA acts as a "molecular glue," firmly and uniformly anchoring MXene nanosheets to the surface of each silk fiber through coordination bonds and hydrogen bonds, forming a stable SFSC-PDA-MXene core-shell structure, completely solving the core problem of material detachment during dynamic vibration. This design not only ensures the consistency of response and structural durability of the sensor during long-term use but also improves the generalization ability of fault feature extraction, achieving stable detection without relying on standardized acoustic signature datasets.
[0031] 3. The sensing mechanism is clear and interpretable, laying the foundation for lightweight diagnostics. This invention constructs a simplified sensing mechanism of "flexible fiber skeleton - continuous MXene conductive path". When sound waves induce minute deformations in the SFSC three-dimensional network, they directly change the contact state of the MXene conductive path, which is then converted into a sensitive and continuous resistance change signal. This direct electromechanical coupling relationship establishes a clear correspondence between electrical signal characteristics (such as response amplitude, waveform timing, and frequency band distribution) and the physical properties of sound wave vibration. This provides hardware support for the construction of lightweight diagnostic models, overcomes the dilemma of complex and poor interpretability of existing diagnostic models, and contributes to transparent decision-making and engineering applications.
[0032] 4. Simple process and excellent performance, adaptable to practical needs in multiple fields. The sensor fabrication process is simple and efficient, requiring no complex equipment for steps such as degumming, ultrasonic bonding, assembly, and packaging. It is convenient to operate and easy to scale up production. The porous structure of SFSC provides ample deformation space for acoustic wave response, while the high conductivity of MXene ensures signal transmission efficiency. Together, they enable the sensor to have a high signal-to-noise ratio, wide bandwidth response, and excellent cycle stability, allowing for long-term reliable operation. With the above comprehensive performance, the voiceprint sensor of this invention can be flexibly adapted to the needs of multiple fields such as power equipment monitoring, medical and biological monitoring, smart homes, and security. It can accurately identify equipment fault voiceprints, stably capture weak physiological voiceprints, and clearly identify voice commands and safety-related voiceprints, making it highly practical and with broad application prospects. Attached Figure Description
[0033] Figure 1 This is an overall structural diagram of the acoustic signature sensor of the present invention;
[0034] Figure 2 This is a flowchart illustrating the fabrication process of the acoustic signature sensor of the present invention.
[0035] Figure 3 This is a sensitivity test diagram of the voiceprint sensor in Embodiment 2 of the present invention;
[0036] Figure 4 This is a current change response diagram of the acoustic fingerprint sensor in Embodiment 2 of the present invention at different frequencies;
[0037] Figure 5 This is a current change response diagram of the acoustic fingerprint sensor in Embodiment 2 of the present invention under different sound pressure levels;
[0038] Figure 6 The results of the repeatability test of the acoustic signature sensor in Embodiment 2 of the present invention;
[0039] In the diagram: 1. PE tape, 2. Copper foil, 3. Fiber network structure sensing layer. Detailed Implementation
[0040] The following embodiments are further illustrations of the present invention and serve as explanations of the technical content of the present invention. However, the essence of the present invention is not limited to the embodiments described below. Those skilled in the art can and should know that any simple changes or substitutions based on the spirit of the present invention should fall within the protection scope claimed by the present invention.
[0041] Example 1
[0042] Combination Figure 1 , Figure 2 As shown, a method for fabricating an acoustic fingerprint sensor based on a fiber network structure includes the following steps:
[0043] S1. Fabrication of fiber network structure sensing layer
[0044] S1-1. Place 60 silkworms in a 20×20 cm 20×20 cm 2 On a flat plate, spinning was carried out for 60 hours to obtain flat yarn; the flat yarn was then cut into 3×3 cm pieces. 2 The silk fabric (SFSC) was fixed in the holder and boiled in 90℃ hot water for 90 minutes to remove the sericin. It was then air-dried at room temperature to obtain the degummed silk fabric.
[0045] S1-2. Mix 50 mg of MXene powder, 50 mg of dopamine hydrochloride and 200 mL of deionized water to form a mixed solution. Immerse SFSC in the mixed solution and sonicate at 30°C for 6 h. Then centrifuge at 500 rpm for 10 min to remove unreacted monomers and finally obtain a fiber network structure sensing layer made of SFSC, MXene and dopamine hydrochloride.
[0046] S2. Two copper foils 2 are tightly adhered to one side of two PE tapes 1 (the side that contacts the fiber network structure sensing layer 3), and one end of each copper foil 2 extends out of the corresponding edge of the PE tape 1 for connecting external wires; the fiber network structure sensing layer 3 is placed between the two copper foils 2, so that the copper foils 2 are in close contact with the upper and lower surfaces of the fiber network structure sensing layer 3 respectively; the edges of the upper and lower PE tapes are aligned and glued together, and the fiber network structure sensing layer 3 and copper foils 2 are wrapped by the sealing effect of the PE tapes 1 to complete the encapsulation of the voiceprint sensor.
[0047] Example 2
[0048] A method for fabricating an acoustic signature sensor based on a fiber network structure includes the following steps:
[0049] Except for adjusting the amount of "dopamine hydrochloride" in steps S1-2 to 100 mg, the other steps are completely consistent with Example 1, and finally a voiceprint sensor based on fiber network structure is obtained.
[0050] The acoustic fingerprint sensor based on the fiber network structure prepared in Example 2 was subjected to sensitivity tests, current change response tests at different frequencies, current change response tests at different sound pressure levels, and stability tests. The specific test procedures and results are as follows:
[0051] Depend on Figure 3 It can be seen that the detection sensitivity of this sensor can reach 0.68491 Hz in the frequency range of 25~2780 Hz. -1This stems from its unique three-dimensional composite fiber network structure: the fluffy, multi-scale network composed of natural silk fibers also functions as an efficient "acoustic resonant cavity," which can efficiently capture and convert sound wave vibrations of different frequencies into significant deformation. Furthermore, MXene nanosheets, firmly anchored to the fiber surface by polydopamine, can sensitively convert this deformation into a strong resistance signal, thereby overcoming the limitations of traditional sensors that have narrow bandwidth and difficulty in handling both high and low frequencies, and achieving wideband, high-sensitivity detection from low to mid-high frequencies.
[0052] The sensitivity and stability of the sensor were verified by repeatedly applying sound stimuli of different frequencies and sound pressure levels. Figure 4 , Figure 5 As shown, the sensor exhibits good dynamic stability under the aforementioned different stimulus conditions; and its cyclic stability was verified by performing 12,000 cycles of testing. The results are as follows. Figure 6 As shown, its conductive network integrity and signal response amplitude show no significant attenuation, and its cycling stability is excellent. This is mainly because the polydopamine interface layer acts like a "molecular glue," tightly integrating MXene and silk fibers into a stable "three-in-one" structure. This ensures the integrity and recoverability of the conductive network during repeated deformation, avoiding material detachment or performance degradation. It fundamentally solves the common problems of easy detachment of nanomaterials, unstable interfaces leading to signal drift and poor durability in conventional sensors, providing a cycle life and dynamic response stability far exceeding those of conventional sensors. This lays a solid hardware foundation for building a reliable voiceprint recognition system.
[0053] Example 3
[0054] A method for fabricating an acoustic signature sensor based on a fiber network structure includes the following steps:
[0055] Except for adjusting the amount of "dopamine hydrochloride" in steps S1-2 to 150 mg, the other steps are completely consistent with Example 1, and finally a voiceprint sensor based on fiber network structure is obtained.
[0056] The amount of dopamine hydrochloride directly determines the thickness and integrity of the polydopamine (PDA) interlayer. Insufficient dosage results in a thin and discontinuous PDA coating, failing to provide sufficient anchoring points for MXene nanosheets. This leads to insufficient anchoring and uneven distribution of MXene on the silk fiber surface, making it difficult to construct a complete conductive network. Consequently, the sensor's response sensitivity and stability are affected. Moderate dosage forms a uniform, dense, and strongly adhesive PDA interface layer. This layer guides the MXene nanosheets to uniformly coat each silk fiber, ultimately constructing a stable and continuous fiber-polymer-nanosheet three-dimensional conductive core-shell structure, ensuring excellent sensor performance. Excessive dosage results in an overly thick PDA layer formed by excessive dopamine hydrochloride polymerization. This can clog the pores of the silk fiber network and cause adhesion between fibers, significantly reducing the specific surface area and porosity of the structure. Consequently, this weakens the sensor's ability to capture acoustic vibrations and respond to deformation, affecting its detection performance.
[0057] The acoustic fingerprint sensor based on a fiber network structure prepared in this invention (Examples 1-3) can, on the one hand, efficiently absorb chaotic interference sound waves in the environment (such as power grid background noise and environmental noise) and directionally focus target acoustic fingerprint signals (such as transformer vibration sound and human physiological voiceprint) by optimizing the porous structure formed after degumming of silk fabric (SFSC) and the layered attachment state of MXene and dopamine hydrochloride in the fiber network. On the other hand, the construction of multiple conductive paths further reduces signal transmission loss, ultimately significantly improving the signal-to-noise ratio and signal fidelity of the sensor, ensuring that the target acoustic fingerprint information can still be accurately captured and restored in complex and noisy environments.
[0058] Based on the aforementioned superior performance, this sensor shows broad application prospects in multiple fields:
[0059] In the field of smart healthcare, it can be made into wearable patches that can accurately monitor and identify laryngeal muscle vibrations, enabling silent speech recognition, sleep apnea syndrome diagnosis, and analysis of early voice changes caused by specific neurological diseases such as Parkinson's disease, providing auxiliary support for clinical diagnosis.
[0060] In the field of industrial safety, it can be integrated into the equipment housing or pipe surface to monitor abnormal sounds of key components such as mechanical bearings and gearboxes in real time, accurately capture early signs of failure, realize early prediction and intelligent diagnosis of equipment failure, and reduce operation and maintenance costs.
[0061] In the fields of human-computer interaction and information security, it can collect and identify the unique voiceprint features of users with high fidelity, providing a new and secure biometric identification unlocking and command input method for devices such as smartphones and smart homes; its flexible properties can also be embedded in the skin of bionic robots to enhance the robot's ability to perceive environmental sounds and expand human-computer interaction scenarios.
[0062] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions or improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. An acoustic signature sensor based on a fiber network structure, characterized in that, It includes a fiber network structure sensing layer and two copper foils; The fiber network structure sensing layer includes silk fabric, MXene powder and dopamine hydrochloride. The silk fabric, which has undergone support and degumming treatment, is immersed in a mixed solution of MXene powder and dopamine hydrochloride. After ultrasonic treatment, it is centrifuged to obtain a fiber network structure sensing layer with a three-dimensional composite fiber network. Two copper foils are respectively attached to the upper and lower surfaces of the fiber network structure sensing layer, and each copper foil is connected to a wire, thus obtaining a voiceprint sensor based on the fiber network structure.
2. The acoustic signature sensor based on a fiber network structure according to claim 1, characterized in that, It also includes two PE tapes, which are respectively applied to the outer sides of the copper foil on the upper and lower surfaces of the fiber network structure sensing layer. The two PE tapes are then bonded together by their edges to encapsulate the fiber network structure sensing layer and the two copper foils.
3. A method for fabricating an acoustic signature sensor based on a fiber network structure as described in any one of claims 1-2, characterized in that, Includes the following steps: S1. Fabrication of fiber network structure sensing layer S1-1. Place the silkworm on a flat plate and spin for a preset time to obtain flat silk. Cut the flat silk to a preset size and fix it in a clamp. Put it in hot water to boil to remove sericin. Then let it air dry at room temperature to obtain degummed silk fabric SFSC. S1-2. Mix MXene powder, dopamine hydrochloride and deionized water to form a mixed solution. Immerse SFSC in the mixed solution and sonicate it. Then remove unreacted monomers by centrifugation to obtain a fiber network structure sensing layer. S2. Two copper foils are respectively attached to the upper and lower surfaces of the fiber network structure sensing layer, and each copper foil is connected to a wire. The fiber network structure sensing layer and the copper foils are then bonded together with PE tape to obtain the voiceprint sensor based on the fiber network structure.
4. The method for fabricating an acoustic signature sensor based on a fiber network structure according to claim 3, characterized in that, In step S1-1, the hot water temperature is 80-100℃ and the boiling time is 80-100 min.
5. The method for fabricating an acoustic signature sensor based on a fiber network structure according to claim 3, characterized in that, In steps S1-2, the ratio of MXene powder, dopamine hydrochloride, and deionized water is 50 mg: (50-150) mg: (190-210) mL.
6. The method for fabricating an acoustic signature sensor based on a fiber network structure according to claim 3, characterized in that, In steps S1-2, the ultrasonic treatment temperature is 25-35 ℃, and the time is 5-7 h; the centrifugation speed is 400-600 rpm, and the time is 8-12 min.
7. The method for fabricating an acoustic signature sensor based on a fiber network structure according to claim 3, characterized in that, Step S1 specifically includes the following steps: S1-1. Place 60 silkworms in a 20×20 cm 20×20 cm 2 On a flat plate, spinning was carried out for 60 hours to obtain flat yarn; the flat yarn was then cut into 3×3 cm pieces. 2 The silk fabric (SFSC) was fixed in the holder and boiled in 90℃ hot water for 90 minutes to remove the sericin. It was then air-dried at room temperature to obtain the degummed silk fabric. S1-2. Mix 50 mg of MXene powder, 100 mg of dopamine hydrochloride and 200 mL of deionized water to form a mixed solution. Immerse SFSC in the mixed solution and sonicate at 30°C for 6 h. Then centrifuge at 500 rpm for 10 min to remove unreacted monomers and obtain a fiber network structure sensing layer.
8. The acoustic fingerprint sensor based on a fiber network structure as described in claim 1, or the acoustic fingerprint sensor based on a fiber network structure prepared by the preparation method according to any one of claims 2-7, in at least one of the following applications: (1) Application in power equipment monitoring equipment; (2) Application in medical equipment; (3) Application in biological monitoring equipment; (4) Applications in smart home devices; (5) Application in security equipment.