A distributed frequency array abrasive grain identification sensor

Abstract

This invention proposes a distributed frequency array abrasive particle identification sensor, relating to the fields of tribology and online oil monitoring of marine power units. It solves the technical problems of existing abrasive particle sensors, such as parameter coupling and identification blind spots in single-frequency detection, and intermodulation interference and signal overlap in multi-frequency detection, which prevent accurate differentiation of dissimilar metal abrasive particle types. This invention achieves the acquisition of multi-dimensional features of abrasive particles through physical spatial arrangement, thereby endowing the sensor with the ability to identify particles of various materials. This invention eliminates multi-frequency intermodulation interference in hardware, can acquire multi-dimensional complex electromagnetic fingerprints of abrasive particles, accurately distinguish dissimilar metal abrasive particles, solve the problems of blind spots and signal overlap in single-frequency identification, eliminates the need for complex frequency sweeping circuits, and significantly improves detection accuracy, anti-interference, and system robustness.

Description

Technical Field

[0001] This invention relates to the field of tribology and online oil monitoring technology for marine power plants, and in particular to a distributed frequency array abrasive particle identification sensor. Background Technology

[0002] In the health management systems of modern large-scale machinery such as ship main engines and gas turbines, the abrasive information carried by lubricating oil is the core basis for assessing the wear condition of friction pairs. Identifying the material type of abrasive particles (such as distinguishing between ferromagnetic iron powder and non-ferromagnetic copper and aluminum powder) plays an irreplaceable role in the accurate location of fault sources. However, existing electromagnetic sensors face physical limitations in material identification.

[0003] First, most traditional inductive sensors employ a single-frequency excitation mode. Their output signal, in essence, represents only the amplitude of the induced voltage or the change in coil inductance at a single frequency. This scalar-value-based analysis method suffers from severe parameter coupling problems. Since the signal generated by abrasive particles is a nonlinear composite result of their geometric volume, conductivity, and permeability at a specific frequency, a single-frequency signal cannot effectively distinguish between the size and physical properties of the abrasive particles. For example, at a single frequency, the flux cancellation effect generated by a large-diameter non-ferromagnetic copper particle may overlap in amplitude with the flux enhancement effect generated by a small-sized ferromagnetic particle, leading to a blind zone in sensor recognition. Furthermore, traditional methods often focus on measuring the amplitude of the induced voltage, neglecting the resistance changes and phase shift information excited by the particles under an alternating magnetic field. This prevents the sensor from using phase polarity to distinguish between magnetization and eddy current effects with opposite polarities, thus hindering the physical classification and identification of dissimilar metals such as iron, copper, and stainless steel.

[0004] Secondly, existing technologies often face the challenge of intermodulation interference at the hardware level when attempting to introduce multi-frequency detection. If a mixed current of multiple frequencies is directly applied to the same set of excitation coils, electromagnetic fields of different frequencies will nonlinearly superimpose within the same physical space. Due to the complexity of the magnetic field distribution within the detection area and the nonlinear characteristics that ferromagnetic materials may exhibit, severe interference noise will be generated between the frequency components. This not only exponentially increases the complexity of the back-end signal processing circuitry but also significantly reduces the signal-to-noise ratio for extracting signals from tiny abrasive particles. Furthermore, to improve detection sensitivity, existing sensors often increase the sensitivity field strength by reducing the inner diameter of the coil or increasing the number of turns. However, this significantly increases the signal pulse width when particles pass through, leading to severe signal overlap problems at high abrasive particle concentrations, thus affecting the accuracy of particle counting. Lacking an architecture that can effectively distribute frequency characteristics in physical space, existing technologies struggle to acquire the complete electromagnetic polarization trajectory of abrasive particles evolving with frequency at a single monitoring node, limiting the in-depth identification of the properties of heterogeneous abrasive particles under complex wear conditions. Summary of the Invention

[0005] This invention proposes a distributed frequency array abrasive particle identification sensor. By connecting multiple physically independent three-coil detection units in series along the axis of a non-magnetic insulated flow tube, applying different excitation frequencies to each unit, and relying on spatiotemporal feature mapping to obtain the multidimensional complex electromagnetic response characteristics of abrasive particles and construct a material physical fingerprint, this invention solves the technical problems of existing abrasive particle sensors, such as parameter coupling and identification blind spots in single-frequency detection, and the easy generation of intermodulation interference and signal overlap in multi-frequency detection, which makes it impossible to accurately distinguish the types of heterogeneous metal abrasive particles.

[0006] A distributed frequency array abrasive particle identification sensor includes a non-magnetic insulated flow tube serving as a sensor channel and multiple detection units arranged in series at equal intervals along the axial direction of the non-magnetic insulated flow tube. Each detection unit is a three-coil structure wound on the non-magnetic insulated flow tube, consisting of excitation coils on both sides and an induction coil in the middle. Multiple detection units are physically independent of each other, and each detection unit is subjected to a different excitation frequency. The excitation coil is energized with an anti-phase harmonic current to excite the alternating main magnetic field. The induction coil is used to collect the secondary scattered field signal generated when the abrasive particles pass through the detection area and output a complex induced voltage signal.

[0007] Furthermore, the magnetic vector potential at any point within the alternating main magnetic field space A Satisfying equation (1): (1) in, oh For the excitation angular frequency, m and s These are the magnetic permeability and electrical conductivity of the medium, respectively. J s The source current density.

[0008] Furthermore, when the radius is a Possessing specific electrical conductivity s With relative permeability m r When abrasive grains pass through the alternating main magnetic field, magnetization and skin eddy current effects occur simultaneously within them, and the secondary scattered magnetic vector potential excited by the abrasive grains in space... A sca Represented as: (2) in, V For abrasive grain volume, r It is a spatial position vector. B exc ( z () indicates the axial position of the excitation coil relative to the particle. z The intensity of the excitation magnetic field generated at the location, χa denoted as the complex magnetic susceptibility under a time-harmonic magnetic field.

[0009] Furthermore, the complex magnetic susceptibility χ under time-harmonic magnetic field a The analytical solution satisfies equation (3): (3) in, m r Let be the relative magnetic permeability of the particle. k For complex wave number, k The expression is: (4) in m 0 represents the permeability of free space.

[0010] Furthermore, according to the reciprocity theorem, the scattering field of the particles generates a small induced electromotive force in the induction coil. u ( z The analytical solution to ) is: (5) in, C Let be the system gain constant. F This is a spatial mapping function characterizing the geometric sensitivity of the detection unit.

[0011] Furthermore, the abrasive grains move at a constant speed v Its axial position when passing through the sensor channel z With time t The following mapping relationship is satisfied: (6) in, z 0 represents the starting position of the first detection unit after the particle enters the sensor channel.

[0012] Furthermore, regarding the first n Each detection unit outputs its induced electromotive force. e n ( t This can be represented as: (7) in, oh n = 2π f n For the first n The excitation angular frequency corresponding to each detection unit u ( oh n , z - z n) represents the complex induced voltage. Based on the quasi-static analytical model of the time-harmonic electromagnetic field of the three-coil sensor unit composed of equations (1) to (5), u ( oh n , z - z n This can be further expressed as a function of the intrinsic parameters of the abrasive particles and the geometric characteristics of the system: (8) in, H n The combined shape factor of this unit is defined as the product of the shape factors of the excitation coil and the induction coil, i.e. H n = h exc · h ind This is used to quantify the geometric sensitivity distribution of the detection unit in axial space. h exc For the form factor of the excitation coil, h ind is the shape factor of the induction coil.

[0013] Furthermore, after quadrature demodulation, the output in-phase component X n ( t ) and orthogonal components Y n ( t Transient waveforms can be directly characterized as the real and imaginary parts of the modulated complex envelope: (9) Furthermore, as the abrasive grains displace axially, when the abrasive grains reach the first... n The spatial shape coefficient reaches its maximum value when the most sensitive area at the center of each detection unit is reached. H n, max At this point, the extreme points of the dual-channel time-domain waveform are extracted to construct the characteristic complex voltage peak M containing hardware modulation information. n : (10).

[0014] Furthermore, there are three detection units.

[0015] Compared with the prior art, the present invention achieves significant beneficial effects through the above technical solution: 1. High-precision identification performance at the physical level: This invention captures the complete physical evolution trajectory of abrasive particles from low-frequency magnetization response to high-frequency skin eddy current response through a space-frequency domain mapping mechanism, solving the problem of heterogeneous abrasive particles that cannot be distinguished by single-dimensional observation from a physical perspective.

[0016] 2. Anti-interference capability and signal-to-noise ratio optimization: The multi-unit physical series layout completely eliminates the intermodulation interference commonly found in multi-frequency circuits, enabling the pure extraction of weak induction signals under the high-frequency skin effect, and significantly improving the recognition accuracy of tiny abrasive particles.

[0017] 3. Multidimensional physical fingerprint acquisition capability: This architecture utilizes particle motion to automatically complete frequency scanning, eliminating the need for complex frequency sweeping circuits. A simple coil array is sufficient to acquire 2D physical fingerprints containing amplitude and phase information. N The 3D feature data stream provides a wealth of physical evidence for identifying a variety of complex materials.

[0018] 4. System robustness and calibration determinism: Since the acquired signal is a complex number containing both real and imaginary parts, the phase lag and gain fluctuations introduced by the hardware circuit can be directly removed by combining the system calibration method, ensuring the physical consistency of material recognition performance during long-term dynamic operation. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the structure and detection principle of a detection unit in a distributed frequency array abrasive particle identification sensor of the present invention; Figure 2 This is a schematic diagram of a multi-frequency cascaded structure and a space-frequency domain mapping mechanism, where, Figure 2 (a) is a diagram of the overall architecture of a multi-frequency sensor with N detection units connected in series. Figure 2 (b) is a schematic diagram of the spatiotemporal feature mapping principle; Figure 2 (c) is a diagram of the construction process of the multidimensional complex observation matrix. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] A distributed frequency array abrasive particle identification sensor includes a non-magnetic insulated flow tube serving as a sensor channel and multiple detection units arranged in series at equal intervals along the axial direction of the non-magnetic insulated flow tube. Each detection unit is a three-coil structure wound on the non-magnetic insulated flow tube, consisting of excitation coils on both sides and an induction coil in the middle. Multiple detection units are physically independent of each other, and each detection unit is subjected to a different excitation frequency. The excitation coil is energized with an anti-phase harmonic current to excite the alternating main magnetic field. The induction coil is used to collect the secondary scattered field signal generated when the abrasive particles pass through the detection area and output a complex induced voltage signal.

[0022] Specifically, this invention employs a non-magnetic insulated flow tube as the sensor channel and arranges multiple spatially independent three-coil detection units in series at equal intervals along the flow tube's axial direction. Each detection unit is given a different excitation frequency. The excitation coil is energized with an anti-phase harmonic current to generate an alternating main magnetic field, and the induction coil collects the secondary scattered field signal as the abrasive grains pass through the detection area and outputs a complex induced voltage. This hardware architecture completely avoids the intermodulation interference problem easily caused by multi-frequency mixed excitation. It also overcomes the limitations of traditional single-frequency sensors, which can only output scalar signals, leading to parameter coupling and recognition blind spots. It can effectively collect the complex electromagnetic response characteristics of abrasive grains, providing core hardware support for the classification of heterogeneous metal abrasive grains. Furthermore, it avoids signal overlap problems caused by reducing the inner diameter of the coil due to sensitivity optimization, ensuring the accuracy of abrasive grain detection and counting.

[0023] Figure 1 This describes the structure and detection principle of a three-coil sensor unit. Two excitation coils on either side are supplied with opposite-phase time-harmonic currents, thereby generating an alternating main magnetic field inside the flow tube. When abrasive particles pass through the detection area, the ferromagnetic and non-ferromagnetic materials are dominated by magnetization and eddy current effects, respectively, producing different secondary scattered fields. This scattered field causes a change in the magnetic flux of the central induction coil, which, after demodulation, outputs the real and imaginary parts of a complex voltage.

[0024] Based on the time-harmonic electromagnetic field theory, the magnetic vector potential at any point in the space of the aforementioned alternating magnetic field is... A The following governing equations must be satisfied: (1) in, oh For the excitation angular frequency, m and s These are the magnetic permeability and electrical conductivity of the medium, respectively. Js The source current density.

[0025] When the radius is a Possessing specific electrical conductivity s With relative permeability m r When abrasive grains pass through this alternating magnetic field, both magnetization and skin eddy current effects occur simultaneously within them. The secondary scattered magnetic vector potential excited by the abrasive grains in space... A sca It can be represented as: (2) in,V This refers to the volume of the abrasive grains. r It is a spatial position vector; B exc ( z The ) indicates the axial position of the excitation coil relative to the particle. z The intensity of the excitation magnetic field generated at that location; χ a Let be the complex magnetic susceptibility under a time-harmonic magnetic field, and its analytical solution is: (3) in, m r Let be the relative magnetic permeability of the particle. k Let be the complex wave number, and its expression is: (4) in m 0 represents the permeability of free space.

[0026] According to the reciprocity theorem, the scattered field of the particle generates a small induced electromotive force in the induction coil. u ( z The analytical solution to ) is: (5) in, C Let be the system gain constant. F This is a spatial mapping function characterizing the geometric sensitivity of the detection unit.

[0027] To address the problem that single-frequency detection cannot effectively decouple abrasive particle size and material information, this invention constructs a... Figure 2 (a) shows N A serially connected architecture for each detection unit is used, by applying different excitation frequencies to each detection unit. f n By utilizing the sequential movement of abrasive particles in axial space, the spatially distributed electromagnetic response of a specific frequency is transformed into multidimensional frequency domain characteristic data unfolded over time. When metal particles move at a constant speed... v Its axial position when passing through the sensor channel z With time t The following mapping relationship is satisfied: (6) in, z 0 represents the starting position of the particle entering the first unit of the series structure. For example... Figure 2 As shown in (b), since the detection units are linearly arranged on the spatial axis, the continuous displacement of the particles maps this spatial distribution characteristic into a sequence of sequentially triggered induction pulses on the time axis. This process effectively transforms the originally discrete distribution on the frequency axis... N Each characteristic response is transformed into a transient signal stream with a definite temporal relationship.

[0028] For the first in the series structure n Each sensor unit outputs its induced electromotive force. e n ( t This can be represented as: (7) in, oh n = 2π f n For the first n The excitation angular frequency corresponding to each detection unit u ( oh n , z - z n () represents the complex induced voltage. Based on the previously established quasi-static analytical model, u ( oh n , z - z n This can be further expressed as a function of the intrinsic parameters of the abrasive particles and the geometric characteristics of the system: (8) in, H n The combined shape factor of this unit is defined as the product of the shape factors of the excitation coil and the induction coil, i.e. H n = h exc · h ind This is used to quantify the geometric sensitivity distribution of the detection unit in axial space. The physical independence of the series structure decouples the frequency components in the spatial field, effectively avoiding intermodulation interference commonly found in multi-frequency mixed excitation.

[0029] Original induction signal e n,t The signal is input to the demodulation module after passing through the front-end conditioning circuit. To eliminate the inherent amplitude deviation and phase shift introduced by the circuit system under different frequency channels, this invention introduces a system characteristic compensation factor. This compensation factor is matched with a preset hardware parameter benchmark, thereby restoring the characteristics of the original signal in the complex domain and outputting complex observation characteristics that only characterize the intrinsic electromagnetic information of the particles.

[0030] To extract the low-frequency baseband envelope carrying intrinsic particle information, the lock-in amplifier internally... v n ( t ) respectively with two cos( oh n t ) and -sin( oh n t The product signal is then mixed and multiplied. The product signal is then filtered out by a low-pass filter. oh n The frequency-doubled carrier component. Because the mathematically introduced constant attenuation term in the mixing process can be absorbed as a whole into the equivalent channel amplification factor of the system hardware. A n Therefore, after quadrature demodulation, the output in-phase component... X n ( t ) and orthogonal components Y n ( t Transient waveforms can be directly characterized as the real and imaginary parts of the modulated complex envelope: (9) As the particles displace axially, when the abrasive grains reach the first... n The spatial shape coefficient reaches its maximum value when the most sensitive area at the center of each detection unit is reached. H n, max At this point, by extracting the extreme points of the dual-channel time-domain waveform, a characteristic complex voltage peak M containing hardware modulation information can be constructed. n : (10) This invention achieves deep perception of abrasive material characteristics through a multi-frequency serial physical architecture, specifically manifested in the construction and analysis of a multi-dimensional feature matrix. For example... Figure 2 As shown in (c), after the above-described orthogonal demodulation and extremum extraction steps, the transient pulse signals of each frequency channel, which were originally triggered sequentially on the time axis, are further discretized and reassembled into a sequence containing... N The multidimensional complex observation matrix M = [a number of independent frequency response information] X 1, Y 1; X 2, Y 2; …; X N , Y N ] N×2 This multidimensional observation matrix fully characterizes the multi-frequency transient response features of the particles, and each of its complex elements M... n = X n + j Y n All correspond to a specific excitation frequency f n Electromagnetic polarization fingerprints of particles.

[0031] The physical significance of this multidimensional characteristic matrix lies in its direct mapping of the evolution of the intrinsic constitutive parameters of the abrasive grains with frequency. According to the time-harmonic electromagnetic field polarization theory, the complex magnetic susceptibility χ of metal abrasive grains of different materials under an alternating magnetic field... a They exhibit significant differences, which lead to drastically different evolutionary trajectories at different excitation frequencies.

[0032] Therefore, the core performance of the multi-frequency multi-series sensor proposed in this invention, which can identify various abrasive materials, lies in: capturing the complex response M of the particles through multiple spatially distributed frequency sensing units. n The multi-point distribution characteristics within the complex plane. Because each metal material (such as iron, stainless steel, copper, aluminum, etc.) corresponds to a unique set of M values ​​distributed with frequency. n The sequence, this multidimensional observation matrix M, actually constitutes the physical "fingerprint" of the abrasive material. This high-dimensional feature redundancy based on the physical architecture enables the sensor to accurately identify and classify the type of heterogeneous metal abrasive particles at the physical level, without relying on complex post-processing, simply by observing the polarity combination and phase evolution law of each frequency component in the observation matrix. This greatly improves the adaptability of the online monitoring system to complex wear conditions.

[0033] The multi-frequency tandem topology metal abrasive particle sensing sensor with material identification capabilities proposed in this invention operates on the principle of mapping the intrinsic physical property differences of abrasive particles of different materials into the polarity and amplitude distribution characteristics of a multi-dimensional feature sequence through multiple independent sensing units connected in series along a spatial axis. This endows the sensor with the ability to identify heterogeneous metals at the hardware level. In a specific implementation, three sets of detection units are equidistantly arranged on the outside of a non-magnetic insulating flow tube with a preset inner diameter, and each unit is subjected to a different excitation frequency. f 1, f 2 and f3. The system utilizes the mechanical displacement of abrasive particles as they flow through each unit to convert spatial frequency into a time-axis pulse sequence. According to time-harmonic electromagnetic field theory, the complex magnetic susceptibility of abrasive particles of different materials under an alternating magnetic field follows a specific evolution law. When the abrasive particle reaches the central sensitive area of ​​a specific detection unit, the system extracts the extreme points of the dual-channel time-domain waveform and reassembles them into a multi-frequency electromagnetic feature set containing multiple independent frequency response information. This feature set completely records the electromagnetic polarization fingerprint of the particle, and its internal characteristic components directly map the evolution law of the abrasive particle's constitutive parameters with frequency. For example, ferromagnetic abrasive particles, due to their high permeability, are dominated by magnetization effects within a preset frequency band, and their induced signals always maintain a specific phase interval in the multi-frequency feature space. Meanwhile, highly conductive non-ferromagnetic abrasive particles, during their movement between low-frequency and high-frequency units, exhibit material-specific phase evolution trajectories in their response components within the feature set due to the difference in eddy current losses induced by changes in skin depth. Since the characteristic value sequence corresponding to each metal material at different frequencies is physically unique, the hardware system only needs to observe the phase deflection law of each frequency response component in the feature set to directly utilize the physical mapping relationship between the material and the feature sequence to achieve real-time classification and identification of heterogeneous abrasive types without relying on complex post-inversion algorithms.

[0034] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A distributed frequency array abrasive grain identification sensor, characterized by, The sensor channel comprises a non-magnetic insulating flow tube and a plurality of detection units arranged in series along the non-magnetic insulating flow tube at equal intervals in the axial direction, each detection unit being a three-coil structure wound on the non-magnetic insulating flow tube, the three-coil structure comprising an excitation coil on each side and an induction coil in the middle, The plurality of detection units are independent of each other in physical space, and each detection unit is applied with a different excitation frequency, the excitation coil is passed through an anti-phase resonant current to excite an alternating main magnetic field, and the induction coil is used to collect a secondary scattering field signal generated when the abrasive particles pass through the detection area and output a complex induction voltage signal.

2. The distributed frequency array abrasive grit identification sensor of claim 1, wherein, Magnetic vector potential at any point in the space of an alternating main magnetic field A satisfies equation (1): （1） where ω is the excitation angular frequency, μ and σ are the magnetic permeability and the electric permittivity of the medium, respectively, J s is the source current density.

3. The distributed frequency array abrasive grit identification sensor of claim 2, wherein, When the radius is a , with a specific electrical conductivity σ and a relative magnetic permeability μ r When the abrasive particles pass through the alternating main magnetic field, the internal magnetization effect and the skin current effect will occur at the same time, and the secondary scattering magnetic potential excited by the abrasive particles in space A sca is represented as: （2） wherein, V is the volume of the abrasive particle, r is the spatial position vector, B exc z represents the excitation magnetic field strength generated by the excitation coil at the axial position of the particle z where χ a is the complex magnetic susceptibility under the time-harmonic magnetic field.​ 4. The distributed frequency array abrasive grit identification sensor of claim 3, wherein, Complex magnetic susceptibility χ under a time-varying magnetic field a The analytical solution of χ satisfies equation (3): （3） wherein, μ r is the relative magnetic permeability of the particle, k is the complex wave number, k The expression for is: （4） wherein μ 0 is the vacuum permeability.

5. The distributed frequency array abrasive grit identification sensor of claim 4, wherein, According to the reciprocity theorem, the small induced electromotive force generated in the induction coil by the scattering field of the particle u ( z ) is given by the analytical solution （5） wherein C is a system gain constant, F is a spatial mapping function characterizing the geometric sensitivity of the detection unit.

6. The distributed frequency array abrasive grit identification sensor of claim 5, wherein, Abrasive particles at a constant speed v Axial position of the abrasive particles as they pass through the sensor channel z With time t Satisfy the following mapping relationship: （6） wherein z 0 is the starting position of the first detection cell after the particle enters the sensor channel.

7. The distributed frequency array abrasive grit identification sensor of claim 6, wherein, For the first n detection unit, the induced electromotive force output e n ( t ) can be expressed as: （7） wherein, ω n = 2π f n is the excitation angular frequency corresponding to the n th detection unit, u ω n , z z n is the complex induced voltage, according to the quasi-static analytical model of the three-coil sensor unit composed of equations (1) to (5), u ω n , z z n is further expressed as a function of the intrinsic parameters of the abrasive particles and the geometrical features of the system:​​​​ （8） wherein H n is the overall shape factor of the unit, defined as the product of the excitation coil shape factor and the induction coil shape factor, i.e. H n = h exc · h ind is used to quantify the geometric sensitivity distribution of the detection unit in the axial space, h exc is the excitation coil shape factor, h ind is the induction coil shape factor.

8. The distributed frequency array abrasive grit identification sensor of claim 7, wherein, After quadrature demodulation, the output in-phase component X n ( t ) and quadrature component Y n t ) Transient waveforms can be directly characterized as real and imaginary parts of a modulated complex envelope:​ （9）。 9. The distributed frequency array abrasive grit identification sensor of claim 8, wherein, As the abrasive particles are displaced axially, the spatial shape factor reaches a maximum value when the abrasive particles reach the center-most sensitive region of the n H n, max At this time, the extreme points of the dual-channel time-domain waveform are extracted, and a feature complex voltage peak M n containing hardware modulation information is constructed.​ （10）。 10. The distributed frequency array abrasive grit identification sensor of claim 9, wherein, The detection unit is provided with three.

Images