Metal processing dust concentration self-adaptive detection method and device

Abstract

The application discloses a metal processing dust concentration self-adaptive detection method and device, comprising: determining the self-adaptive calibration parameter of dust concentration according to dust scattering signals and processing working condition information; determining dust mass concentration; the self-adaptive calibration parameter of dust concentration comprises: determining a particle size distribution characteristic parameter; determining a basic calibration coefficient; determining a humidity compensation factor according to the environmental humidity information and the particle size distribution characteristic parameter, and determining the self-adaptive calibration coefficient according to the basic calibration coefficient and the humidity compensation factor; wherein the material-particle size calibration mapping information comprises a mapping relationship between the particle size distribution characteristic parameter, the material optical characteristic parameter and the basic calibration coefficient. The application can realize dynamic self-adaptive adjustment of the calibration coefficient through coupling and fusion of three parameters of the particle size distribution characteristic, the material optical characteristic and the environmental humidity, and effectively solves the concentration distortion problem caused by the fixed calibration coefficient.

Description

Technical Field

[0001] This application relates to the field of dust detection technology, and in particular to an adaptive detection method and device for metal processing dust concentration. Background Technology

[0002] With the rapid development of the metal processing and manufacturing industry, dust collection systems are being used more and more widely in metal cutting, grinding, polishing, welding, laser cutting and other processing scenarios. The measurement accuracy of the dust concentration detection module in the dust collection system directly affects the operational efficiency and safety protection level of the entire dust collection system.

[0003] In relevant metal processing dust removal system signal acquisition technology solutions, light scattering dust sensors are usually used to collect the scattered light intensity signal at a single scattering angle, and then linearly convert the scattered light intensity signal into a dust mass concentration value using a fixed calibration coefficient set at the factory. However, the technology solution using a fixed calibration coefficient lacks the ability to adapt to changes in dust particle size distribution, differences in the optical properties of metal materials, and fluctuations in ambient humidity under different processing conditions. This may cause serious distortion in the dust concentration detection results, reduce the control accuracy and safety reliability of the dust removal system when dealing with various processing scenarios, and affect the operating efficiency and energy-saving performance of the dust removal system. Summary of the Invention

[0004] In view of the above problems, this application provides an adaptive detection method and device for metal processing dust concentration, which can achieve dynamic adaptive adjustment of calibration coefficient by coupling and integrating three parameters: particle size distribution characteristics, material optical properties and environmental humidity, effectively solving the concentration distortion problem caused by fixed calibration coefficient.

[0005] In some embodiments, according to a first aspect of this application, an adaptive detection method for metal processing dust concentration is provided, comprising: determining adaptive calibration parameters for dust concentration based on dust scattering signals and processing condition information; determining dust mass concentration based on the adaptive calibration parameters and the dust scattering signals; wherein, particle size distribution characteristic parameters are determined based on scattered light intensity signals collected at at least two different scattering angles; a basic calibration coefficient is determined using material-particle size calibration mapping information, based on the particle size distribution characteristic parameters and material optical property parameters determined by processing process identification information; a humidity compensation factor is determined based on environmental humidity information and particle size distribution characteristic parameters, and an adaptive calibration coefficient is determined based on the basic calibration coefficient and the humidity compensation factor.

[0006] In this embodiment, particle size distribution characteristics are extracted using multi-angle scattering signals, and a material-particle size calibration mapping relationship is established by combining material optical property parameters to determine the basic calibration coefficients. At the same time, humidity compensation is performed based on the coupling relationship between ambient humidity and particle size, which improves the accuracy of dust concentration detection. It can adaptively match the working conditions of different metal materials, different particle size distributions, and different humidity conditions. It can reduce concentration distortion and control misjudgment caused by fixed calibration coefficients. It improves the operating efficiency and safety level of the dust removal system in various processing scenarios.

[0007] In some embodiments, the scattered light intensity signals acquired at at least two different scattering angles include: a first scattered light intensity signal acquired at a forward scattering angle and a second scattered light intensity signal acquired at a side scattering angle.

[0008] In this embodiment, by collecting scattered light intensity signals at both the forward scattering angle and the side scattering angle, the differences in scattering characteristics of particles of different sizes at different angles provide sufficient information dimensions for the extraction of particle size distribution characteristics.

[0009] In some embodiments, determining the particle size distribution characteristic parameters includes: calculating the ratio of the first scattered light intensity signal to the second scattered light intensity signal to obtain the scattering angle intensity ratio; and performing sliding window filtering on the scattering angle intensity ratio to obtain the particle size distribution characteristic parameters.

[0010] In this embodiment, the particle size distribution characteristics are characterized by calculating the ratio of scattered light intensity at two angles, and the signal noise is suppressed by using a sliding window filter, thereby improving the stability and reliability of the particle size characteristic parameters.

[0011] In some embodiments, based on a pre-stored scattering simulation database, the particle size distribution characteristic parameters are mapped to an equivalent median particle size estimate by lookup table interpolation, and the equivalent median particle size estimate is used as one of the input parameters for material-particle size calibration mapping information.

[0012] In this embodiment, a precise mapping from scattering characteristic parameters to physical particle size is achieved through a scattering simulation database, providing more intuitive and physically meaningful input parameters for the subsequent determination of basic calibration coefficients, and improving the interpretability of the calibration model.

[0013] In some embodiments, the material optical property parameters determined by the processing technology identification information include: extracting the complex refractive index parameter of the corresponding metal dust from a pre-stored material optical property database according to the processing technology identification information, wherein the complex refractive index parameter includes the real part of the refractive index reflecting the scattering intensity and the imaginary part of the refractive index reflecting the absorption characteristics.

[0014] In this embodiment, the processing technology identification information is converted into optical parameters with physical meaning through the material optical property database, so that the determination of the calibration coefficient can accurately match the scattering characteristics of different metal materials, thereby improving the detection accuracy under cross-material working conditions.

[0015] In some embodiments, determining the basic calibration coefficients using material-particle size calibration mapping information includes one of the following methods: querying a pre-established material-particle size multidimensional calibration coefficient lookup table and obtaining the basic calibration coefficients through interpolation; or using a multivariate regression model that includes particle size distribution characteristic parameters and material optical property parameters to calculate the basic calibration coefficients.

[0016] In this embodiment, two implementation methods are provided: lookup table and regression model. This allows the system to select the appropriate calibration coefficient determination method based on the complexity of the actual application scenario and computing resources, thereby improving the flexibility and applicability of the solution.

[0017] In some embodiments, the multivariate regression model includes a cross-product term of particle size distribution characteristic parameters and material optical property parameters, which is used to characterize the coupling effect of particle size distribution and material optical properties on the basic calibration coefficients.

[0018] In this embodiment, by introducing a cross-product term to explicitly model the coupling effect between particle size and material, the regression model can capture the interaction relationship that a single-parameter model cannot express, further improving the accuracy of the basic calibration coefficients.

[0019] In some embodiments, determining the humidity compensation factor includes: using a humidity-particle size coupling compensation model, determining the humidity compensation factor based on the difference between the current ambient relative humidity and the reference humidity, and the ratio of the reference particle size to the current equivalent median particle size; wherein, the humidity-particle size coupling compensation model includes the product relationship of a nonlinear power function of the difference between the current ambient relative humidity and the reference humidity, and a power function of the ratio of the reference particle size to the current equivalent median particle size.

[0020] In this embodiment, by establishing a coupled compensation model of humidity and particle size, the influence of the hygroscopic expansion effect of particles of different sizes on the scattering signal under different humidity conditions can be accurately quantified, effectively suppressing the problem of artificially high concentration in high humidity environments, and reducing the energy waste caused by unnecessary excessive response of the dust removal system.

[0021] In some embodiments, determining the adaptive calibration coefficient based on the basic calibration coefficient and the humidity compensation factor includes: when the current ambient relative humidity is higher than the reference humidity, dividing the basic calibration coefficient by the humidity compensation factor to obtain the adaptive calibration coefficient; when the current ambient relative humidity is not higher than the reference humidity, using the basic calibration coefficient as the adaptive calibration coefficient.

[0022] In this embodiment, humidity compensation is applied on demand through condition judgment, avoiding overcompensation in low humidity environments and improving the accuracy and stability of the calibration coefficient across the entire humidity range.

[0023] In some embodiments, determining the dust mass concentration based on adaptive calibration parameters and dust scattering signals includes multiplying the adaptive calibration coefficient by the digitally sampled value of the second scattered light intensity signal to obtain the dust mass concentration.

[0024] In this embodiment, by selecting the scattered light intensity signal under the lateral scattering angle as the main measurement channel and combining it with the adaptive calibration coefficients after multi-parameter compensation, high-precision dust mass concentration calculation is achieved, improving the reliability of the concentration detection results.

[0025] In some embodiments, the method further includes: constructing a feature vector corresponding to the current working condition, the feature vector including particle size distribution feature parameters and material optical property parameters; calculating the feature space distance between the feature vector and the nearest neighbor working condition point in the calibration database; and reducing the measurement confidence flag and generating calibration supplementary alarm information when the feature space distance exceeds a preset distance threshold.

[0026] In this embodiment, the deviation between the current operating condition and the calibrated operating condition is evaluated by the feature spatial distance. This can automatically identify the extrapolation area of ​​the model, avoid blindly relying on the measurement results under uncalibrated operating conditions, improve the safety and reliability of the system, and reduce false alarm events caused by concentration misjudgment.

[0027] In some embodiments, when the measurement confidence flag is lower than a preset confidence threshold, the sampling frequency of the dust scattering signal is increased, and the window length of the sliding window filter in the calculation of the particle size distribution characteristic parameter is shortened.

[0028] In this embodiment, the sampling and filtering strategies are adaptively adjusted when the confidence level decreases, so as to cope with unknown operating conditions with faster response speed and denser data collection, thereby improving the robustness and adaptability of the system under abnormal operating conditions.

[0029] In some embodiments, the method further includes: monitoring the rate of change of particle size distribution characteristic parameters within a continuous time window; and initiating a fast recalibration process when the rate of change exceeds a preset rate of change threshold. The fast recalibration process includes shortening the window length of the sliding window filtering process to accelerate the update response speed of the adaptive calibration coefficients.

[0030] In this embodiment, the drastic changes in dust characteristics are detected by monitoring the rate of change of particle size characteristics, and rapid recalibration is triggered, which shortens the update delay of calibration coefficients when operating conditions change abruptly and improves the dynamic response capability of the system.

[0031] In some embodiments, when a change in processing technology identification information is detected, the expected range of the basic calibration coefficient is pre-calculated based on the material optical property parameters corresponding to the changed processing technology identification information and the historical typical particle size range of the processing technology; before the particle size distribution characteristic parameter converges to a stable value, the median of the expected range is used as the transitional basic calibration coefficient.

[0032] In this embodiment, the calibration coefficient range is predicted in advance through the operating condition switching prediction mechanism, which can provide approximately accurate transition calibration before the dust field is stable. This significantly shortens the calibration response time after the operating condition switch and reduces the concentration distortion during the switching transition period.

[0033] In some embodiments, the dust scattering signal further includes scattered light intensity signals acquired at at least two different light source wavelengths; determining the particle size distribution characteristic parameter further includes: calculating the wavelength intensity ratio between the scattered light intensity signals at at least two different light source wavelengths; and using the wavelength intensity ratio as a supplementary dimension of the particle size distribution characteristic parameter.

[0034] In this embodiment, by introducing dual-wavelength scattering signals and utilizing the differences in scattering sensitivity of different wavelengths to particles of different sizes, a two-dimensional particle size feature space is constructed, which significantly improves the differentiation accuracy between submicron and micron-sized particles and enhances the system's ability to identify mixed particle size dust fields.

[0035] In some embodiments, the method further includes: outputting the dust mass concentration to the dust removal system controller so that the dust removal system controller can adjust the dust removal operating parameters according to the dust mass concentration.

[0036] In this embodiment, the high-precision concentration value after adaptive calibration compensation is output to the downstream control system, enabling the dust removal system to make reasonable operation and control decisions based on accurate concentration information, thereby improving the overall operating efficiency and energy saving level of the dust removal system.

[0037] In some embodiments, according to a second aspect of this application, an adaptive detection device for metal processing dust concentration is provided, comprising: an adaptive calibration parameter determination module and a dust mass concentration determination module. The adaptive calibration parameter determination module includes: a particle size feature extraction submodule, a basic calibration coefficient determination submodule, and a humidity compensation submodule.

[0038] In this embodiment, the modular device design enables the efficient execution of the aforementioned adaptive detection method for metal processing dust concentration. Each sub-module has a clear division of labor and a clear data flow, facilitating engineering implementation and maintenance.

[0039] In some embodiments, according to a third aspect of this application, an electronic device is provided, comprising: a memory; and a processor coupled to the memory, the processor being configured to perform the method described above based on instructions stored in the memory.

[0040] In some embodiments, according to a fourth aspect of this application, a dust removal control system is provided, including the electronic device described above.

[0041] In some embodiments, according to a fifth aspect of this application, a dust removal system is provided, comprising: a dust removal actuator and a dust removal control system as described above.

[0042] In some embodiments, according to a sixth aspect of this application, a metal processing apparatus is provided, the metal processing apparatus including one or more dust removal systems as described above, the dust removal systems being used to control dust generated during the metal processing process.

[0043] In some embodiments, according to a seventh aspect of this application, a computer-readable storage medium is provided that stores computer instructions which, when executed by a processor, implement the method described above.

[0044] In some embodiments, according to an eighth aspect of this application, a computer program product is provided, the computer program product storing computer instructions that, when executed by a processor, implement the method described above.

[0045] In summary, the present invention has the following beneficial effects: This invention addresses the problem of severe distortion in traditional fixed calibration coefficients caused by the three factors of variable dust material, wide particle size distribution, and fluctuating environmental humidity in metal processing scenarios. It proposes an adaptive calibration method that couples and fuses three parameters: particle size, material, and humidity. The method extracts particle size distribution characteristics in real time using the ratio of dual-angle scattered light intensity, automatically obtains the complex refractive index parameter of the corresponding metal dust by combining processing technology identifiers, determines the basic calibration coefficients using a material-particle size joint calibration mapping—a cross-product coupling term, and then uses a humidity-particle size coupling compensation model to compensate for the moisture absorption and expansion of small-diameter particles in high-humidity environments. The system performs nonlinear correction on the artificially high scattering signal, ultimately achieving dynamic adaptive adjustment of the calibration coefficient according to the working conditions. This allows the same sensor to maintain high-precision concentration output when switching between vastly different working conditions, such as aluminum grinding (coarse particles, high scattering) and steel laser cutting (ultrafine particles, strong absorption). This effectively eliminates concentration deviations of up to several times that can occur when the calibration coefficient is fixed across different materials, particle sizes, and humidity levels. At the same time, through the reliability assessment of characteristic spatial distance and the working condition switching prediction transition mechanism, the system ensures measurement safety and response speed during uncalibrated working conditions and process switching transition periods. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating some embodiments of the adaptive detection method for metalworking dust concentration of this application; Figure 2 This is a flowchart illustrating the process of determining particle size distribution characteristic parameters in some embodiments of the adaptive detection method for metal processing dust concentration of this application. Figure 3 This is a flowchart illustrating the determination of the basic calibration coefficient in some embodiments of the adaptive detection method for metal processing dust concentration of this application; Figure 4 This is a flowchart illustrating the determination of the humidity compensation factor and adaptive calibration coefficient in some embodiments of the adaptive detection method for metal processing dust concentration of this application. Figure 5 This is a schematic diagram of the material-particle size calibration mapping relationship; Figure 6 This is a schematic diagram of the humidity-particle size coupling compensation model; Figure 7 This is a flowchart illustrating the reliability assessment and dynamic adjustment process in some embodiments of the adaptive detection method for metal processing dust concentration in this application; Figure 8 This is a schematic diagram of the working condition switching prediction process in some embodiments of the adaptive detection method for metal processing dust concentration of this application. Figure 9A The diagram shows some embodiments of the adaptive detection device for metalworking dust concentration of this application. Figure 9B A schematic diagram of the submodules for determining adaptive calibration parameters; Figure 10 The diagram shows some embodiments of the electronic device of this application. Figure 11 This is a schematic diagram of some embodiments of the metal processing equipment of this application. Detailed Implementation

[0047] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0048] The terms "first," "second," etc., used in this application's specification and the foregoing drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0049] It should be noted that the relevant terms used in this application are defined as follows: Dust scattering signal refers to the light intensity signal received by a photodetector at a specific angle after dust particles are irradiated by a light source, resulting from the scattering of incident light by the particles. Forward scattering angle refers to a scattering direction with a small deflection angle relative to the incident light direction (e.g., 10° to 30°), in which the scattering intensity of large-diameter particles is much higher than that of small-diameter particles. Lateral scattering angle refers to a scattering direction with a deflection angle close to 90° relative to the incident light direction (e.g., 80° to 100°), in which the difference in scattering intensity between particles of different diameters is relatively small. Particle size distribution characteristic parameters refer to numerical parameters that characterize the particle size distribution law of a dust particle group; in this application, they are mainly obtained by filtering the ratio of scattered light intensity from multiple angles. Equivalent median particle size (d...) 50 The particle size distribution (PBD) refers to the particle size in a dust particle group where 50% of the particles by mass have a diameter smaller than this value, expressed in micrometers (μm). Complex refractive index is a parameter describing the optical properties of a material; its real part reflects the material's ability to refract and scatter light, while its imaginary part reflects the material's ability to absorb light. Material-particle size calibration mapping information refers to the correspondence between particle size distribution characteristic parameters, material optical property parameters, and basic calibration coefficients, which can be expressed through lookup tables or mathematical models. Mie scattering theory is a classical theory describing the interaction between spherical particles and electromagnetic waves, and can be used to predict the intensity distribution of scattered light at various scattering angles for particles of different sizes and refractive indices. An ADC (Analog-to-Digital Converter) is an electronic device that converts analog signals into digital signals; its bit depth determines the conversion accuracy. The least significant bit (LSB) is the change in analog quantity corresponding to the least significant bit in the ADC output value.

[0050] Figure 1 This is a flowchart illustrating some embodiments of the adaptive detection method for metalworking dust concentration of this application, such as... Figure 1 As shown, the method includes steps S101 to S102.

[0051] Step S101: Determine the adaptive calibration parameters for dust concentration based on the dust scattering signal and processing condition information.

[0052] In this step, the dust scattering signal refers to the scattered light intensity signal collected at at least two different scattering angles after dust particles are illuminated by a light source in the dust detection area of ​​the metal processing site. Processing condition information refers to comprehensive information reflecting the current processing site status, including environmental humidity information and processing technology identification information. Environmental humidity information is the current relative humidity (RH) and ambient temperature (T) value collected in real time by a temperature and humidity sensor integrated into the sensing module. Processing technology identification information is an identification code obtained from the equipment control system or manufacturing execution system (MES) used to characterize the type of metal material and processing technology being processed, such as "aluminum - grinding," "steel - laser cutting," and "copper - polishing."

[0053] Adaptive calibration parameters are calibration transformation parameters dynamically determined based on the current operating conditions, with the adaptive calibration coefficients at their core. The process of determining the adaptive calibration coefficients includes: first, determining the particle size distribution characteristic parameters based on the multi-angle scattered light intensity signal; then, using the material-particle size calibration mapping information, determining the basic calibration coefficients based on the particle size distribution characteristic parameters and the material optical property parameters determined by the processing technology identification information; finally, determining the humidity compensation factor based on the environmental humidity information and the particle size distribution characteristic parameters, and determining the adaptive calibration coefficients based on the basic calibration coefficients and the humidity compensation factor.

[0054] The material-particle size calibration mapping information includes the mapping relationship between particle size distribution characteristic parameters and material optical property parameters and basic calibration coefficients. This mapping relationship can be expressed by a pre-established multidimensional calibration coefficient lookup table or a multiple regression model, reflecting the optimal conversion ratio from scattered light intensity signal to dust mass concentration under different particle size distributions and different material optical properties.

[0055] Step S102: Determine the dust mass concentration based on the adaptive calibration parameters and the dust scattering signal.

[0056] In this step, the adaptive calibration coefficient determined in step S101 is used to calibrate and convert the scattered light intensity signal of the main measurement channel in the dust scattering signal, obtaining the dust mass concentration value after triple compensation for particle size, material, and humidity. The unit of this concentration value is milligrams per cubic meter (mg / m³). 3 It can be directly used for decision-making and interlocking of downstream dust removal system controllers.

[0057] Through steps S101 to S102, a dynamic calibration mechanism with multi-parameter coupling is established by extracting particle size distribution characteristics using multi-angle scattering signals and combining material optical property parameters and environmental humidity information, thereby achieving adaptive adjustment of calibration coefficients. Compared with the existing technology that uses fixed calibration coefficients, this method can adaptively match the changing operating conditions under different metal materials, particle size distributions, and humidity levels, reducing concentration measurement errors and minimizing control misjudgments and false alarms caused by concentration distortion. This improves the operational efficiency and safety level of the dust removal system in various processing scenarios.

[0058] The following provides a detailed explanation of each sub-process in step S101 for determining the adaptive calibration parameters.

[0059] Figure 2 This is a flowchart illustrating the process of determining particle size distribution characteristic parameters in some embodiments of the adaptive detection method for metal processing dust concentration of this application, such as... Figure 2 As shown, determining the particle size distribution characteristic parameters includes steps S201 to S203.

[0060] Step S201: Acquire scattered light intensity signals at at least two different scattering angles.

[0061] In some embodiments, the scattered light intensity signals acquired at at least two different scattering angles include: a first scattered light intensity signal acquired at a forward scattering angle and a second scattered light intensity signal acquired at a side scattering angle.

[0062] Specifically, a detector group containing at least two photodetectors is set up in the detection area. The first photodetector is positioned at the forward scattering angle, for example, at a deflection of θ1 = 15° ± 2° relative to the incident light direction, to collect the first scattered light intensity signal I(θ1). The second photodetector is positioned at the side scattering angle, for example, at a deflection of θ2 = 90° ± 2° relative to the incident light direction, to collect the second scattered light intensity signal I(θ2). The analog output signals of the two photodetectors are digitally sampled by a 16-bit ADC at a sampling rate of not less than 10Hz to ensure real-time tracking capability of dynamic changes in dust concentration.

[0063] The selection of forward scattering angle and side scattering angle is based on the following: According to Mie scattering theory, the scattered light intensity of large-diameter particles (e.g., particles larger than 10 μm) is much greater in the forward direction than in the side direction, while the scattered light intensity of small-diameter particles (e.g., particles smaller than 2 μm) shows less difference between different angles. Therefore, the ratio of the two-angle scattering signals can effectively reflect the particle size distribution characteristics of the dust particle group.

[0064] Step S202: Calculate the ratio of scattered angle to light intensity.

[0065] In some embodiments, the ratio of the first scattered light intensity signal to the second scattered light intensity signal is calculated to obtain the scattering angle light intensity ratio.

[0066] Specifically, the ratio of the first scattered light intensity signal I(θ1) to the second scattered light intensity signal I(θ2) at each sampling time is calculated to obtain the scattering angle intensity ratio R. θ =I(θ1) / I(θ2). This ratio reflects the particle size distribution characteristics of the current dust field: R θ The higher the value, the higher the proportion of large-diameter particles in the dust; R θ The closer the value is to 1, the higher the proportion of small-diameter particles in the dust.

[0067] For example, under the condition of grinding aluminum parts, the forward scattering angle is measured to be I(θ1) = 3200 LSB, and the side scattering angle is measured to be I(θ2) = 400 LSB, then R θ =3200 / 400=8.0, indicating that the dust is mainly composed of larger particle sizes. Under the condition of laser cutting of steel parts, the forward scattering angle is measured as I(θ1)=1800LSB, and the side scattering angle is measured as I(θ2)=1200LSB, then R θ =1800 / 1200=1.5, indicating that the dust is mainly composed of smaller particle sizes.

[0068] Step S203: Perform sliding window filtering on the ratio of scattered angle to light intensity to obtain the particle size distribution characteristic parameters.

[0069] In some embodiments, the ratio of scattered angle light intensity R obtained by successive sampling θ Perform sliding window filtering. The recommended sliding window length is 10 to 30 sampling points, corresponding to a time window of 1 to 3 seconds. Within the window, evaluate R... θ The arithmetic mean is taken to obtain a stable particle size distribution characteristic parameter, which is denoted as the filtered scattering angle-to-intensity ratio. The sliding window filter is used to suppress random noise and transient disturbances in a single measurement, ensuring that the particle size characteristic parameter can stably and reliably reflect the particle size distribution of the current dust field.

[0070] For example, in the above-mentioned aluminum part grinding process, R at 20 consecutive sampling points θ The value fluctuated between 7.5 and 8.5, and after sliding window filtering, the particle size distribution characteristic parameter was 7.8.

[0071] In some embodiments, after obtaining the particle size distribution characteristic parameters, the method further includes: mapping the particle size distribution characteristic parameters to an equivalent median particle size estimate d by interpolation through a lookup table based on a pre-stored scattering simulation database. 50 The equivalent median grain size estimate is used as one of the input parameters for the material-grain size calibration mapping information.

[0072] This scattering simulation database was pre-established based on Mie scattering theory through numerical simulations, containing the correspondence between particle size distribution characteristic parameters and equivalent median particle size. The lookup tables are set at logarithmic intervals, with a total of 32 to 64 interpolation nodes. The correspondence is as follows: when the particle size distribution characteristic parameter is in the range of 1.0 to 2.0, the corresponding d... 50 Approximately 0.5 to 2 μm; when the particle size distribution characteristic parameter is in the range of 2.0 to 6.0, the corresponding d 50 Approximately 2 to 10 μm; when the particle size distribution characteristic parameter is in the range of 6.0 to 20.0, the corresponding d 50 It is approximately 10 to 50 μm.

[0073] For example, in the above-mentioned aluminum grinding process, the particle size distribution characteristic parameter is 7.8. The equivalent median particle size estimate d is determined by interpolation from a table. 50 Approximately 12 μm. In the above-mentioned laser cutting of steel parts, assuming the particle size distribution characteristic parameter is 1.5 after filtering, d is determined by interpolation from a table. 50 It is approximately 1.5 μm.

[0074] Through steps S201 to S203 and the mapping of the scattering simulation database, the extraction and quantization of physical particle size characteristics from the original dual-angle scattered light intensity signal were realized, laying the foundation for the accurate determination of the subsequent basic calibration coefficients.

[0075] Figure 3 This is a flowchart illustrating the determination of the basic calibration coefficients in some embodiments of the adaptive detection method for metalworking dust concentration of this application, such as... Figure 3 As shown, determining the basic calibration coefficients includes steps S301 to S303.

[0076] Step S301: Determine the optical property parameters of the material based on the processing technology identification information.

[0077] In some embodiments, the material optical property parameters determined by the processing technology identification information include: extracting the complex refractive index parameter of the corresponding metal dust from a pre-stored material optical property database based on the processing technology identification information, wherein the complex refractive index parameter includes the real part n of the refractive index that reflects the scattering intensity. r and the imaginary part of the refractive index n, which reflects absorption characteristics i .

[0078] The material optical property database pre-stores the optical parameters of various common metallic dust particles. For example, the real part of the refractive index n of aluminum dust. r =1.37, imaginary part of refractive index n i =7.6; Real part of the refractive index n of iron dust r =2.87, imaginary part of refractive index ni =3.35; Real part of the refractive index n of copper dust r =0.62, imaginary part of refractive index n i =2.57; Real part of the refractive index n of stainless steel dust r =2.75, imaginary part of refractive index n i =3.80. The above values ​​are typical values ​​under the condition of incident light wavelength of 880nm.

[0079] When the processing technology identifier information "aluminum-grinding" is received, the complex refractive index parameter (n) of aluminum dust is extracted from the database. r =1.37, n i =7.6); When "steel-laser cutting" is received, the complex refractive index parameter (n) of the iron dust is extracted. r =2.87, n i =3.35).

[0080] Step S302: Construct the material-grain size joint feature vector.

[0081] The material optical property parameters obtained in step S301 are combined with the particle size distribution characteristic parameters and equivalent median particle size estimates obtained in steps S201 to S203 to construct a joint feature vector F=[d 50 ,n r ,n i [Particle size distribution characteristic parameter]. This feature vector integrates the physical particle size information and material optical information of the dust, and is a multi-dimensional input for determining the basic calibration coefficients.

[0082] For example, in the case of grinding aluminum parts, the joint feature vector is F=[12,1.37,7.6,7.8]; in the case of laser cutting steel parts, the joint feature vector is F=[1.5,2.87,3.35,1.5].

[0083] Step S303: Determine the basic calibration coefficient using material-particle size calibration mapping information.

[0084] In some embodiments, determining the basic calibration coefficients using material-grain size calibration mapping information includes one of the following methods.

[0085] Figure 5 This is a schematic diagram of the material-particle size calibration mapping relationship, such as... Figure 5 As shown, this mapping relationship can be achieved in the following two ways.

[0086] The first method is a lookup table approach: A pre-established material-particle size multidimensional calibration coefficient lookup table is consulted, and the basic calibration coefficient is obtained through interpolation. This lookup table is divided into 5 to 8 levels based on the equivalent median particle size (e.g., 1μm, 2μm, 5μm, 10μm, 20μm, 50μm) and 4 to 6 categories based on the metal material (e.g., aluminum, iron, copper, stainless steel, titanium alloy), totaling 20-48 working condition combinations. For working conditions that do not precisely fall into discrete nodes, the basic calibration coefficient K is obtained through multidimensional linear interpolation. base .

[0087] The second method is the multiple regression model approach: This method uses a multiple regression model that includes parameters related to particle size distribution and material optical properties to calculate the basic calibration coefficients. The general form of this model is K... base =a0+a1·d 50 +a2·n r +a3·n i +a4·d 50 ·n r +a5·Particle size distribution characteristic parameter, where a0 to a5 are regression coefficients, obtained by fitting offline calibration experimental data (covering at least 4 kinds of metal materials multiplied by 5 kinds of particle size ranges, totaling 20 sets of standard working conditions).

[0088] In some embodiments, the multivariate regression model includes a4·d, which is a cross-product term of particle size distribution characteristic parameters and material optical property parameters. 50 ·n r The cross-product term characterizes the coupling effect between particle size distribution and material optical properties on the basic calibration coefficients. The same metal material exhibits different scattering behaviors at different particle sizes. For example, aluminum powder at a particle size of 20 μm is dominated by geometric scattering with a scattering efficiency factor of approximately 2.1, but when the particle size decreases to 1 μm, it enters the Mie scattering region, and the scattering efficiency factor can rise to 3.8. The cross-product term can capture this interaction between particle size and material properties, making the model's fitting accuracy higher than that of independent linear combinations of parameters.

[0089] Basic calibration coefficient K base The typical value range is 0.005 to 0.05 (mg / m³). 3 ) / LSB. For example, in aluminum grinding conditions (d 50 =12μm, n r =1.37, n i =7.6), K is determined by lookup table or regression model. base Approximately 0.015 (mg / m³) 3 ) / LSB. In the laser cutting of steel parts (d 50 =1.5μm, n r =2.87, n i =3.35), determine Kbase Approximately 0.038 (mg / m³) 3 ) / LSB. K under two operating conditions base The difference of approximately 2.5 times indicates that using the same fixed calibration coefficient would result in severe distortion of concentration readings under at least one operating condition.

[0090] Figure 4 This is a flowchart illustrating the determination of the humidity compensation factor and adaptive calibration coefficient in some embodiments of the adaptive detection method for metal processing dust concentration of this application, as shown below. Figure 4 As shown, determining the humidity compensation factor and adaptive calibration coefficient includes steps S401 to S403.

[0091] Step S401: Obtain ambient humidity information.

[0092] The current ambient relative humidity (RH) value is acquired in real time from a digital temperature and humidity sensor integrated into the sensing module. The sensor's measurement range covers 0% to 100%RH, with a measurement accuracy of no less than ±2%RH, and the sampling frequency is synchronized with the scattered light intensity signal.

[0093] Step S402: Determine the humidity compensation factor using the humidity-particle size coupling compensation model.

[0094] In some embodiments, determining the humidity compensation factor includes: using a humidity-particle-size coupled compensation model, determining the humidity compensation factor based on the difference between the current ambient relative humidity and the reference humidity, and the ratio of the reference particle size to the current equivalent median particle size.

[0095] Figure 6 This is a schematic diagram of the humidity-particle size coupling compensation model, as shown below. Figure 6 As shown, the humidity compensation factor α H The formula for calculating α is: H =1+β1×((RH-RH0) / 100)^γ×(d ref / d 50 )^β2. Where RH is the current ambient relative humidity, RH0 is the reference humidity (i.e., the calibration ambient humidity, typically 45%), and d ref For reference particle size (typically 10 μm), d 50 β1, β2, and γ are the current equivalent median particle size estimates, and β1, β2, and γ are the model parameters obtained through offline calibration experiments.

[0096] The core of this model is that the humidity-particle size coupling compensation model includes the product relationship of the non-linear power function of the difference between the current environmental relative humidity and the reference humidity and the power function of the ratio of the reference particle size to the current equivalent median particle size. The non-linear power exponent γ (typical value range 1.5 to 2.5) reflects the non-linear growth characteristic of the influence of humidity on the scattering signal: when the humidity is slightly higher than the reference value, the compensation amount is small, and when the humidity significantly exceeds the reference value, the compensation amount increases sharply. The power exponent β2 of the particle size ratio (typical value range 0.3 to 0.8) reflects the physical law that small particle size particles absorb moisture and expand more significantly due to their large specific surface area. The coefficient β1 (typical value range 0.8 to 2.5) is the scaling parameter of the overall compensation intensity.

[0097] The working mechanism of the humidity-particle size coupling compensation model is illustrated by the following specific numerical examples. Take β1 = 1.5, β2 = 0.5, γ = 2.0, RH0 = 45%, d ref = 10μm.

[0098] Example 1: Aluminum part grinding, high humidity condition. RH = 80%, d 50 = 12μm. Calculate α H = 1 + 1.5×((80 - 45) / 100)^2×(10 / 12)^0.5 = 1 + 1.5×0.1225×0.913 ≈ 1.168. The humidity compensation amount is about 16.8%.

[0099] Example 2: Steel part laser cutting, low humidity condition. RH = 40%, d 50 = 1.5μm. Since RH < RH0 = 45%, the expansion effect of humidity on the scattering signal is not significant, and α H approaches 1.0.

[0100] Example 3: Steel part grinding, high humidity condition. RH = 85%, d 50 = 5μm. Calculate α H = 1 + 1.5×((85 - 45) / 100)^2×(10 / 5)^0.5 = 1 + 1.5×0.16×1.414 ≈ 1.339. The humidity compensation amount is about 33.9%. In this condition, the particle size is small and the humidity is high, and the moisture absorption and expansion effect is significant, indicating the coupling relationship between humidity and particle size.

[0101] Example 4: The condition of RH = 85% but large particle size d 50 = 20μm under the same high humidity condition. Calculate α H = 1 + 1.5×0.16×(10 / 20)^0.5 = 1 + 1.5×0.16×0.707 ≈ 1.170. The compensation amount is only about 17.0%, much smaller than that in Example 3 where d 50=33.9% at 5μm. This verifies the physical law that small-diameter particles require stronger compensation under the same humidity conditions.

[0102] Step S403: Determine the adaptive calibration coefficient based on the basic calibration coefficient and the humidity compensation factor.

[0103] In some embodiments, determining the adaptive calibration coefficient based on the base calibration coefficient and the humidity compensation factor includes: when the current ambient relative humidity is higher than the reference humidity, dividing the base calibration coefficient by the humidity compensation factor to obtain the adaptive calibration coefficient, i.e., K. adapt =K base / α H When the relative humidity of the current environment is not higher than the reference humidity, the basic calibration coefficient is directly used as the adaptive calibration coefficient, i.e., K. adapt =K base .

[0104] Taking the aluminum part grinding and high humidity conditions in Example 1 as an example, K base =0.015, α H =1.168, then K adapt =0.015 / 1.168≈0.0128 (mg / m³) 3 ) / LSB. Taking the laser cutting of steel parts under low humidity conditions in Example 2 as an example, K base =0.038, α H =1.0, then K adapt =0.038 (mg / m³) 3 ) / LSB.

[0105] Through steps S401 to S403, the coupling relationship between humidity and particle size is used to perform nonlinear compensation on the basic calibration coefficient, which effectively suppresses the distortion of concentration readings caused by the falsely high scattering signal due to the moisture absorption and expansion of dust particles under high humidity conditions.

[0106] In some embodiments, the method for determining the dust mass concentration is as follows: determining the dust mass concentration based on adaptive calibration parameters and dust scattering signals includes multiplying the adaptive calibration coefficients by the digitally sampled value of the second scattered light intensity signal to obtain the dust mass concentration. The second scattered light intensity signal I(θ2) at the side scattering angle is selected as the main measurement channel because the side scattering signal is less sensitive to particle size changes than the forward scattering signal, making it more stable as the basic measurement signal, while the influence of particle size and material has been compensated for by the adaptive calibration coefficients.

[0107] The formula for calculating dust mass concentration is: C dust =K adapt ×I(θ2).

[0108] Taking the aluminum grinding and high humidity conditions in Example 1 as an example, I(θ2) = 400 LSB, K adapt =0.0128, then C dust =0.0128×400=5.13mg / m 3 If the existing technology uses a fixed calibration coefficient of 0.020, then C dust =0.020×400=8.0mg / m 3 High concentration readings trigger unnecessary high-concentration alarms and accelerate the operation of the dust removal system, resulting in energy waste.

[0109] Taking the laser cutting of steel parts under low humidity conditions in Example 2 as an example, I(θ2) = 1200 LSB, K adapt =0.038, then C dust =0.038×1200=45.6mg / m 3 If a fixed calibration coefficient of 0.020 is used, then C dust =0.020×1200=24.0mg / m 3 The concentration reading was about 47% lower than normal, which may cause the dust removal system to fail to respond in time when the dust concentration exceeds the standard, posing a safety hazard.

[0110] The above comparative examples clearly demonstrate that the adaptive calibration coefficients using this scheme can accurately reproduce the true dust concentration under different operating conditions, while the fixed calibration coefficients will produce deviations in opposite directions and with significant amplitudes under different operating conditions.

[0111] Figure 7 This is a flowchart illustrating the reliability assessment and dynamic adjustment process in some embodiments of the adaptive detection method for metal processing dust concentration of this application, such as... Figure 7 As shown, the credibility assessment and dynamic adjustment includes steps S501 to S503.

[0112] Step S501: Construct the feature vector corresponding to the current working condition.

[0113] In some embodiments, a feature vector corresponding to the current operating condition is constructed. This feature vector includes particle size distribution feature parameters and material optical property parameters. Specifically, the feature vector F=[d 50 ,n r ,n i The particle size distribution characteristic parameters are the same as the joint feature vector constructed in step S302.

[0114] Step S502: Calculate the feature space distance.

[0115] Calculate the Euclidean distance between the feature vector of the current operating condition and all calibrated operating condition points in the calibration database, and take the minimum value as the feature space distance D. FThe calibration database stores all the operating condition feature vectors used in offline calibration experiments and their corresponding calibration coefficients.

[0116] For example, if the calibration database is closest to the current operating condition (aluminum-grinding, d) 50 The calibrated operating condition for (=12μm) is (aluminum-grinding, d) 50 =15μm), the characteristic spatial distance D between the two F A smaller value indicates that the current operating condition is within the calibrated range. If the current operating condition is a newly introduced titanium alloy polishing process, and there is no close calibrated operating condition point in the calibration database, then D... F Relatively large.

[0117] Step S503: Determine the credibility and adjust the strategy based on the distance results.

[0118] In some embodiments, when the feature space distance exceeds a preset distance threshold D th In this case, the measurement confidence flag is lowered and a calibration supplementary alarm message is generated. Preset distance threshold D th The settings can be configured based on the coverage density of the calibration data and the security requirements of the application scenario. The trustworthiness flag uses a multi-level system (e.g., high, medium, and low levels). When D... F ≤0.5×D th For high confidence, when 0.5×D th <D F ≤D th When the confidence level is medium, when D F >D th The system is considered to have low confidence. In this low confidence state, the system generates a calibration supplement alarm, prompting operators to supplement the offline calibration data for the corresponding operating condition.

[0119] In some embodiments, when the measurement confidence flag is lower than a preset confidence threshold, the sampling frequency of the dust scattering signal is increased, and the window length of the sliding window filter in the calculation of the particle size distribution characteristic parameters is shortened. For example, under normal high confidence conditions, the sampling frequency is 10Hz and the sliding window length is 20 sampling points; when the confidence level drops to a low level, the sampling frequency is increased to 20Hz and the sliding window length is shortened to 8 sampling points, so as to cope with unknown working conditions with faster response speed and denser data acquisition.

[0120] In some embodiments, the method further includes: monitoring the rate of change of particle size distribution characteristic parameters within a continuous time window; and initiating a rapid recalibration process when the rate of change exceeds a preset rate of change threshold. The rapid recalibration process includes shortening the window length of the sliding window filtering process to accelerate the update response speed of the adaptive calibration coefficients. For example, under normal conditions, the sliding window length is 20 sampling points (2-second time window). When the rate of change of particle size distribution characteristic parameters exceeds 30% per second, the sliding window length is shortened to 5 sampling points (0.5-second time window), enabling the calibration coefficients to respond to sudden changes in dust characteristics within 1 second.

[0121] Through steps S501 to S503, the system can automatically assess the reliability of the current measurement results and adaptively adjust the acquisition strategy when the reliability decreases or the dust characteristics change drastically. This avoids safety accidents or false alarms caused by blindly relying on measurement results in the extrapolation region of the model. It is estimated that this mechanism can reduce false alarm events caused by concentration misjudgment by more than 70%.

[0122] Figure 8 This is a schematic diagram of the working condition switching prediction process in some embodiments of the adaptive detection method for metal processing dust concentration of this application, such as... Figure 8 As shown, the working condition switching prediction process includes steps S601 to S602.

[0123] Step S601: Detect the switching event of the processing technology identification information.

[0124] The system continuously monitors the processing technology identification information received from the equipment control system or MES. When a change in the identification code is detected (e.g., switching from "aluminum-grinding" to "steel-laser cutting"), it determines that a working condition switching event has occurred.

[0125] Step S602: Pre-calculate the expected range of the basic calibration coefficients and set the transition calibration coefficients.

[0126] In some embodiments, when a change in processing technology identification information is detected, the expected range of the basic calibration coefficient is pre-calculated based on the material optical property parameters corresponding to the changed processing technology identification information and the historical typical particle size range of the processing technology; before the particle size distribution characteristic parameter converges to a stable value, the median of the expected range is used as the transitional basic calibration coefficient.

[0127] Specifically, when the process is switched from "aluminum-grinding" to "steel-laser cutting", the system extracts the material optical property parameters (n) of the iron dust based on the new process identifier. r =2.87, n i =3.35), and look up the historical typical particle size range of this process (e.g., d). 50(Range 0.5 to 3 μm). Substituting the two ends of this particle size range and the material parameters into the calibration mapping model, K is calculated. base The expected range, for example, is [0.032, 0.045], and the median value of 0.0385 is taken as the transitional baseline calibration coefficient. For the first 5 to 10 seconds after the operating condition switch (before the particle size distribution characteristic parameter converges to a stable value), the system uses this transitional value; after the particle size distribution characteristic parameter converges, it automatically switches to the precisely calculated K. base value.

[0128] Through the operating condition switching prediction mechanism, the convergence time of the calibration coefficients is shortened from 30 to 60 seconds in the existing technology to 3 to 5 seconds, reducing the concentration distortion window during the switching transition period by about 85% to 90%.

[0129] In some embodiments, the dust scattering signal further includes scattered light intensity signals acquired at at least two different light source wavelengths; determining the particle size distribution characteristic parameter further includes: calculating the wavelength intensity ratio between the scattered light intensity signals at at least two different light source wavelengths; and using the wavelength intensity ratio as a supplementary dimension of the particle size distribution characteristic parameter.

[0130] Specifically, in addition to the original first light source (e.g., an infrared light source with a wavelength of 880nm), a second light source (e.g., a blue light source with a wavelength of 450nm) is added to the light scattering sensing module. The scattered light intensity signals of the two wavelengths are collected at the same scattering angle, and the wavelength-to-intensity ratio R is calculated. λ = I(880nm) / I(450nm). According to Mie scattering theory, particles of different sizes have different scattering efficiencies for different wavelengths of light: submicron-sized particles have significantly higher scattering efficiency at short wavelengths (450nm) than at long wavelengths (880nm), while particles larger than micrometers show less difference in scattering efficiency between the two wavelengths. Therefore, the wavelength-to-intensity ratio R... λ Together with the ratio of scattering angle to light intensity, they form a two-dimensional particle size feature space, which can significantly improve the distinction accuracy between submicron and micron-sized particles, thus enhancing particle size identification accuracy.

[0131] This embodiment is applicable to mixed processing workshops where both ultrafine dust (such as laser cutting fumes) and coarse dust (such as grinding iron powder) are present.

[0132] In some embodiments, the method further includes: outputting the dust mass concentration to the dust removal system controller so that the dust removal system controller can adjust the dust removal operating parameters according to the dust mass concentration.

[0133] Specifically, the dust mass concentration C after adaptive calibration compensation is... dustThe high-precision concentration value is output to the dust collection system controller via a standard communication interface (e.g., 4-20mA analog output, RS485 digital interface, or industrial Ethernet interface). The dust collection system controller can optimize dust collection operation parameters such as adjusting the dust collector fan speed, controlling valve opening, and controlling filter cartridge pulse cleaning based on the received high-precision concentration value. Because the accuracy of the output concentration value in this solution is significantly improved compared to existing technologies, the dust collection system controller can more accurately match the actual dust load, avoiding energy waste caused by over-dust collection or safety hazards caused by under-dust collection.

[0134] Figure 9A These are schematic diagrams of some embodiments of the adaptive detection device for metalworking dust concentration of this application, such as... Figure 9A As shown, the metal processing dust concentration adaptive detection device 700 includes: an adaptive calibration parameter determination module 701 and a dust mass concentration determination module 702.

[0135] The adaptive calibration parameter determination module 701 is used to determine the adaptive calibration parameters of dust concentration based on the dust scattering signal and processing condition information. The dust scattering signal includes scattered light intensity signals collected at at least two different scattering angles; the processing condition information includes environmental humidity information and processing technology identification information; and the adaptive calibration parameters include adaptive calibration coefficients.

[0136] The dust mass concentration determination module 702 is used to determine the dust mass concentration based on adaptive calibration parameters and dust scattering signals.

[0137] Figure 9B A schematic diagram of the submodules for the adaptive calibration parameter determination module is shown below. Figure 9B As shown, the adaptive calibration parameter determination module 701 includes: a particle size feature extraction submodule 7011, a basic calibration coefficient determination submodule 7012, and a humidity compensation submodule 7013.

[0138] The particle size feature extraction submodule 7011 is used to determine particle size distribution feature parameters based on scattered light intensity signals acquired at at least two different scattering angles. In some embodiments, the particle size feature extraction submodule 7011 calculates the ratio of the first scattered light intensity signal to the second scattered light intensity signal to obtain the scattering angle intensity ratio, and performs sliding window filtering on the scattering angle intensity ratio to obtain the particle size distribution feature parameters. In some embodiments, the particle size feature extraction submodule 7011 further includes a scattering simulation database lookup unit, used to map the particle size distribution feature parameters to an equivalent median particle size estimate through table lookup interpolation.

[0139] The basic calibration coefficient determination submodule 7012 is used to determine basic calibration coefficients based on material-particle size calibration mapping information, according to particle size distribution characteristic parameters and material optical property parameters determined by processing technology identification information. The material-particle size calibration mapping information includes the mapping relationship between particle size distribution characteristic parameters, material optical property parameters, and basic calibration coefficients. In some embodiments, the basic calibration coefficient determination submodule 7012 includes a material optical property database lookup unit, used to extract the complex refractive index parameter of the corresponding metal dust based on the processing technology identification information. In some embodiments, the basic calibration coefficient determination submodule 7012 also includes a calibration coefficient calculation unit, which can determine the basic calibration coefficients using a lookup table or a multiple regression model.

[0140] The humidity compensation submodule 7013 is used to determine a humidity compensation factor based on environmental humidity information and particle size distribution characteristic parameters, and to determine an adaptive calibration factor based on the basic calibration coefficient and the humidity compensation factor. In some embodiments, the humidity compensation submodule 7013 utilizes a humidity-particle size coupled compensation model to determine the humidity compensation factor based on the difference between the current relative humidity and the reference humidity and the ratio of the reference particle size to the current equivalent median particle size, and further determines the adaptive calibration factor based on the basic calibration coefficient and the humidity compensation factor.

[0141] In some embodiments, the metal processing dust concentration adaptive detection device 700 further includes a reliability assessment module 703, which is used to construct a feature vector corresponding to the current working condition, calculate the feature space distance between the feature vector and the nearest neighbor working condition point in the calibration database, and reduce the measurement reliability flag and generate calibration supplementary alarm information when the feature space distance exceeds a preset distance threshold.

[0142] In some embodiments, the metal processing dust concentration adaptive detection device 700 further includes a working condition switching prediction module 704, which is used to pre-calculate the expected range of the basic calibration coefficient when a switching of the processing process identification information is detected, and to use the median of the expected range as the transition basic calibration coefficient before the particle size distribution characteristic parameter converges to a stable value.

[0143] With the modular device design described above, each sub-module has a clear division of labor and a clear data flow, which can efficiently execute the above-mentioned adaptive detection method for metal processing dust concentration, making it easy to implement and maintain in engineering.

[0144] Figure 10 This is a schematic diagram of modules for some embodiments of the electronic device of this application, such as... Figure 10As shown, the electronic device 800 includes a memory 801, a processor 802, a communication interface 803, and a bus 804. The memory 801 stores computer instructions. The processor 802 is connected to the memory 801 via the bus 804, and the communication interface 803 is connected to the processor 802 via the bus 804. When the processor 802 executes the computer instructions stored in the memory 801, it implements the aforementioned adaptive detection method for metal processing dust concentration. The communication interface 803 is used to receive digitized data of dust scattering signals, environmental humidity information, and processing technology identification information, and to output dust mass concentration values ​​and confidence flags to the dust removal system controller.

[0145] The processor 802 can be an embedded microcontroller (such as an MCU based on the ARM Cortex-M series), a digital signal processor (DSP), or a field-programmable gate array (FPGA). The memory 801 can include Flash memory (for storing program instructions, calibration coefficient lookup tables, and material optical property databases) and RAM memory (for storing intermediate variables and sliding window data buffers during runtime).

[0146] Dust removal control system and dust removal system examples In some embodiments, the dust removal control system includes the electronic device 800 as described above. In addition to including the electronic device for adaptive dust concentration detection, the dust removal control system may also include a dust removal operation control unit for adjusting dust removal operation parameters such as fan speed control, valve control, and dust removal control based on the dust mass concentration value output by the electronic device.

[0147] In some embodiments, the dust removal system includes a dust removal actuator and a dust removal control system as described above. The dust removal actuator includes a dust collection hood, a conveying pipe, a dust removal fan, a cartridge or bag filter, a pulse cleaning mechanism, and a dust collection bin, etc., for performing dust adsorption, conveying, filtering, and collection under the command of the dust removal control system.

[0148] Figure 11 This is a schematic diagram of some embodiments of the metal processing equipment of this application, such as... Figure 11 As shown, the metal processing equipment 900 includes one or more dust removal systems 901, which are used to control the dust generated during the metal processing process. The metal processing equipment 900 can be, but is not limited to, CNC grinding machines, CNC milling machines, laser cutting machines, plasma cutting machines, welding workstations, polishing equipment, or integrated processing lines containing multiple processing stations. Each processing station can be configured with an independent dust removal system, or multiple stations can share a centralized dust removal system. The dust concentration adaptive detection devices in the system can be distributed and deployed at various detection points.

[0149] In some embodiments, a computer-readable storage medium is provided, which stores computer instructions that, when executed by a processor, implement the aforementioned adaptive detection method for metalworking dust concentration. The computer-readable storage medium may include, but is not limited to, read-only memory (ROM), random access memory (RAM), disk storage, optical disk storage, flash memory, etc.

[0150] In some embodiments, a computer program product is provided, which stores computer instructions that, when executed by a processor, implement the aforementioned adaptive detection method for metal processing dust concentration. The computer program product can be written in any combination of one or more programming languages ​​to execute the aforementioned method, including but not limited to C, C++, or Python.

[0151] It should be noted that in the above method embodiments, the sequence number of each step does not represent the execution order. The execution order of each step should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0152] The above-described 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, and should all be included within the protection scope of this application.

Claims

1. An adaptive detection method for metal processing dust concentration, characterized in that, include: Based on dust scattering signals and processing condition information, adaptive calibration parameters for dust concentration are determined. The dust mass concentration is determined based on the adaptive calibration parameters and the dust scattering signal. The dust scattering signal includes scattered light intensity signals collected at at least two different scattering angles; the processing condition information includes environmental humidity information and processing technology identification information; the adaptive calibration parameters include adaptive calibration coefficients; and the adaptive calibration parameters for determining dust concentration based on the dust scattering signal and processing condition information include: Based on the scattered light intensity signals at at least two different scattering angles, determine the particle size distribution characteristic parameters; Using material-particle size calibration mapping information, the basic calibration coefficients are determined based on the particle size distribution characteristic parameters and the material optical property parameters determined by the processing technology identification information; Based on the environmental humidity information and the particle size distribution characteristic parameters, a humidity compensation factor is determined, and an adaptive calibration factor is determined based on the basic calibration coefficient and the humidity compensation factor. The material-particle size calibration mapping information includes the mapping relationship between the particle size distribution characteristic parameters and the material optical property parameters and the basic calibration coefficients.

2. The method as described in claim 1, characterized in that, The scattered light intensity signals acquired at at least two different scattering angles include: The intensity signal of the first scattered light acquired at the forward scattering angle; and The intensity signal of the second scattered light acquired at the side scattering angle; The step of determining the particle size distribution characteristic parameter based on the scattered light intensity signals at at least two different scattering angles includes: calculating the ratio of the first scattered light intensity signal to the second scattered light intensity signal to obtain the scattering angle intensity ratio; and performing sliding window filtering on the scattering angle intensity ratio to obtain the particle size distribution characteristic parameter.

3. The method as described in claim 2, characterized in that, Also includes: Based on the pre-stored scattering simulation database, the particle size distribution characteristic parameters are mapped to an equivalent median particle size estimate through table lookup interpolation. The equivalent median particle size estimate is used as one of the input parameters of the material-particle size calibration mapping information. The material optical property parameters determined by the processing technology identification information include: extracting the complex refractive index parameter of the corresponding metal dust from the pre-stored material optical property database based on the processing technology identification information. The complex refractive index parameter includes the real part of the refractive index reflecting the scattering intensity and the imaginary part of the refractive index reflecting the absorption characteristics.

4. The method as described in claim 1, characterized in that, The determination of the basic calibration coefficients using material-particle size calibration mapping information includes one of the following methods: querying a pre-established material-particle size multidimensional calibration coefficient lookup table and obtaining the basic calibration coefficients through interpolation; or using a multivariate regression model that includes the particle size distribution characteristic parameters and the material optical property parameters to calculate the basic calibration coefficients, wherein the multivariate regression model includes a cross-product term of the particle size distribution characteristic parameters and the material optical property parameters, and the cross-product term is used to characterize the coupling effect of particle size distribution and material optical properties on the basic calibration coefficients.

5. The method as described in claim 1, characterized in that, The step of determining the humidity compensation factor based on the environmental humidity information and the particle size distribution characteristic parameters includes: The humidity compensation factor is determined using a humidity-particle size coupling compensation model based on the difference between the current relative humidity and the reference humidity, and the ratio of the reference particle size to the current equivalent median particle size. The humidity-particle size coupling compensation model includes the product of a nonlinear power function of the difference between the current ambient relative humidity and the reference humidity and a power function of the ratio of the reference particle size to the current equivalent median particle size. The step of determining the adaptive calibration coefficient based on the basic calibration coefficient and the humidity compensation factor includes: When the current relative humidity is higher than the reference humidity, the basic calibration coefficient is divided by the humidity compensation factor to obtain the adaptive calibration coefficient; When the relative humidity of the current environment is not higher than the reference humidity, the basic calibration coefficient is used as the adaptive calibration coefficient.

6. The method as described in claim 2, characterized in that, The step of determining the dust mass concentration based on the adaptive calibration parameters and the dust scattering signal includes: multiplying the adaptive calibration coefficient by the digitally sampled value of the second scattered light intensity signal to obtain the dust mass concentration; and further includes: outputting the dust mass concentration to the dust removal system controller so that the dust removal system controller can adjust the dust removal operation parameters according to the dust mass concentration.

7. The method as described in claim 1, characterized in that, Also includes: Construct a feature vector corresponding to the current working condition, the feature vector including the particle size distribution feature parameter and the material optical property parameter; calculate the feature space distance between the feature vector and the nearest neighbor working condition point in the calibration database; if the feature space distance exceeds a preset distance threshold, reduce the measurement confidence flag and generate calibration supplementary alarm information; if the measurement confidence flag is lower than the preset confidence threshold, increase the sampling frequency of the dust scattering signal and shorten the window length of the sliding window filter in the calculation of the particle size distribution feature parameter.

8. The method as described in claim 2, characterized in that, Also includes: The rate of change of the particle size distribution characteristic parameter within a continuous time window is monitored. If the rate of change exceeds a preset rate of change threshold, a fast recalibration process is initiated. The fast recalibration process includes shortening the window length of the sliding window filtering process to accelerate the update response speed of the adaptive calibration coefficients. When a change in the processing technology identification information is detected, the expected range of the basic calibration coefficients is pre-calculated based on the material optical property parameters corresponding to the changed processing technology identification information and the historical typical particle size range of the processing technology. Before the particle size distribution characteristic parameter converges to a stable value, the median of the expected range is used as the transitional basic calibration coefficient.

9. An adaptive detection device for metal processing dust concentration, characterized in that, include: An adaptive calibration parameter determination module is used to determine the adaptive calibration parameters of dust concentration based on dust scattering signals and processing condition information; a dust mass concentration determination module is used to determine the dust mass concentration based on the adaptive calibration parameters and the dust scattering signals. The dust scattering signal includes scattered light intensity signals collected at at least two different scattering angles; the processing condition information includes environmental humidity information and processing technology identification information; the adaptive calibration parameters include adaptive calibration coefficients; and the adaptive calibration parameter determination module includes: The particle size feature extraction submodule is used to determine the particle size distribution feature parameters based on the scattered light intensity signals at at least two different scattering angles. The basic calibration coefficient determination submodule is used to determine the basic calibration coefficients by utilizing the material-particle size calibration mapping information, based on the particle size distribution characteristic parameters and the material optical property parameters determined by the processing technology identification information. The humidity compensation submodule is used to determine the humidity compensation factor based on the environmental humidity information and the particle size distribution characteristic parameters, and to determine the adaptive calibration factor based on the basic calibration coefficient and the humidity compensation factor. The material-particle size calibration mapping information includes the mapping relationship between the particle size distribution characteristic parameters and the material optical property parameters and the basic calibration coefficients.

10. An electronic device, characterized in that, include: Memory; And a processor coupled to the memory, the processor being configured to perform the method as described in any one of claims 1 to 8 based on instructions stored in the memory.

Images