Multi-spectral dynamic sensing laser welding mask self-adaptive adjusting method and multi-spectral dynamic sensing laser welding mask self-adaptive adjusting system

Abstract

The invention provides a multi-spectral dynamic sensing laser welding mask self-adaptive adjusting method and system, and relates to the technical field of spectral analysis, and the method comprises the steps that a plurality of spectral sensing units are arranged on a laser welding mask; collecting a plurality of groups of current spectrum sensing data of the plurality of spectrum sensing units; collecting a current head posture to construct a main view area; screening main view spectrum sensing data of the main view area from the multiple groups of spectrum sensing data; and inputting the main view spectrum sensing data into the filtering adaptive adjustment model, and outputting filtering adjustment parameters to control the adjustable liquid crystal filter. According to the invention, the technical problem that a filtering adjustment mechanism only depends on a single photosensitive device and cannot identify the wavelength of a light source in the prior art can be solved, and the technical targets of accurate identification and adaptive filtering adjustment of a high-risk laser band in the main view direction of a wearer are achieved. The technical effects of improving the intelligent level of laser operation protection and enhancing the visual safety and the filtering adjustment precision are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of spectral analysis, and in particular to a laser welding mask adaptive adjustment method and system based on multi-spectrum dynamic perception. BACKGROUND

[0002] In industrial laser application scenarios such as welding, cutting, and engraving, the face protection of the operator depends on the laser welding mask, and one of the core functions is to effectively filter harmful spectral bands to prevent eye and skin damage.

[0003] At present, although some existing automatic light-changing masks can automatically adjust the light transmittance according to the light intensity, the adjustment mechanism is mainly based on the response of a single photosensitive device to brightness changes, lacks accurate identification of wavelength information, and cannot distinguish different types of light sources, which may cause false triggering under high-intensity ordinary white light irradiation, and slow or ineffective response to low-intensity high-risk laser bands. Furthermore, these automatic light-changing systems generally do not have directional recognition capability, that is, they are not accurate in determining the laser risk area in the actual gaze direction of the wearer, and are prone to problems such as mispositioning and response lag, which cannot meet the requirements of high safety and high precision for operation. At the same time, existing devices generally lack a feedback mechanism for filtering effect, and cannot issue warnings or optimize parameters when filtering is ineffective or the effect is reduced, making it difficult to form a closed-loop control system.

[0004] In summary, the existing technology has the technical problem that the filtering adjustment mechanism only relies on a single photosensitive device and cannot identify the wavelength of the light source, lacks gaze direction judgment and filtering effect feedback mechanism, resulting in a lack of pertinence and real-time protection for laser bands, which further affects the visual safety and operation accuracy of the wearer in laser operation. SUMMARY

[0005] The purpose of the present application is to provide a laser welding mask adaptive adjustment method and system based on multi-spectrum dynamic perception, to solve the technical problem in the prior art that the filtering adjustment mechanism only relies on a single photosensitive device and cannot identify the wavelength of the light source, lacks gaze direction judgment and filtering effect feedback mechanism, resulting in a lack of pertinence and real-time protection for laser bands, which further affects the visual safety and operation accuracy of the wearer in laser operation.

[0006] In view of the above problems, the present application provides a laser welding mask adaptive adjustment method and system based on multi-spectrum dynamic perception.

[0007] In a first aspect, the application provides a multi-spectrum dynamic perception laser welding mask adaptive adjustment method, which is implemented by a multi-spectrum dynamic perception laser welding mask adaptive adjustment system, and includes: arranging a plurality of spectrum perception units on a laser welding mask, wherein each spectrum perception unit is provided with a micro spectrum sensor, and each spectrum perception unit corresponds to a region on the laser welding mask; collecting a plurality of groups of current spectrum perception data of the plurality of spectrum perception units, wherein each group of spectrum perception data includes an incident wavelength and an illumination intensity; collecting a current head posture by an IMU module built-in to construct a main view region; screening main view spectrum perception data of the main view region from the plurality of groups of spectrum perception data; inputting the main view spectrum perception data into a filter adaptive adjustment model to output filter adjustment parameters to control an adjustable liquid crystal filter.

[0008] Preferably, the multi-spectrum dynamic perception laser welding mask adaptive adjustment method further includes: collecting a current head posture including a pitch angle, a yaw angle and a roll angle according to the IMU module; defining a coordinate system of the laser welding mask, and converting the pitch angle, the yaw angle and the roll angle into a main view unit vector under the coordinate system; projecting the main view unit vector to obtain a main view region by a field of view region mapping relationship, wherein the main view region includes one or more regions.

[0009] Preferably, the multi-spectrum dynamic perception laser welding mask adaptive adjustment method further includes: obtaining a plurality of regions of the laser welding mask, and extracting a center unit vector of each region; defining a field of view region mapping relationship based on a mapping relationship between each region and the center unit vector of each region; calculating an included angle between the main view unit vector and the center unit vector of each region by using the field of view region mapping relationship to output an included angle value set; selecting a region with an included angle greater than a preset included angle threshold value from the included angle value set as a main view region output.

[0010] Preferably, the multi-spectrum dynamic perception laser welding mask adaptive adjustment method further includes: fusing the main view spectrum perception data to output fused spectrum perception data; identifying a risk wavelength of the main view region according to a preset laser hazard wavelength feature library based on the fused spectrum perception data; obtaining a risk level of the risk wavelength of the main view region according to a wavelength type and an illumination intensity of each wavelength in the risk wavelength of the main view region; extracting a dominant wavelength of the main view region according to a size of the risk level, and obtaining a filter adjustment parameter corresponding to the dominant wavelength.

[0011] Preferably, the multi-spectrum dynamic perception adaptive adjustment method of the laser welding mask further comprises: the light filtering adaptive adjustment model further comprises a dynamic weight network layer; a plurality of main view weights are configured according to the angle values of the main view area; the plurality of main view weights are input as dynamic input data into the dynamic weight network layer, the risk level output by the light filtering adaptive adjustment model is weighted calculated, and the dominant wave band of the main view area is re-extracted according to the updated risk level.

[0012] Preferably, the multi-spectrum dynamic perception adaptive adjustment method of the laser welding mask further comprises: the light filtering adjustment parameters output by the light filtering adaptive adjustment model include a light filtering center wavelength, a light filtering width, a light filtering DIN intensity, and a response duration.

[0013] Preferably, the multi-spectrum dynamic perception adaptive adjustment method of the laser welding mask further comprises: outputting the light filtering adjustment parameters to control the adjustable liquid crystal filter, the adjustable liquid crystal filter comprising at least one adjustable liquid crystal layer for transmittance adjustment according to the light filtering adjustment parameters; the adjustable liquid crystal layer comprises a plurality of electrically controlled pixel areas, each electrically controlled pixel area being used for responding to adjustment of a corresponding wavelength range.

[0014] Preferably, the multi-spectrum dynamic perception adaptive adjustment method of the laser welding mask further comprises: collecting light filtered spectrum perception data according to the plurality of spectrum perception units; performing light filtering absorption effect evaluation on the light filtered spectrum perception data to obtain an absorption effect index; if the absorption effect index is less than a preset threshold, updating the light filtering adjustment parameters to obtain optimized light filtering adjustment parameters.

[0015] Preferably, the multi-spectrum dynamic perception adaptive adjustment method of the laser welding mask further comprises: establishing an absorption correlation degree between the light filtering adjustment parameters and the non-dominant wave band; obtaining a light filtering adjustment parameter interval corresponding to the dominant wave band; and optimizing the light filtering adjustment parameters in the light filtering adjustment parameter interval to maximize the sum of the absorption correlation degrees, and outputting the optimized light filtering adjustment parameters.

[0016] Secondly, this application also provides a multispectral dynamic sensing adaptive adjustment system for laser welding masks, used to execute the multispectral dynamic sensing adaptive adjustment method for laser welding masks as described in the first aspect, comprising: a spectral sensing unit arrangement module for arranging multiple spectral sensing units on the laser welding mask, wherein each spectral sensing unit is equipped with a miniature spectral sensor, and each spectral sensing unit corresponds to a region on the laser welding mask; a spectral sensing data acquisition module for acquiring multiple sets of current spectral sensing data from the multiple spectral sensing units, each set of spectral sensing data including incident wavelength and illumination intensity; a main viewing region construction module for constructing a main viewing region by acquiring the current head posture through a built-in IMU module; a main viewing spectral sensing data filtering module for filtering the main viewing spectral sensing data of the main viewing region from the multiple sets of spectral sensing data; and a filter adjustment parameter output module for inputting the main viewing spectral sensing data into a filter adaptive adjustment model and outputting filter adjustment parameters to control an adjustable liquid crystal filter.

[0017] The technical solution provided in this application has at least the following technical effects or advantages: by achieving the technical goal of accurately identifying and adaptively adjusting the high-risk laser band in the wearer's main viewing direction, it achieves the technical effects of improving the level of intelligent protection for laser operations, enhancing visual safety and the accuracy of filter adjustment.

[0018] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the adaptive adjustment method for laser welding masks based on multispectral dynamic sensing, as described in this application.

[0021] Figure 2 This is a schematic diagram of the structure of the multispectral dynamic sensing adaptive adjustment system for laser welding masks in this application.

[0022] Figure labeling: 11 Spectral sensing unit arrangement module, 12 Spectral sensing data acquisition module, 13 Main view area construction module, 14 Main view spectral sensing data filtering module, 15 Filter adjustment parameter output module. Detailed Implementation

[0023] This application provides a multispectral dynamic sensing adaptive adjustment method and system for laser welding masks. It addresses the technical problems in existing technologies where the filter adjustment mechanism relies solely on a single photosensitive device and cannot identify the light source wavelength. Furthermore, the lack of gaze direction judgment and filter effect feedback mechanisms results in a lack of targeted and real-time protection against laser wavelengths, further impacting the wearer's visual safety and operational accuracy during laser operations. The method achieves the technical goal of accurately identifying and adaptively adjusting high-risk laser wavelengths in the wearer's primary gaze direction, thereby improving the intelligence level of laser operation protection and enhancing visual safety and filter adjustment accuracy.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to the appendix. Figure 1 This application provides a multispectral dynamic sensing adaptive adjustment method for laser welding masks, applied to a multispectral dynamic sensing adaptive adjustment system for laser welding masks, specifically including the following steps: S1: Multiple spectral sensing units are arranged on the laser welding mask, wherein each spectral sensing unit is equipped with a miniature spectral sensor, and each spectral sensing unit corresponds to a region on the laser welding mask.

[0026] Specifically, multiple spectral sensing units are uniformly arranged on the surface of the laser welding mask. These units are distributed at different locations within the mask to sense the light environment in different areas. These spectral sensing units form the basic module capable of capturing the wavelength and intensity information of incident light. Each unit contains a miniature spectral sensor, which can be embedded within the laser welding mask structure. This sensor possesses high sensitivity and resolution, enabling it to capture the hazardous spectra caused by high-energy laser reflection or scattering in complex welding environments. The miniature spectral sensors typically cover the wavelength range from visible to near-infrared, with an accuracy down to 1 nanometer, facilitating precise identification of specific laser wavelengths. The entire laser welding mask is divided into several sensing areas, with each spectral sensing unit corresponding to one of these areas, allowing for comprehensive and detailed sensing of the spectral information at different locations within the welding area.

[0027] S2: Collect multiple sets of current spectral sensing data from the multiple spectral sensing units. Each set of spectral sensing data includes the incident wavelength and the light intensity.

[0028] Specifically, the system collects current data from multiple spectral sensing units and reads multiple sets of spectral sensing data detected by these units, which are distributed across different areas of the laser welding mask, in real time. Each set of spectral sensing data includes the incident wavelength and the light intensity. The incident wavelength refers to the wavelength unit of light. The light intensity is a parameter describing the strength of light energy within a certain wavelength range and is used to assess the degree of stimulation of the photosensitive element or the human eye by the light.

[0029] S3: Constructs the main view area by collecting the current head posture through the built-in IMU module.

[0030] Specifically, an inertial measurement unit (IMU) installed inside the laser welding mask monitors the wearer's head position and orientation in real time. The IMU module includes a gyroscope, accelerometer, and magnetometer to measure pitch (the angle of head tilting up and down), yaw (the angle of head turning left and right), and roll (the angle of head rotation around its front-to-back axis), thus comprehensively reflecting the wearer's orientation and posture. This angular information is processed and converted into a spatial direction vector to determine the wearer's current primary gaze direction. Based on this direction, one or more areas are designated as the primary viewing area within the laser welding mask's multiple sensing regions—the area the wearer is currently focusing on observing.

[0031] S4: Select the main view spectral sensing data of the main view area from the multiple sets of spectral sensing data.

[0032] Specifically, after collecting spectral data from all areas, the main view spectral sensing data for the main viewing area is selected from multiple sets of spectral sensing data. Multiple sets of spectral sensing data refer to the information collection containing different wavelengths and corresponding light intensities acquired by each spectral sensing unit, originating from different directions of the lighting environment during the welding process. The main view spectral sensing data is data selected from these multiple sets of spectral sensing data based on the sensing unit corresponding to the previously identified main view area, focusing on data from a specific direction. Prioritizing the extraction of the main view spectral sensing data during data processing helps to focus on the most critical field of view in the welding operation.

[0033] S5: Input the main viewing spectrum sensing data into the filter adaptive adjustment model, and output the filter adjustment parameters to control the adjustable liquid crystal filter.

[0034] Specifically, the main visual spectral sensing data is input into the adaptive filter adjustment model. This model is a functional network built based on artificial intelligence or rule-based algorithms. It determines potentially dangerous wavelengths and their intensities based on the input spectral data and automatically generates corresponding filter control strategies. The adaptive filter adjustment model outputs filter adjustment parameters based on the spectral analysis results to control the tunable liquid crystal filter, which drives the tunable liquid crystal filter in the mask to adjust its transmittance within the required protection wavelength range. The tunable liquid crystal filter is an optical element that can change its optical properties according to external electronic control signals, featuring fast response and wavelength selectivity. By adjusting the pixel voltage within the liquid crystal layer, the tunable liquid crystal filter can enhance absorption or blocking effects within a specific wavelength range, thereby achieving dynamic protection.

[0035] Furthermore, this application also includes: acquiring the current head attitude, including pitch angle, yaw angle and roll angle, according to the IMU module; defining the coordinate system of the laser welding mask, and converting the pitch angle, yaw angle and roll angle into a main view unit vector in the coordinate system; projecting the main view unit vector through the field of view mapping relationship to obtain the main view region, wherein the main view region includes one or more regions.

[0036] Specifically, the current head posture is acquired by the IMU module, meaning that the wearer's head movement is acquired in real time using the inertial measurement unit inside the laser-welded mask. The IMU module is a composite sensor consisting of an accelerometer, a gyroscope, and sometimes a magnetometer, capable of capturing the rotation and acceleration information of an object in three-dimensional space. The current head posture includes pitch, yaw, and roll angles. Pitch angle represents the angle at which the head rises or falls, yaw angle reflects the angle at which the head turns left or right, and roll angle describes the angle at which the head tilts left or right. Pitch, yaw, and roll angles constitute a complete description of the current head posture.

[0037] After acquiring the current head posture, a unified analysis needs to be performed within the coordinate system of the laser-welded mask. The mask coordinate system is defined to provide a reference frame for the posture angles, facilitating the calculation of the wearer's actual facing direction in three-dimensional space. The acquired pitch, yaw, and roll angles are converted into a primary view unit vector, a spatial vector of length 1, used to accurately represent the head's orientation. The primary view unit vector points in the wearer's forward-looking direction and is a key parameter for subsequent identification of the primary view area.

[0038] To map the main viewing unit vector onto the actual area defined by the laser welding mask, a field-of-view mapping relationship was further introduced. By matching spatial directions with specific sensing areas on the mask and assigning a central unit vector to each sensing area, the angle between the main viewing vector and each area can be calculated. Finally, one or more areas meeting the criteria are selected as the main viewing area. The main viewing area is most likely to appear directly in front of the wearer and is the key monitoring area for laser protection.

[0039] Furthermore, this application also includes: obtaining multiple regions of the laser welding mask and extracting the central unit vector of each region; defining a field-of-view region mapping relationship based on the mapping relationship between each region and the central unit vector of each region; using the field-of-view region mapping relationship, calculating the angle between the main viewing unit vector and the central unit vector of each region, and outputting a set of angle values; and selecting regions in the set of angle values ​​that are greater than a preset angle threshold as the main viewing region for output.

[0040] Specifically, multiple areas of the laser welding mask are acquired and divided according to the field of view coverage, with each area representing a portion of the wearer's field of vision. The division can be a regular grid or a functional division based on common observation directions encountered in actual welding operations. For example, the mask could be divided into 16 equal-area areas, each capable of independently acquiring spectral data and undergoing computational control.

[0041] Next, extracting the central unit vector for each region involves calculating the spatial direction corresponding to the center point of each defined region. A unit vector is a vector representing direction in three-dimensional space, with a length of 1 and a direction consistent with the target. Based on the mapping relationship between each region and its central unit vector, a field-of-view mapping relationship is defined, associating each specific region with its spatial direction to form a set of standard mapping rules. This mapping relationship constitutes the internal spatial perception model, used for subsequent identification of the relationship between the main viewing direction and the regions. The field-of-view mapping relationship not only includes matching region numbers with directions but can also be extended to include spatial logic such as region overlap and boundary fusion to improve the continuity and accuracy of identification.

[0042] Then, using the field-of-view mapping relationship, the angle between the dominant unit vector and the central unit vector of each region is calculated, and a set of angle values ​​is output. Using the formula for calculating the angle between spatial vectors, the dominant unit vector and the central vector of each region are compared one by one. The smaller the angle value, the closer the direction of that region is to the wearer's actual gaze direction.

[0043] Finally, the regions with angle values ​​greater than a preset angle threshold are selected from the set of angle values ​​and output as the main viewing area. The angle threshold is a range set by those skilled in the art based on experience or accuracy requirements; for example, a range of less than 20 degrees is considered to be within the main viewing range. By filtering the angle values ​​of all regions and retaining only those regions within the threshold range as the main viewing area, the system can automatically identify the mask area of ​​interest based on the wearer's head posture.

[0044] Furthermore, this application also includes: fusing the main view spectral sensing data to output fused spectral sensing data; identifying risk wavelengths in the fused spectral sensing data according to a preset laser hazard band feature library to output risk bands in the main view region; obtaining the risk level of the risk bands in the main view region based on the band type and illumination intensity of each band; and extracting the dominant band of the main view region based on the magnitude of the risk level to obtain the filter adjustment parameters corresponding to the dominant band.

[0045] Specifically, data collected by different spectral sensing units in multiple main viewing areas are aggregated, averaged, or weighted to form a unified spectral description, outputting fused spectral sensing data. The main viewing spectral sensing data includes light intensity information for specific wavelengths in each direction. Fusion improves the stability and representativeness of the data, reducing interference caused by individual sensor errors or localized reflections.

[0046] Next, the fused spectral sensing data is analyzed for risk wavelengths based on a pre-defined laser hazard band feature library. The risk bands for the main visual area are then output, and the fused spectral data is compared with a pre-built laser hazard database. The pre-defined laser hazard band feature library contains dangerous wavelength ranges that different types of lasers may cause damage to the human eye or visual organs. For example, near-infrared lasers with wavelengths between 780 nm and 1400 nm belong to the retinal hazard band. The identification process involves finding components in the fused spectral data that fall within the dangerous band range, thereby pinpointing the spectral segments in the current main visual area that may pose a visual risk.

[0047] Subsequently, based on the band type and light intensity of each band in the main visual area risk band, the risk level of the main visual area risk band is obtained, and the potential degree of damage is assessed according to the physical properties and irradiation intensity of different bands. The band type determines the location of the damage (such as cornea, lens, or retina), and the higher the light intensity, the greater the risk of thermal or photochemical damage to ocular tissues.

[0048] Finally, based on the risk level, the dominant wavelength band of the main viewing area is extracted, and the corresponding filter adjustment parameters are obtained. Among multiple risk wavelength bands, the band with the highest risk level is selected as the dominant target, and the filter adjustment method is determined. The dominant wavelength band is the wavelength range that must be filtered in the current key area. Based on its characteristics, parameters such as the cutoff frequency and attenuation intensity of the liquid crystal filter are set to effectively suppress light of this wavelength band from entering the field of view. Table 1 shows a partial record of the most recent risk level acquisition.

[0049] Table 1: Partial Records of the Most Recent Risk Level Acquisition Waveband λi I_i (intensity of light) T_i (type of waveband) R_i (risk level) b1 808nm 0.85 1.1 7 b2 532nm 0.92 1.0 6 b3 1064nm 0.75 1.2 8 If we take the weighting coefficients α=0.5, β=0.3, and γ=0.2, then: C(b1)=0.5×7+0.3×0.85+0.2×1.1=3.5+0.255+0.22=3.975; C(b2)=0.5×6+0.3×0.92+0.2×1.0=3+0.276+0.2=3.476; C(b3)=0.5×8+0.3×0.75+0.2×1.2=4+0.225+0.24=4.465. Therefore, the dominant band is selected as b3 (1064nm band).

[0050] Furthermore, this application also includes: the adaptive filter adjustment model further includes a dynamic weight network layer; multiple main view weights are configured according to the included angle value of the main view region; the multiple main view weights are input as dynamic input data into the dynamic weight network layer to perform weighted calculation on the risk level output by the adaptive filter adjustment model, and the dominant band of the main view region is re-extracted according to the updated risk level.

[0051] Specifically, the adaptive filter adjustment model also includes a dynamic weight network layer, which can dynamically adjust the importance of each region based on different input conditions. The dynamic weight network layer consists of a group of neurons that can process multiple sets of input data and calculate weight factors. The weights can continuously change with parameters such as head pose and environmental variations. Through the dynamic weight network layer, more refined and real-time adjustment strategies can be implemented in the spectral risk assessment process.

[0052] Next, multiple primary view weights are assigned based on the angle value of the primary view area. This means that the visual attention level of each area is evaluated based on the angle between the user's gaze direction and the center direction of each area. The smaller the angle value, the closer the area is to the user's gaze direction, and the higher its visual attention level, thus it is assigned a higher primary view weight; while areas with larger angles may be in the peripheral field of vision, and their weights are relatively lower.

[0053] Then, multiple gaze weights are input as dynamic input data into the dynamic weight network layer to calculate the risk level output by the adaptive filter adjustment model. That is, the gaze-related weights are used as input features for further processing by the dynamic weight network layer, and a new weighted level result is output by combining the original risk level. This can avoid the system from over-responding due to a single high-intensity light source at the visual edge, thereby improving the overall rationality and stability of the filter strategy.

[0054] Finally, the dominant bands of the main visual area are re-extracted based on the updated risk level. Then, based on the weighted risk results, the risk bands requiring the highest priority are reassessed. The dominant band refers to the light wave in the current spectrum that has the greatest impact on visual safety, determining the adjustment direction and intensity of subsequent filters. Because the weighting incorporates visual attention, even if multiple bands have equal light intensity, their priority may change depending on the user's gaze direction.

[0055] Furthermore, this application also includes: the filter adjustment parameters output by the filter adaptive adjustment model include the filter center wavelength, filter width, filter DIN intensity, and response duration.

[0056] Specifically, the filter adjustment parameters output by the adaptive filter adjustment model include the filter center wavelength, filter width, filter DIN intensity, and response duration. The filter center wavelength is the working center position of the liquid crystal filter or other optical element, set according to the identified dominant risk band; it is the position where light is primarily reduced. The filter width refers to the suppression bandwidth covered around the center wavelength, meaning it can simultaneously filter light within a range above and below the center wavelength. The filter DIN intensity refers to the filter intensity level expressed according to the DIN (Deutsches Institut für Normung) standard, used to measure the filter's ability to block different laser intensities. A higher DIN value indicates a stronger filter intensity and can block higher light power. The response duration indicates the length of time the filter state is maintained after adjustment, used to control the timeliness of the filter behavior and avoid prolonged visibility degradation due to short-term interference.

[0057] Furthermore, this application also includes: outputting filter adjustment parameters to control an adjustable liquid crystal filter, the adjustable liquid crystal filter including at least one adjustable liquid crystal layer for adjusting transmittance according to the filter adjustment parameters; the adjustable liquid crystal layer including a plurality of electrically controlled pixel regions, each electrically controlled pixel region being used to respond to adjustment of a corresponding wavelength range.

[0058] Specifically, output filter adjustment parameters are used to control the tunable liquid crystal filter. Parameters such as center wavelength, filter width, DIN intensity, and response duration are sent as commands to the tunable liquid crystal filter, causing it to selectively filter external light according to these parameters. A tunable liquid crystal filter is an optical device that can dynamically change its light transmittance based on an input electrical signal. It alters the arrangement of liquid crystal molecules through electronic control, thereby absorbing or reflecting light of different wavelengths to varying degrees.

[0059] Next, the tunable liquid crystal filter includes at least one tunable liquid crystal layer, which is an optical filter film with intelligent adjustment function. The tunable liquid crystal layer is composed of liquid crystal material, and its molecular arrangement changes under the action of an external electric field, resulting in changes in its polarization characteristics and transmittance of incident light. According to the filter adjustment parameters, the tunable liquid crystal layer can switch the light transmission state within a millisecond response time.

[0060] Next, the adjustable liquid crystal layer includes multiple electrically controlled pixel regions. Each electrically controlled pixel region can independently adjust its transmittance based on the electrical signals it receives, achieving regionalized filtering control and improving accuracy. For example, when a laser threat is detected in the main viewing area but there is no risk in the side areas, only the electrically controlled pixel region corresponding to the front will activate its light-blocking function, thereby improving overall visual efficiency.

[0061] Furthermore, this application also includes: acquiring filtered spectral sensing data based on the plurality of spectral sensing units; evaluating the filtering absorption effect of the filtered spectral sensing data to obtain an absorption effect index; and updating the filtering adjustment parameters if the absorption effect index is less than a preset threshold to obtain optimized filtering adjustment parameters.

[0062] Specifically, after filtering, spectral sensing units installed in different areas of the laser welding mask will operate again, collecting spectral sensing data after filtering from multiple spectral sensing units to gather light information processed by the tunable liquid crystal filter. The spectral sensing units rely on built-in miniature spectral sensors to record the spectral characteristics after filtering, including the transmitted light intensity at various wavelengths, as a basic data source for evaluating the filtering effect.

[0063] Next, the absorption effect of the filtered spectral sensing data is evaluated to obtain absorption effect indices, and the intensity attenuation of each band before and after filtering is calculated. The absorption effect index can be measured by the ratio of light intensity before and after filtering or the attenuation ratio, reflecting the blocking ability of the filtering system at a specific wavelength. The higher the value, the more ideal the filtering effect.

[0064] Subsequently, if the absorption effect index is lower than the preset threshold, the filter adjustment parameters are updated, and the absorption effect is compared with the set safety standard. For example, if the preset absorption rate threshold is 85%, but only 76% is reached in a certain wavelength band, it indicates that the current filter adjustment parameter settings are insufficient to cope with the risks in that wavelength band. Therefore, parameters such as center wavelength, filter width, and DIN intensity need to be readjusted to enhance the filtering capability. The automatic adjustment mechanism ensures the system's continuous response and stable protection capability in variable laser environments.

[0065] Finally, the optimized filter adjustment parameters are obtained, and new control values ​​are output after updating the parameter model to further adjust the response behavior of the liquid crystal filter. The optimized parameters may include widening the filter width to cover a wider wavelength range, increasing the DIN intensity to improve the darkening effect, or extending the response duration to cover the duration of laser interference.

[0066] Furthermore, this application also includes: establishing the absorption correlation between the filter adjustment parameters and the non-dominant wavelength bands; obtaining the filter adjustment parameter range corresponding to the dominant wavelength bands; optimizing the filter adjustment parameters in the filter adjustment parameter range with the goal of maximizing the sum of the absorption correlations, and outputting the optimized filter adjustment parameters.

[0067] Specifically, the absorption correlation between filter adjustment parameters and non-dominant wavelength bands is established. After identifying the dominant wavelength band, the masking or suppression effects of the filter adjustment parameters used in that dominant wavelength band on other non-dominant wavelength bands are further analyzed. The absorption correlation is an index representing the correlation between the adjustment parameters and the absorption effect on non-dominant wavelength bands, calculated through a spectral response model. For example, when the dominant wavelength band is 1064 nm, using a specific filter center wavelength and filter width may also produce a certain filtering effect in the nearby 900 nm to 1000 nm range. The "indirect benefit" relationship between wavelength bands is expressed by the absorption correlation; the higher the value, the better the absorption capability of a filter parameter scheme for more wavelength bands simultaneously.

[0068] Next, the filter adjustment parameter range corresponding to the dominant wavelength band is obtained. For the determined dominant wavelength band, a reasonable range of variation for the filter parameters is set, including upper and lower limits for multiple dimensions such as center wavelength, filter width, DIN intensity, and response duration. For example, for a dominant wavelength band of 1064 nm, the allowable range for filter center wavelength variation might be 1060 to 1070 nm, filter width 30 to 50 nm, DIN intensity 3 to 6, and response duration 50 to 300 milliseconds. Parameter range constraints provide boundary conditions for optimized calculations, preventing the output of unfeasible extreme filter schemes.

[0069] Finally, the filter adjustment parameters are optimized within the specified range with the goal of maximizing the sum of absorption correlations. Within the allowable parameter variation range, an optimization algorithm is used to find an optimal parameter combination that achieves satisfactory absorption in the dominant wavelength band while also generating high absorption correlations in as many non-dominant wavelength bands as possible. The optimization objective is a multi-objective function, with the core goal of improving the overall protection effect of the filter scheme. After optimization, updated filter center wavelength and filter width are output. For example, optimizing from the original settings of 1064 nm center wavelength and 30 nm width to 1067 nm center wavelength and 45 nm width ensures effective blocking of 1064 nm while enhancing absorption in non-dominant wavelength bands near 1080 nm, thus improving the overall filter quality.

[0070] In summary, the multispectral dynamic sensing adaptive adjustment method for laser welding masks provided in this application has the following technical effects: by achieving the technical goal of accurately identifying and adaptively adjusting the high-risk laser bands in the wearer's main viewing direction, it achieves the technical effects of improving the intelligent level of laser operation protection, enhancing visual safety, and improving the accuracy of filter adjustment.

[0071] Example 2: Based on the same inventive concept as the multispectral dynamic sensing adaptive adjustment method for laser welding masks in the foregoing examples, this application also provides a multispectral dynamic sensing adaptive adjustment system for laser welding masks. Please refer to the appendix. Figure 2The system includes: a spectral sensing unit arrangement module 11, used to arrange multiple spectral sensing units on the laser welding mask, wherein each spectral sensing unit is equipped with a miniature spectral sensor and each spectral sensing unit corresponds to a region on the laser welding mask; a spectral sensing data acquisition module 12, used to acquire multiple sets of current spectral sensing data from the multiple spectral sensing units, each set of spectral sensing data including incident wavelength and light intensity; a main viewing region construction module 13, used to construct the main viewing region by acquiring the current head posture through the built-in IMU module; a main viewing spectral sensing data filtering module 14, used to filter the main viewing spectral sensing data of the main viewing region from the multiple sets of spectral sensing data; and a filter adjustment parameter output module 15, used to input the main viewing spectral sensing data into the filter adaptive adjustment model and output filter adjustment parameters to control the adjustable liquid crystal filter.

[0072] Furthermore, the multispectral dynamic sensing adaptive adjustment system for the laser welding mask is also used to: acquire the current head posture, including pitch angle, yaw angle, and roll angle, according to the IMU module; define the coordinate system of the laser welding mask, and convert the pitch angle, yaw angle, and roll angle into a main view unit vector in the coordinate system; project the main view unit vector through the field of view mapping relationship to obtain the main view region, wherein the main view region includes one or more regions.

[0073] Furthermore, the multispectral dynamic sensing adaptive adjustment system for the laser welding mask is also used to: acquire multiple regions of the laser welding mask and extract the central unit vector of each region; define a field-of-view mapping relationship based on the mapping relationship between each region and the central unit vector of each region; calculate the angle between the main viewing unit vector and the central unit vector of each region using the field-of-view mapping relationship, and output a set of angle values; and select regions in the set of angle values ​​that are greater than a preset angle threshold as the main viewing region for output.

[0074] Furthermore, the multispectral dynamic sensing laser welding mask adaptive adjustment system is also used for: fusing the main view spectral sensing data and outputting fused spectral sensing data; identifying risk wavelengths in the fused spectral sensing data according to a preset laser hazard band feature library and outputting risk bands in the main view area; obtaining the risk level of the risk bands in the main view area based on the band type and illumination intensity of each band; and extracting the dominant band in the main view area based on the magnitude of the risk level and obtaining the filter adjustment parameters corresponding to the dominant band.

[0075] Furthermore, the multispectral dynamic sensing laser welding mask adaptive adjustment system is also used for: the filter adaptive adjustment model further includes a dynamic weight network layer; configuring multiple main view weights according to the included angle value of the main view area; inputting the multiple main view weights as dynamic input data into the dynamic weight network layer to perform weighted calculation on the risk level output by the filter adaptive adjustment model, and re-extracting the dominant band of the main view area according to the updated risk level.

[0076] Furthermore, the multispectral dynamic sensing laser welding mask adaptive adjustment system is also used for: the filter adjustment parameters output by the filter adaptive adjustment model include the filter center wavelength, filter width, filter DIN intensity, and response duration.

[0077] Furthermore, the multispectral dynamic sensing laser welding mask adaptive adjustment system is also used to: output filter adjustment parameters to control an adjustable liquid crystal filter, the adjustable liquid crystal filter including at least one adjustable liquid crystal layer for adjusting transmittance according to the filter adjustment parameters; the adjustable liquid crystal layer including multiple electrically controlled pixel regions, each electrically controlled pixel region for responding to adjustment of a corresponding wavelength range.

[0078] Furthermore, the multispectral dynamic sensing laser welding mask adaptive adjustment system is also used for: collecting filtered spectral sensing data based on the multiple spectral sensing units; evaluating the filtering absorption effect of the filtered spectral sensing data to obtain an absorption effect index; and updating the filtering adjustment parameters if the absorption effect index is less than a preset threshold to obtain optimized filtering adjustment parameters.

[0079] Furthermore, the multispectral dynamic sensing adaptive adjustment system for laser welding masks is also used to: establish the absorption correlation between the filter adjustment parameters and the non-dominant bands; obtain the filter adjustment parameter range corresponding to the dominant bands; optimize the filter adjustment parameters in the filter adjustment parameter range with the goal of maximizing the sum of the absorption correlations, and output the optimized filter adjustment parameters.

[0080] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The multispectral dynamic sensing laser welding mask adaptive adjustment method and specific examples in the foregoing embodiment one are also applicable to the multispectral dynamic sensing laser welding mask adaptive adjustment system of this embodiment. Through the foregoing detailed description of the multispectral dynamic sensing laser welding mask adaptive adjustment method, those skilled in the art can clearly understand the multispectral dynamic sensing laser welding mask adaptive adjustment system of this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0081] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0082] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A multispectral dynamic sensing adaptive adjustment method for laser welding masks, characterized in that, The method includes: Multiple spectral sensing units are arranged on the laser welding mask, each of which is equipped with a miniature spectral sensor and corresponds to a region on the laser welding mask. Collect multiple sets of current spectral sensing data from the multiple spectral sensing units, each set of spectral sensing data including incident wavelength and light intensity; The main view area is constructed by collecting the current head posture through the built-in IMU module; Filter the main view spectral sensing data of the main view area from the multiple sets of spectral sensing data; The main viewing spectral sensing data is input into the filter adaptive adjustment model, and the filter adjustment parameters are output to control the adjustable liquid crystal filter.

2. The method as described in claim 1, characterized in that, The main view region is constructed by acquiring the current head pose through the built-in IMU module, and the methods include: The current head attitude, including pitch angle, yaw angle, and roll angle, is collected by the IMU module. Define the coordinate system of the laser welding mask, and convert the pitch angle, yaw angle and roll angle into main view unit vectors in the coordinate system; The main view unit vector is projected using the field of view region mapping relationship to obtain the main view region, which includes one or more regions.

3. The method as described in claim 2, characterized in that, The main view unit vector is projected using the view region mapping relationship to obtain the main view region. The method includes: Multiple regions of the laser welding mask are obtained, and the central unit vector of each region is extracted; The field-of-view mapping relationship is defined by the mapping relationship between each region and the central unit vector of each region; Using the field of view mapping relationship, the angle between the main view unit vector and the center unit vector of each region is calculated, and the set of angle values ​​is output. The region with an angle value greater than a preset angle threshold is selected from the set of angle values ​​and output as the main view region.

4. The method as described in claim 3, characterized in that, The method of inputting the main viewing spectral sensing data into the filter adaptive adjustment model and outputting filter adjustment parameters includes: The main view spectral sensing data is fused to output fused spectral sensing data; Based on a preset laser hazard band feature library, the fused spectral sensing data is used to identify risk wavelengths and output the risk bands for the main viewing area. Based on the band type and illumination intensity of each band in the main viewing area risk band, the risk level of the main viewing area risk band is obtained; Based on the magnitude of the risk level, the dominant wavelength of the main viewing area is extracted, and the filter adjustment parameters corresponding to the dominant wavelength are obtained.

5. The method as described in claim 4, characterized in that, The adaptive filter adjustment model also includes a dynamic weighted network layer; Configure multiple main view weights according to the included angle value of the main view area; The multiple main view weights are input as dynamic input data into the dynamic weight network layer to perform weighted calculation on the risk level output by the filter adaptive adjustment model, and the dominant band of the main view region is re-extracted based on the updated risk level.

6. The method as described in claim 4, characterized in that, The filter adjustment parameters output by the adaptive filter adjustment model include the filter center wavelength, filter width, filter DIN intensity, and response duration.

7. The method as described in claim 1, characterized in that, Output filter adjustment parameters to control an adjustable liquid crystal filter, the adjustable liquid crystal filter including at least one adjustable liquid crystal layer for adjusting transmittance according to the filter adjustment parameters; The adjustable liquid crystal layer includes multiple electrically controlled pixel regions, each of which is used to respond to adjustments within a corresponding wavelength range.

8. The method as described in claim 1, characterized in that, After outputting filter adjustment parameters to control the adjustable liquid crystal filter, the method further includes: Filtered spectral sensing data are acquired based on the multiple spectral sensing units; The filtered spectral sensing data is evaluated for the absorption effect of the filter to obtain an absorption effect index. If the absorption effect index is less than a preset threshold, the filter adjustment parameters are updated to obtain optimized filter adjustment parameters.

9. The method as described in claim 8, characterized in that, The method for updating the filter adjustment parameters to obtain optimized filter adjustment parameters includes: Establish the correlation between the filter adjustment parameters and the absorption of the non-dominant wavelength band; Obtain the filter adjustment parameter range corresponding to the dominant wavelength band; The filter adjustment parameters are optimized within the filter adjustment parameter range with the goal of maximizing the sum of the absorption correlations, and the optimized filter adjustment parameters are output.

10. A multispectral dynamic sensing adaptive adjustment system for laser welding masks, characterized in that, The steps for implementing the multispectral dynamic sensing adaptive adjustment method for laser welding masks according to any one of claims 1 to 9 include: The spectral sensing unit arrangement module is used to arrange multiple spectral sensing units on the laser welding mask. Each spectral sensing unit is equipped with a miniature spectral sensor, and each spectral sensing unit corresponds to a region on the laser welding mask. The spectral sensing data acquisition module is used to acquire multiple sets of current spectral sensing data from the multiple spectral sensing units. Each set of spectral sensing data includes the incident wavelength and the light intensity. The main view region construction module is used to construct the main view region by acquiring the current head posture through the built-in IMU module; The main view spectral sensing data filtering module is used to filter the main view spectral sensing data of the main view area from the multiple sets of spectral sensing data; The filter adjustment parameter output module is used to input the main viewing spectral sensing data into the filter adaptive adjustment model and output filter adjustment parameters to control the adjustable liquid crystal filter.

Images