A gradient control method and system based on powder feeding

Abstract

The application discloses a gradient control method and system based on powder feeding, relates to the technical field of feeding control, and solves the technical problem that the existing feeding adopts a fixed vibration mode, has low adaptability, causes poor uniformity of the powder, and further reduces the mixing effect of the whole feeding system; the method comprises the following steps: acquiring feeding data of the feeding material; generating an initial vibration scheme based on the feeding data; collecting first infrared data when the initial vibration scheme is executed; correcting the initial vibration scheme based on the first infrared data to obtain a second vibration scheme; collecting second infrared data when the second vibration scheme is executed; confirming the feeding vibration scheme based on the second infrared data; and controlling the vibration plate based on the feeding vibration frequency in the feeding vibration scheme; the dynamic adjustment of the vibration frequency of the vibration plate is realized, the adaptability of the powder feeding is improved, the problem of uneven feeding caused by the difference in material characteristics is solved, and the mixing effect of the whole feeding system is improved.

Description

Technical Field

[0001] This application belongs to the field of material feeding control technology, specifically a gradient control method and system based on powder feeding. Background Technology

[0002] In industrial production, the uniformity of powder material feeding has a crucial impact on product quality. Existing powder feeding equipment typically uses vibrating feeders, with current control methods employing fixed vibration frequencies for feeding. However, in practical applications, due to differences in the physical properties of powder materials, such as particle size and density, these differences lead to varying degrees of agglomeration during the feeding process, resulting in uneven powder distribution. Traditional fixed-frequency or simple variable-frequency control methods are insufficient to guarantee uniform material distribution on the vibrating plate. Especially for different batches or types of powders, using fixed vibration parameters often results in localized accumulation or intermittent feeding, severely affecting the stability of subsequent mixing or reaction processes and leading to poor mixing effects throughout the feeding system. Therefore, a gradient control method based on powder feeding is needed. Summary of the Invention

[0003] This application provides a gradient control method and system based on powder feeding, which solves the technical problem that the existing feeding technology uses a fixed vibration mode for feeding, which has low adaptability, resulting in poor uniformity of the fed powder, and thus reducing the mixing effect of the entire feeding system.

[0004] To achieve the above objectives, this application adopts the following technical solution: Firstly, a gradient control method based on powder feeding is provided, including: The feeding data of the material is obtained, including the material reference particle size and the material density; the material reference particle size can be the mesh size of the pre-filter of the powder feeder, and the material density is the density of the substance corresponding to the feeding material. An initial vibration scheme is generated based on the deployment data, the initial vibration scheme including an initial vibration amplitude and several first vibration frequencies; The first infrared data collected during the initial vibration scheme is obtained. The first infrared data is image data collected by an infrared thermal imaging camera installed in the direction directly opposite the vibration plate. The initial vibration scheme is corrected based on the first infrared data to obtain the second vibration scheme. The second infrared data is acquired during the second vibration scheme, which is image data acquired by an infrared thermal imaging camera installed in the direction of the material outlet of the vibrating plate; the feeding vibration scheme is confirmed based on the second infrared data; and the vibrating plate is controlled based on the feeding vibration frequency in the feeding vibration scheme. The second infrared image is acquired in real time. A correction frequency is generated based on the second infrared image and the auxiliary control frequency data in the feeding vibration scheme. The frequency of the vibrating plate is adjusted based on the correction frequency.

[0005] Based on the above technical solution, in the gradient control method and system for powder feeding provided in this application, the following steps are taken: First, the initial vibration scheme is generated by acquiring the feeding data of the material; second, the initial vibration scheme is corrected based on the first infrared data to obtain a second vibration scheme; third, the second vibration scheme is confirmed based on the second infrared data; fourth, the vibrating plate is controlled based on the feeding vibration frequency in the feeding vibration scheme; the first vibration scheme is determined by the feeding material, the initial scheme is corrected by the first infrared data to narrow the optimal frequency search range, and then fine screening is performed by combining the second infrared data with the uniformity of the first and second directions. This achieves gradient approximation of the optimal vibration frequency, realizes dynamic adjustment of the vibration frequency of the vibrating plate, effectively increases the adaptability of powder feeding, solves the problem of uneven feeding caused by differences in material characteristics, and improves the mixing effect of the entire feeding system.

[0006] In conjunction with the first aspect above, in one possible implementation, generating an initial vibration scheme based on the deployed data includes: Extract the material reference particle size and material density from the delivery data; determine the initial frequency range based on the material reference particle size; determine the initial vibration amplitude based on the material reference particle size and material density; The corresponding frequency step size is obtained by substituting the material reference particle size, material density, and vibration amplitude into the set step size confirmation function; one form of the step size confirmation function is as follows: ; in, This is the frequency step size; As the reference particle size, The standard particle density is a dimensionless quantity used to remove the reference particle density. This represents the initial vibration amplitude; The set unit vibration amplitude is used to adjust the initial vibration amplitude. Application ratio and removal of initial vibration amplitude Dimensions; The density of the material; The set reference density is used to remove the dimension of material density; The initial frequency is set. Using the lower limit of the initial frequency range as the first vibration frequency and the frequency step size as the step size, select several vibration frequencies in the initial frequency range to obtain several vibration frequencies; record the several vibration frequencies as the first vibration frequencies; integrate the several first vibration frequencies and the initial vibration amplitude into an initial vibration scheme.

[0007] In conjunction with the first aspect above, in one possible implementation, the acquisition method of the first infrared data includes: S1: Extract several first vibration frequencies and initial vibration amplitudes from the initial vibration scheme; number each first vibration frequency in ascending order of its corresponding frequency value; S2: Select the first vibration frequency with the smallest number as the control vibration frequency; S3: Control the vibrating plate using the initial vibration amplitude and the control vibration frequency. S4: Acquire the infrared image corresponding to the first infrared camera, and select the first frame of the infrared image as the reference image; the first frame of the infrared image is the infrared image acquired at a set time after the start of S3, such as the image at the 5th second after the start of S3; S5: Obtain the infrared image captured by the first infrared camera at the current moment, i.e., the infrared image of the current frame; S6: Calculate the similarity between the infrared image and the reference image; S7: Integrate the similarity between the current frame infrared image and a predetermined number of previous infrared images into an evaluation array; calculate the stability of the evaluation array, where stability is one of variance and standard deviation, etc. S8: Determine whether the stability is greater than the set stability threshold. If yes, then record the infrared images corresponding to each similarity in the evaluation array as the infrared image group corresponding to the first vibration frequency, and proceed to S9; otherwise, obtain the next frame of infrared image and proceed to S6. S9: Determine whether the number corresponding to the controlled vibration frequency is equal to the number of first vibration frequencies. If yes, obtain the infrared image group corresponding to each first vibration frequency and integrate each first vibration frequency and its corresponding infrared image group into first infrared data. If no, use the first vibration frequency whose number corresponding to the controlled vibration frequency is increased by one as the controlled vibration frequency.

[0008] In conjunction with the first aspect above, in one possible implementation, a second vibration scheme is obtained by modifying the initial vibration scheme based on the first infrared data, including: Extract infrared image groups corresponding to each first vibration frequency from the first infrared data; extract several infrared images from the groups of infrared images; input the infrared images into the region annotation model for region extraction to obtain the region to be analyzed, the region to be analyzed is the region in the infrared image where the material is placed; map the region to be analyzed corresponding to the infrared image into a temperature matrix, calculate the uniformity corresponding to the infrared image based on the temperature matrix, the uniformity is the distribution uniformity of the material placed in the region to be analyzed in the infrared image, the higher the value of the uniformity of the dry powder distribution on the vibrating plate surface, the more uniform the dry powder is distributed on the vibrating plate; Calculate the average uniformity of each infrared image in the infrared image group, and record the first vibration frequency of the infrared image group whose average uniformity is greater than the set uniformity threshold as the undetermined center frequency; select the undetermined center frequency whose average uniformity is greater than the average value of the first vibration frequency on both sides as the candidate center frequency. The candidate center frequency with the smallest corresponding frequency is selected as the center frequency; the first vibration frequencies adjacent to the center frequency are recorded as reference frequencies; several second vibration frequencies are generated based on the center frequency, reference frequencies and their corresponding uniformity; and several second vibration frequencies are integrated into a second vibration scheme.

[0009] In conjunction with the first aspect above, in one possible implementation, calculating the uniformity corresponding to the infrared image based on the temperature matrix includes: Obtain the temperature matrix and calculate the temperature gradient corresponding to each element in the temperature matrix. One method for calculating the temperature gradient includes: ; in, Let be the temperature gradient of the element in the i-th row and j-th column of the temperature matrix. This refers to the element in the (i+1)th row and jth column of the temperature matrix, i.e., the temperature value. This refers to the element in the (i-1)th row and jth column of the temperature matrix; This refers to the element in the i-th row and j+1-th column of the temperature matrix; This refers to the element in the i-th row and j-1-th column of the temperature matrix; This refers to the element in the (i+1)th row and (j+1)th column of the temperature matrix; This refers to the element in the (i-1)th row and (j-1)th column of the temperature matrix; This refers to the element in the (i+1)th row and (j-1)th column of the temperature matrix; This refers to the element in the (i-1)th row and (j+1)th column of the temperature matrix; A set temperature gradient threshold is obtained. When the temperature gradient is greater than the set temperature gradient threshold, the element corresponding to the temperature gradient is recorded as an aberrant element; otherwise, the element corresponding to the temperature gradient is recorded as a uniform element. The ratio of the total number of uniform elements to the total number of all elements in the temperature matrix is ​​denoted as the uniformity.

[0010] In conjunction with the first aspect mentioned above, one possible implementation method for training the region labeling model includes: Acquire several infrared images and the corresponding regions to be analyzed, wherein the regions to be analyzed are the areas in the infrared images that correspond to the material being placed in the images; integrate the infrared images and their corresponding regions to be analyzed into several training data and test data. The training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model. Finally, the input is an infrared image, and the output is a region labeling model of the region to be analyzed corresponding to the infrared image. The artificial intelligence model includes a deep neural network model. It is understood that this region labeling technology is a relatively existing technology, and will not be elaborated on in detail here.

[0011] In conjunction with the first aspect above, in one possible implementation, generating several second vibration frequencies based on the center frequency, the reference frequency, and their corresponding uniformity includes: The uniformity corresponding to the center frequency is denoted as the pseudo-center uniformity, and the uniformity corresponding to the two reference frequencies is denoted as the reference edge uniformity. A uniformity variation function is constructed based on the center frequency, the pseudo-center uniformity, and the uniformity of the two reference frequencies and the reference edge uniformity. This construction method can be quadratic fitting or other function fitting methods that conform to the uniformity variation characteristics. It is understood that at the optimal vibration frequency, the corresponding uniformity should be the highest, and the uniformity corresponding to frequencies distributed on both sides of it should be lower than its corresponding uniformity, and generally uniformly distributed. One way to construct the uniformity variation function includes: obtaining the reference frequencies... and and center frequency Obtain the uniformity of the reference variable and and pseudo-center uniformity ,in, , , < < Therefore, a uniformity variation function is constructed: ; in, , and These are the quadratic coefficient, the linear coefficient, and the constant coefficient, respectively. ; ; ; The quadratic coefficient, linear coefficient, and constant coefficient can be calculated using the above formula. It is understood that this embodiment uses the simplest quadratic fitting method. Other fitting methods can be used. The higher the fitting accuracy, the higher the probability of the optimal frequency appearing in the subsequent second vibration frequency. A density function is constructed based on a uniformity variation function, and one form of the density function is as follows: ; in, Let be the sampling density function. Let uniformity change function be used. The set parameter coefficients are generally very small positive numbers to ensure that the probability density value is positive; typically 0.01. The number of frequencies of the set second vibration frequency is obtained, and a frequency selection objective function corresponding to the second vibration frequency is constructed based on the number of frequencies and the density function. One expression of the frequency selection objective function is as follows: ; in, The objective function is... For the corresponding frequency variable in the objective function; The sampling density function; The reference frequency to the left of the center frequency; The reference frequency is to the right of the center frequency. For frequency quantity, The objective function for frequency selection is minimized to obtain several second vibration frequencies. The expression for minimization is as follows: ; in, This is the kth second vibration frequency.

[0012] In conjunction with the first aspect above, in one possible implementation, confirming the feeding vibration scheme based on the second infrared data includes: Extract infrared image groups corresponding to each second vibration frequency from the second infrared data; extract several infrared images from the infrared image groups; input the infrared images into a region annotation model to obtain the corresponding regions to be identified; map the regions to be analyzed corresponding to the infrared images into a temperature matrix; and generate the first directional uniformity and the second directional uniformity corresponding to the infrared images based on the temperature matrix; the first directional uniformity is parallel to the direction of the vibrating plate, i.e., the uniformity of the material thrown from the vibrating plate at the same time; the second directional uniformity is perpendicular to the direction of the vibrating plate, i.e., the continuous uniformity of the material thrown from the vibrating plate at different times; one method for calculating the first directional uniformity includes: Calculate the row gradient for each element. : ; When the directional gradient is greater than a set directional gradient threshold (set according to expert experience), the element corresponding to the directional gradient is recorded as a variant element; otherwise, the element corresponding to the directional gradient is recorded as a uniform element; the ratio of the total number of uniform elements to the total number of all elements in the temperature matrix is ​​recorded as the first directional uniformity. One method for calculating the uniformity in the second direction includes: Calculate the column gradient for each element. : ; in, This is the set correction coefficient; because the material's temperature decreases as it falls from the vibrating plate into the mixing device, but this temperature decrease is generally not significant, a temperature change correction coefficient is set. To mitigate the impact of material temperature variations on gradient coefficient imbalance, the correction coefficient is used. The specific coefficients can be set by experts based on experience, and can be determined by the difference between two temperatures in the same column of history, separated by one row. The comprehensive uniformity corresponding to the infrared image is obtained by weighted summation of the uniformity in the first direction and the uniformity in the second direction. It can be understood that the weight coefficients corresponding to the uniformity in the first direction and the uniformity in the second direction in the weighted summation are set by experts based on experience. The average value of the comprehensive uniformity corresponding to each infrared image in the infrared image group is calculated. The second vibration frequency corresponding to the infrared image group with the largest average value is selected and recorded as the feeding vibration frequency. Each second vibration frequency and its corresponding first-direction uniformity and second-direction uniformity are integrated into auxiliary control frequency data; The feeding vibration frequency and auxiliary control frequency data are integrated into a feeding vibration scheme.

[0013] In conjunction with the first aspect above, in one possible implementation, a method for acquiring the second infrared data includes: S1: Extract several second vibration frequencies from the second vibration scheme; S2: Control the vibrating plate to run at the current second vibration frequency and initial vibration amplitude in order of frequency from small to large; S3: At each second vibration frequency, use an infrared thermal imaging camera installed in the direction of the vibrating plate outlet to continuously acquire infrared images for a set time and a set frame rate. S4: For each second vibration frequency, select the infrared images with a set ratio within that time period as the infrared image group; S5: Integrate each second vibration frequency and its corresponding infrared image group into second infrared data.

[0014] In another embodiment, in one possible implementation, generating the corrected frequency based on the second infrared image and the auxiliary control frequency data in the feeding vibration scheme includes: The second infrared image is input into the region annotation model to obtain the corresponding region to be identified; the region to be analyzed corresponding to the infrared image is mapped into a temperature matrix; the strongly defined rows of the temperature matrix are truncated to obtain the leading edge temperature matrix, and the first and second direction uniformity corresponding to the infrared image are generated based on the leading edge temperature matrix; it is understood that the amount of material falling into the vibrating plate through the filter device per unit time may have slight variations, which may cause the vibration frequency to deviate from the actual situation, and may result in uneven material falling into the mixing device through the vibrating plate. Therefore, the vibration frequency needs to be fine-tuned; the first and second uniformity are normalized to obtain normalized values. and , ,Will As a weight value for the uniformity in the first direction, As a weight value for the uniformity in the second direction; Obtain the first and second directional uniformity corresponding to several second vibration frequencies in the auxiliary control frequency data; based on the weight value, perform a weighted summation of the first and second directional uniformity corresponding to each second vibration frequency to obtain the comprehensive uniformity corresponding to the infrared image; select the second vibration frequency corresponding to the infrared image group with the largest average value and record it as the updated feeding vibration frequency.

[0015] Secondly, this application provides a gradient control system based on powder feeding, comprising: a data acquisition module, a data analysis module, and a vibration plate control module; wherein, The data acquisition module includes a material data acquisition unit and an infrared data acquisition unit; The material data acquisition unit is used to acquire the delivery data of the delivered material, which includes the material's baseline particle size and material density. The infrared data acquisition unit is used to acquire first infrared data acquired by the first infrared camera and second infrared data acquired by the second infrared camera. The data analysis module includes a frequency coarse adjustment unit, a frequency fine adjustment unit, and a frequency correction unit; The frequency coarse adjustment unit is used to generate an initial vibration scheme based on the projection data, the initial vibration scheme including a number of vibration amplitudes and a number of first vibration frequencies; and to acquire first infrared data collected during the initial vibration scheme; and to correct the initial vibration scheme based on the first infrared data to obtain a second vibration scheme. The frequency fine-tuning unit is used to acquire second infrared data during the second vibration scheme; to confirm the feeding vibration scheme based on the second infrared data; and to control the vibrating plate based on the feeding vibration frequency in the feeding vibration scheme. The frequency correction unit is used to acquire the second infrared image in real time, generate a correction frequency based on the second infrared image and the auxiliary control frequency data in the feeding vibration scheme, and adjust the frequency of the vibrating plate based on the correction frequency. The vibration plate control module controls the vibration plate based on the vibration frequency and vibration amplitude. The vibration frequency includes a first vibration frequency, a second vibration frequency, a feeding vibration frequency, and a correction frequency. The vibration amplitude includes the initial vibration amplitude.

[0016] This application provides a gradient control method and system based on powder feeding. It can acquire material feeding data, generate an initial vibration scheme based on the data, acquire first infrared data collected during the initial vibration scheme execution, correct the initial vibration scheme based on the first infrared data to obtain a second vibration scheme, acquire second infrared data collected during the second vibration scheme execution, confirm the feeding vibration scheme based on the second infrared data, and control the vibrating plate based on the feeding vibration frequency in the feeding vibration scheme. By determining the first vibration scheme through material feeding, correcting the initial scheme using the first infrared data to narrow the optimal frequency search range, and then using the second infrared data combined with the uniformity of the first and second directions for fine screening, a gradient approximation of the optimal vibration frequency is achieved. This enables dynamic adjustment of the vibrating plate vibration frequency, effectively increasing the adaptability of powder feeding, solving the problem of uneven feeding caused by differences in material characteristics, and improving the mixing effect of the entire feeding system.

[0017] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of 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 only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram illustrating the steps of the gradient control method in this application; Figure 2 This is a schematic diagram of the module connections of the gradient control system in this application. Detailed Implementation

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

[0021] Please see Figure 1 The first aspect of this application provides a gradient control method based on powder feeding, comprising: The feeding data of the material is obtained, including the material reference particle size and material density. The material reference particle size reflects the average size of the powder particles, which can be determined by the mesh size of the pre-filter of the powder feeder. For example, if the pre-filter is a 200-mesh sieve, the reference particle size corresponds to the particle size range of 200 mesh. The material density refers to the bulk density or true density of the material itself. The powder is first dried, then screened by the filter device, falls into the vibrating plate, and is finally evenly fed into the mixing device. An initial vibration scheme is generated based on the delivery data. The initial vibration scheme includes an initial vibration amplitude and several first vibration frequencies. In this embodiment, the initial vibration scheme is the first stage of gradient control, i.e., the "coarse adjustment" stage. The system determines an initial vibration amplitude and a scanning range containing multiple frequency points by querying a preset database or through a calculation model based on the material's reference particle size and density. Several first vibration frequencies are distributed in a certain rule within this range. The purpose is to perform rapid scanning within a wide frequency range to initially detect the material's response state at different frequencies, thereby narrowing the search range for subsequent fine adjustment. It should be understood that the initial vibration scheme does not directly pursue the optimal solution, but rather aims to cover the possible optimal solution range to ensure the comprehensiveness of the search.

[0022] The first infrared data, acquired during the initial vibration scheme, is image data collected by an infrared thermal imaging camera mounted directly opposite the vibrating plate. Based on this first infrared data, the initial vibration scheme is corrected to obtain a second vibration scheme. This step is an intermediate transition stage in gradient control. During the operation of the vibrating plate according to the initial vibration scheme, the temperature distribution image of the material on the vibrating plate is acquired using an infrared thermal imaging device; this is the first infrared data. Since the uniformity of material distribution directly affects the temperature field distribution, analyzing the first infrared data allows for the evaluation of the feeding uniformity effect at each first vibration frequency. The correction process involves screening and evaluating these effects, eliminating frequency points with poor performance, and retaining or focusing on frequency ranges with higher uniformity, thereby generating a second vibration scheme containing a more precise frequency range. This process achieves a transition from wide-range scanning to focusing on the potential optimal range, effectively eliminating interference from invalid frequency points.

[0023] The system acquires second infrared data during the second vibration scheme acquisition process. This second infrared data consists of image data captured by an infrared thermal imaging camera installed directly opposite the discharge port of the vibrating plate. The feeding vibration scheme is confirmed based on this second infrared data. The vibrating plate is then controlled based on the feeding vibration frequency within the scheme. This step represents the final "fine-tuning" stage of gradient control. Within a narrow frequency range defined by the second vibration scheme, the system acquires infrared data again, this time with higher accuracy and greater specificity. Through in-depth analysis of the second infrared data, such as calculating more refined uniformity indices, the system can precisely lock onto an optimal feeding vibration frequency from candidate frequencies. Finally, the control device controls the operation of the vibrating plate according to this feeding vibration frequency, ensuring the most uniform flow and distribution of material on the vibrating plate.

[0024] The second infrared image is acquired in real time. A correction frequency is generated based on the second infrared image and the auxiliary control frequency data in the feeding vibration scheme. The frequency of the vibrating plate is adjusted based on the correction frequency.

[0025] Based on the above technical solution, in the gradient control method and system for powder feeding provided in this application, the following steps are taken: First, the initial vibration scheme is generated by acquiring the feeding data of the material; second, the initial vibration scheme is corrected based on the first infrared data to obtain a second vibration scheme; third, the second vibration scheme is confirmed based on the second infrared data; fourth, the vibrating plate is controlled based on the feeding vibration frequency in the feeding vibration scheme; the first vibration scheme is determined by the feeding material, the initial scheme is corrected by the first infrared data to narrow the optimal frequency search range, and then fine screening is performed by combining the second infrared data with the uniformity of the first and second directions. This achieves gradient approximation of the optimal vibration frequency, realizes dynamic adjustment of the vibration frequency of the vibrating plate, effectively increases the adaptability of powder feeding, solves the problem of uneven feeding caused by differences in material characteristics, and improves the mixing effect of the entire feeding system.

[0026] In one possible implementation, an initial vibration scheme is generated based on the deployed data, including: Extract the material baseline particle size and density from the delivery data; determine the initial frequency range based on the material baseline particle size. Specifically, this is done by consulting a preset particle size frequency confirmation table. The particle size confirmation table includes each material baseline particle and its corresponding initial frequency range. It should be understood that the larger the particle size, the lower its corresponding optimal vibration frequency is usually. Therefore, by consulting the table, a wide range containing the possibility of the optimal frequency can be quickly locked, avoiding blind scanning across the entire frequency band. Determine the initial vibration amplitude based on the material baseline particle size and density. Specifically, this can be done by consulting a preset vibration amplitude table. The vibration amplitude table includes the material baseline particle size and density, as well as their corresponding initial vibration amplitudes. The vibration amplitude table is constructed by experts. The higher the density and the coarser the particles, the greater the inertia of the particles, requiring a larger amplitude to effectively throw, flow, and disperse them. The corresponding frequency step size is obtained by substituting the material reference particle size, material density, and vibration amplitude into the set step size confirmation function; one form of the step size confirmation function is as follows: ; in, This is the frequency step size; As a reference particle size, in this embodiment, the particle size is based on the particle diameter. The standard particle density is a dimensionless quantity used to remove the reference particle density. This represents the initial vibration amplitude; The density of the material; The set unit vibration amplitude is used to adjust the vibration amplitude. Action ratio and vibration removal amplitude Dimensions; The density of the material; The set reference density is used to remove the dimension of material density; The initial frequency is set. The higher the particle density, the greater the particle inertia, and the less sensitive the particles are to frequency changes. Therefore, the corresponding frequency step size can be set appropriately larger to save computation. That is, the higher the density, the smaller the material response caused by the same frequency change, so a larger step size can be used to quickly scan the frequency range. The coarser the particles, the less likely they are to agglomerate, and the wider the optimal frequency region. Therefore, a larger frequency step size can be set. The smaller the particles, the more easily they are affected by friction and cohesion, causing agglomeration. They are more sensitive to frequency, and the optimal frequency region is narrower. That is, the smaller the particles, the more sensitive the uniformity of distribution is to frequency, requiring a smaller step size to capture the optimal frequency point. The larger the vibration amplitude, the more a small change in frequency will cause a significant change in the material's jumping state. Therefore, a smaller step size is needed for fine adjustment. It can be understood that formulas that conform to the above logic are applicable. It is understandable that equal frequency step sizes can also be used in other implementations; Using the lower limit of the initial frequency range as the first vibration frequency and the frequency step size as the step size, select several vibration frequencies in the initial frequency range to obtain several vibration frequencies; record the several vibration frequencies as the first vibration frequencies; integrate the several first vibration frequencies and the initial vibration amplitude into an initial vibration scheme.

[0027] In one possible implementation, the first infrared data is acquired in one manner, including: S1: Extract several first vibration frequencies and initial vibration amplitudes from the initial vibration scheme; number each first vibration frequency in ascending order of its corresponding frequency value; for example, number the lowest frequency as 1, the second lowest as 2, and so on. This sorting provides the logical basis for subsequent sequential scanning; S2: Select the first vibration frequency with the smallest number as the control vibration frequency; for example, select the frequency with the smallest number 1 when executing for the first time. S3: Control the vibrating plate using the initial vibration amplitude and the control vibration frequency. S4: Acquire the infrared image corresponding to the first infrared camera and select the first frame of infrared image as the reference image; the first frame of infrared image is the infrared image acquired at a set time after the start of S3, such as the image at the 5th second after the start of S3; it should be understood that at this time the material distribution has initially formed a specific shape, which serves as a reference for subsequent judgment on whether it has entered a steady state; S5: Obtain the infrared image captured by the first infrared camera at the current moment, i.e., the infrared image of the current frame; S6: Calculate the similarity between the infrared image and the reference image; S7: Integrate the similarity of the current frame infrared image and the previous set number of infrared images into an evaluation array; calculate the stability of the evaluation array, where stability is one of variance and standard deviation; if the similarity value fluctuates drastically, it indicates that the material distribution is still unstable; if the similarity value tends to be stable, it indicates that the material distribution has entered a steady state. S8: Determine whether the stability is greater than the set stability threshold. If yes, it is determined that the material distribution at the current frequency has reached a steady state. The infrared images corresponding to each similarity in the evaluation array are recorded as the infrared image group corresponding to the first vibration frequency, and proceed to S9. If no, it indicates that it is still in a transitional state. Then, the next frame of infrared image is obtained, and proceed to S6. This effectively eliminates the misjudgment of infrared data caused by transitional phenomena such as material accumulation and instantaneous agglomeration, and ensures the reliability of the data source for subsequent uniformity calculation. S9: Determine whether the number corresponding to the controlled vibration frequency is equal to the number of first vibration frequencies. If yes, acquire the infrared image group corresponding to each first vibration frequency and integrate each first vibration frequency and its corresponding infrared image group into first infrared data. If no, use the first vibration frequency whose number is increased by one as the controlled vibration frequency. Through the above steps, this embodiment constructs a rigorous infrared data acquisition logic, which not only achieves full-frequency range coverage scanning, but more importantly, by introducing a steady-state determination mechanism, eliminates invalid transitional state data from the source, providing high-quality data support for subsequent vibration scheme correction based on infrared data.

[0028] In one possible implementation, a second vibration scheme is obtained by modifying the initial vibration scheme based on the first infrared data, including: The infrared image group corresponding to each first vibration frequency in the first infrared data is extracted; several infrared images from the group are extracted; since the images acquired by the infrared camera usually include non-material areas such as the background of the vibrating plate and the edge support, directly analyzing the entire image would introduce huge errors. Therefore, the infrared image is input into the region annotation model for region extraction to obtain the region to be analyzed, which is the area in the infrared image where the material is placed; the region to be analyzed corresponding to the infrared image is mapped into a temperature matrix, where each element in the temperature matrix corresponds to the temperature value of a pixel in the infrared image; the uniformity corresponding to the infrared image is calculated based on the temperature matrix. Different packing densities of powder materials on the vibrating plate will lead to different heat capacities, which will be manifested as temperature differences under infrared thermal imaging. If the material is uniformly distributed, the temperature field distribution is gentle and the temperature gradient is small; if the material is piled up or has voids, the temperature field will change drastically and the temperature gradient will be large; the uniformity is the distribution uniformity of the material placed in the region to be analyzed in the infrared image. The higher the value of the uniformity of the dry powder distribution on the vibrating plate surface, the more uniform the dry powder is distributed on the vibrating plate. Calculate the average uniformity of each infrared image in the infrared image group, and record the first vibration frequency of the infrared image group whose average uniformity is greater than the set uniformity threshold as the undetermined center frequency; select the undetermined center frequency whose average uniformity is greater than the average value of the first vibration frequency on both sides as the candidate center frequency. The candidate center frequency with the smallest corresponding frequency is selected as the center frequency. The first vibration frequencies adjacent to the center frequency on both sides are recorded as reference frequencies. Several second vibration frequencies are generated based on the center frequency, reference frequencies, and their corresponding uniformity, and these second vibration frequencies are integrated into a second vibration scheme. After determining the center frequency, the system records the first vibration frequencies adjacent to the center frequency on both sides as reference frequencies. At this point, the search range has been focused from the entire initial frequency range to a very small range of the center frequency and its neighborhood. The system generates several second vibration frequencies based on the center frequency, reference frequencies, and their corresponding uniformity, and integrates these frequencies into a second vibration scheme. It should be understood that the number of second vibration frequencies is usually less than the number of first vibration frequencies, and they are distributed near the center frequency, thereby improving search accuracy and narrowing the search range.

[0029] In one possible implementation, calculating the uniformity of the infrared image based on the temperature matrix includes: obtaining the temperature matrix and calculating the temperature gradient corresponding to each element in the temperature matrix. One method for calculating the temperature gradient includes: ; in, Let be the temperature gradient of the element in the i-th row and j-th column of the temperature matrix. This refers to the element in the (i+1)th row and jth column of the temperature matrix, i.e., the temperature value. This refers to the element in the (i-1)th row and jth column of the temperature matrix; This refers to the element in the i-th row and j+1-th column of the temperature matrix; This refers to the element in the i-th row and j-1-th column of the temperature matrix; This refers to the element in the (i+1)th row and (j+1)th column of the temperature matrix; This refers to the element in the (i-1)th row and (j-1)th column of the temperature matrix; This refers to the element in the (i+1)th row and (j-1)th column of the temperature matrix; This refers to the element in the (i-1)th row and (j+1)th column of the temperature matrix; A set temperature gradient threshold is obtained. When the temperature gradient is greater than the set temperature gradient threshold, the element corresponding to the temperature gradient is recorded as an aberrant element; otherwise, the element corresponding to the temperature gradient is recorded as a uniform element. The ratio of the total number of uniform elements to the total number of all elements in the temperature matrix is ​​denoted as the uniformity.

[0030] In conjunction with the first aspect mentioned above, one possible implementation method for training the region labeling model includes: Acquire several infrared images and the corresponding regions to be analyzed, wherein the regions to be analyzed are the areas in the infrared images that correspond to the material being placed in the images; integrate the infrared images and their corresponding regions to be analyzed into several training data and test data. The training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model. Finally, the input is an infrared image, and the output is a region labeling model of the region to be analyzed corresponding to the infrared image. The artificial intelligence model includes a deep neural network model. It is understood that this region labeling technology is a relatively existing technology, and will not be elaborated on in detail here.

[0031] In one possible implementation, several second vibration frequencies are generated based on the center frequency, reference frequencies, and their corresponding uniformity. This includes: obtaining the uniformity corresponding to the center frequency and denoteing it as pseudo-center uniformity; denoteing the uniformity corresponding to the two reference frequencies as reference edge uniformity; constructing a uniformity variation function based on the center frequency, pseudo-center uniformity, two reference frequencies, and reference edge uniformity. The construction method can be quadratic fitting or other function fitting methods that conform to the uniformity variation characteristics. It is understood that, physically, the influence of vibration frequency on the uniformity of material distribution usually exhibits a unimodal characteristic, that is, the uniformity is highest at the optimal frequency and gradually decreases on both sides. One way to construct the uniformity variation function includes: obtaining the reference frequency... and and center frequency Obtain the uniformity of the reference variable and and pseudo-center uniformity ,in, , , < < Therefore, a uniformity variation function is constructed: ; in, , and These are the quadratic coefficient, the linear coefficient, and the constant coefficient, respectively. ; ; ; The quadratic coefficient, linear coefficient, and constant coefficient can be calculated using the above formula. It is understood that this embodiment uses the simplest quadratic fitting method. Other fitting methods can be used. The higher the fitting accuracy, the higher the probability of the optimal frequency appearing in the subsequent second vibration frequency. A density function is constructed based on a uniformity variation function, and one form of the density function is as follows: ; in, Let be the sampling density function. Let uniformity change function be used. The set parameter coefficients are generally very small positive numbers to ensure that the probability density value is positive; typically 0.01. To minimize the function; obtain the number of frequencies of the set second vibration frequency, and construct the frequency selection objective function corresponding to the second vibration frequency based on the number of frequencies and the density function; one expression of the frequency selection objective function is as follows: ; in, The objective function is... For the corresponding frequency variable in the objective function; The sampling density function; The reference frequency to the left of the center frequency; The reference frequency is to the right of the center frequency. For frequency quantity, The objective function for frequency selection is minimized to obtain several second vibration frequencies. The expression for minimization is as follows: ; in, This is the kth second vibration frequency.

[0032] Due to the inherent influence of the equipment in actual vibration feeding, the actual optimal vibration frequency deviates from the theoretical one to a certain extent. This embodiment adaptively sets several second vibration frequencies at non-equidistant intervals using the above method. The density of the second vibration frequencies is higher when they are close to the maximum theoretical uniformity value, and lower when they are far from the maximum theoretical uniformity value. The nonlinear programming effect of dense sampling near the optimal frequency and sparse sampling in non-optimal regions is different from the traditional equidistant sampling method. It can capture the global optimal frequency with the highest probability with a limited number of samplings, which significantly improves the search efficiency and accuracy in the fine-tuning stage.

[0033] In one possible implementation, confirming the feeding vibration scheme based on the second infrared data includes: extracting infrared image groups corresponding to each second vibration frequency in the second infrared data; extracting several infrared images from the infrared image groups; inputting the infrared images into a region annotation model to obtain the corresponding region to be identified; mapping the region to be analyzed corresponding to the infrared images into a temperature matrix; and generating a first directional uniformity and a second directional uniformity corresponding to the infrared images based on the temperature matrix; the first directional uniformity is parallel to the direction of the vibrating plate, i.e., the uniformity of the material fed from the vibrating plate at the same time in spatial distribution; the second directional uniformity is perpendicular to the direction of the vibrating plate, i.e., the continuous uniformity of the material fed from the vibrating plate at different times; one method for calculating the first directional uniformity includes: Calculate the row gradient for each element. : ; When the directional gradient is greater than a set directional gradient threshold (set according to expert experience), the element corresponding to the directional gradient is recorded as a variant element; otherwise, the element corresponding to the directional gradient is recorded as a uniform element; the ratio of the total number of uniform elements to the total number of all elements in the temperature matrix is ​​recorded as the first directional uniformity. One method for calculating the uniformity in the second direction includes: Calculate the column gradient for each element. : ; in, This is the set correction coefficient; because the material's temperature decreases as it falls from the vibrating plate into the mixing device, but this temperature decrease is generally not significant, a temperature change correction coefficient is set. To mitigate the impact of material temperature variations on gradient coefficient imbalance, the correction coefficient is used. The specific coefficients can be set by experts based on experience, and can be determined by the difference between two temperatures in the same column of history, separated by one row. When the column gradient is greater than the set column gradient threshold, the column gradient threshold is set according to expert experience; the element corresponding to the column gradient is recorded as the variant element; otherwise, the element corresponding to the column gradient is recorded as the uniform element; the ratio of the total number of uniform elements to the total number of all elements in the temperature matrix is ​​recorded as the second directional uniformity. The comprehensive uniformity corresponding to the infrared image is obtained by weighted summation of the uniformity in the first direction and the uniformity in the second direction. It can be understood that the weight coefficients corresponding to the uniformity in the first direction and the uniformity in the second direction in the weighted summation are set by experts based on experience. The average value of the comprehensive uniformity corresponding to each infrared image in the infrared image group is calculated. The second vibration frequency corresponding to the infrared image group with the largest average value is selected and recorded as the feeding vibration frequency. Each second vibration frequency and its corresponding first-direction uniformity and second-direction uniformity are integrated into auxiliary control frequency data; The feeding vibration frequency and auxiliary control frequency data are integrated into a feeding vibration scheme.

[0034] It is understandable that, in order to ensure the accuracy of the final feeding vibration frequency confirmation, in another embodiment, the first infrared data and the second infrared data are collected during the second vibration scheme, and the feeding vibration scheme is confirmed based on the first infrared data and the second infrared data. In one possible implementation, a method for acquiring the second infrared data includes: S1: Extract several second vibration frequencies from the second vibration scheme; S2: Control the vibrating plate to run at the current second vibration frequency and initial vibration amplitude in order of frequency from small to large; S3: At each second vibration frequency, use an infrared thermal imaging camera installed in the direction of the vibrating plate outlet to continuously acquire infrared images for a set time and a set frame rate. S4: For each second vibration frequency, select the infrared images of the last set proportion within that time period as the infrared image group; for example, if the set proportion is the last 30%, then select the last 30% of the images in the acquisition time series; the purpose of this design is to exclude the transitional data during the start-up or frequency switching process of the vibration plate, ensuring that the data used for analysis is in the state after the vibration plate has stabilized, thereby ensuring the accuracy of the evaluation results; finally, integrate each second vibration frequency and its corresponding infrared image groups into second infrared data; S5: Integrate each second vibration frequency and its corresponding infrared image group into second infrared data.

[0035] In another embodiment, in one possible implementation, generating the corrected frequency based on the second infrared image and the auxiliary control frequency data in the feeding vibration scheme includes: The second infrared image is input into the region annotation model to obtain the corresponding region to be identified; the region to be analyzed corresponding to the infrared image is mapped into a temperature matrix; the strongly defined rows of the temperature matrix are truncated to obtain the leading edge temperature matrix, and the first and second direction uniformity corresponding to the infrared image are generated based on the leading edge temperature matrix; it is understood that the amount of material falling into the vibrating plate through the filter device per unit time may have slight variations, which may cause the vibration frequency to deviate from the actual situation, and may result in uneven material falling into the mixing device through the vibrating plate. Therefore, the vibration frequency needs to be fine-tuned; the first and second uniformity are normalized to obtain normalized values. and , ,Will As a weight value for the uniformity in the first direction, As a weight value for the uniformity in the second direction; Obtain the first and second directional uniformity corresponding to several second vibration frequencies in the auxiliary control frequency data; based on the weight value, perform a weighted summation of the first and second directional uniformity corresponding to each second vibration frequency to obtain the comprehensive uniformity corresponding to the infrared image; select the second vibration frequency corresponding to the infrared image group with the largest average value and record it as the updated feeding vibration frequency.

[0036] Please see Figure 2 Secondly, this application provides a gradient control system based on powder feeding, comprising: a data acquisition module, a data analysis module, and a vibration plate control module; wherein, The data acquisition module includes a material data acquisition unit and an infrared data acquisition unit; The material data acquisition unit is used to acquire the delivery data of the delivered material, which includes the material's baseline particle size and material density. The infrared data acquisition unit is used to acquire first infrared data acquired by the first infrared camera and second infrared data acquired by the second infrared camera; The data analysis module includes a frequency coarse adjustment unit, a frequency fine adjustment unit, and a frequency correction unit; The frequency coarse adjustment unit is used to generate an initial vibration scheme based on the projection data, the initial vibration scheme including a number of vibration amplitudes and a number of first vibration frequencies; and to acquire first infrared data collected during the initial vibration scheme; and to correct the initial vibration scheme based on the first infrared data to obtain a second vibration scheme. The frequency fine-tuning unit is used to acquire second infrared data during the second vibration scheme; to confirm the feeding vibration scheme based on the second infrared data; and to control the vibrating plate based on the feeding vibration frequency in the feeding vibration scheme. The frequency correction unit is used to acquire the second infrared image in real time, generate a correction frequency based on the second infrared image and the auxiliary control frequency data in the feeding vibration scheme, and adjust the frequency of the vibrating plate based on the correction frequency. The vibration plate control module controls the vibration plate based on the vibration frequency and vibration amplitude. The vibration frequency includes a first vibration frequency, a second vibration frequency, a feeding vibration frequency, and a correction frequency. The vibration amplitude includes the initial vibration amplitude.

[0037] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

[0038] How this application works: By acquiring the material feeding data, an initial vibration scheme is generated based on the feeding data; first infrared data is collected during the initial vibration scheme; the initial vibration scheme is corrected based on the first infrared data to obtain a second vibration scheme; second infrared data is collected during the second vibration scheme; the feeding vibration scheme is confirmed based on the second infrared data; the vibrating plate is controlled based on the feeding vibration frequency in the feeding vibration scheme; the first vibration scheme is determined by the feeding material, the initial scheme is corrected by the first infrared data to narrow the optimal frequency search range, and then the second infrared data is combined with the uniformity of the first and second directions for fine screening, realizing gradient approximation of the optimal vibration frequency, realizing dynamic adjustment of the vibration frequency of the vibrating plate, effectively increasing the adaptability of powder feeding, solving the problem of uneven feeding caused by differences in material characteristics, and improving the mixing effect of the entire feeding system.

[0039] The above embodiments are only used to illustrate the technical methods of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of this application without departing from the spirit and scope of the technical methods of this application.

Claims

1. A gradient control method based on powder feeding, characterized in that, include: Obtain the delivery data of the material to be delivered, wherein the delivery data includes the material's baseline particle size and material density; An initial vibration scheme is generated based on the deployment data, the initial vibration scheme including an initial vibration amplitude and several first vibration frequencies; Acquire the first infrared data collected during the initial vibration scheme implementation; based on the first infrared data, correct the initial vibration scheme to obtain the second vibration scheme; Acquire second infrared data while the second vibration scheme is in progress; confirm the feeding vibration scheme based on the second infrared data; control the vibrating plate based on the feeding vibration frequency in the feeding vibration scheme.

2. The gradient control method based on powder feeding according to claim 1, characterized in that, An initial vibration scheme is generated based on the deployed data, including: Extract the material reference particle size and material density from the delivery data; determine the initial frequency range based on the material reference particle size; determine the initial vibration amplitude based on the material reference particle size and material density; Substitute the material reference particle size, material density, and vibration amplitude into the set step size confirmation function to obtain the corresponding frequency step size; Using the lower limit of the initial frequency range as the first vibration frequency and the frequency step size as the step size, select several vibration frequencies in the initial frequency range to obtain several vibration frequencies; record the several vibration frequencies as the first vibration frequencies; integrate the several first vibration frequencies and the initial vibration amplitude into an initial vibration scheme.

3. The gradient control method based on powder feeding according to claim 1, characterized in that, One method for acquiring the first infrared data includes: S1: Extract several first vibration frequencies and initial vibration amplitudes from the initial vibration scheme; number each first vibration frequency in ascending order of its corresponding frequency value; S2: Select the first vibration frequency with the smallest number as the control vibration frequency; S3: Control the vibrating plate using the initial vibration amplitude and the control vibration frequency. S4: Obtain the infrared image corresponding to the first infrared camera, and select the first frame of the infrared image as the reference image; S5: Obtain the infrared image captured by the first infrared camera at the current moment, i.e., the infrared image of the current frame; S6: Calculate the similarity between the infrared image and the reference image; S7: Integrate the similarity between the current frame infrared image and the infrared images of the previous set number into an evaluation array; calculate the stability of the evaluation array; S8: Determine whether the stability is greater than the set stability threshold. If yes, then record the infrared images corresponding to each similarity in the evaluation array as the infrared image group corresponding to the first vibration frequency, and proceed to S9; otherwise, obtain the next frame of infrared image and proceed to S6. S9: Determine whether the number corresponding to the controlled vibration frequency is equal to the number of first vibration frequencies. If yes, obtain the infrared image group corresponding to each first vibration frequency and integrate each first vibration frequency and its corresponding infrared image group into first infrared data. If no, use the first vibration frequency whose number corresponding to the controlled vibration frequency is increased by one as the controlled vibration frequency.

4. The gradient control method based on powder feeding according to claim 1, characterized in that, A second vibration scheme is obtained by correcting the initial vibration scheme based on the first infrared data, including: Extract infrared image groups corresponding to each first vibration frequency from the first infrared data; extract several infrared images from the groups of infrared images; input the infrared images into the region annotation model to extract the region to be analyzed; map the region to be analyzed corresponding to the infrared images into a temperature matrix, and calculate the uniformity corresponding to the infrared images based on the temperature matrix. Calculate the average uniformity of each infrared image in the infrared image group, and record the first vibration frequency of the infrared image group whose average uniformity is greater than the set uniformity threshold as the undetermined center frequency; select the undetermined center frequency whose average uniformity is greater than the average value of the first vibration frequency on both sides as the candidate center frequency. The candidate center frequency with the smallest corresponding frequency is selected as the center frequency; the first vibration frequencies adjacent to the center frequency are recorded as reference frequencies; several second vibration frequencies are generated based on the center frequency, reference frequencies and their corresponding uniformity; and several second vibration frequencies are integrated into a second vibration scheme.

5. The gradient control method based on powder feeding according to claim 4, characterized in that, Calculating the uniformity of the infrared image based on the temperature matrix includes: Obtain the temperature matrix and calculate the temperature gradient corresponding to each element in the temperature matrix; A set temperature gradient threshold is obtained. When the temperature gradient is greater than the set temperature gradient threshold, the element corresponding to the temperature gradient is recorded as an aberrant element; otherwise, the element corresponding to the temperature gradient is recorded as a uniform element. The ratio of the total number of uniform elements to the total number of all elements in the temperature matrix is ​​denoted as the uniformity.

6. The gradient control method based on powder feeding according to claim 4, characterized in that, One training method for the region labeling model includes: Acquire several infrared images and the corresponding regions to be analyzed, wherein the regions to be analyzed are the areas in the infrared images that correspond to the material being placed in the images; integrate the infrared images and their corresponding regions to be analyzed into several training data and test data. The training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model. Finally, the input is an infrared image, and the output is a region labeling model of the region to be analyzed corresponding to the infrared image.

7. The gradient control method based on powder feeding according to claim 4, characterized in that, The generation of several second vibration frequencies based on the center frequency, reference frequency, and their corresponding uniformity includes: The uniformity corresponding to the center frequency is denoted as the pseudo-center uniformity, and the uniformity corresponding to the two reference frequencies is denoted as the reference edge uniformity; a uniformity variation function is constructed based on the center frequency, the pseudo-center uniformity, the two reference frequencies, and the reference edge uniformity. A density function is constructed based on the uniformity variation function; the number of frequencies at the set second vibration frequency is obtained, and a frequency selection target function corresponding to the second vibration frequency is constructed based on the number of frequencies and the density function; one expression of the frequency selection target function is as follows: ； in, The objective function is... For the corresponding frequency variable in the objective function; The sampling density function; The reference frequency to the left of the center frequency; The reference frequency is to the right of the center frequency. For frequency quantity, The objective function for frequency selection is minimized to obtain several second vibration frequencies. The expression for minimization is as follows: ； in, This is the kth second vibration frequency.

8. The gradient control method based on powder feeding according to claim 1, characterized in that, The feeding vibration scheme is confirmed based on the second infrared data, including: Extract infrared image groups corresponding to each second vibration frequency from the second infrared data; extract several infrared images from the infrared image groups; input the infrared images into the region annotation model to obtain the corresponding region to be identified; map the region to be analyzed corresponding to the infrared images into a temperature matrix; and generate the first direction uniformity and the second direction uniformity corresponding to the infrared images based on the temperature matrix. The comprehensive uniformity corresponding to the infrared image is obtained by weighted summation of the first direction uniformity and the second direction uniformity; the average value of the comprehensive uniformity corresponding to each infrared image in the infrared image group is calculated; the second vibration frequency corresponding to the infrared image group with the largest average value is selected as the feeding vibration frequency; Each second vibration frequency and its corresponding first-direction uniformity and second-direction uniformity are integrated into auxiliary control frequency data; The feeding vibration frequency and auxiliary control frequency data are integrated into a feeding vibration scheme.

9. The gradient control method based on powder feeding according to claim 1, characterized in that, One method for acquiring the second infrared data includes: S1: Extract several second vibration frequencies from the second vibration scheme; S2: Control the vibrating plate to run at the current second vibration frequency and initial vibration amplitude in order of frequency from small to large; S3: At each second vibration frequency, acquire several infrared images; S4: For each second vibration frequency, select infrared images of a set proportion after the time period corresponding to each second vibration frequency as an infrared image group; S5: Integrate each second vibration frequency and its corresponding infrared image group into second infrared data.

10. A gradient control system based on powder feeding, characterized in that, include: The system comprises a data acquisition module, a data analysis module, and a vibration plate control module; among which, The data acquisition module includes a material data acquisition unit and an infrared data acquisition unit; The material data acquisition unit is used to acquire the delivery data of the delivered material, which includes the material's baseline particle size and material density. The infrared data acquisition unit is used to acquire first infrared data acquired by the first infrared camera and second infrared data acquired by the second infrared camera. The data analysis module includes a coarse frequency tuning unit and a fine frequency tuning unit; The frequency coarse adjustment unit is used to generate an initial vibration scheme based on the projection data, the initial vibration scheme including a number of vibration amplitudes and a number of first vibration frequencies; and to acquire first infrared data collected during the initial vibration scheme; and to correct the initial vibration scheme based on the first infrared data to obtain a second vibration scheme. The frequency fine-tuning unit is used to acquire second infrared data during the second vibration scheme; to confirm the feeding vibration scheme based on the second infrared data; and to control the vibrating plate based on the feeding vibration frequency in the feeding vibration scheme. The vibration plate control module controls the vibration plate based on the vibration frequency and vibration amplitude. The vibration frequency includes a first vibration frequency, a second vibration frequency, and a feeding vibration frequency, and the vibration amplitude includes the initial vibration amplitude.

Images