Industrial robot accurate positioning and motion control method based on visual perception

Abstract

The invention discloses an industrial robot accurate positioning and motion control method based on visual perception, and relates to the industrial robot related field, and the method comprises the following steps: carrying out primary spraying operation on a target storage tank according to initial spraying parameters; a multi-view camera is used for carrying out image collection on a target storage tank, and image splicing is carried out to generate a primary spraying storage tank image; gray processing is carried out, the storage tank gray image is segmented according to a preset gray threshold, and a plurality of storage tank local areas and a plurality of area gray values are determined; a target gray value is obtained, deviation calculation is conducted on the multiple area gray values with the target gray value as the reference, and multiple secondary spraying thicknesses are determined; and a secondary spraying path sequence and a secondary spraying parameter sequence are generated, and secondary spraying operation is conducted on the target storage tank. The technical problems of inaccurate positioning and poor spraying uniformity existing in spraying control of an existing spraying operation robot are solved, and the technical effects of improving the positioning precision and achieving uniform distribution of the spraying thickness are achieved.

Description

Technical Field

[0001] The present application relates to the field of industrial robots, and in particular to a method for precise positioning and motion control of industrial robots based on visual perception. Background Art

[0002] Spraying is an important process that is widely used in many industries such as automobile manufacturing, aerospace, and home appliance manufacturing. In existing methods, spraying operations usually rely on preset spraying parameters and fixed spraying paths. Due to differences in the shape, size, and surface condition of workpieces, fixed spraying paths often cannot accurately match the actual position of each workpiece, resulting in poor spraying effects. At the same time, operations based on preset spraying parameters cannot adjust the spraying thickness in real time to adapt to the actual conditions of different areas, which may cause some areas to be over-sprayed, wasting materials, and some areas to be under-sprayed, failing to achieve the expected protective effect.

[0003] In the current related technologies, the spraying control of the spraying robot has technical problems such as inaccurate positioning and poor spraying uniformity. Summary of the invention

[0004] The present application provides a method for precise positioning and motion control of an industrial robot based on visual perception, performs a spraying operation on a target tank according to initial spraying parameters, uses a multi-eye camera to capture images of the target tank after the first spraying, and performs image stitching to generate a sprayed tank image, performs grayscale processing on the first sprayed tank image, segments the tank grayscale image according to a predetermined grayscale threshold, determines multiple local areas of the tank and multiple regional grayscale values, obtains a target grayscale value of a target spraying thickness, calculates deviations of multiple regional grayscale values ​​based on the value, determines multiple secondary spraying thicknesses, generates a secondary spraying path sequence and a secondary spraying parameter sequence according to multiple local areas of the tank and multiple secondary spraying thicknesses, controls the spraying robot to perform a secondary spraying operation on the target tank, and other technical means, thereby achieving the technical effect of improving positioning accuracy and achieving uniform distribution of spraying thickness.

[0005] This application provides a method for precise positioning and motion control of an industrial robot based on visual perception, including: According to the initial spraying parameters, the spraying robot is controlled to perform a spraying operation on the target tank, wherein the initial spraying parameters are set based on the primary spraying thickness, and the primary spraying thickness is 70% of the target spraying thickness; a multi-eye camera is used to collect images of the target tank after the primary spraying operation, and the images are stitched to generate a primary spraying tank image; the primary spraying tank image is gray-scale processed, the tank gray-scale image is segmented according to a predetermined gray-scale threshold, and multiple local areas of the tank and multiple regional gray-scale values ​​are determined; the target gray-scale value of the target spraying thickness is obtained, and the deviation of the multiple regional gray-scale values ​​is calculated based on the target gray-scale value to determine multiple secondary spraying thicknesses; a secondary spraying path sequence and a secondary spraying parameter sequence are generated according to the multiple local areas of the tank and the multiple secondary spraying thicknesses, and based on the secondary spraying path sequence and the secondary spraying parameter sequence, the spraying robot is controlled to perform a secondary spraying operation on the target tank.

[0006] In a possible implementation, the grayscale image of the storage tank is segmented according to a predetermined grayscale threshold, a plurality of local areas of the storage tank and a plurality of regional grayscale values ​​are determined, and the following processing is performed: Obtain a predetermined grayscale threshold, wherein the predetermined grayscale threshold is set based on a minimum spraying accuracy of the spray robot; extract grayscale values ​​from the grayscale image of the storage tank to obtain a minimum grayscale value and a maximum grayscale value; segment the grayscale image of the storage tank based on the predetermined grayscale threshold, the minimum grayscale value, and the maximum grayscale value to obtain a plurality of local areas of the storage tank and a plurality of regional grayscale values.

[0007] In a possible implementation, based on the predetermined grayscale threshold, the minimum grayscale value and the maximum grayscale value, the grayscale image of the storage tank is segmented and the following processing is performed: Taking the position coordinates of the minimum grayscale value as the starting point, simulate water filling into the grayscale image of the storage tank, record the first grayscale value that the water surface covers, and perform deviation calculation between the minimum grayscale value and the first grayscale value to obtain a first grayscale difference; if the first grayscale difference is greater than or equal to the predetermined grayscale threshold, set a first segmentation line in this area; if the first grayscale difference is less than the predetermined grayscale threshold, continue to simulate water filling; perform iterative water filling until the water surface covers the maximum grayscale value, stop water filling and obtain multiple segmentation lines; use the multiple segmentation lines to segment the grayscale image of the storage tank, and obtain multiple local areas of the storage tank and multiple regional grayscale values, wherein the regional grayscale value is the grayscale median value of the local area of ​​the storage tank.

[0008] In a possible implementation, the target grayscale value of the target spraying thickness is obtained, and the following processing is performed: The standard grayscale value of the target spray thickness under standard illumination parameters is obtained; the ambient illumination parameters during the image acquisition of the spray tank are obtained, deviation calculation is performed on the ambient illumination parameters and the standard illumination parameters, and the illumination parameter difference is determined; the illumination parameter difference is input into an illumination corrector, and a grayscale correction coefficient is output, wherein the illumination corrector is constructed based on a BP neural network and is trained to convergence through sample data; parameter correction is performed on the standard grayscale value according to the grayscale correction coefficient to obtain the target grayscale value.

[0009] In a possible implementation, the target grayscale value is used as a reference, the deviation of the grayscale values ​​of the multiple regions is calculated respectively, multiple secondary spraying thicknesses are determined, and the following processing is performed: Based on the target grayscale value, deviation calculations are performed on the grayscale values ​​of the multiple regions to determine multiple regional grayscale differences; the multiple regional grayscale differences are input into a predetermined grayscale difference-secondary spraying thickness comparison table for matching, and the multiple secondary spraying thicknesses are output.

[0010] In a possible implementation, a secondary spraying path sequence and a secondary spraying parameter sequence are generated according to the multiple storage tank local areas and the multiple secondary spraying thicknesses, and the following processing is performed: The multiple tank local areas are arranged from large to small in terms of area area to generate a tank local area sequence; with the purpose of covering the tank local area sequence, the operation simulation of the spraying robot is performed in the spraying operation twin space, and the secondary spraying path sequence is generated based on the simulation results; based on the tank local area sequence, a secondary spraying thickness sequence is generated according to the multiple secondary spraying thickness mappings; in order to meet the secondary spraying thickness as an operation constraint, the secondary spraying parameter sequence is generated according to the secondary spraying thickness sequence.

[0011] In a possible implementation, the secondary spraying parameter sequence is generated according to the secondary spraying thickness sequence, and the following processing is performed: In order to meet the secondary spraying thickness as the operation constraint, the operation simulation is performed in the spraying operation twin space based on the secondary spraying thickness sequence to generate an initial secondary spraying parameter sequence; a primary spraying operation deviation analysis is performed according to the grayscale values ​​of the multiple local areas of the tanks and the multiple areas to determine the primary spraying thickness deviation; the primary spraying thickness deviation is input into the spraying parameter compensation model, and the spraying parameter compensation coefficient is output to compensate the initial secondary spraying parameter sequence to obtain the secondary spraying parameter sequence.

[0012] In a possible implementation, a spraying operation deviation analysis is performed based on the multiple local areas of the tanks and the multiple regional grayscale values ​​to determine a spraying thickness deviation, and the following processing is performed: Acquire multiple area areas of the multiple local areas of the tanks, and calculate multiple area ratios; perform product operations on the multiple area ratios and multiple area grayscale values, and sum them to obtain a single spraying grayscale mean; input the single spraying grayscale mean into a predetermined grayscale-spraying thickness comparison table for matching, and output a single spraying thickness mean; perform deviation calculation on the single spraying thickness and the single spraying thickness mean, and obtain the single spraying thickness deviation.

[0013] The method for precise positioning and motion control of an industrial robot based on visual perception proposed in this application is proposed. First, according to the initial spraying parameters, the spraying robot is controlled to perform a spraying operation on the target tank. The initial spraying parameters are set based on the primary spraying thickness, and the primary spraying thickness is 70% of the target spraying thickness. Then, a multi-eye camera is used to collect images of the target tank after the primary spraying operation is completed, and image stitching is performed to generate a primary spraying tank image. Then, the primary spraying tank image is gray-scale processed, and the tank gray-scale image is segmented according to a predetermined gray-scale threshold to determine multiple local areas of the tank and multiple regional gray-scale values. Then, the target gray-scale value of the target spraying thickness is obtained. Based on the target gray-scale value, the deviation of the gray-scale values ​​of the multiple regions is calculated respectively to determine multiple secondary spraying thicknesses. Finally, a secondary spraying path sequence and a secondary spraying parameter sequence are generated according to the multiple local areas of the tank and the multiple secondary spraying thicknesses. Based on the secondary spraying path sequence and the secondary spraying parameter sequence, the spraying robot is controlled to perform a secondary spraying operation on the target tank, thereby achieving the technical effect of improving positioning accuracy and achieving uniform distribution of spraying thickness. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0015] Figure 1 A flow chart of a method for precise positioning and motion control of an industrial robot based on visual perception provided in an embodiment of the present application.

[0016] Figure 2 A schematic diagram of the process of generating a secondary spraying path sequence and a secondary spraying parameter sequence in the method for precise positioning and motion control of an industrial robot based on visual perception provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0018] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0019] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0020] The present application provides a method for precise positioning and motion control of an industrial robot based on visual perception, such as Figure 1 As shown, the method includes: Step S100, according to the initial spraying parameters, control the spraying robot to perform a spraying operation on the target tank, wherein the initial spraying parameters are set based on the primary spraying thickness, and the primary spraying thickness is 70% of the target spraying thickness. Specifically, the initial spraying parameters refer to various parameters during a spraying operation, such as spraying speed, spraying pressure, spraying angle, etc. These parameters are set based on 70% of the target spraying thickness (the final spraying thickness expected to be achieved, which is the quality standard of the spraying operation), and jointly determine the thickness and uniformity of the spraying. The spraying robot performs a spraying operation on the target tank according to the preset primary spraying parameters.

[0021] Step S200, using a multi-camera to collect images of a target tank after a spraying operation, and performing image stitching to generate a spray tank image. Specifically, using a multi-camera to collect all-round images of the target tank after a spraying operation. Using image stitching technology, images collected by multiple cameras are stitched into a complete spray tank image.

[0022] Step S300, grayscale processing is performed on the image of the sprayed tank once, and the grayscale image of the tank is segmented according to a predetermined grayscale threshold, and multiple local areas of the tank and multiple regional grayscale values ​​are determined. Specifically, grayscale processing is performed on the image of the sprayed tank once, that is, the color image is converted into a grayscale image, and each pixel in the grayscale image has only one brightness value. According to a predetermined grayscale threshold (a threshold used for image segmentation, when the grayscale value of a pixel exceeds or falls below the threshold, it will be classified as a different area), the grayscale image of the tank is segmented, multiple local areas of the tank are determined (different parts of the tank surface obtained by image segmentation), and the grayscale value of each area is calculated.

[0023] In a possible implementation, the grayscale image of the storage tank is segmented according to a predetermined grayscale threshold, and multiple local areas of the storage tank and multiple regional grayscale values ​​are determined. Step S300 further includes step S310, obtaining a predetermined grayscale threshold, wherein the predetermined grayscale threshold is set based on the minimum spraying accuracy of the spraying robot. Specifically, the predetermined grayscale threshold is read from the system parameter or configuration file, and this threshold is set according to the minimum spraying accuracy of the spraying robot and is used as a benchmark for image segmentation. Among them, the predetermined grayscale threshold is a grayscale value limit for image segmentation. When the grayscale value of a part of the image exceeds or falls below this threshold, the part will be regarded as a different area or object. The minimum spraying accuracy is the minimum spraying thickness adjustment unit that can be achieved by the spraying robot, reflecting the spraying control capability of the robot. For example, N times (N is a positive integer) of the minimum spraying accuracy is set as the minimum spraying adjustment thickness to ensure that the adjustment of the spraying thickness has practical significance. Step S320, grayscale value extraction is performed on the grayscale image of the storage tank to obtain the minimum grayscale value and the maximum grayscale value. Specifically, traverse each pixel of the grayscale image of the storage tank and extract its grayscale value (the brightness value of each pixel in the grayscale image, ranging from 0 to 255). During the traversal process, record and save the minimum and maximum grayscale values. Step S330, based on the predetermined grayscale threshold, the minimum grayscale value and the maximum grayscale value, segment the grayscale image of the storage tank to obtain multiple local areas of the storage tank and multiple regional grayscale values. Specifically, according to the predetermined grayscale threshold, the minimum grayscale value and the maximum grayscale value, the threshold segmentation method is used to divide the grayscale image of the storage tank into different areas. For each segmented area, the average value or median of its grayscale value is calculated as the grayscale value of the area. This implementation method divides different spraying areas according to the grayscale value difference of the grayscale image of the storage tank by setting a predetermined grayscale threshold, thereby improving the accuracy of the division and thus improving the control accuracy of the spraying thickness.

[0024] In a possible implementation, the grayscale image of the storage tank is segmented based on the predetermined grayscale threshold, the minimum grayscale value and the maximum grayscale value, and step S330 further includes step S331, taking the position coordinates of the minimum grayscale value as the starting point, simulating water filling to the grayscale image of the storage tank, recording the first grayscale value that is submerged by the water surface, and performing deviation calculation between the minimum grayscale value and the first grayscale value to obtain a first grayscale difference. Specifically, in the grayscale image of the storage tank, find the pixel corresponding to the minimum grayscale value, and record its position coordinates as the starting point of the simulated water filling. Starting from the starting point, "simulated water filling" is performed to the surrounding pixels, that is, the consideration range of the grayscale value is gradually expanded. During the water filling process, the grayscale value of each pixel point passed is recorded, and the deviation from the current minimum grayscale value is calculated. Step S332, if the first grayscale difference is greater than or equal to the predetermined grayscale threshold, a first segmentation line is set in this area, and if the first grayscale difference is less than the predetermined grayscale threshold, the simulated water filling continues. Specifically, when the deviation (first grayscale difference) between the first grayscale value and the minimum grayscale value during the water filling process reaches or exceeds the predetermined grayscale threshold, it is considered that this is a significant change point of the grayscale value, and a segmentation line is set at this significant change point to divide the image into different areas. If the first grayscale difference is less than the predetermined grayscale threshold, continue to fill water to find the next possible significant change point. Step S333, iteratively fill water until the water surface exceeds the maximum grayscale value, stop filling water and harvest multiple segmentation lines. Specifically, continue to simulate filling water, gradually expand the range of consideration until the water surface (i.e., the grayscale value range considered) covers the maximum grayscale value. In this process, the change of grayscale value is continuously recorded, and the segmentation line is set according to the predetermined grayscale threshold. Step S334, the grayscale image of the tank is segmented using the multiple segmentation lines, and multiple local areas of the tank and multiple regional grayscale values ​​are harvested, wherein the regional grayscale value is the grayscale intermediate value of the local area of ​​the tank. Specifically, according to the multiple segmentation lines, the grayscale image of the tank is segmented into multiple local areas. For each local area, the median value of its grayscale value is calculated as the regional grayscale value of the area. This implementation adopts the image segmentation method of simulating water filling, which can adaptively process the distribution of different grayscale values ​​and effectively segment images with complex grayscale value changes, ensuring the accuracy of the segmentation results and providing a reliable basis for the subsequent calculation and adjustment of the spraying thickness.

[0025] Step S400, obtain the target grayscale value of the target spraying thickness, and use the target grayscale value as a reference to perform deviation calculations on the grayscale values ​​of the multiple regions respectively, and determine multiple secondary spraying thicknesses. Specifically, the target grayscale value of the target spraying thickness is obtained through experiments or standard data. The target grayscale value is the grayscale value corresponding to the target spraying thickness and is the reference standard for the spraying operation. Based on the target grayscale value, the grayscale value of each local area of ​​the storage tank is calculated for deviation, and multiple regional grayscale differences (the difference between the actual grayscale value and the target grayscale value reflects the deviation of the spraying thickness) are obtained. Based on these grayscale differences, the secondary spraying thickness of each area (the thickness of the additional spraying required to achieve the target spraying thickness after the first spraying) is determined to compensate for the shortcomings of the first spraying.

[0026] In a possible implementation, the target grayscale value of the target spraying thickness is obtained, and step S400 further includes step S410, obtaining the standard grayscale value of the target spraying thickness under standard lighting parameters. Specifically, according to experimental or historical data, the grayscale value corresponding to the target spraying thickness under standard lighting conditions (such as specific lighting intensity, lighting angle and lighting color, etc.) is determined using standard spraying equipment and materials. This grayscale value is called the standard grayscale value. Step S420, obtain the ambient lighting parameters when the image of the spraying tank is collected once, perform deviation calculation on the ambient lighting parameters and the standard lighting parameters, and determine the difference in lighting parameters. Specifically, at the actual spraying operation site, use a lighting sensor or other equipment to measure the ambient lighting parameters when collecting the image of the spraying tank once. Compare the measured ambient lighting parameters with the standard lighting parameters, and calculate the difference between the two, that is, the difference in lighting parameters. Step S430, input the illumination parameter difference into the illumination corrector, and output the grayscale correction coefficient, wherein the illumination corrector is constructed based on the BP neural network and trained to convergence through sample data. Specifically, a lighting corrector based on the BP neural network is constructed. The BP neural network is a multi-layer feedforward neural network that is trained through a back propagation algorithm and can approximate complex nonlinear functions. The illumination corrector is trained using sample data (including grayscale values ​​under different illumination conditions and corresponding illumination parameter differences) until the network converges. In actual applications, the calculated illumination parameter difference is input into the trained illumination corrector, and the corresponding grayscale correction coefficient is output. The grayscale correction coefficient is a coefficient used to correct the grayscale value to eliminate the influence of the illumination conditions on the grayscale value measurement. Step S440, parameter correction is performed on the standard grayscale value according to the grayscale correction coefficient to obtain the target grayscale value. Specifically, the grayscale correction coefficient obtained in step 430 is used to correct the standard grayscale value obtained in step 410. The corrected grayscale value is the target grayscale value, which is used to compare with the grayscale values ​​of multiple regions of the primary spray tank image, calculate the deviation and determine the secondary spray thickness. This implementation method corrects the grayscale value through the illumination corrector, eliminates the influence of illumination conditions on the grayscale value measurement, and thus improves the accuracy of the grayscale value measurement.

[0027] In a possible implementation, the target grayscale value is used as a reference, and the deviation calculation is performed on the grayscale values ​​of the multiple regions respectively to determine multiple secondary spraying thicknesses. Step S400 further includes step S450, which is based on the target grayscale value and the deviation calculation is performed on the grayscale values ​​of the multiple regions respectively to determine multiple regional grayscale differences. Specifically, for each local area, the grayscale value is used as a reference to calculate the difference with the target grayscale value. This difference reflects the degree of deviation between the current grayscale of the area and the target grayscale, that is, the regional grayscale difference. Step S460, input the multiple regional grayscale differences into a predetermined grayscale difference-secondary spraying thickness comparison table for matching, and output the multiple secondary spraying thicknesses. Specifically, a grayscale difference-secondary spraying thickness comparison table is prepared in advance, which records the secondary spraying thicknesses corresponding to different grayscale differences. This table is obtained through experiments or historical data, and reflects the corresponding relationship between grayscale differences and secondary spraying thicknesses. For each area grayscale difference, the corresponding spraying thickness is found in the grayscale difference-secondary spraying thickness comparison table, and all the spraying thicknesses found are output as secondary spraying thicknesses to guide the secondary spraying operation. This implementation method directly converts the grayscale difference into the secondary spraying thickness through the grayscale difference-secondary spraying thickness comparison table, avoiding the error accumulation problem that may occur in traditional methods and improving the accuracy of the spraying operation.

[0028] Step S500, generates a secondary spraying path sequence and a secondary spraying parameter sequence according to the multiple local areas of the storage tanks and the multiple secondary spraying thicknesses, and based on the secondary spraying path sequence and the secondary spraying parameter sequence, controls the spraying robot to perform secondary spraying operations on the target storage tanks. Specifically, according to the determined multiple local areas of the storage tanks and the secondary spraying thicknesses, a path sequence and a parameter sequence for secondary spraying are generated. Among them, the secondary spraying path sequence refers to the order and path that the spraying robot needs to follow when performing secondary spraying. The secondary spraying parameter sequence refers to the various parameters that the spraying robot needs to follow when performing secondary spraying, such as spraying speed, spraying pressure, etc. According to these paths and parameters, the spraying robot performs secondary spraying operations on the target storage tank to achieve the target spraying thickness. The embodiment of the present application adopts a method of performing a spraying operation on a target tank according to initial spraying parameters, using a multi-eye camera to collect images of the target tank after the first spraying, and performing image stitching to generate a sprayed tank image, performing grayscale processing on the first sprayed tank image, segmenting the tank grayscale image according to a predetermined grayscale threshold, determining multiple local areas of the tank and multiple regional grayscale values, obtaining a target grayscale value of a target spraying thickness, and using this value as a reference to perform deviation calculation on the grayscale values ​​of multiple regions, determining multiple secondary spraying thicknesses, generating a secondary spraying path sequence and a secondary spraying parameter sequence according to multiple local areas of the tank and multiple secondary spraying thicknesses, and controlling a spraying robot to perform a secondary spraying operation on the target tank, and other technical means, thereby achieving the technical effect of improving positioning accuracy and achieving uniform distribution of spraying thickness.

[0029] like Figure 2As shown, in a possible implementation, a secondary spraying path sequence and a secondary spraying parameter sequence are generated according to the multiple local areas of the tanks and the multiple secondary spraying thicknesses, and step S500 further includes step S510, arranging the multiple local areas of the tanks according to the area from large to small, and generating a sequence of local areas of the tanks. Specifically, the area of ​​each local area of ​​the tank is measured by image processing technology (such as edge detection, contour extraction, etc.). According to the size of the area, these areas are sorted, and the principle of sorting is to arrange them according to the area from large to small, so as to ensure that in the secondary spraying operation, large areas are processed first, and then small areas are processed. The sorted areas are sequentially generated into a sequence of local areas of the tank, and this sequence is used to guide the secondary spraying path planning and parameter setting. Step S520, with the purpose of covering the sequence of local areas of the tanks, the operation simulation of the spraying robot is performed in the twin space of the spraying operation, and the secondary spraying path sequence is generated based on the simulation results. Specifically, a twin space of the spraying operation corresponding to the real spraying environment is constructed, which includes the geometric model of the tank, the motion model of the spraying robot, and environmental parameters. In the twin space, the operation of the spraying robot is simulated for the purpose of covering the local area sequence of the tank, including determining the parameters such as the robot's motion trajectory to ensure that all areas can be fully covered. According to the simulation results, a spraying path sequence is generated, which includes the motion trajectory of the spraying robot in each local area and related control instructions. Step S530, based on the local area sequence of the tank, a secondary spraying thickness sequence is generated according to the multiple secondary spraying thickness mappings. Specifically, each local area of ​​the tank is mapped to its corresponding secondary spraying thickness by establishing a mapping table or database, which includes the identification of each area and the corresponding spraying thickness. According to the order of the local area sequence of the tank, the spraying thickness of each area is extracted in turn, and a secondary spraying thickness sequence is generated, which is used to guide the setting of secondary spraying parameters. Step S540, the secondary spraying parameter sequence is generated according to the secondary spraying thickness sequence to meet the secondary spraying thickness as the operation constraint. Specifically, it is clear that the secondary spraying thickness is the main constraint of the spraying operation, that is, when generating the spraying parameters, it is necessary to ensure that the spraying thickness of each area meets the requirements. According to the secondary spraying thickness sequence and the performance parameters of the spraying robot (such as spraying speed, spraying flow, etc.), a set of suitable spraying parameters is generated for each area. These parameters can ensure that the required spraying thickness is achieved in each area. The spraying parameters of all areas are arranged in the order of the local area sequence of the tank to generate a secondary spraying parameter sequence, which is directly used to control the operation process of the spraying robot. This implementation method reduces the number of movements and spraying time of the spraying robot by processing large areas first and then small areas, thereby improving the spraying efficiency.

[0030] In a possible implementation, the secondary spraying parameter sequence is generated according to the secondary spraying thickness sequence, and step S540 further includes step S541, so as to meet the secondary spraying thickness as an operation constraint, and perform operation simulation in the spraying operation twin space based on the secondary spraying thickness sequence to generate an initial secondary spraying parameter sequence. Specifically, in the spraying operation twin space, according to the secondary spraying thickness sequence, the operation process of the spraying robot is simulated, including the setting of parameters such as spraying path, speed, and flow rate. According to the simulation results, a set of spraying parameters that can meet the secondary spraying thickness requirements is adjusted and determined to form an initial secondary spraying parameter sequence. Step S542, a spraying operation deviation analysis is performed based on the grayscale values ​​of the multiple local areas of the storage tanks and the multiple areas, and a spraying thickness deviation is determined. Specifically, the grayscale values ​​of multiple areas are compared with the expected target grayscale values, and the spraying thickness deviation of each area is calculated. According to the spraying thickness deviation of all areas, the overall deviation of a spraying operation is analyzed. Step S543, input the primary spraying thickness deviation into the spraying parameter compensation model, output the spraying parameter compensation coefficient, compensate the initial secondary spraying parameter sequence, and obtain the secondary spraying parameter sequence. Specifically, the primary spraying thickness deviation calculated in step 542 is used as input and input into the spraying parameter compensation model. The spraying parameter compensation model is a model based on machine learning technology, which is used to calculate the compensation coefficient of the spraying parameter according to the spraying thickness deviation. Using the machine learning algorithm, the model calculates the corresponding spraying parameter compensation coefficient based on the input deviation data. The calculated compensation coefficient is applied to the initial secondary spraying parameter sequence, and each parameter is adjusted accordingly to obtain the final secondary spraying parameter sequence. This implementation method accurately understands the problems existing in the spraying process through the deviation analysis after the primary spraying operation, and optimizes the secondary spraying parameters using the spraying parameter compensation model, thereby improving the spraying accuracy and spraying quality.

[0031] In a possible implementation, a spraying operation deviation analysis is performed based on the multiple local areas of the multiple tanks and the multiple regional grayscale values ​​to determine the deviation of the spraying thickness. Step S542 further includes step S5421, obtaining multiple regional areas of the multiple local areas of the multiple tanks and calculating multiple regional area ratios. Specifically, from the result of image segmentation, the area data of each local area of ​​the tank is extracted. The total area is obtained by adding up all the regional areas, and then the proportion of each regional area to the total area is calculated respectively to obtain multiple regional area ratios. Step S5422, multiplying the multiple regional area ratios and the multiple regional grayscale values, and summing them to obtain the grayscale mean of the spraying. Specifically, each regional area ratio is multiplied by the grayscale value of the region to obtain a weighted grayscale value. All weighted grayscale values ​​are added to obtain the grayscale mean of the spraying. Step S5423, the grayscale mean of the spraying is input into a predetermined grayscale-spraying thickness comparison table for matching, and the average of the spraying thickness is output. Specifically, in a predetermined grayscale-spraying thickness comparison table, the grayscale value closest to the grayscale mean of the first spraying and its corresponding spraying thickness are searched, and the spraying thickness found is output as the average of the first spraying thickness. Step S5424, the deviation between the first spraying thickness and the average of the first spraying thickness is calculated to obtain the deviation of the first spraying thickness. Specifically, the first spraying thickness is subtracted from the average of the first spraying thickness to obtain the deviation of the first spraying thickness, and the deviation of the first spraying thickness is output for parameter compensation of the second spraying. This implementation method takes into account the contribution of different areas to the overall grayscale when calculating the grayscale mean of the first spraying, so that the result is more accurate and reliable.

[0032] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for precise positioning and motion control of industrial robots based on visual perception, characterized in that: Methods include: According to the initial spraying parameters, the spraying robot is controlled to perform a spraying operation on the target storage tank, wherein the initial spraying parameters are set based on a spraying thickness, and the spraying thickness is 70% of the target spraying thickness; Use a multi-camera to collect images of the target tank after a spraying operation, and perform image stitching to generate a sprayed tank image; grayscale processing is performed on the once-sprayed tank image, and the tank grayscale image is segmented according to a predetermined grayscale threshold value to determine a plurality of tank local regions and a plurality of regional grayscale values; Obtaining a target grayscale value of the target spraying thickness, and using the target grayscale value as a reference, respectively performing deviation calculations on the grayscale values ​​of the multiple regions to determine multiple secondary spraying thicknesses; A secondary spraying path sequence and a secondary spraying parameter sequence are generated according to the multiple local areas of the tanks and the multiple secondary spraying thicknesses, and based on the secondary spraying path sequence and the secondary spraying parameter sequence, the spraying robot is controlled to perform a secondary spraying operation on the target tank.

2. The method for precise positioning and motion control of an industrial robot based on visual perception according to claim 1 is characterized in that: The grayscale image of the storage tank is segmented according to a predetermined grayscale threshold, and multiple local areas of the storage tank and multiple regional grayscale values ​​are determined, including: Obtaining a predetermined grayscale threshold, wherein the predetermined grayscale threshold is set based on a minimum spraying accuracy of the spraying robot; Extracting grayscale values ​​from the grayscale image of the storage tank to obtain a minimum grayscale value and a maximum grayscale value; Based on the predetermined grayscale threshold, the minimum grayscale value and the maximum grayscale value, the grayscale image of the storage tank is segmented to obtain a plurality of local areas of the storage tank and a plurality of regional grayscale values.

3. The method for precise positioning and motion control of an industrial robot based on visual perception according to claim 2 is characterized in that: Segmenting the tank grayscale image based on the predetermined grayscale threshold, the minimum grayscale value, and the maximum grayscale value includes: Taking the position coordinates of the minimum grayscale value as the starting point, simulate water filling into the grayscale image of the storage tank, record the first grayscale value submerged by the water surface, and calculate the deviation between the minimum grayscale value and the first grayscale value to obtain a first grayscale difference value; If the first grayscale difference is greater than or equal to the predetermined grayscale threshold, a first dividing line is set in this area; if the first grayscale difference is less than the predetermined grayscale threshold, the simulated irrigation is continued; Iteratively flooding the water until the water surface exceeds the maximum grayscale value, stopping flooding and harvesting multiple segmentation lines; The tank grayscale image is segmented using the multiple segmentation lines to obtain multiple tank local regions and multiple regional grayscale values, wherein the regional grayscale value is the grayscale median value of the local region of the tank.

4. The method for precise positioning and motion control of an industrial robot based on visual perception according to claim 1, characterized in that: Obtaining a target grayscale value of the target spraying thickness includes: Obtaining a standard grayscale value of the target spraying thickness under standard illumination parameters; Obtaining the ambient lighting parameters during the first spray tank image acquisition, performing deviation calculation on the ambient lighting parameters and standard lighting parameters, and determining the lighting parameter difference; Inputting the illumination parameter difference into an illumination corrector and outputting a grayscale correction coefficient, wherein the illumination corrector is constructed based on a BP neural network and trained to convergence through sample data; The standard grayscale value is parameter-corrected according to the grayscale correction coefficient to obtain the target grayscale value.

5. The method for precise positioning and motion control of an industrial robot based on visual perception according to claim 1, characterized in that: Based on the target grayscale value, deviation calculations are performed on the grayscale values ​​of the multiple regions respectively to determine multiple secondary spraying thicknesses, including: Based on the target grayscale value, respectively performing deviation calculations on the grayscale values ​​of the multiple regions to determine multiple regional grayscale difference values; The grayscale differences of the multiple regions are input into a predetermined grayscale difference-secondary spraying thickness comparison table for matching, and the multiple secondary spraying thicknesses are output.

6. The method for precise positioning and motion control of an industrial robot based on visual perception according to claim 1, characterized in that: Generating a secondary spraying path sequence and a secondary spraying parameter sequence according to the multiple local areas of the storage tank and the multiple secondary spraying thicknesses, including: Arranging the plurality of storage tank local regions according to the area from large to small to generate a storage tank local region sequence; For the purpose of covering the local area sequence of the storage tank, the operation simulation of the spraying robot is performed in the twin space of the spraying operation, and the secondary spraying path sequence is generated based on the simulation result; Based on the tank local area sequence, generating a secondary spraying thickness sequence according to the plurality of secondary spraying thickness mappings; Taking the secondary spraying thickness as an operation constraint, the secondary spraying parameter sequence is generated according to the secondary spraying thickness sequence.

7. The method for precise positioning and motion control of an industrial robot based on visual perception according to claim 6, characterized in that: Generating the secondary spraying parameter sequence according to the secondary spraying thickness sequence includes: Taking the secondary spraying thickness as the operation constraint, performing operation simulation in the spraying operation twin space based on the secondary spraying thickness sequence to generate an initial secondary spraying parameter sequence; Perform a spraying operation deviation analysis based on the grayscale values ​​of the multiple local areas and multiple areas of the tanks to determine a spraying thickness deviation; The primary spraying thickness deviation is input into a spraying parameter compensation model, a spraying parameter compensation coefficient is output, the initial secondary spraying parameter sequence is compensated, and the secondary spraying parameter sequence is obtained.

8. The method for precise positioning and motion control of an industrial robot based on visual perception according to claim 7 is characterized in that: A spraying operation deviation analysis is performed according to the gray values ​​of the multiple local areas of the tanks and the multiple areas to determine the spraying thickness deviation, including: Acquire multiple area areas of the multiple storage tank local areas, and calculate multiple area ratios; Performing a product operation on the area ratios of the multiple regions and the grayscale values ​​of the multiple regions, and summing them up to obtain a spraying grayscale mean value; Inputting the grayscale mean value of the spraying once into a predetermined grayscale-spraying thickness comparison table for matching, and outputting the grayscale mean value of the spraying once; The deviation between the primary spraying thickness and the primary spraying thickness mean is calculated to obtain the primary spraying thickness deviation.

Images