Automatic Valve Assembly and Control System Based on Intelligent Vision
By using intelligent vision-assisted detection and control, the problems of the angle between the valve stem axis and the valve body flow channel axis and uneven force distribution are solved, ensuring the perpendicularity of the valve stem to the valve body axis and uniform force distribution, thereby improving the valve assembly quality and reliability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG UNIV
- Filing Date
- 2026-01-29
- Publication Date
- 2026-05-05
AI Technical Summary
During valve assembly, the valve stem axis and the valve body flow channel axis are at an angle, which causes the actuator flange to be non-perpendicular to the valve body mounting surface. This results in the valve stem being tilted under stress, affecting the vertical movement of the valve core and easily leading to jamming.
An automatic control system for valve assembly based on intelligent vision is adopted, including a vertical judgment module, a uniformity analysis module, a superposition evaluation module, and an automatic control module. The system detects the angle between the valve stem axis and the valve body flow channel axis through intelligent vision, analyzes the uniformity of bolt stress, and performs automatic control to ensure that the valve stem axis coincides with the valve body flow channel axis and that the stress is uniform.
This achieves the alignment of the valve stem axis with the valve body flow channel axis and ensures uniform force distribution, avoiding the superposition effect caused by uneven force distribution, improving valve assembly quality and reliability, and reducing the risk of valve core jamming.
Smart Images

Figure CN121613796B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of industrial automatic assembly technology, specifically an automatic control system for valve assembly based on intelligent vision. Background Technology
[0002] In industrial production, valves, as key equipment for fluid control, directly affect the operational stability and reliability of the entire system through their assembly quality. Traditional valve assembly methods primarily rely on manual operation, with operators using their experience to judge and adjust the installation position and tightness of various components. However, with the continuous development of intelligent vision technology, its application in industrial assembly is becoming increasingly widespread. Intelligent vision technology, with its advantages of high precision and non-contact detection, has brought new solutions to valve assembly. Through intelligent vision systems, various image information during the valve assembly process can be acquired in real time and analyzed, thereby achieving precise monitoring and control of the assembly process.
[0003] In actual valve assembly, a critical issue that urgently needs to be addressed is the perpendicularity of the actuator to the valve body mounting surface. Due to the complexity of the valve's internal structure, the valve stem axis and the valve body flow channel axis must be highly aligned. However, during the tightening of the bolts on both sides, uneven force often occurs, causing the actuator flange to be non-perpendicular to the valve body mounting surface. This can also lead to mutual influence and even a superimposed effect, resulting in an angle between the valve stem axis and the valve body flow channel axis. This causes the valve stem to tilt under stress. When the valve is subsequently opened or closed, the valve core relies on the valve stem for linear reciprocating motion. However, the tilted valve stem causes the valve core's actual movement trajectory to be an inclined straight line, rather than a perpendicular straight line, ultimately leading to valve core jamming.
[0004] To address this, the present invention provides an automatic control system for valve assembly based on intelligent vision. Summary of the Invention
[0005] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.
[0006] The technical solution adopted by this invention to solve its technical problem is:
[0007] The valve assembly automatic control system based on intelligent vision includes:
[0008] Vertical judgment module: During the intelligent vision-assisted assembly of valves, it detects whether there is an angle between the valve stem axis and the valve body flow channel axis, and determines whether the actuator is perpendicular to the valve body mounting surface;
[0009] Uniformity Analysis Module: If not perpendicular, perform uniformity analysis on the bolts distributed on the same side of the actuator flange and the bolts on the symmetrical sides to determine whether the bolts distributed on the flange are subjected to uniform force.
[0010] Superposition evaluation module: If the bolts distributed on the flange are not under uneven stress, the tilt analysis is performed on the bolts distributed on the same side and the bolts distributed on both sides symmetrically, and the superposition effect of uneven stress is evaluated in combination.
[0011] Automatic control module: Based on the superposition effect of uneven force, analyze the correlation between the superposition tilt value and the angle between the axis, obtain the tilt angle correlation coefficient, and obtain the tilt angle adjustment amount based on the tilt angle correlation coefficient to control the current superposition tilt.
[0012] Preferably, the process for determining whether the actuator is perpendicular to the valve body mounting surface is as follows:
[0013] The positions of the valve stem and valve body flow channel inside the valve are obtained by using intelligent vision, spatial transformation is performed, and the axes of the valve stem and valve body flow channel are extracted respectively to obtain the valve stem axis and valve body flow channel axis.
[0014] If the valve stem axis and the valve body flow channel axis are not coincident and not parallel, the signal will be displayed as "valve installation is not perpendicular".
[0015] Preferably, a uniformity analysis is performed on the bolts distributed on the same side of the actuator flange, as follows:
[0016] Obtain the force value corresponding to each bolt distributed on the same side of the actuator flange, and sort the force values corresponding to each bolt according to the arrangement of the bolts on the actuator flange to obtain the force curve of the bolts on the same side;
[0017] Connect the starting and ending points of the bolt stress curves on the same side with straight lines to construct a uniform stress fitting line on the same side.
[0018] Obtain the vertical distance from each coordinate point on the bolt force curve on the same side to the uniform force fitting line on the same side, and calculate the average value by summing them up, and output the uniform force value on one side.
[0019] The uniform force values on each side of the actuator flange corresponding to the bolts are subtracted, and the absolute value is taken to obtain the uniform force value on the same side.
[0020] Preferably, a uniformity analysis is performed on the bolts on both symmetrical sides, as follows:
[0021] By combining the bolts symmetrically on both sides of the actuator flange, multiple sets of symmetrical force analysis groups are obtained;
[0022] Within the symmetrical force analysis group, the force value corresponding to each bolt is obtained, and the difference is calculated. The absolute value is then taken to obtain the force difference value of a single group.
[0023] The average value of the single-group force difference corresponding to each group of symmetrical force analysis is calculated by summing the values, and the uniform value of symmetrical force is output.
[0024] Preferably, the process for determining whether the bolts distributed on the flange are under uniform stress is as follows:
[0025] The uniform force value on the same side is summed with the uniform force value on the symmetrical side, and the uniform force distribution value is output.
[0026] If the uniformity of force distribution is greater than the threshold for uniform force distribution, it will be displayed as a non-uniform force signal on the bolt.
[0027] Preferably, the forces on the bolts on the same side of the actuator flange are analyzed to obtain the local tilt angle, as follows:
[0028] Extract the uniform force value on the same side corresponding to the bolts on each side of the actuator flange, compare the magnitudes, and select the side corresponding to the box with the largest uniform force on one side as the side with the most prominent force.
[0029] Mark each bolt on the stress-protruding monitoring point to obtain the stress-protruding bolts. Randomly select four stress-protruding monitoring points on each bolt, and use intelligent vision to obtain the three-dimensional coordinates of each monitoring point on the stress-protruding bolt in its uninstalled state. These coordinates are as follows: , , (), , , (), , , )and( , , );
[0030] Using intelligent vision, the three-dimensional coordinates of each stress monitoring point on the stressed protruding bolt in the installed state are obtained, respectively ( , , (), , , (), , , )and( , , );
[0031] Based on the three-dimensional coordinates of each stress monitoring point on the stressed protruding bolt in the uninstalled state, the least squares method is used to fit and construct a local tilt plane. The angle between the local tilt plane and the valve body mounting surface is obtained as the local tilt angle.
[0032] Preferably, an overall tilt analysis is performed on the bolts distributed on both symmetrical sides to obtain the overall tilt angle. The process is as follows:
[0033] Extract the force values corresponding to the bolts on the left and right sides within each symmetrical force analysis group, sum them separately to obtain the force values on the left and right sides, and then subtract them to obtain the force tendency value. If the force tendency value is positive, it is marked as overall force tilting; if the force tendency value is negative, it is marked as overall force tilting.
[0034] Obtain all bolts on the overall tilt of the load-bearing structure, extract the center of each circle on each bolt, and use a circle fitting algorithm to take each circle center as the core feature point;
[0035] Four edge feature points are randomly selected at the edge of the actuator flange. The entire circumference of the flange edge is scanned using intelligent vision to obtain the initial three-dimensional coordinates of each edge feature point before installation and stress. , , (), , , (), , , )and( , , );
[0036] The flange edge is scanned around its entire circumference using intelligent vision, and the three-dimensional coordinates of each edge feature point after installation under stress are obtained. , , (), , , (), , , )and( , , );
[0037] Based on the coordinates of all valid feature points, an overall tilt plane is constructed using the least squares method, and the angle between the overall tilt plane and the valve body mounting surface is obtained as the overall tilt angle.
[0038] Preferably, the evaluation of the cumulative effect of uneven force distribution is carried out as follows:
[0039] If the overall tilt angle and the local tilt angle are in the same direction, it will be displayed as an uneven superposition signal;
[0040] If the overall tilt angle and the local tilt angle are not in the same direction, it will be displayed as an uneven non-superimposed signal.
[0041] Preferably, the method for obtaining the correlation coefficient of the tilt angle is as follows:
[0042] The overall tilt angle and the local tilt angle are summed to obtain the superimposed tilt value. Based on the non-vertical installation signal of the valve, the angle between the valve stem axis and the valve body flow channel axis is obtained as the axis angle.
[0043] Extract the included angle of the axis that shows the non-perpendicular signal of valve installation in each historical assembly cycle, as well as the corresponding superimposed tilt value, and combine them into a set of coupling correlation analysis groups to obtain multiple coupling correlation analysis groups;
[0044] The angle between axes and the degree of superposition of tilt within all coupling correlation analysis groups are calculated using the Pearson correlation coefficient formula, and the absolute value is taken to output the correlation analysis value.
[0045] If the correlation analysis value is close to 1, it is considered a closely correlated signal, and the correlation analysis value is used as the tilt angle correlation coefficient.
[0046] Preferably, the tilt angle adjustment amount is obtained based on the tilt angle correlation coefficient, and the current superimposed tilt is adjusted accordingly, as follows:
[0047] The difference between the current angle between the axes and the extreme value of the angle between the axes is taken as the absolute value to obtain the angle difference between the axes.
[0048] The ratio of the difference in the included angle of the axis to the correlation coefficient of the included angle of tilt is calculated, and the amount of tilt angle adjustment is output.
[0049] The tilt angle adjustment amount is summed with the current superimposed tilt degree value, and the adjusted tilt degree value is output.
[0050] The beneficial effects of this invention are as follows:
[0051] 1. In the process of intelligent vision-assisted assembly of valves, this invention detects whether there is an angle between the valve stem axis and the valve body flow channel axis, and determines whether the actuator is perpendicular to the valve body mounting surface. If not perpendicular, it performs a uniformity analysis on the bolts distributed on the actuator flange to assess whether the force is uniform. From the perspective of installation accuracy, combined with the real-time monitoring and feedback of the installation angle by intelligent vision technology, the bolt tightening force can be adjusted in time to ensure that the actuator flange is perpendicular to the valve body mounting surface, thereby ensuring that the valve stem axis and the valve body flow channel axis coincide. Moreover, uniform force can reduce the superposition effect caused by uneven force, prevent the angle between the valve stem axis and the valve body flow channel axis, ensure the consistency of the valve stem and valve body flow channel axes, make the internal mechanical structure of the valve meet the design requirements, and improve the assembly quality of the valve.
[0052] 2. If the bolts distributed on the flange are not under uneven stress, the present invention performs an analysis of the uneven superposition effect of the bolt stress on the actuator flange, evaluates the uneven superposition effect, analyzes the correlation between the superposition tilt value and the axial angle based on the uneven superposition effect, obtains the tilt angle correlation coefficient, and obtains the tilt angle adjustment amount based on the tilt angle correlation coefficient, and performs adjustment operation on the current superposition tilt, so that the intelligent vision system can accurately calculate the angle that needs to be adjusted according to the specific axial angle, avoiding blind adjustment and effectively avoiding the valve stem tilt problem caused by valve body assembly tilt. Attached Figure Description
[0053] The invention will now be further described with reference to the accompanying drawings.
[0054] Figure 1 This is a schematic diagram of the valve assembly automatic control system based on intelligent vision according to the present invention;
[0055] Figure 2 This is a flowchart of the steps in the valve assembly automatic control system based on intelligent vision of the present invention;
[0056] Figure 3 This is a flowchart of the judgment steps in the valve assembly automatic control system based on intelligent vision according to the present invention. Detailed Implementation
[0057] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.
[0058] Example 1
[0059] Please see Figure 1 - Figure 3As shown, when installing the actuator flange to the valve body mounting surface, bolts are used to connect the actuator flange to the valve body. However, during the tightening of the bolts on both sides, it is important to note that if the bolts are subjected to uneven force during tightening, the actuator flange and the valve body mounting surface will not be perpendicular. Moreover, when the bolts are tightened unevenly, there will be mutual influence and even superimposed effects, resulting in an angle between the valve stem axis and the valve body flow channel axis. This causes the valve stem inside the valve to be tilted under force. When the valve is subsequently opened and closed, the valve core inside the valve relies on the valve stem for linear reciprocating motion. However, the tilted force on the valve stem causes the actual movement trajectory of the valve core inside the valve to be an inclined straight line, not a perpendicular straight line, ultimately leading to valve core jamming. Therefore, the automatic control system for valve assembly based on intelligent vision described in this embodiment of the invention includes the following steps:
[0060] Vertical judgment module: During the intelligent vision-assisted assembly of valves, it detects whether there is an angle between the valve stem axis and the valve body flow channel axis, and determines whether the actuator is perpendicular to the valve body mounting surface;
[0061] In some embodiments, intelligent vision is used to obtain the positions of the valve stem and valve body flow channel inside the valve, spatial transformation is performed, and the axes of the valve stem and valve body flow channel are extracted respectively to obtain the valve stem axis and valve body flow channel axis.
[0062] The valve stem axis and the valve body flow channel axis are overlapped and compared in parallel. The process is as follows:
[0063] If the valve stem axis and the valve body flow channel axis are completely coincident and parallel, it indicates that the axis and the valve body flow channel axis are on a straight line, which shows the valve installation vertical signal;
[0064] If the valve stem axis and the valve body flow channel axis are not coincident and not parallel, it indicates that there is an angle between the axis and the valve body flow channel axis, which is displayed as a signal that the valve is not installed perpendicularly.
[0065] It should be noted that the reason for checking whether there is an angle between the valve stem axis and the valve body flow channel axis is as follows: When installing the actuator flange to the valve body mounting surface, bolts are used to connect the actuator flange and the valve body. If the force is uneven when tightening the bolts on both sides, it will have several adverse effects: First, it will cause the actuator flange and the valve body mounting surface to be non-perpendicular; second, this uneven force will also have mutual influence and even superimposed effects, which will cause an angle between the valve stem axis and the valve body flow channel axis. Since the valve core inside the valve relies on the valve stem to make linear reciprocating motion, when there is an angle between the valve stem axis and the valve body flow channel axis, the valve stem will be tilted under force, and the actual movement trajectory of the valve core will become an inclined straight line, instead of a vertical straight line.
[0066] Uniformity Analysis Module: If not perpendicular, perform uniformity analysis on the bolts distributed on the same side of the actuator flange and the bolts on the symmetrical sides to determine whether the bolts distributed on the flange are subjected to uniform force.
[0067] In some embodiments, the force value corresponding to each bolt distributed on the same side of the actuator flange is obtained, and the force value corresponding to each bolt is sorted according to the arrangement of the bolts on the actuator flange to obtain the force curve of the bolts on the same side.
[0068] Connect the starting and ending points of the bolt stress curves on the same side with straight lines to construct a uniform stress fitting line on the same side.
[0069] Obtain the vertical distance from each coordinate point on the bolt force curve on the same side to the uniform force fitting line on the same side, and calculate the average value by summing them up, and output the uniform force value on one side.
[0070] The uniform force values on each side of the actuator flange corresponding to the bolts are subtracted, and the absolute value is taken to obtain the uniform force value on the same side.
[0071] By combining the bolts symmetrically on both sides of the actuator flange, multiple sets of symmetrical force analysis groups are obtained;
[0072] Within the symmetrical force analysis group, the force value corresponding to each bolt is obtained, and the difference is calculated. The absolute value is then taken to obtain the force difference value of a single group.
[0073] The average value of the single force difference corresponding to each group of symmetrical force analysis is calculated by summing the values, and the uniform value of symmetrical force is output.
[0074] The uniform force value on the same side is summed with the uniform force value on the symmetrical side, and the uniform force distribution value is output.
[0075] It is understandable that the uniformity of force distribution means that it is a comprehensive index calculated to evaluate whether the bolt force is uniform during the bolt connection process between the flange of the valve actuator and the valve body mounting surface. Specifically, on the one hand, the uniformity of the force on the same side reflects the uniformity of the bolt force on the same side; on the other hand, the uniformity of the force on the symmetrical bolts on both sides of the flange reflects the uniformity of the bolt force on both sides of the flange.
[0076] If the uniform force distribution value is greater than the uniform force distribution threshold, it indicates that the bolts on the same side of the flange are not uniformly stressed, and the bolts on both symmetrical sides are also not uniformly stressed, which is displayed as a non-uniform bolt force signal.
[0077] If the uniform force distribution value is less than or equal to the uniform force distribution threshold, it indicates that the bolts on the same side of the flange are subjected to uniform force, and the bolts on both symmetrical sides are subjected to uniform force, which is displayed as a bolt force uniformity signal.
[0078] It should be noted that the purpose of assessing whether the force is uniform is as follows: From the perspective of installation accuracy, by assessing the uniformity of bolt force, combined with real-time monitoring and feedback of the installation angle using intelligent vision technology, the bolt tightening force can be adjusted in a timely manner to ensure that the actuator flange is perpendicular to the valve body mounting surface, thereby ensuring that the valve stem axis coincides with the valve body flow channel axis. Moreover, uniform force can reduce the superposition effect caused by uneven force, prevent the angle between the valve stem axis and the valve body flow channel axis, ensure the consistency of the valve stem and valve body flow channel axes, make the internal mechanical structure of the valve meet the design requirements, and improve the assembly quality of the valve.
[0079] From the perspective of intelligent vision installation and control, the data obtained by evaluating the uniformity of bolt force, such as the uniformity of force on the same side, the uniformity of force on the symmetry, and the uniformity of force distribution, can provide important feedback information for the intelligent vision valve assembly automatic control system, and adjust the visual detection parameters and control strategies in real time based on this data.
[0080] The specific solution in this embodiment is as follows: During the intelligent vision-assisted assembly of the valve, the system detects whether there is an angle between the valve stem axis and the valve body flow channel axis, and determines whether the actuator is perpendicular to the valve body mounting surface. If not perpendicular, the uniformity of the bolts distributed on the actuator flange is analyzed to assess whether the force is uniform. From the perspective of installation accuracy, combined with the real-time monitoring and feedback of the installation angle using intelligent vision technology, the bolt tightening force can be adjusted in a timely manner to ensure that the actuator flange is perpendicular to the valve body mounting surface, thereby ensuring that the valve stem axis coincides with the valve body flow channel axis. Moreover, uniform force can reduce the superposition effect caused by uneven force, prevent the angle between the valve stem axis and the valve body flow channel axis, ensure the consistency of the valve stem and valve body flow channel axes, make the internal mechanical structure of the valve meet the design requirements, and improve the assembly quality of the valve.
[0081] Example 2
[0082] Please see Figure 1 - Figure 3 As shown in the figure, the automatic control system for valve assembly based on intelligent vision according to an embodiment of the present invention includes the following steps:
[0083] Superposition evaluation module: If the bolts on the flange are not under uneven stress, the superposition effect analysis of the uneven stress on the bolts on the actuator flange is performed to evaluate the superposition effect of uneven stress.
[0084] In some embodiments, a local tilt analysis is performed on bolts distributed on the same side to obtain the local tilt angle. The process is as follows:
[0085] Extract the uniform force value on the same side corresponding to the bolts on each side of the actuator flange, compare the magnitudes, and select the side corresponding to the box with the largest uniform force on one side as the side with the most prominent force.
[0086] It should be noted that if the uniform force on one side of the bolts on the right side of the actuator flange is greater than the uniform force on one side of the bolts on the left side of the actuator flange, it indicates that the force unevenness on the bolts on the right side of the actuator flange is greater, and the side with greater force unevenness is taken as the side with the greater force.
[0087] If the uniform force on the left side of the actuator flange is greater than the uniform force on the right side of the actuator flange, it indicates that the uneven force distribution on the bolts on the left side of the actuator flange is greater, and the side with greater uneven force distribution is taken as the side with the more prominent force distribution.
[0088] Mark each bolt on the stress protrusion test to obtain the stress protrusion bolt. Randomly select four stress protrusion monitoring points on the stress protrusion bolt, wherein these four stress protrusion monitoring points are symmetrical about the center of the bolt cross section (circular cross section) in pairs;
[0089] It should be noted that the four stress protrusion monitoring points are arranged symmetrically with respect to the center of the bolt cross section (circular cross section). The purpose is that if the four stress protrusion monitoring points are not distributed evenly, they will be concentrated on the same side, resulting in fitting deviation. In addition, the four stress protrusion monitoring points are located between the actuator flange and the bolt.
[0090] Using intelligent vision (binocular camera + structured light), the three-dimensional coordinates of each stress monitoring point on a stressed protruding bolt in an uninstalled state are obtained, respectively ( , , (), , , (), , , )and( , , );
[0091] in,( , , () represents the three-dimensional coordinates of the first stress-protruding monitoring point in the uninstalled state. , , () represents the three-dimensional coordinates of the second stress-protruding monitoring point in the uninstalled state. , , ) represents the three-dimensional coordinates of the third stress-protruding monitoring point in the uninstalled state, and ( , , () represents the three-dimensional coordinates of the fourth stress-protruding monitoring point in the uninstalled state;
[0092] Similarly, using intelligent vision (binocular camera + structured light), the three-dimensional coordinates of each stress monitoring point on the stressed protruding bolt in the completed installation state are obtained, respectively ( , , (), , , (), , , )and( , , );
[0093] in,( , , () represents the three-dimensional coordinates of the first stress-protrusion monitoring point after installation is complete. , , () represents the three-dimensional coordinates of the second stress-protrusion monitoring point after installation is complete. , , ) represents the three-dimensional coordinates of the third stress-protrusion monitoring point after installation is complete, and ( , , The coordinates of the fourth stress-protruding monitoring point are shown in the three-dimensional coordinates after installation.
[0094] Based on the three-dimensional coordinates of each stress monitoring point on the stressed protruding bolt in its uninstalled state, the initial plane fitting equation is constructed using the least squares method: ;
[0095] The specific fitting steps are as follows: S1, calculate the mean coordinate value of all force monitoring points ( , , );
[0096] S2, construct the covariance matrix, solve for the eigenvalues and eigenvectors of the matrix, where the eigenvector corresponding to the smallest eigenvalue is the normal vector of the plane, and obtain (A, B, C).
[0097] S3, the mean coordinates ( , , Substitute the values into the plane equation, solve for the value of D, determine the complete equation of the locally tilted plane, and obtain the locally tilted plane.
[0098] Obtain the angle between the locally tilted plane and the valve body mounting surface as the locally tilted angle;
[0099] A global tilt analysis was performed on the bolts distributed on both symmetrical sides to obtain the overall tilt angle. The process is as follows:
[0100] Extract the force values corresponding to the bolts on the left and right sides within each symmetrical force analysis group, sum them up to obtain the force values on the left and right sides, and then subtract them to obtain the force tendency value.
[0101] If the force tendency value is positive, it indicates that the right side of the whole body is inclined, and it is marked as the whole body tilting under force.
[0102] If the force tendency value is negative, it indicates that the left side is inclined as a whole, and it is marked as the force-bearing body tilting.
[0103] Obtain all bolts on the overall tilt of the load-bearing structure, extract the center of each circle on each bolt, and use a circle fitting algorithm to take each circle center as the core feature point;
[0104] Four edge feature points are randomly selected at the edge of the actuator flange. The entire circumference of the flange edge is scanned using intelligent vision (binocular camera + structured light) to obtain the initial three-dimensional coordinates of each edge feature point before installation and stress. , , (), , , (), , , )and( , , );
[0105] in,( , , ) represents the three-dimensional coordinates of the first edge feature point before the installation force is applied. , , ) represents the three-dimensional coordinates of the second edge feature point before the installation force was applied. , , ) represents the three-dimensional coordinates of the third edge feature point before the installation force was applied. , , () represents the three-dimensional coordinates of the fourth edge feature point before installation and application of force;
[0106] Similarly, intelligent vision (binocular camera + structured light) is used to scan the entire circumference of the flange edge, and the three-dimensional coordinates of each edge feature point after installation under stress are obtained. , , (), , , (), , , )and( , , );
[0107] in,( , , ) represents the three-dimensional coordinates of the first edge feature point after installation under force. , , ) represents the three-dimensional coordinates of the second edge feature point after installation under force. , , ) represents the three-dimensional coordinates of the third edge feature point after installation under force. , , The coordinates of the fourth edge feature point after installation under stress are represented as follows:
[0108] Using the three-dimensional plane general equation ax + by + cz + d = 0, based on the coordinates of all effective feature points, the plane coefficients a, b, c, and d are solved by minimizing the sum of squares of the perpendicular distances from each point to the plane using the least squares method.
[0109] Specifically, J1 calculates the mean coordinates of all valid feature points to obtain ( , , );
[0110] J2, construct the covariance matrix M, solve for the eigenvalues and eigenvectors of the covariance matrix M, and the eigenvector corresponding to the smallest eigenvalue is the plane normal vector (a,b,c).
[0111] J3, the mean coordinates ( , , Substituting the equation ax + by + cz + d = 0 into the plane equation, we solve for d and finally determine the complete plane equation of the overall end face, thus obtaining the overall inclined plane.
[0112] Obtain the angle between the overall tilt plane and the valve body mounting surface as the overall tilt angle;
[0113] If the overall tilt angle and the local tilt angle are in the same direction, it means that the overall tilt and the local tilt are in the same direction, which is displayed as an uneven superposition signal. The overall tilt angle and the local tilt angle are summed to obtain the superposition tilt degree value.
[0114] If the overall tilt angle and the local tilt angle are not in the same direction, it means that the overall tilt and the local tilt are not in the same direction, which is displayed as an uneven non-superimposed signal;
[0115] It should be noted that the purpose of assessing the degree of superimposed effect of uneven force is to: from the perspective of positioning and installation accuracy, by performing local tilt analysis on bolts distributed on the same side to obtain the local tilt angle, and by performing overall tilt analysis on bolts distributed on both sides symmetrically to obtain the overall tilt angle, it is possible to clarify whether the problem is caused by uneven force on local bolts or uneven force on the whole, which helps to adjust the bolt tightening force in a targeted manner and improve installation accuracy.
[0116] From the perspective of intelligent vision installation and control, the data obtained by assessing the degree of superimposed effect of uneven force, such as local tilt angle, overall tilt angle, and the degree of superimposed uneven force, can provide important feedback information for the intelligent vision system. This provides a basis for adjusting the parameters of visual detection in real time, and automatically adjusts the installation process based on the data detected in real time, continuously optimizing the installation quality until the preset installation standards are met. This improves the automation level and production efficiency of valve assembly and reduces the need for manual intervention.
[0117] Automatic control module: Based on the superposition effect of uneven force, analyze the correlation between the superposition tilt value and the angle between the axis, obtain the tilt angle correlation coefficient, and obtain the tilt angle adjustment amount based on the tilt angle correlation coefficient to control the current superposition tilt.
[0118] In some embodiments, the included angle of the axes is obtained as follows:
[0119] Based on the non-vertical installation signal of the valve, the angle between the valve stem axis and the valve body flow channel axis is obtained as the axis angle;
[0120] Extract the included angle of the axis that shows the non-perpendicular signal of valve installation in each historical assembly cycle, as well as the corresponding superimposed tilt value, and combine them into a set of coupling correlation analysis groups to obtain multiple coupling correlation analysis groups;
[0121] It should be noted that the historical assembly cycle refers to the time previously required to assemble the valve;
[0122] The angle between axes and the degree of superposition of tilt within all coupling correlation analysis groups are calculated using the Pearson correlation coefficient formula, and the absolute value is taken to output the correlation analysis value.
[0123] If the correlation analysis value is close to 0, it indicates that the correlation between the axial angle and the superimposed tilt value is not strong, showing a weak correlation signal.
[0124] If the correlation analysis value is close to 1, it indicates that there is a close correlation between the axial angle and the superimposed tilt value, which is a signal of close correlation. The correlation analysis value is used as the correlation coefficient of the tilt angle.
[0125] The difference between the current angle between the axes and the extreme value of the angle between the axes is taken as the absolute value to obtain the angle difference between the axes.
[0126] The ratio of the difference in the included angle of the axis to the correlation coefficient of the included angle of tilt is calculated, and the amount of tilt angle adjustment is output.
[0127] The tilt angle adjustment is summed with the current superimposed tilt value, and the adjusted tilt value is output.
[0128] It should be noted that the purpose of adjusting the current overlay tilt is as follows: when the axial angle is closely related to the overlay tilt degree, the correlation analysis value is used as the tilt angle correlation to reflect the degree of influence of the axial angle change on the overlay tilt degree. The axial angle difference is obtained by subtracting the current axial angle from its limit value, and then the ratio of the axial angle difference to the tilt angle correlation coefficient is calculated to output the tilt angle adjustment amount. This quantitative calculation method enables the intelligent vision system to accurately calculate the angle to be adjusted based on the specific axial angle, avoiding blind adjustments and improving the accuracy and efficiency of the adjustment.
[0129] From the perspective of optimized assembly, it effectively avoids the problem of valve stem tilting caused by valve body tilting. When the valve stem is subjected to vertical force, the valve core inside the valve can move back and forth in a vertical straight line, reducing the risk of valve core jamming, ensuring the valve's opening and closing control performance, and improving the valve's reliability and service life.
[0130] The specific solution in this embodiment is as follows: If the bolts distributed on the flange are subjected to uneven force, the uneven superposition effect of the bolt force on the actuator flange is analyzed to evaluate the uneven superposition effect. Based on the uneven superposition effect, the correlation between the superposition tilt value and the axial angle is analyzed to obtain the tilt angle correlation coefficient. Based on the tilt angle correlation coefficient, the tilt angle adjustment amount is obtained, and the current superposition tilt is adjusted. This allows the intelligent vision system to accurately calculate the angle that needs to be adjusted based on the specific axial angle, avoiding blind adjustment and effectively avoiding the valve stem tilt problem caused by valve body assembly tilt.
[0131] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.
Claims
1. An automatic control system for valve assembly based on intelligent vision, characterized in that: include: Vertical judgment module: During the intelligent vision-assisted assembly of valves, it detects whether there is an angle between the valve stem axis and the valve body flow channel axis, and determines whether the actuator is perpendicular to the valve body mounting surface; Uniformity Analysis Module: If not perpendicular, perform uniformity analysis on the bolts distributed on the same side of the actuator flange and the bolts on the symmetrical sides to determine whether the bolts distributed on the flange are subjected to uniform force. Superposition evaluation module: If the bolts distributed on the flange are not subjected to uneven stress, the tilt analysis is performed on the bolts distributed on the same side and the bolts distributed on both sides symmetrically, and the superposition effect of uneven stress is evaluated. Automatic control module: Based on the superposition effect of uneven force, analyze the correlation between the superposition tilt value and the angle between the axis, obtain the tilt angle correlation coefficient, and obtain the tilt angle adjustment amount based on the tilt angle correlation coefficient to control the current superposition tilt.
2. The automatic control system for valve assembly based on intelligent vision according to claim 1, characterized in that: The procedure for determining whether the actuator is perpendicular to the valve body mounting surface is as follows: The positions of the valve stem and valve body flow channel inside the valve are obtained by using intelligent vision, spatial transformation is performed, and the axes of the valve stem and valve body flow channel are extracted respectively to obtain the valve stem axis and valve body flow channel axis. If the valve stem axis and the valve body flow channel axis are not coincident and not parallel, the signal will be displayed as "valve installation is not perpendicular".
3. The automatic control system for valve assembly based on intelligent vision according to claim 1, characterized in that: The uniformity analysis of bolts distributed on the same side of the actuator flange is performed as follows: Obtain the force value corresponding to each bolt distributed on the same side of the actuator flange, and sort the force values corresponding to each bolt according to the arrangement of the bolts on the actuator flange to obtain the force curve of the bolts on the same side; Connect the starting and ending points of the bolt stress curves on the same side with straight lines to construct a uniform stress fitting line on the same side. Obtain the vertical distance from each coordinate point on the bolt force curve on the same side to the uniform force fitting line on the same side, and calculate the average value by summing them up, and output the uniform force value on one side. The uniform force values on each side of the actuator flange corresponding to the bolts are subtracted, and the absolute value is taken to obtain the uniform force value on the same side.
4. The automatic control system for valve assembly based on intelligent vision according to claim 3, characterized in that: The uniformity analysis of the bolts on both symmetrical sides is performed as follows: By combining the bolts symmetrically on both sides of the actuator flange, multiple sets of symmetrical force analysis groups are obtained; Within the symmetrical force analysis group, the force value corresponding to each bolt is obtained, and the difference is calculated. The absolute value is then taken to obtain the force difference value of a single group. The average value of the single-group force difference corresponding to each group of symmetrical force analysis is calculated by summing the values, and the uniform value of symmetrical force is output.
5. The automatic control system for valve assembly based on intelligent vision according to claim 4, characterized in that: The process for determining whether the bolts distributed on the flange are under uniform stress is as follows: The uniform force value on the same side is summed with the uniform force value on the symmetrical side, and the uniform force distribution value is output. If the uniformity of force distribution is greater than the threshold for uniform force distribution, it will be displayed as a non-uniform force signal on the bolt.
6. The automatic control system for valve assembly based on intelligent vision according to claim 1, characterized in that: The local tilt angle is obtained by analyzing the forces on the bolts on the same side of the actuator flange, as follows: Extract the uniform single-sided force value corresponding to the bolts on each side of the actuator flange, compare the magnitudes, and select the side corresponding to the largest uniform single-sided force value as the side with the most prominent force. Mark each bolt on the stress-protruding monitoring point to obtain the stress-protruding bolts. Randomly select four stress-protruding monitoring points on each bolt, and use intelligent vision to obtain the three-dimensional coordinates of each monitoring point on the stress-protruding bolt in its uninstalled state. These coordinates are as follows: , , (), , , (), , , )and( , , ); Using intelligent vision, the three-dimensional coordinates of each stress monitoring point on the stressed protruding bolt in the installed state are obtained, respectively ( , , (), , , (), , , )and( , , ); Based on the three-dimensional coordinates of each stress monitoring point on the stressed protruding bolt in the uninstalled state, the least squares method is used to fit and construct a local tilt plane. The angle between the local tilt plane and the valve body mounting surface is obtained as the local tilt angle.
7. The automatic control system for valve assembly based on intelligent vision according to claim 1, characterized in that: A global tilt analysis was performed on the bolts distributed on both symmetrical sides to obtain the overall tilt angle. The process is as follows: Extract the force values corresponding to the bolts on the left and right sides within each symmetrical force analysis group, sum them separately to obtain the force values on the left and right sides, and then subtract them to obtain the force tendency value. If the force tendency value is positive, it is marked as overall force tilting; if the force tendency value is negative, it is marked as overall force tilting. Obtain all bolts on the overall tilt of the load-bearing structure, extract the center of each circle on each bolt, and use a circle fitting algorithm to take each circle center as the core feature point; Four edge feature points are randomly selected at the edge of the actuator flange. The entire circumference of the flange edge is scanned using intelligent vision to obtain the initial three-dimensional coordinates of each edge feature point before installation and stress. , , (), , , (), , , )and( , , ); The flange edge is scanned around its entire circumference using intelligent vision, and the three-dimensional coordinates of each edge feature point after installation under stress are obtained. , , (), , , (), , , )and( , , ); Based on the coordinates of all valid feature points, an overall tilt plane is constructed using the least squares method, and the angle between the overall tilt plane and the valve body mounting surface is obtained as the overall tilt angle.
8. The automatic control system for valve assembly based on intelligent vision according to claim 1, characterized in that: The process for assessing the cumulative effect of uneven force distribution is as follows: If the overall tilt angle and the local tilt angle are in the same direction, it will be displayed as an uneven superposition signal; If the overall tilt angle and the local tilt angle are not in the same direction, it will be displayed as an uneven non-superimposed signal.
9. The automatic control system for valve assembly based on intelligent vision according to claim 1, characterized in that: The method for obtaining the correlation coefficient of the tilt angle is as follows: The overall tilt angle and the local tilt angle are summed to obtain the superimposed tilt value. Based on the non-vertical installation signal of the valve, the angle between the valve stem axis and the valve body flow channel axis is obtained as the axis angle. Extract the included angle of the axis that shows the non-perpendicular signal of valve installation in each historical assembly cycle, as well as the corresponding superimposed tilt value, and combine them into a set of coupling correlation analysis groups to obtain multiple coupling correlation analysis groups; The angle between axes and the degree of superposition of tilt within all coupling correlation analysis groups are calculated using the Pearson correlation coefficient formula, and the absolute value is taken to output the correlation analysis value. If the correlation analysis value is close to 1, it is considered a closely correlated signal, and the correlation analysis value is used as the tilt angle correlation coefficient.
10. The automatic control system for valve assembly based on intelligent vision according to claim 9, characterized in that: The tilt angle adjustment amount is obtained based on the tilt angle correlation coefficient, and the current superimposed tilt is adjusted accordingly. The process is as follows: The difference between the current angle between the axes and the extreme value of the angle between the axes is taken as the absolute value to obtain the angle difference between the axes. The ratio of the difference in the included angle of the axis to the correlation coefficient of the included angle of tilt is calculated, and the amount of tilt angle adjustment is output. The tilt angle adjustment amount is summed with the current superimposed tilt degree value, and the adjusted tilt degree value is output.
Citation Information
Patent Citations
Sealing valve
CN115978217A
Automatic assembling system and method for large nut of water turbine ball valve
CN120395406A