Full-position intelligent positioning system for large rotating equipment

Abstract

The present application relates to the technical field of device positioning, in particular to a large rotating device full-position intelligent positioning system, comprising: determining a three-dimensional coordinate set of each search direction according to a reference coordinate system of a current rotating device; calculating an attitude angle deviation of a current search direction and a preset search direction, correlating a position deviation through the attitude angle deviation, and configuring a position deviation matrix of each reference point; calculating a deviation probability of each reference point under a combination of multiple search directions, and combining into a deviation probability matrix; correlating the deviation probability of each reference point with each search direction according to the deviation probability matrix, coordinating and optimizing the probability distribution of the search direction through the relative spatial coordinates between each reference point; mapping the optimized probability distribution of each search direction into the reference coordinate system, correcting the three-dimensional coordinates of each search direction, and obtaining the final output positioning result. The accuracy and efficiency of rotating device positioning are realized.

Description

Technical Field

[0001] This invention relates to the field of equipment positioning technology, specifically to an intelligent positioning system for all positions of large rotating equipment. Background Technology

[0002] The positioning of large rotating equipment mainly relies on precision measuring equipment such as laser trackers, total stations, and laser displacement sensors. By detecting the three-dimensional coordinates of key reference points such as couplings, bearing seats, and cylinder flanges in multiple search directions, the position and attitude of the equipment can be identified.

[0003] For example, Chinese Patent Publication No. CN119045027A discloses a target positioning method based on multi-source information fusion, which relates to the field of radio direction finding technology. The method includes the following steps: acquiring the target's location information based on the collected GPS positioning system; simultaneously acquiring the location information of two sets of airborne radio ranging devices; acquiring the distance information between each airborne radio ranging device and the target location; calculating the actual data of the target location based on the distance data measured and collected by the airborne radio ranging devices; outputting the fused target positioning data; and collecting the distance information between the two sets of airborne radio ranging devices to verify the acquired target positioning data.

[0004] For example, Chinese Patent Publication No. CN118566842A discloses a target positioning method, device, medium, equipment, and product, which relates to the field of underwater positioning technology. The method includes: obtaining the initial position of the target to be positioned based on the positions of multiple first array elements and the circular intersection method; if the distance between the initial position and the center of the array elements of the multiple first array elements is less than a preset distance, using the initial position as the initial position value of the target to be positioned in the adjustment model of the target to be positioned; and correcting the initial position value based on the adjustment model to obtain the actual position of the target to be positioned.

[0005] Existing technologies primarily use distance data for target localization, completing the localization of multiple sets of radio wave ranging devices; or they solve the equation of a circle under the intersection of horizontal baselines to achieve target localization by finding the intersection of regions under multi-angle refraction. However, in the localization scenario of rotating equipment, in addition to determining the coordinates of its corresponding position based on angle and distance values, it is also necessary to perform correlation and deviation retrieval based on the search direction during localization to reduce deviations in target localization and thus improve the accuracy and precision of localization. Summary of the Invention

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a large rotating equipment all-position intelligent positioning system, including: an initial coordinate layout module, used to perform parallel retrieval of the three-dimensional coordinates of each reference point according to the reference coordinate system of the current rotating equipment, and determine the three-dimensional coordinate set of each search direction.

[0007] The deviation identification module is used to calculate the attitude angle deviation between the current search direction and the preset search direction based on the three-dimensional coordinate set of each search direction, associate the attitude angle deviation with the position deviation, and configure the position deviation matrix of each reference point.

[0008] The probability configuration module is used to calculate the deviation probability of each reference point under multiple search direction combinations based on the position deviation matrix of each reference point, and combine the deviation probabilities of all reference points into a deviation probability matrix using the interval range corresponding to the deviation probability.

[0009] The pointing analysis module is used to associate the deviation probability of each reference point with each search direction based on the deviation probability matrix, and to coordinate and optimize the probability distribution of the search direction using the relative spatial coordinates between each reference point.

[0010] The positioning correction module is used to map the optimized probability distribution of each search direction to the reference coordinate system, correct the three-dimensional coordinates of each search direction, and obtain the final output positioning result.

[0011] The beneficial effects of this invention are as follows: First, this invention establishes a unified reference coordinate system with the axis of the rotating equipment as the origin, uses a preset normal vector as the fitting reference, and employs the least squares method to fit the spatial plane of the reference point in a single search direction. The propagation region of the search direction is defined by the fitted normal vector, and the attitude angle deviation between the fitted plane and the preset search direction is calculated and correlated with the position deviation, ultimately generating a position deviation matrix for each reference point. This improves the usability and accuracy of the deviation data, allowing the relative deviation of each plane to be mapped to the global coordinate system, further conforming to the planar distribution of the rotating equipment at various structural locations, and improving the fitting effect of the current data.

[0012] Second, this invention performs consistency verification by calculating the cosine similarity of the fitted normal vector and the preset normal vector. After the verification is passed, the symmetry deviation between the reference point and the center of the fitted plane is calculated, and a symmetry calibration matrix is ​​generated. Principal component analysis is used to extract the principal axis direction of the calibration matrix to determine the propagation area of ​​the search direction in three-dimensional space. This ensures the consistency of the calibration of the current propagation area, making the obtained deviation data more reliable and further reducing the positional deviation of the identification at each search direction, so that the subsequent adjustment can fit the rotating equipment scenario as closely as possible.

[0013] Third, this invention determines the dominant search direction when multiple search directions point to the same reference point, compares the positional deviations of the dominant direction with those of other search directions to prevent cross-offsets during data comparison, and uses the minimum sum of positional deviations as a constraint to ensure that the output deviation data conforms to the rigidity characteristics of the current rotating equipment. After removing excessively large deviation data, the positional deviations are used as random variables to fit the probability distribution of the measurement deviations in multiple search directions, calculating the deviation probability of each reference point under multiple search directions, and generating a deviation probability matrix based on the interval range of the deviation probabilities. The deviation probability matrix can effectively assess the confidence level and effectiveness of multiple search directions, and thus correlate the optimized probability distributions of each search direction.

[0014] Fourth, this invention verifies whether the changing trends of the deviation probability differences among the reference points within the reference point set are consistent. When the trends are consistent, the round data with the smallest sum of deviation probability differences is selected as valid pointing data. When the trends are inconsistent, offset status labels are set according to the range of deviation probability differences, and valid pointing data is filtered based on the labels. This quickly identifies local abnormal deviations within the reference point set, locates the distortion points in the probability data, and thus aligns with the actual measurement state of the current rotating equipment, further improving the optimization effect of the pointing analysis process. Attached Figure Description

[0015] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0016] Figure 1 This is a system framework diagram of a large rotating equipment all-position intelligent positioning system.

[0017] Figure 2 This is a flowchart illustrating the deviation identification module of a large rotating equipment all-position intelligent positioning system.

[0018] Figure 3 This is a flowchart illustrating the probability configuration module of a large rotating equipment all-position intelligent positioning system.

[0019] Figure 4 This is a flowchart illustrating the pointing analysis module of a large rotating equipment all-position intelligent positioning system. Detailed Implementation

[0020] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Where specific techniques or conditions are not specified in the embodiments, they shall be performed in accordance with the techniques or conditions described in the literature in the art or in accordance with the product manual.

[0021] See Figure 1The large rotating equipment all-position intelligent positioning system includes: an initial coordinate layout module, a deviation identification module, a probability configuration module, a pointing analysis module, and a positioning correction module; wherein, the output of the initial coordinate layout module is connected to the deviation identification module, the output of the deviation identification module is connected to the probability configuration module, the output of the probability configuration module is connected to the pointing analysis module, and the output of the pointing analysis module is connected to the positioning correction module.

[0022] The initial coordinate layout module is used to perform parallel retrieval of the three-dimensional coordinates of each reference point based on the reference coordinate system of the current rotating equipment, and determine the three-dimensional coordinate set for each search direction.

[0023] The deviation identification module is used to calculate the attitude angle deviation between the current search direction and the preset search direction based on the three-dimensional coordinate set of each search direction, associate the attitude angle deviation with the position deviation, and configure the position deviation matrix of each reference point.

[0024] The probability configuration module is used to calculate the deviation probability of each reference point under multiple search direction combinations based on the position deviation matrix of each reference point, and combine the deviation probabilities of all reference points into a deviation probability matrix using the interval range corresponding to the deviation probability.

[0025] The pointing analysis module is used to associate the deviation probability of each reference point with each search direction based on the deviation probability matrix, and to coordinate and optimize the probability distribution of the search direction using the relative spatial coordinates between each reference point.

[0026] The positioning correction module is used to map the optimized probability distribution of each search direction to the reference coordinate system, correct the three-dimensional coordinates of each search direction, and obtain the final output positioning result.

[0027] The search direction mentioned above refers to the spatial orientation detected in real time by measuring equipment such as laser trackers, total stations, laser displacement sensors, and laser rangefinders. Essentially, it is the spatial orientation of the measuring axis of the measuring equipment in the global coordinate system. It directly determines whether the equipment can effectively collect data such as the length, angle, and coordinates of the target point. It is also the basic data commonly used in angle interactive calculation. By forming dual-station interactive detection of the same endpoint through two search directions, and completing the calculation under distance interaction through the interaction of three search directions.

[0028] For example, laser equipment aims at target points such as reflective targets and rigid feature surfaces of the equipment by emitting a laser beam. The axial direction of the laser beam is the search direction. Then, it can combine horizontal and vertical angles to complete the angle intersection. At the same time, it can complete the processing of horizontal angle pointing around the Z-axis and vertical angle pointing around the X / Y axes. Here, the Z-axis refers to the axial direction, and the X / Y axes represent the horizontal / vertical directions. This can further demonstrate a horizontal search direction of 30 degrees around the Z-axis and a vertical search direction of 15 degrees around the X-axis.

[0029] When performing full-position positioning of rotating equipment, the initial coordinate layout module is implemented by taking the axis of the rotating equipment as the origin and the components of the rotating equipment at each position as reference points to calibrate the reference coordinate system for data acquisition.

[0030] Based on the distribution of the reference coordinate system, the starting and ending points of each search direction are collected in parallel. The starting point is generally the center of the station / sensor, representing the beginning of data acquisition, and the ending point is generally the reference point of the rotating equipment, to explain the global coordinate acquisition method in a multi-station deployment.

[0031] Generally, three-dimensional coordinates are achieved through a combination of angle positioning and length measurement. First, the horizontal and vertical angles are measured by sensors to determine the direction of the reference point. Then, the distance between the sensor and the reference point is calculated using the signal phase difference (or time of flight). Based on the angle and distance calculations, the three-dimensional coordinates of the reference point are set. However, considering the operating characteristics of large rotating equipment in a large space, it is necessary to predetermine the corresponding reference coordinates. Then, multiple stations are used to collect and record parameters such as angle and distance measured in parallel and synchronously to realize attitude changes under interactive processing and synchronize the global positioning processing of the rotating equipment.

[0032] Based on the angle and distance between the starting point and the ending point of each search direction, the three-dimensional coordinates are calculated for each measurement and synchronized to the reference coordinate system to obtain global coordinate statistics.

[0033] It should be noted that the current 3D coordinates are preliminary measurement results obtained from the reference points identified in each search direction. Subsequently, it is necessary to check the position coordinates under multiple pose changes and the intersection of multiple search directions to update the currently acquired coordinate values. This is to ensure the accuracy of the measurement of each position of the rotating device relative to the global context and to prevent the identification of coordinates in some search directions from being deviated due to the excessive size or size of some components.

[0034] The reference points can be rigid components such as couplings, bearing housings, and cylinder flanges. Then, each reference point is synchronized to the corresponding global coordinate system of the equipment to interpret the position of multiple components in the equipment body in the global coordinate system. At least three stations / sensors are set up to record the corresponding three-dimensional coordinates.

[0035] During data acquisition, it is necessary to synchronize the timestamps of all stations and normalize the acquired data, synchronizing the acquired coordinates, angles, distances, and other values ​​to the values ​​under the global coordinate system. These data will also be grouped according to the search direction to interpret the coordinate data obtained by each station.

[0036] like Figure 2As shown, the implementation of the deviation identification module includes: for the three-dimensional coordinates of multiple reference points under the same search direction, using the preset normal vector of each reference point as the fitting reference, the least squares method is used to fit the spatial plane corresponding to the reference point to obtain the fitting plane after fitting for each search direction.

[0037] This preset normal vector is a unit normal vector set based on the global coordinates of the rotating equipment. The normal vector is perpendicular to or coincides with the axial direction of the rigid plane / axis of the component where the reference point is located, to conform to the geometry of the rotating equipment under calibration. In this case, the preset normal vector is a vector pre-set from the rotating equipment design drawings, etc., and the relevant coordinates on the corresponding normal vector are identified during the actual calibration process to verify errors generated after the rotating equipment is installed or has been in operation for a long time. At this time, the direction of the preset normal vector is consistent with the target detected in the corresponding search direction, ensuring that the currently fitted attitude angle can reflect the equipment attitude at the corresponding location.

[0038] Using the normal vector of the fitted plane as the basis for verification, the propagation region corresponding to the current search direction is calibrated.

[0039] The attitude angles are represented by Euler angles around the X, Y, and Z axes, which are the angles between the starting and ending points in the search direction corresponding to the initial coordinate layout module. The least squares fitting data will be based on multiple sets of coordinate data collected in the same search direction. The right singular vector corresponding to the minimum singular value will be obtained through singular value decomposition as the fitting normal vector to eliminate random errors during data collection. The Euler angles corresponding to the corresponding reference points will be obtained through the spatial plane where the current fitted data is located to interpret the corresponding attitude information.

[0040] Singular value decomposition essentially decomposes the covariance matrix of a reference point and extracts the normal vector of the best-fit plane from the variance distribution of the data. Since the reference point is approximately coplanar, the variance of its data at the plane normal is minimized, that is, the normal vector in the three-dimensional coordinate direction is minimized. Finally, the right singular vector of the decomposition is used as the normal vector of the current fitting plane.

[0041] When calibrating the propagation area corresponding to the current search direction, the implementation method also includes: calculating the cosine similarity by performing a dot product between the normal vector of the fitted plane and the preset normal vector; the normal vector of the current fitted plane is required to be consistent with the content of the preset normal vector annotation. 0.8 can be set as the cosine similarity threshold at this time. The part greater than this value is regarded as meeting the pre-similarity standard. Otherwise, it means that the planes where the current multiple reference points are located are significantly different and it is difficult to use them as coordinate data to be identified in the same search direction. It is necessary to re-acquire the relevant data to reduce the jump point situation that occurs during laser scanning.

[0042] When the cosine similarity meets the standard, the symmetry deviation between each reference point and the center coordinate of the fitting plane is calculated to generate a symmetry calibration matrix; the symmetry deviation between each point in the three-dimensional coordinate set and the left / right / up / down coordinates of the fitting center is calculated to explain the spatial distribution relative to the fitting center, and the symmetry calibration matrix is ​​formed by combining the values ​​of the symmetry deviation; then, the principal axis direction of the symmetry deviation is determined by principal component analysis to determine the effective range that the current search direction can cover in the actual three-dimensional interval.

[0043] Principal component analysis is used to extract the principal axis direction of the calibration matrix, which is then used as the principal axis of the fitting plane normal vector to determine the propagation region of the search direction in three-dimensional space.

[0044] Principal component analysis decomposes a symmetry calibration matrix into eigenvalues ​​and eigenvectors, and defines the propagation area of ​​the effective coverage of a laser beam based on the values ​​of the eigenvectors.

[0045] At this point, the decomposition of eigenvalues ​​will yield three eigenvalues ​​A1, A2, and A3 in descending order of magnitude, along with corresponding eigenvectors B1, B2, and B3. First, B1 is taken as the main propagation direction of the laser beam. If eigenvalues ​​A2 and A3 are much smaller than eigenvalue A1 (e.g., A2 / A1 < 0.1), the laser beam is approximately a narrow beam, and the propagation area is a cylinder with B1 as its axis. If eigenvalues ​​A2 and A3 are large, the angles between the coordinates of all reference points and eigenvector B1 are calculated, and the largest angle is taken as the divergence angle of the laser beam. This yields a cone with eigenvector B1 as its axis and a half-angle as the largest angle, thus clarifying the corresponding area for current processing and identification.

[0046] Using the current propagation area as a reference, obtain the attitude angle of the fitted plane after verification, calculate the attitude angle deviation between the attitude angle of the fitted plane and the attitude angle of the preset search direction, and derive the position deviation of the current search direction based on the attitude angle deviation.

[0047] Based on the positional deviation of each search direction, a positional deviation matrix corresponding to each reference point is generated.

[0048] The position deviation matrix is ​​determined based on the attitude angle deviation of the fitted plane relative to the global coordinates. First, the coordinates of the fitted plane are transformed into the deviation state of the global coordinates. Then, the rotation matrix of each fitted plane around each axis and the translation matrix of the current plane relative to the global coordinates are determined. Using a first-order Taylor expansion, the relative coordinates of each reference point on the fitted plane are multiplied by the rotation matrix. After transforming each reference point to the global coordinates, it is multiplied by the translation matrix to obtain the position deviation relative to the global coordinates.

[0049] Preferably, if the current output reference point corresponds to multiple search directions, the method for deriving the positional deviation of the current search direction further includes: when there are multiple search directions corresponding to the reference point in the propagation area, determining the current dominant search direction, which represents the search direction with the highest fitting accuracy and the largest amount of data; that is, judging by the amount of data and fitting accuracy when fitting the three-dimensional coordinates for each search direction. The fitting accuracy can be selected by using the root mean square error method, calculating the average square root of the distance between the fitted point and the actual point. When this error is the smallest, it is considered to have the highest fitting accuracy, thereby selecting the dominant search direction.

[0050] Calculate the positional deviation between the dominant search direction and other search directions at the same reference point. If the difference between other search directions and the dominant search direction exceeds a preset threshold, it is determined that there is a cross offset of the reference point. The three-dimensional coordinates of the reference point are re-acquired, and the search direction is marked as abnormal. Otherwise, the corresponding positional deviation is output directly.

[0051] At this point, by identifying multiple search directions, the different parts of the positional deviation under the intersection of multiple search directions are examined. Parts exceeding a preset threshold are removed, such as ±0.05mm. The configured preset threshold is used to filter parts where there are obvious differences in data calculation. If the positional deviation of the dominant search direction is inconsistent with other search directions, it means that the coordinates of the corresponding reference point measurement have an intersection offset, indicating a certain degree of measurement anomaly, or an accuracy error in the installation or use of the equipment. If it does not exceed the preset threshold, it means that the positional deviation identification of multiple directions is relatively consistent, and the currently identified positional deviation can be output. Then, it is determined whether each three-dimensional coordinate corresponds to a change in the operating state of the rotating equipment. The positional deviation matrix is ​​output according to the search direction, thereby synchronizing the relative coordinate position of each reference point.

[0052] Preferably, in addition to outputting a position deviation matrix with smaller errors in multiple search directions, it is also necessary to determine the constraint conditions of the plane where each reference point is located so that the acquired data conforms to the geometric structure of each component in the rotating device.

[0053] The implementation method for configuring the position deviation matrix of each reference point also includes: determining the sum of the position deviations of each reference point, using the minimum sum of position deviations as a constraint condition; and outputting the position deviation matrix that satisfies the constraint condition.

[0054] Limiting the minimum positional deviation is to ensure that each search direction produces a sufficiently small deviation after identifying the coordinates of the reference point, so as to facilitate subsequent verification of the relative deviation of the rotating device under multiple structures, thereby assisting in the coupling output of the final result.

[0055] In the probability configuration module, angle interaction and distance interaction in multiple search directions are taken as the calculation center. Since the attitude angle fitting and position deviation derivation in the previous section have already checked and processed the content of angle and distance interaction, the current step will focus on the generation and processing of deviation probability to further deepen the measurement scenario of angle interaction and distance interaction, thereby determining the deviation probability generated by parallel search in the corresponding scenario, and finally obtaining the deviation probability matrix of each position that conforms to the global distribution.

[0056] like Figure 3 As shown, the implementation of the probability configuration module includes: obtaining the position deviation sequence corresponding to each benchmark point based on the benchmark points corresponding to each search direction. This benchmark point represents the data position comparison processed by comparing data from multiple stations using multiple search directions. Based on the values ​​of each benchmark point in multiple search directions, the position deviation is standardized to obtain the position deviation sequence for each benchmark point.

[0057] Using the positional deviation of the reference point as a random variable, the probability distribution of the positional deviation measured in at least two search directions is fitted, and the deviation probability of each reference point in multiple search directions is calculated.

[0058] Based on the deviation probability of each benchmark point in multiple search directions, the deviation probability interval range corresponding to each benchmark point is divided, and the deviation probabilities and interval ranges of all benchmark points are combined into a deviation probability matrix.

[0059] When calculating the deviation probability of each benchmark point in multiple search directions, the deviation probability of each benchmark point can be calculated based on kernel density estimation; or a Gaussian mixture model can be used to calculate the deviation probability corresponding to each benchmark point. Then, according to the range of deviation probability values, the probabilities of all benchmark points and intervals are combined into a matrix to finally obtain the deviation probability matrix of the global description.

[0060] In the pointing analysis module, the state of deviation in each search direction is judged based on the deviation probability obtained for each reference point, and the probability distribution under effective pointing is calculated according to the change in the relative spatial coordinates between reference points.

[0061] like Figure 4As shown, the implementation of the pointing analysis module includes: for reference points under the same search direction, density clustering is performed based on the relative spatial coordinates between the reference points to form multiple geometric association clusters associated with the rotating equipment. Each geometric association cluster corresponds to a set of reference points; this set is a combination of associated reference points measured by the same station within a specific angle range. Furthermore, when setting the reference point set according to the relative spatial coordinates between the reference points, spatial clustering is performed based on the distance between points to set multiple geometric association clusters associated with the rotating equipment. During geometric association clustering, a neighborhood radius of 50mm can be selected, with a minimum of 4 reference points, or the neighborhood radius and minimum number of points can be adjusted according to the current geometric structure of the reference points to achieve density clustering processing of the reference points.

[0062] The deviation probability difference between each benchmark point in the current round and the measurement in the previous round is calculated. Based on this difference, the changing trend of each search direction is determined, and a corresponding offset status label is configured for each benchmark point set. The offset status at this time represents the consistency of each benchmark point set under multiple measurements, whether there is a situation where some search directions show the measurement offset of the round, and the value of the overall deviation probability under each combination, so as to track the changes of each benchmark point in local spatial position. Through the synchronization and verification of this part of the data, the subsequent measurement results are more accurate.

[0063] The set of reference points with valid pointing directions is filtered using offset status labels, and the filtered data is regarded as the optimized probability distribution for each search direction.

[0064] When determining the changing trend of each search direction, the implementation method includes: for any reference point, for the difference in deviation probability between the current round and the previous round, check whether the changing trend of each reference point in the current reference point set is consistent; checking the trend change is to verify the reference points located in adjacent structures. If the deviation probability of the reference points is consistent, such as all increasing, all decreasing, or all remaining unchanged; generally, it represents a positive signal that both the measurement end and the equipment operation end are tending to be stable, indicating that the dynamic deviation at this position is gradually decreasing, the measurement result will tend to be stable, and thus obtain the relevant global positioning data.

[0065] If the trends are inconsistent, it indicates that there is a local anomaly at the measurement end or the device end. For example, a certain reference point in the global positioning is blocked, some lenses of the sensor are dirty, or the spot of the laser beam is offset. This will cause some results to be biased and the probability calculation is wrong, thus affecting the overall accuracy.

[0066] If the changing trends of all benchmarks in the current benchmark set are consistent, determine the round with the smallest sum of deviation probability differences among all deviation points, and output the data of the corresponding round as valid pointing data.

[0067] If the changing trends of the reference points within the current reference point set are inconsistent, an offset status label is set based on the range of the deviation probability difference, and the valid data to be pointed to is determined by the offset status label. The output data is adjusted only in scenarios where the trends are inconsistent, based on the value of the deviation probability difference.

[0068] For example, the average difference of the probability of effective downward deviation is selected as the current classification standard. Data with a value less than or equal to this value is considered normal data, while other data is considered invalid data. Offset status labels are set sequentially according to their values ​​to prevent errors when updating the positioning data later.

[0069] In the positioning correction module, the initial 3D coordinate set is optimized based on the coordinate data corresponding to the optimized probability distribution. The coordinate data of the optimized probability distribution is then fitted with the 3D coordinate set of the corresponding search direction to obtain the final output coordinates. For example, by fitting the coordinates of the optimized probability distribution with the original 3D coordinates, a coordinate center is obtained. This coordinate center is then considered as the updated data to complete the global coordinate positioning update of the large rotating equipment.

[0070] The implementation of the positioning correction module also includes: when the coordinates of any reference point in the reference coordinate system change, updating the reference constraints of each reference point with the distance between adjacent reference points, and using the correction amount of each reference point under the reference constraints as the output positioning result.

[0071] The reference constraint is used to check whether the adjustment of the three-dimensional coordinates will lead to an excessive change in the coordinates. At this time, the distance between adjacent references is introduced to quantify the situation after the coordinate correction, so as to know whether the coordinate adjustment of multiple reference points meets the rigid geometric constraints.

[0072] An allowable error in the distance between adjacent reference points can be introduced as a reference constraint when updating the current reference point. If the distance between the corrected coordinates is greater than this allowable error, it is considered that the modification is too large and the correction amount needs to be obtained again to complete the correction of the corresponding coordinates.

[0073] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered within the protection scope of the present invention.

Claims

1. A large-scale rotating equipment all-position intelligent positioning system, characterized in that, include: The initial coordinate layout module is used to perform parallel retrieval of the three-dimensional coordinates of each reference point based on the reference coordinate system of the current rotating equipment, and determine the three-dimensional coordinate set for each search direction; The deviation identification module is used to calculate the attitude angle deviation between the current search direction and the preset search direction based on the three-dimensional coordinate set of each search direction, associate the attitude angle deviation with the position deviation, and configure the position deviation matrix of each reference point. The probability configuration module is used to calculate the deviation probability of each reference point under the combination of multiple search directions based on the position deviation matrix of each reference point, and to combine the deviation probabilities of all reference points into a deviation probability matrix based on the interval range corresponding to the deviation probability. The pointing analysis module is used to associate the deviation probability of each reference point with each search direction based on the deviation probability matrix, and to coordinate and optimize the probability distribution of the search direction using the relative spatial coordinates between each reference point. The positioning correction module is used to map the optimized probability distribution of each search direction to the reference coordinate system, correct the three-dimensional coordinates of each search direction, and obtain the final output positioning result.

2. The intelligent positioning system for all positions of large rotating equipment according to claim 1, characterized in that, The initial coordinate layout module is implemented in the following ways: The reference coordinate system for data acquisition is calibrated using the axis of the rotating equipment as the origin and the components of the rotating equipment at various positions as reference points. Based on the distribution location of the reference coordinate system, the starting point and ending point corresponding to each search direction are collected in parallel; Based on the angle and distance between the starting point and the ending point of each search direction, the three-dimensional coordinates are calculated for each measurement and synchronized to the reference coordinate system.

3. The intelligent positioning system for all positions of large rotating equipment according to claim 1, characterized in that, The implementation methods of the deviation recognition module include: For the three-dimensional coordinates of multiple reference points under the same search direction, the preset normal vector of each reference point is used as the fitting reference, and the least squares method is used to fit the spatial plane corresponding to the reference point to obtain the fitting plane after fitting for each search direction. Using the normal vector of the fitted plane as the basis for verification, the propagation region corresponding to the current search direction is calibrated; Using the current propagation area as a reference, obtain the attitude angle of the fitted plane after verification, calculate the attitude angle deviation between the attitude angle of the fitted plane and the attitude angle of the preset search direction, and derive the position deviation of the current search direction based on the attitude angle deviation. Based on the positional deviation of each search direction, a positional deviation matrix corresponding to each reference point is generated.

4. The intelligent positioning system for all positions of large rotating equipment according to claim 3, characterized in that, When defining the propagation region corresponding to the current search direction, the implementation methods also include: The cosine similarity is calculated by performing a dot product between the fitted plane normal vector and the preset normal vector. When the cosine similarity meets the standard, the symmetry deviation between each reference point and the center coordinate of the fitted plane is calculated to generate a symmetry calibration matrix; Principal component analysis is used to extract the principal axis direction of the calibration matrix, which is then used as the principal axis of the fitting plane normal vector to determine the propagation region of the search direction in three-dimensional space.

5. The intelligent positioning system for all positions of large rotating equipment according to claim 3, characterized in that, Other methods for deriving the positional deviation of the current search direction include: When multiple search directions correspond to a reference point within the propagation area, determine the current dominant search direction; Calculate the positional deviation between the dominant search direction and other search directions at the same reference point. If the difference between other search directions and the dominant search direction exceeds a preset threshold, it is determined that there is a cross offset of the reference point. The three-dimensional coordinates of the reference point are re-acquired, and the search direction is marked as abnormal. Otherwise, the corresponding positional deviation is output directly.

6. The intelligent positioning system for all positions of large rotating equipment according to claim 1, characterized in that, The implementation methods for configuring the position deviation matrix of each reference point also include: Determine the total positional deviation of each reference point, using the minimum total positional deviation as a constraint; output the positional deviation matrix that satisfies the constraint.

7. The intelligent positioning system for all positions of large rotating equipment according to claim 1, characterized in that, The probability configuration module can be implemented in the following ways: Based on the reference points corresponding to each search direction, obtain the position deviation sequence corresponding to each reference point; Using the position deviation of the reference point as a random variable, the probability distribution of the position deviation measured in at least two search directions is fitted, and the deviation probability of each reference point in multiple search directions is calculated. Based on the deviation probability of each benchmark point in multiple search directions, the deviation probability interval range corresponding to each benchmark point is divided, and the deviation probabilities and interval ranges of all benchmark points are combined into a deviation probability matrix.

8. The intelligent positioning system for all positions of large rotating equipment according to claim 1, characterized in that, The implementation methods for the analysis module include: For reference points under the same search direction, density clustering is performed based on the relative spatial coordinates between each reference point to form multiple geometric association clusters associated with the rotating equipment. Each geometric association cluster corresponds to a set of reference points. Calculate the probability difference between the deviation of each benchmark point in the current round and the measurement in the previous round, determine the trend of change in each search direction based on the difference, and configure the corresponding offset status label for each benchmark point set; The set of reference points with valid pointing directions is filtered using offset status labels, and the filtered data is regarded as the optimized probability distribution for each search direction.

9. The intelligent positioning system for all positions of large rotating equipment according to claim 8, characterized in that, When determining the changing trends of each search direction, the implementation methods include: For any benchmark point, check the probability difference between the deviation in the current round and the previous round, and then check whether the changing trends of each benchmark point in the current benchmark point set are consistent. If the changing trends of all benchmarks in the current set of benchmarks are consistent, determine the round with the smallest sum of the deviation probability differences of all deviation points, and output the data of the corresponding round as the data with valid pointers. If the changing trends of the reference points in the current reference point set are inconsistent, the offset status label is set according to the range of the deviation probability difference, and the data to be effectively pointed to is determined by the offset status label.

10. The intelligent positioning system for all positions of large rotating equipment according to claim 1, characterized in that, The implementation methods of the positioning correction module also include: When the coordinates of any reference point in the reference coordinate system change, the reference constraints of each reference point are updated based on the distance between adjacent reference points, and the correction amount of each reference point under the reference constraints is used as the output positioning result.

Images