Valve hall wall casing displacement monitoring method, device, equipment and storage medium

Abstract

The embodiments of this specification provide a method, device, equipment and storage medium for monitoring the displacement of a valve hall through-the-wall casing, and relate to the field of computer technology. The method includes: obtaining multiple current initial point clouds at the current time, and the multiple current initial point clouds are obtained by scanning the corresponding target scanning area by the acquisition module at the current time; downsampling each current initial point cloud according to multiple sampling resolutions to obtain multiple downsampled point clouds for each current initial point cloud; aligning and splicing the multiple downsampled point clouds of each of the multiple current initial point clouds to obtain the current spliced ​​point cloud; extracting feature points from the current spliced ​​point cloud to obtain the current target point cloud, and the through-the-wall casing area in the current spliced ​​point cloud is the area of ​​interest; determining the displacement deviation information of the through-the-wall casing according to the current target point cloud and the historical target point cloud at the historical time. In this way, it is possible to timely determine whether the through-the-wall casing of the valve hall has been displaced, thereby improving the accuracy and efficiency of monitoring the through-the-wall casing of the valve hall.

Description

Technical Field

[0001] The embodiments of the present specification relate to the field of computer technology, and specifically to a method, device, equipment and storage medium for monitoring the displacement of a valve hall wall casing. Background Art

[0002] Converter valves or converter transformers are key equipment in the process of electric energy transmission. They can realize the mutual conversion of AC and DC and realize the mutual electrical isolation of AC and DC. With the rapid development of power systems, they have been widely used in converter stations, substations or power plants. The valve hall is a building mainly used to arrange converter valves or converter transformers and related equipment. The valve hall usually uses a steel structure and adopts a full metal shielding to shield external electromagnetic interference and external radio interference generated by valve commutation. The valve hall wall bushing, also known as the converter transformer bushing, is the passage between the valve hall and the DC field, and plays the role of current flow, insulation and mechanical support. The stability of the wall bushing is crucial to the safe and reliable operation of the entire transmission project.

[0003] In the related art, in order to ensure the stability of the wall bushing, the wall bushing is usually regularly inspected and maintained. However, this method lacks specificity and easily leads to problems such as excessive maintenance or untimely maintenance, and the monitoring accuracy and efficiency are low.

[0004] Therefore, there is an urgent need to provide a valve hall wall sleeve displacement monitoring method to timely determine whether the valve hall wall sleeve is displaced, thereby improving the monitoring efficiency of the valve hall wall sleeve. Summary of the invention

[0005] In view of this, multiple embodiments of this specification are dedicated to providing a valve hall wall bushing displacement monitoring method, device, equipment and storage medium to improve the monitoring efficiency of the valve hall wall bushing.

[0006] An embodiment of the present specification provides a method for monitoring the displacement of a wall sleeve in a valve hall, the method comprising: obtaining a plurality of current initial point clouds at a current time, wherein the plurality of current initial point clouds are obtained by scanning corresponding target scanning areas in the valve hall by at least one acquisition module at the current time, and the acquisition module corresponds to one or more target scanning areas; downsampling each of the current initial point clouds according to a plurality of sampling resolutions to obtain a plurality of downsampled point clouds of each of the current initial point clouds, wherein the plurality of downsampled point clouds of each of the current initial point clouds correspond one-to-one to the plurality of sampling resolutions; aligning and splicing the plurality of downsampled point clouds of each of the plurality of current initial point clouds to obtain a current spliced ​​point cloud; extracting feature points from the current spliced ​​point cloud to obtain a current target point cloud, wherein the wall sleeve area in the current spliced ​​point cloud is an area of ​​interest; determining the displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at a historical time.

[0007] In some embodiments, the registering and stitching of the multiple downsampled point clouds based on each of the multiple current initial point clouds to obtain the current stitched point cloud includes: taking the sampling resolutions among the multiple sampling resolutions as the current sampling resolution in order from low to high, and iteratively registering the multiple downsampled point clouds at the current sampling resolution; and stitching the multiple downsampled point clouds after registration at the current sampling resolution to obtain the current stitched point cloud when the current sampling resolution reaches the highest sampling resolution or the iterative registration result satisfies a specified convergence condition.

[0008] In some embodiments, the iterative registration of the multiple downsampled point clouds at the current sampling resolution includes: based on the multiple downsampled point clouds at the current sampling resolution, obtaining multiple feature descriptors corresponding to the current sampling resolution, wherein the feature descriptors can describe data points in the corresponding downsampled point clouds; performing feature matching based on the multiple feature descriptors to obtain a transformation matrix at the current sampling resolution; and determining the iterative registration result based on the transformation matrix at the current sampling resolution.

[0009] In some embodiments, each of the current initial point clouds includes multiple initial data points; before downsampling each of the current initial point clouds according to multiple sampling resolutions, the method also includes: obtaining a first filtering parameter and a second filtering parameter; for each of the current initial point clouds, determining a distance mean and a distance standard deviation between all initial data points in the current initial point cloud and all corresponding neighborhood points; taking the initial data points in the current initial point cloud as current initial data points in turn, and when the distance between the current initial data point and any corresponding neighborhood point is greater than or equal to an outlier threshold, removing the corresponding neighborhood point from the current initial point cloud, wherein the outlier threshold is determined based on the first filtering parameter, the second filtering parameter, the distance mean and the distance standard deviation.

[0010] In some embodiments, the current target point cloud includes multiple first data points, and the historical target point cloud includes multiple second data points; determining the displacement deviation information of the wall sleeve based on the current target point cloud and the historical target point cloud at historical time includes: taking the first data point in the current target point cloud as the current data point in turn, and determining the square value of the distance between the current data point and the corresponding point based on the current data point and the historical target point cloud, wherein the corresponding point is the one with the smallest square value of the distance between the current data point and the multiple second data points; determining the displacement deviation information of the wall sleeve based on the square value of the distance between each of the multiple first data points in the current target point cloud and the corresponding point.

[0011] In some embodiments, the displacement deviation information is determined by the following formula:

[0012] ;in, Indicates the displacement deviation information of the wall bushing, represents the current target point cloud, represents any first data point in the current target point cloud, represents the number of the first data points in the current target point cloud, represents the historical target point cloud, represents the number of second data points in the historical target point cloud, represents the first The second data point.

[0013] In some embodiments, the current time is the current moment, and any of the acquisition modules corresponds to one of the target scanning areas at the current moment; or, the current time is the current time period, and any of the acquisition modules corresponds to one or more of the target scanning areas in the current time period.

[0014] The embodiment of the present specification provides a displacement monitoring device for a wall sleeve in a valve hall, the device comprising: a point cloud acquisition module, used to acquire a plurality of current initial point clouds at a current time, wherein the plurality of current initial point clouds are obtained by scanning the corresponding target scanning area in the valve hall by at least one acquisition module at the current time, and the acquisition module corresponds to one or more target scanning areas; a point cloud downsampling module, used to downsample each of the current initial point clouds according to a plurality of sampling resolutions to obtain a plurality of downsampled point clouds of each of the current initial point clouds, wherein the plurality of downsampled point clouds of each of the current initial point clouds correspond one-to-one to the plurality of sampling resolutions; a point cloud registration module, used to perform registration and splicing based on the plurality of downsampled point clouds of each of the plurality of current initial point clouds to obtain a current spliced ​​point cloud; a feature extraction module, used to perform feature point extraction on the current spliced ​​point cloud to obtain a current target point cloud, wherein the wall sleeve area in the current spliced ​​point cloud is an area of ​​interest; a displacement determination module, used to determine the displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at a historical time.

[0015] In some embodiments, the point cloud registration module is also used to: according to the multiple sampling resolutions from low to high, take the sampling resolutions among the multiple sampling resolutions as the current sampling resolution in turn, and iteratively register the multiple downsampled point clouds under the current sampling resolution; when the current sampling resolution reaches the highest sampling resolution or the iterative registration result satisfies the specified convergence condition, splice the multiple downsampled point clouds after registration under the current sampling resolution to obtain the current spliced ​​point cloud.

[0016] In some embodiments, the point cloud registration module is also used to: obtain a plurality of feature descriptors corresponding to the current sampling resolution based on the plurality of downsampled point clouds at the current sampling resolution, wherein the feature descriptors can describe the data points in the corresponding downsampled point clouds; perform feature matching based on the plurality of feature descriptors to obtain a transformation matrix at the current sampling resolution; and determine the iterative registration result based on the transformation matrix at the current sampling resolution.

[0017] In some embodiments, each of the current initial point clouds includes multiple initial data points; the device also includes a filtering module, which is used to: obtain a first filtering parameter and a second filtering parameter; for each of the current initial point clouds, determine the distance mean and distance standard deviation between all the initial data points in the current initial point cloud and all the corresponding neighborhood points; take the initial data points in the current initial point cloud as the current initial data points in turn, and when the distance between the current initial data point and any corresponding neighborhood point is greater than or equal to an outlier threshold, remove the corresponding neighborhood point from the current initial point cloud, wherein the outlier threshold is determined based on the first filtering parameter, the second filtering parameter, the distance mean and the distance standard deviation.

[0018] In some embodiments, the current target point cloud includes multiple first data points, and the historical target point cloud includes multiple second data points; the displacement determination module is also used to: take the first data point in the current target point cloud as the current data point in turn, and determine the square distance value between the current data point and the corresponding point based on the current data point and the historical target point cloud, wherein the corresponding point is the one with the smallest square distance value between the current data point and the multiple second data points; based on the square distance value between each of the multiple first data points in the current target point cloud and the corresponding point, determine the displacement deviation information of the wall sleeve.

[0019] An embodiment of the present specification provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the displacement monitoring method described in any of the above embodiments is implemented.

[0020] An embodiment of the present specification provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the displacement monitoring method described in any of the above embodiments is implemented.

[0021] In multiple implementations provided in this specification, the acquisition module corresponds to one or more target scanning areas in the valve hall. First, at the current time, the current initial point cloud corresponding to the target scanning area is acquired through the acquisition module to obtain multiple current initial point clouds. Then, each current initial point cloud is downsampled according to multiple sampling resolutions to obtain downsampled point clouds of each current initial point cloud at multiple sampling resolutions. Then, based on multiple sampling resolutions, the downsampled point clouds of the multiple current initial point clouds are aligned and spliced ​​to obtain the current spliced ​​point cloud. Then, for the current spliced ​​point cloud, feature points are extracted with the wall casing area in the current spliced ​​point cloud as the area of ​​interest to obtain the current target point cloud. Then, based on the current target point cloud and the historical target point cloud at historical time, the displacement deviation information of the wall casing is determined. In this way, it is possible to timely determine whether the wall casing of the valve hall is displaced, thereby improving the monitoring accuracy and efficiency of the wall casing of the valve hall. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of a valve hall wall bushing displacement monitoring system provided in an embodiment of this specification;

[0023] Figure 2 A flow chart of a method for monitoring displacement of a wall sleeve in a valve hall provided in an embodiment of this specification;

[0024] Figure 3 A schematic diagram of a valve hall wall sleeve displacement monitoring device provided in an embodiment of this specification;

[0025] Figure 4 A schematic diagram of a computer device provided for embodiments of the present specification. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solution of this specification, the technical solution in the implementation of this specification will be clearly and completely described below in conjunction with the drawings in the implementation of this specification. Obviously, the described implementation is only a part of the implementation of this specification, not all of the implementations. Based on the implementation in this specification, all other implementations obtained by ordinary technicians in this field without creative work are within the scope of protection of this specification.

[0027] The valve hall wall bushing is an important accessory of the converter valve or converter transformer, and is also one of the most important power transmission and transformation equipment. In power systems such as converter stations, substations or power plants, the reliability or failure of the wall bushing is affected by many factors, such as improper installation, insulation degradation, external force or mechanical influence, etc.

[0028] In some related technologies, vibration sensors, temperature sensors or tilt sensors are usually installed on the surface of the casing to monitor it. This method requires the relevant sensors to be attached to the surface of the casing, which is prone to looseness during the construction of the symbiotic structure of new and old facilities, and has low accuracy. In some other related technologies, the casing is monitored by using cameras, ultrasonic detection, etc., but these methods are easily interfered by environmental noise.

[0029] Therefore, it is necessary to conduct online or real-time monitoring of the bushing to timely and effectively grasp the condition of the bushing, thereby reducing the possibility of substation risks caused by bushing displacement and providing safety protection for the power system.

[0030] This specification provides an example of a scenario for a valve hall wall bushing displacement monitoring system. Figure 1 , Figure 1 Schematic diagram of a valve hall wall sleeve displacement monitoring system provided in an embodiment of the present specification. The valve hall wall sleeve displacement monitoring system may include at least one acquisition module 110 and a data processing module 120, and the acquisition module 110 is connected to the data processing module 120, for example, by wired connection or wireless connection.

[0031] In this scenario example, the valve hall can be provided with one or more wall bushings, and the acquisition module 110 and the data processing module 120 can be respectively provided inside and outside the valve hall. The acquisition module can use the laser radar to perform high-precision scanning in the target scanning area in the valve hall to obtain the three-dimensional point cloud data of the target scanning area, wherein the target scanning area can cover at least part of the wall bushing of the converter transformer on the valve hall side. In this way, the position data of all bushings on the valve hall side can be collected through one or more acquisition modules without affecting the operation of the bushings and other equipment in the valve hall. In addition, the installation method of the acquisition module is relatively simple, and the three-dimensional point cloud data collected by it can also improve the displacement monitoring accuracy of the wall bushing.

[0032] In this scenario example, before performing displacement monitoring on the valve hall wall sleeve, the target scanning area of ​​each acquisition module can be determined first, that is, the scanning range of each acquisition module can be determined. Exemplarily, each acquisition module can correspond to one or more target scanning areas. Exemplarily, reference points can be arranged for the target scanning area of ​​each acquisition module. Reference points, also known as targets, target points, marking points or control points, can be points set on or around the surface of the wall sleeve to improve scanning accuracy and assist the scanning process. By setting the reference point, the acquisition module can scan the target scanning area centered on the reference point, and record the coordinate information of the reference point in the scanned point cloud data. In the subsequent data processing, on the one hand, these reference points can be used to calibrate and align the point cloud data, thereby improving the measurement accuracy; on the other hand, by analyzing and comparing the reference points, accurate measurement data of the wall casing can be obtained, and matching and correction of multiple target scanning areas can be achieved; on the other hand, when monitoring the displacement of the wall casing, it may be necessary to splice the point cloud data multiple times, and the cumulative error will gradually increase. By reasonably setting the reference points, accurate coordinate alignment can be performed during the registration and splicing process, thereby reducing the splicing error and improving the overall scanning accuracy. As an example, the reference points can include one or more of a spherical target, a plane target, and a checkerboard target. Of course, it is understandable that the reference points can be specifically set according to the actual situation, for example, avoiding obstacles as much as possible to reduce the number of reference points, etc., which is not limited here.

[0033] In this scenario example, the acquisition module may be a 3D laser scanner, and the acquisition module may acquire point cloud data by laser pulse ranging. For example, the point cloud or point cloud data may include multiple data points. For example, any data point may be represented as , the point cloud can be represented as ,in, Point Cloud The corresponding three-dimensional coordinate set, Point Cloud Any data point in .

[0034] As an example, the acquisition module can determine the data points in the point cloud through Formula 1 and Formula 2.

[0035] Formula 1;

[0036] Formula 2;

[0037] in, Indicates the scanning distance or measurement distance of any data point or scanning electron in the point cloud. represents the speed of light, Indicates the round-trip time difference of a laser signal or a pulse signal. , , The data points are axis, axis, The coordinates on the axis, For data points and The angle of the axis, is the horizontal scanning angle of the data point, which can refer to the horizontal scanning angle of the data point. Plane projection and The angle of the axis.

[0038] Exemplarily, the round-trip time difference can be obtained based on the pulse method (time of flight, TOF) or based on the phase method (phase difference). Among them, the pulse method uses the diffuse reflection effect of the laser pulse on the target to obtain the round-trip time difference through the timing of the receiving and transmitting ends. The phase method indirectly determines the propagation time of the laser pulse by measuring the phase difference generated by the continuous modulated wave propagating back and forth once over the measuring distance or scanning distance. As an example, the period information of the laser pulse can be obtained to measure the round-trip time difference based on the period information, thereby performing phase method ranging.

[0039] This specification provides a method for monitoring the displacement of a valve hall wall bushing. Figure 2 , Figure 2 It is a flow chart of a valve hall wall sleeve displacement monitoring method provided by this embodiment. This embodiment provides method operation steps as shown in the flow chart, but may include more or fewer operation steps based on conventional or non-creative labor. The sequence of steps listed in the embodiment is only one execution method among many step execution sequences, and does not represent the only execution sequence. When the actual system or server product is executed, it can be executed in sequence or in parallel according to the method shown in the embodiment (for example, in a parallel processor or multi-threaded processing environment). The displacement monitoring method can be applied to the valve hall wall sleeve displacement monitoring system, specifically Figure 2 As shown, the displacement monitoring method may include the following steps.

[0040] Step S210: Acquire multiple current initial point clouds at the current time, wherein the multiple current initial point clouds are obtained by scanning corresponding target scanning areas in the valve hall by at least one acquisition module at the current time, and the acquisition module corresponds to one or more target scanning areas.

[0041] In some cases, the displacement monitoring system may include one or more acquisition modules, each of which may correspond to one or more target scanning areas to obtain point clouds corresponding to the target scanning areas, that is, any acquisition module may acquire one or more point clouds at a certain time.

[0042] In this embodiment, the point clouds collected by one or more acquisition modules in the valve hall at the current time can be recorded as current initial point clouds to obtain multiple current initial point clouds at the current time, and each current initial point cloud corresponds to a target scanning area at the current time.

[0043] The data processing module can obtain the point cloud from the acquisition module to monitor the displacement of the valve hall wall sleeve. Specifically, the point cloud obtained at the current time can be recorded as the current initial point cloud.

[0044] Step S220: downsampling each current initial point cloud according to multiple sampling resolutions to obtain multiple downsampled point clouds of each current initial point cloud, wherein the multiple downsampled point clouds of each current initial point cloud correspond one-to-one to the multiple sampling resolutions.

[0045] In some cases, the acquired point cloud often contains a large number of data points, especially the point cloud acquired by high-precision or high-resolution acquisition modules, the data volume may reach hundreds of thousands or millions or even more. In order to reduce the amount of data and calculation, it can be downsampled to improve the calculation efficiency. At the same time, downsampling can also adjust the point cloud to a suitable processing range, thereby improving the accuracy and efficiency of subsequent registration.

[0046] The sampling resolution may refer to the distribution density of data points in the point cloud or the number of data points per unit volume, which reflects the average spacing between these data points and the ability to represent the details of the wall casing.

[0047] Specifically, for any current initial point cloud among the multiple current initial point clouds, the current initial point cloud can be downsampled according to multiple sampling resolutions to obtain downsampled point clouds of the current initial point cloud at multiple sampling resolutions. In other words, the current initial point cloud can be downsampled according to multiple sampling resolutions to obtain multiple downsampled point clouds of the current initial point cloud. In this way, after downsampling each current initial point cloud according to multiple sampling resolutions, downsampled point clouds of each current initial point cloud at multiple resolutions can be obtained, thereby obtaining the final multiple downsampled point clouds.

[0048] Exemplarily, when downsampling any current initial point cloud according to multiple sampling resolutions, the current initial point cloud may be downsampled by using any one of uniform grid sampling, voxel grid sampling, or uniform interval sampling.

[0049] Step S230: performing registration and stitching based on the multiple downsampled point clouds of the multiple current initial point clouds to obtain a current stitched point cloud.

[0050] Specifically, after determining the multiple downsampled point clouds of each current initial point cloud, the multiple downsampled point clouds are registered and spliced, which may refer to aligning the multiple downsampled point clouds through rigid body transformations such as translation and rotation, so as to unify the point clouds of different target scanning areas at the current time into the same coordinate system, and merge the aligned multiple downsampled point clouds into a whole to obtain the current spliced ​​point cloud, so as to more comprehensively reflect the three-dimensional shape and structure of the wall casing.

[0051] Step S240: extracting feature points from the current spliced ​​point cloud to obtain the current target point cloud, wherein the wall casing area in the current spliced ​​point cloud is the area of ​​interest.

[0052] Among them, the current target point cloud is the target point cloud corresponding to the current time.

[0053] Specifically, the current spliced ​​point cloud may include a wall casing area and a non-wall casing area. In order to monitor the displacement of the wall casing, feature points can be extracted for the wall casing area in the current spliced ​​point cloud with the wall casing area as the area of ​​interest to obtain the current target point cloud corresponding to the wall casing area in the current spliced ​​point cloud.

[0054] Step S250: Determine the displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at the historical time.

[0055] Specifically, the historical time may refer to the casing setting time corresponding to the beginning of the wall casing setting in the valve hall, or may refer to any time after the setting time and before the current time. The historical target point cloud may refer to the target point cloud corresponding to the historical time.

[0056] In the above implementation, multiple current initial point clouds at the current time are acquired, and each current initial point cloud is downsampled according to multiple sampling resolutions to obtain multiple downsampled point clouds of each current initial point cloud. Then, the multiple downsampled point clouds of the multiple current initial point clouds are aligned and spliced ​​to obtain the current spliced ​​point cloud. The wall casing area is used as the area of ​​interest to extract feature points of the current spliced ​​point cloud to obtain the current target point cloud. Then, the displacement deviation information of the wall casing is determined based on the current target point cloud and the historical target point cloud at the historical time. In this way, it is possible to determine in real time and timely whether the valve hall wall casing is displaced, thereby improving the monitoring efficiency of the valve hall wall casing.

[0057] In some implementations, registering and stitching multiple downsampled point clouds of each of the multiple current initial point clouds to obtain a current stitched point cloud may include the following steps S310-S320.

[0058] Step S310: according to the multiple sampling resolutions from low to high, the sampling resolutions among the multiple sampling resolutions are sequentially used as the current sampling resolution, and the multiple downsampled point clouds under the current sampling resolution are iteratively registered.

[0059] In some cases, by downsampling any current initial point cloud according to multiple sampling resolutions, multiple downsampled point clouds corresponding to multiple sampling resolutions of the current initial point cloud can be obtained, and multiple downsampled point clouds at any sampling resolution correspond to multiple current initial point clouds. In order to improve the registration and stitching accuracy of point clouds while improving the registration efficiency and registration real-time performance, iterative registration can be performed in order of sampling resolution from low to high.

[0060] Specifically, the sampling resolutions among the multiple sampling resolutions that are pre-set from low to high can be used as the current sampling resolution in turn, and then the multiple downsampled point clouds at the current sampling resolution can be iteratively aligned to obtain the iterative alignment result at the current sampling resolution, so that it can be determined whether the alignment is successful or meets the stitching requirements based on the iterative alignment result at the current sampling resolution.

[0061] Step S320: when the current sampling resolution reaches the highest sampling resolution or the iterative registration result satisfies the specified convergence condition, the multiple registered downsampled point clouds at the current sampling resolution are spliced ​​to obtain the current spliced ​​point cloud.

[0062] Specifically, it is possible to determine whether the current sampling resolution is the highest sampling resolution. If the current sampling rate reaches the highest resolution among multiple sampling resolutions, the multiple downsampled point clouds after alignment at the current sampling resolution can be directly spliced ​​to obtain the current spliced ​​point cloud. If the current sampling resolution does not reach the highest resolution among multiple sampling resolutions, it is possible to determine whether the iterative alignment result at the current sampling resolution satisfies the specified convergence condition. If so, the multiple downsampled point clouds after alignment can be determined based on the iterative alignment result, and the multiple downsampled point clouds after alignment at the current sampling resolution can be spliced ​​to obtain the current spliced ​​point cloud.

[0063] In the above implementation, iterative registration is performed in sequence according to multiple sampling resolutions from low to high, and when the highest sampling resolution is reached or the iterative registration result satisfies the specified convergence condition, the multiple downsampled point clouds after registration are spliced ​​to obtain the current spliced ​​point cloud. In this way, the registration efficiency and real-time performance can be improved while improving the point cloud registration and stitching accuracy.

[0064] In some implementations, iteratively registering a plurality of downsampled point clouds at a current sampling resolution may include the following steps S410 - S430 .

[0065] Step S410: Based on a plurality of downsampled point clouds at a current sampling resolution, a plurality of feature descriptors corresponding to the current sampling resolution are obtained.

[0066] The feature descriptor can describe the data points in the corresponding downsampled point cloud, or in other words, can describe the geometric features of each data point in the corresponding downsampled point cloud.

[0067] Specifically, for multiple downsampled point clouds at the current sampling resolution, a feature descriptor corresponding to each downsampled point cloud that can describe the data points in the downsampled point cloud is obtained, so as to perform feature matching based on the feature descriptors to achieve data alignment. Exemplarily, the feature descriptor can be a PFH (Point Feature Histogram) feature descriptor, or can be a FPFH (Fast Point Feature Histogram) feature descriptor.

[0068] Step S420: Perform feature matching based on multiple feature descriptors to obtain a transformation matrix at the current sampling resolution.

[0069] Specifically, multiple feature descriptors may be matched in pairs in sequence to determine the transformation matrix between two downsampled point clouds corresponding to any two feature descriptors, thereby determining the transformation matrix at the current sampling resolution. Exemplarily, the transformation matrix may include a translation matrix and a rotation matrix.

[0070] Step S430: determining an iterative registration result based on the transformation matrix at the current sampling resolution.

[0071] Specifically, at the current sampling resolution, the iterative registration result at the current sampling resolution may be determined based on the transformation matrix at the current sampling resolution.

[0072] Exemplarily, the iterative registration result can be determined by Formula 3.

[0073] Formula 3;

[0074] in, is the objective function, which represents the iterative registration result. represents the number of data points in any downsampled point cloud, Indicates the current sampling resolution. and Represents the rotation matrix and translation matrix in the transformation matrix at the current sampling resolution. and Represent the data points in any downsampled point cloud and the other downsampled point cloud respectively.

[0075] In the above implementation, by obtaining multiple feature descriptors corresponding to the current sampling resolution, and performing feature matching based on the multiple feature descriptors, the transformation matrix at the current sampling resolution is obtained, and the transformation matrix at the current sampling resolution is used as the iterative registration result. In this way, the iterative registration result at each sampling resolution can be quickly determined, thereby improving the registration efficiency.

[0076] In some implementations, before downsampling each current initial point cloud according to multiple sampling resolutions, the displacement monitoring method may further include the following steps S510 - S530 .

[0077] Step S510: Acquire a first filtering parameter and a second filtering parameter.

[0078] In some cases, the current initial point cloud may contain a large number of abnormal data points, noise points or isolated points, etc. For example, data points beyond the target scanning area, overlapping data points generated by different angles in multiple target scanning areas, noise points caused by defects in the acquisition module itself, noise points caused by surface factors of the wall casing, sparse discrete points or isolated points far away from the wall casing or floating above the wall casing, noise points caused by other unpredictable factors, etc. Therefore, before performing down-sampling point cloud registration, it is necessary to filter and denoise the collected current initial point cloud to remove noise points or outlier data points, and smooth the current initial point cloud to provide a more accurate basic point cloud for feature matching and displacement analysis.

[0079] The first filtering parameter and the second filtering parameter can be used to adjust or influence the filtering effect, and the first filtering parameter and the second filtering parameter can be used to achieve coarse adjustment and precise adjustment of the filtering capability.

[0080] Step S520: for each current initial point cloud, determine the mean value and standard deviation of the distance between all initial data points in the current initial point cloud and all corresponding neighborhood points.

[0081] In some cases, each data point in the current initial point cloud may be recorded as an initial data point, that is, the current initial point cloud includes multiple initial data points.

[0082] Among them, the neighborhood point corresponding to any initial data point may refer to the r neighborhood point of the initial data point. All neighborhood points corresponding to any initial data point are all data points within the r neighborhood of the initial data point. Specifically, for any initial data point and a given positive number r, the r neighborhood of the initial data point may refer to the set of all data points whose distance from the initial data point is less than or equal to r, and the data points in the set are the r neighborhood points of the initial data point.

[0083] Specifically, for any current initial point cloud, the Gaussian distribution model corresponding to the current initial point cloud can be determined based on the mean and standard deviation of the distances between all initial data points in the current initial point cloud and all their neighboring points: ,in, Represents the mean distance between all initial data points in the current initial point cloud and all their corresponding neighborhood points, Represents the standard deviation of the distances between all initial data points in the current initial point cloud and all their corresponding neighborhood points. Represents any initial data point in the current initial point cloud Any of its neighboring points The distance between.

[0084] Step S530: The initial data points in the current initial point cloud are taken as the current initial data points in sequence. When the distance between the current initial data point and any corresponding neighborhood point is greater than or equal to the outlier threshold, the corresponding neighborhood point is removed from the current initial point cloud, wherein the outlier threshold is determined based on the first filtering parameter, the second filtering parameter, the distance mean and the distance standard deviation.

[0085] In the above implementation, by obtaining the first filter parameter and the second filter parameter for coarse adjustment and fine adjustment, and for each current initial point cloud, determining the distance mean and distance standard deviation between all initial data points in the current initial point cloud and all corresponding neighborhood points, then determining the outlier threshold based on the first filter parameter, the second filter parameter, the distance mean and the distance standard deviation, and eliminating outliers based on the outlier threshold, in this way, adaptive filtering can be achieved, the filtering effect can be adjusted flexibly and accurately, and the impact of noise points on displacement monitoring can be reduced.

[0086] In some embodiments, the outlier threshold may be determined by Equation 4.

[0087] Formula 4;

[0088] in, represents the outlier threshold, and represent the first filter parameter and the second filter parameter respectively, Represents the mean distance between all initial data points in the current initial point cloud and all their corresponding neighborhood points, Represents the standard deviation of the distances between all initial data points in the current initial point cloud and all their corresponding neighborhood points.

[0089] In some embodiments, data points in the current target point cloud can be recorded as first data points, and data points in the historical target point cloud can be recorded as second data points. In other words, the current target point cloud can include multiple first data points, and the historical target point cloud can include multiple second data points.

[0090] In this embodiment, determining the displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at the historical time may include the following steps S610-S620.

[0091] Step S610: Take the first data point in the current target point cloud as the current data point in turn, and determine the square distance value between the current data point and the corresponding point based on the current data point and the historical target point cloud, wherein the corresponding point is the one with the smallest square distance value between the current data point and the multiple second data points.

[0092] Step S620: Determine the displacement deviation information of the wall sleeve based on the square value of the distance between each of the plurality of first data points and the corresponding point in the current target point cloud.

[0093] In some implementations, the displacement deviation information may be determined using Formula 5.

[0094] Formula 5;

[0095] in, Indicates the displacement deviation information of the wall bushing. Represents the current target point cloud, represents any first data point in the current target point cloud, Indicates the number of the first data point in the current target point cloud, represents the historical target point cloud, represents the number of second data points in the historical target point cloud, Represents the first The second data point.

[0096] In some embodiments, the current time may be the current moment. In this embodiment, any acquisition module may correspond to a target scanning area at the current moment, or in other words, any acquisition module scans once at the current moment, and multiple acquisition modules may scan simultaneously at the current moment. Correspondingly, the historical time may be a historical moment, and any acquisition module may correspond to a target scanning area at the historical moment.

[0097] In some embodiments, the current time may be a current time period. In this embodiment, any acquisition module may correspond to one or more target scanning areas in the current time period, or in other words, any acquisition module scans once or multiple times in the current time period, and multiple acquisition modules may scan in the current time period. Accordingly, the historical time may be a historical time period, and any acquisition module may correspond to one or more target scanning areas in the historical time period.

[0098] The embodiment of this specification provides a displacement monitoring device for a valve hall wall bushing. The displacement monitoring device can be applied to a displacement monitoring system for a valve hall wall bushing. Figure 3 The displacement monitoring device may include a point cloud acquisition module 310 , a point cloud downsampling module 320 , a point cloud registration module 330 , a feature extraction module 340 and a displacement determination module 350 .

[0099] The point cloud acquisition module 310 is used to acquire multiple current initial point clouds at the current time, wherein the multiple current initial point clouds are obtained by scanning the corresponding target scanning areas in the valve hall by at least one acquisition module at the current time, and the acquisition module corresponds to one or more target scanning areas;

[0100] A point cloud downsampling module 320 is used to downsample each current initial point cloud according to multiple sampling resolutions to obtain multiple downsampled point clouds of each current initial point cloud, wherein the multiple downsampled point clouds of each current initial point cloud correspond one-to-one to the multiple sampling resolutions;

[0101] A point cloud registration module 330 is used to register and stitch multiple downsampled point clouds based on the multiple current initial point clouds to obtain a current stitched point cloud;

[0102] The feature extraction module 340 is used to extract feature points from the current spliced ​​point cloud to obtain the current target point cloud, wherein the wall casing area is the area of ​​interest;

[0103] The displacement determination module 350 is used to determine the displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at the historical time.

[0104] In some embodiments, the point cloud registration module 330 is also used to: according to multiple sampling resolutions from low to high, take the sampling resolution among the multiple sampling resolutions as the current sampling resolution in turn, and iteratively align multiple downsampled point clouds at the current sampling resolution; when the current sampling resolution reaches the highest sampling resolution or the iterative registration result meets the specified convergence condition, splice the multiple downsampled point clouds after registration at the current sampling resolution to obtain the current spliced ​​point cloud.

[0105] In some embodiments, the point cloud registration module 330 is also used to: obtain multiple feature descriptors corresponding to the current sampling resolution based on multiple downsampled point clouds at the current sampling resolution, wherein the feature descriptors can describe the data points in the corresponding downsampled point clouds; perform feature matching based on multiple feature descriptors to obtain the transformation matrix at the current sampling resolution; and determine the iterative registration result based on the transformation matrix at the current sampling resolution.

[0106] In some embodiments, each current initial point cloud includes multiple initial data points; the displacement monitoring device may also include a filtering module; the filtering module is used to: obtain a first filtering parameter and a second filtering parameter; for each current initial point cloud, determine the distance mean and distance standard deviation between all initial data points in the current initial point cloud and all corresponding neighborhood points; take the initial data points in the current initial point cloud as the current initial data points in turn, and when the distance between the current initial data point and any corresponding neighborhood point is greater than or equal to an outlier threshold, remove the corresponding neighborhood point from the current initial point cloud, wherein the outlier threshold is determined based on the first filtering parameter, the second filtering parameter, the distance mean and the distance standard deviation.

[0107] In some embodiments, the current target point cloud includes multiple first data points, and the historical target point cloud includes multiple second data points; the displacement determination module 350 is also used to: sequentially take the first data point in the current target point cloud as the current data point, and determine the square distance value between the current data point and the corresponding point based on the current data point and the historical target point cloud, wherein the corresponding point is the one with the smallest square distance value between the current data point and the multiple second data points; based on the square distance value between each of the multiple first data points in the current target point cloud and the corresponding point, determine the displacement deviation information of the wall sleeve.

[0108] The specific functions and effects achieved by the displacement monitoring device can be explained in other embodiments of this specification, and will not be repeated here. Each module in the displacement monitoring device can be implemented in whole or in part by software, hardware, and a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0109] The embodiments of the present specification also provide a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a computer, the computer executes the displacement monitoring method in any of the above embodiments.

[0110] The embodiments of this specification also provide a computer program product including instructions, which, when executed by a computer, enables the computer to execute the displacement monitoring method in any of the above embodiments.

[0111] The embodiment of the present specification also provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the displacement monitoring method in the above embodiment is implemented.

[0112] In some embodiments, see Figure 4 The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, and a communication interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, the displacement monitoring method is implemented.

[0113] It should be understood that the specific examples herein are only intended to help those skilled in the art to better understand the embodiments of the present specification, rather than to limit the scope of the present invention.

[0114] It can be understood that in the various implementations of this specification, the size of the sequence number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the implementation methods of this specification.

[0115] It can be understood that the various embodiments described in this specification can be implemented individually or in combination, and the embodiments of this specification are not limited to this.

[0116] Unless otherwise specified, all technical and scientific terms used in the embodiments of this specification have the same meaning as those generally understood by those skilled in the art of the technical field of this specification. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the scope of this specification. The term "and / or" used in this specification includes any and all combinations of one or more related listed items. The singular forms of "a", "above", and "the" used in the embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0117] It can be understood that the processor of the embodiment of this specification can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method implementation can be completed by the hardware integrated logic circuit or software instructions in the processor. The above processor can be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiment of this specification can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiment of this specification can be directly embodied as a hardware decoding processor to perform, or the hardware and software modules in the decoding processor are combined and executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0118] It is understood that the memory in the embodiments of this specification may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (programmable ROM, PROM), an erasable programmable read-only memory (erasablePROM, EPROM), an electrically erasable programmable read-only memory (EEPROM) or a flash memory. The volatile memory may be a random access memory (RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0119] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this specification.

[0120] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method implementation methods and will not be repeated here.

[0121] The above is only a specific implementation of this specification, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in this specification, which should be included in the protection scope of this specification. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for monitoring displacement of a valve hall wall casing, characterized in that: The method comprises: Acquire multiple current initial point clouds at the current time, wherein the multiple current initial point clouds are obtained by scanning corresponding target scanning areas in the valve hall by at least one acquisition module at the current time, and the acquisition module corresponds to one or more target scanning areas; each current initial point cloud includes multiple initial data points; For each of the current initial point clouds, determine the mean value and standard deviation of the distances between all initial data points in the current initial point cloud and all corresponding neighboring points; Obtaining a first filtering parameter and a second filtering parameter, taking the initial data points in the current initial point cloud as the current initial data points in turn, and removing the corresponding neighborhood points from the current initial point cloud when the distance between the current initial data point and any corresponding neighborhood point is greater than or equal to an outlier threshold; wherein the outlier threshold is determined based on the first filtering parameter, the second filtering parameter, the distance mean, and the distance standard deviation; and the first filtering parameter and the second filtering parameter are used to respectively realize coarse adjustment and precise adjustment of filtering capability; Downsampling each of the current initial point clouds according to a plurality of sampling resolutions to obtain a plurality of downsampled point clouds of each of the current initial point clouds, wherein the plurality of downsampled point clouds of each of the current initial point clouds correspond one-to-one to the plurality of sampling resolutions; Performing registration and stitching based on the multiple downsampled point clouds of the multiple current initial point clouds, respectively, to obtain a current stitched point cloud; Extracting feature points from the current spliced ​​point cloud to obtain a current target point cloud, wherein the wall casing region in the current spliced ​​point cloud is the region of interest; Determining displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at the historical time; Wherein, the current target point cloud includes a plurality of first data points, and the historical target point cloud includes a plurality of second data points; the determining of the displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at a historical time comprises: taking the first data point in the current target point cloud as the current data point in turn, and determining the square value of the distance between the current data point and the corresponding point according to the current data point and the historical target point cloud, wherein the corresponding point is the one with the smallest square value of the distance between the current data point and the plurality of second data points; determining the displacement deviation information of the wall sleeve based on the square value of the distance between each of the plurality of first data points in the current target point cloud and the corresponding point; The displacement deviation information is determined by the following formula: ； in, Indicates the displacement deviation information of the wall bushing, represents the current target point cloud, represents any first data point in the current target point cloud, represents the number of the first data points in the current target point cloud, represents the historical target point cloud, represents the number of second data points in the historical target point cloud, represents the first The second data point.

2. The method according to claim 1, characterized in that: The registering and stitching the multiple downsampled point clouds based on the multiple current initial point clouds to obtain the current stitched point cloud includes: According to the plurality of sampling resolutions from low to high, the sampling resolutions among the plurality of sampling resolutions are sequentially used as the current sampling resolution, and the plurality of downsampled point clouds under the current sampling resolution are iteratively registered; When the current sampling resolution reaches the highest sampling resolution or the iterative registration result satisfies the specified convergence condition, the multiple downsampled point clouds after registration at the current sampling resolution are spliced ​​to obtain the current spliced ​​point cloud.

3. The method according to claim 2, characterized in that The iterative registration of the multiple downsampled point clouds at the current sampling resolution includes: Based on the multiple downsampled point clouds at the current sampling resolution, obtaining multiple feature descriptors corresponding to the current sampling resolution, wherein the feature descriptors can describe data points in the corresponding downsampled point clouds; Perform feature matching based on the multiple feature descriptors to obtain a transformation matrix at the current sampling resolution; The iterative registration result is determined based on the transformation matrix at the current sampling resolution.

4. The method according to claim 1, characterized in that: The current time is the current moment, and any of the acquisition modules corresponds to one of the target scanning areas at the current moment; or, The current time is the current time period, and any of the acquisition modules corresponds to one or more of the target scanning areas in the current time period.

5. A valve hall wall sleeve displacement monitoring device, characterized in that: The device comprises: A point cloud acquisition module, used to acquire multiple current initial point clouds at the current time, wherein the multiple current initial point clouds are obtained by scanning the corresponding target scanning areas in the valve hall by at least one acquisition module at the current time, and the acquisition module corresponds to one or more target scanning areas; each of the current initial point clouds includes multiple initial data points; and is also used to determine, for each current initial point cloud, the mean distance and the standard deviation of the distance between all the initial data points in the current initial point cloud and all the corresponding neighboring points; acquire a first filtering parameter and a second filtering parameter, and sequentially use the initial data points in the current initial point cloud as the current initial data points, and when the distance between the current initial data point and any corresponding neighboring point is greater than or equal to an outlier threshold, remove the corresponding neighboring point from the current initial point cloud; wherein the outlier threshold is determined based on the first filtering parameter, the second filtering parameter, the mean distance and the standard deviation of the distance; the first filtering parameter and the second filtering parameter are used to respectively realize coarse adjustment and precise adjustment of filtering capability; A point cloud downsampling module, configured to downsample each of the current initial point clouds according to a plurality of sampling resolutions to obtain a plurality of downsampled point clouds of each of the current initial point clouds, wherein the plurality of downsampled point clouds of each of the current initial point clouds correspond one-to-one to the plurality of sampling resolutions; A point cloud registration module, configured to register and stitch the multiple downsampled point clouds based on the respective multiple current initial point clouds to obtain a current stitched point cloud; A feature extraction module is used to extract feature points from the current spliced ​​point cloud to obtain a current target point cloud, wherein the wall casing area in the current spliced ​​point cloud is an area of ​​interest; A displacement determination module, used to determine the displacement deviation information of the wall sleeve according to the current target point cloud and the historical target point cloud at the historical time; Wherein, the current target point cloud includes a plurality of first data points, and the historical target point cloud includes a plurality of second data points; the displacement determination module is further used to: sequentially take the first data point in the current target point cloud as the current data point, and determine the square value of the distance between the current data point and the corresponding point according to the current data point and the historical target point cloud, wherein the corresponding point is the one with the smallest square value of the distance between the current data point and the plurality of second data points; based on the square value of the distance between each of the plurality of first data points in the current target point cloud and the corresponding point, determine the displacement deviation information of the wall sleeve; The displacement deviation information is determined by the following formula: ； in, Indicates the displacement deviation information of the wall bushing, represents the current target point cloud, represents any first data point in the current target point cloud, represents the number of the first data points in the current target point cloud, represents the historical target point cloud, represents the number of second data points in the historical target point cloud, represents the first The second data point.

6. The device according to claim 5, characterized in that The point cloud registration module is further used to: in order from low to high sampling resolutions, sequentially use the sampling resolutions among the multiple sampling resolutions as the current sampling resolution, and iteratively register the multiple downsampled point clouds at the current sampling resolution; When the current sampling resolution reaches the highest sampling resolution or the iterative registration result satisfies the specified convergence condition, the multiple downsampled point clouds after registration at the current sampling resolution are spliced ​​to obtain the current spliced ​​point cloud.

7. The device according to claim 6, characterized in that The point cloud registration module is also used to: obtain a plurality of feature descriptors corresponding to the current sampling resolution based on the plurality of downsampled point clouds at the current sampling resolution, wherein the feature descriptors can describe the data points in the corresponding downsampled point clouds; perform feature matching based on the plurality of feature descriptors to obtain a transformation matrix at the current sampling resolution; and determine the iterative registration result based on the transformation matrix at the current sampling resolution.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the displacement monitoring method according to any one of claims 1 to 4 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the displacement monitoring method according to any one of claims 1 to 4 is implemented.

Images