A navigation point correction method, system and photovoltaic cleaning device

Abstract

The present application relates to the field of photovoltaic cleaning, and discloses a navigation point correction method, system, and photovoltaic cleaning device, wherein the method includes: using a multi-line laser radar set on a photovoltaic cleaning device to collect point cloud data of a target photovoltaic panel; constructing a circular point cloud map with the multi-line laser radar as the center, and partitioning the circular point cloud map; obtaining representative points of each partition; connecting the representative points of adjacent partitions to filter out the edge point cloud of the target photovoltaic panel; performing straight line fitting on the filtered edge point cloud; using the fitted edge straight line to calculate in real time the position of the latest target navigation point in the coordinate system centered on the multi-line laser radar. The present application uses point cloud partitioning technology to detect the expression equation of the edge of the photovoltaic panel in real time. No matter how the tilt angle of the photovoltaic panel changes, the cleaning device can always maintain a fixed cleaning distance from the photovoltaic panel, thereby improving cleaning efficiency and safety.

Description

Technical Field

[0001] The present application relates to the photovoltaic field, and in particular, to a navigation point correction method, system and photovoltaic cleaning device. Background Art

[0002] Photovoltaic cleaning devices usually rely on manually preset fixed navigation points. Usually, the photovoltaic cleaning device is parked at a fixed navigation point at a target distance away from the lower edge of the photovoltaic panel to achieve photovoltaic panel cleaning work. However, with the continuous development of photovoltaic technology, the photovoltaic tracking system monitors the position of the sun through sensors, and then drives the photovoltaic panel to rotate through the control system. As the angle of the photovoltaic panel changes due to the tracking system, the lower edge of the photovoltaic panel is constantly changing. The coordinate position of the fixed navigation point cannot dynamically and effectively adapt to the requirements of the photovoltaic cleaning device. Therefore, the photovoltaic cleaning device may not be able to reach the accurate position when performing the task, and even cause damage to the photovoltaic cleaning equipment. Summary of the invention

[0003] In order to solve the above technical problems, the present application provides a navigation point correction method, system and photovoltaic cleaning device. Through point cloud partitioning technology, the point cloud of the edge of the photovoltaic panel is screened and fitted with a straight line to obtain the edge straight line of the photovoltaic panel. By real-time detection of the expression equation of the edge straight line of the photovoltaic panel, the coordinate position of the new target navigation point in the coordinate system of the photovoltaic cleaning device is generated in real time, ensuring that no matter how the inclination angle of the photovoltaic panel changes, the photovoltaic cleaning device can always maintain a fixed cleaning distance from the photovoltaic panel, thereby improving cleaning efficiency and safety.

[0004] Specifically, the technical solution of this application is as follows:

[0005] In a first aspect, the present application discloses a navigation point correction method for navigation and positioning of a photovoltaic cleaning device, comprising:

[0006] Using a multi-line laser radar disposed on the photovoltaic cleaning device to collect point cloud data of the target photovoltaic panel;

[0007] Taking the multi-line laser radar as the center, constructing a ring point cloud map, and partitioning the ring point cloud map; obtaining a representative point of each partition;

[0008] Connecting the representative points of the adjacent partitions to filter out the edge point cloud of the target photovoltaic panel;

[0009] The screened edge point cloud is subjected to straight line fitting to obtain the edge straight line of the target photovoltaic panel; the fitted edge straight line is used to calculate in real time the position of the latest target navigation point in the coordinate system centered on the multi-line laser radar.

[0010] In some embodiments, constructing a ring point cloud map with the multi-line laser radar as the center and partitioning the ring point cloud map includes:

[0011] The environment is scanned in all directions by the multi-line laser radar to obtain all-round point cloud data and generate a ring point cloud map;

[0012] A three-dimensional coordinate system is established with the multi-line laser radar as the center; the annular point cloud image is divided according to a set angle to form a plurality of partitions.

[0013] In some embodiments, the step of obtaining a representative point of each partition comprises the following steps: dividing each partition into blocks according to the Euclidean distance of the xy plane at a preset resolution in the xy plane of the three-dimensional coordinate system at the center of the multi-line laser radar, obtaining the z value of all point clouds in each block, and taking the maximum z value of the point cloud in the block as the height of the block;

[0014] Depicting points corresponding to the heights of the respective blocks in a two-dimensional graph according to the distances from the multi-line laser radar;

[0015] The points corresponding to the height values ​​of the blocks are connected in sequence, and a representative point of each partition is determined from the points.

[0016] In some implementations, determining a representative point of each of the partitions from the points comprises the following steps:

[0017] Calculating the gradient change value between the points connected in sequence, and when the gradient change value is greater than a first gradient threshold, taking the points corresponding to both ends of the gradient change value connection line as candidate representative points of the partition;

[0018] Among the candidate representative points, the candidate representative points whose heights are within a second height threshold are screened as representative points of the partition.

[0019] In some embodiments, performing straight line fitting on the screened edge point cloud to obtain the edge straight line of the target photovoltaic panel comprises the following steps:

[0020] Performing outlier filtering on the edge point cloud;

[0021] The filtered edge point cloud is clustered; after clustering, straight line detection is performed to obtain the expression equation of the edge straight line.

[0022] In some implementations, the detecting obtains the expression equation of the edge line; specifically comprises the following steps:

[0023] Calculating a centroid of a set of edge point cloud data and performing centralization processing on the edge point cloud;

[0024] Translating the edge point cloud to a coordinate system with the centroid as the origin, and calculating the covariance matrix of the centralized edge point cloud;

[0025] Performing eigenvalue decomposition on the covariance matrix, and selecting an eigenvector corresponding to the maximum eigenvalue in the covariance matrix as the direction vector of the edge line;

[0026] An expression equation of the edge straight line is determined based on the direction vector and the centroid.

[0027] In some embodiments, the method of using the fitted edge straight line to calculate in real time the position of the latest target navigation point in a coordinate system centered on the multi-line laser radar comprises the following steps:

[0028] Identify the position coordinates of the initially set fixed navigation point; identify the position coordinates of the vertical projection point of the fixed navigation point on the edge straight line;

[0029] Calculate the direction vector of the line between the fixed navigation point and the vertical projection point;

[0030] Taking the vertical projection point as the starting point, the standard working distance is extended outward along the direction vector; and the new point position obtained after the extension is used as the target navigation point.

[0031] In some embodiments, the navigation point correction method further includes: generating a navigation path from the current position of the photovoltaic cleaning device to the target navigation point.

[0032] In a second aspect, the present application also discloses a navigation point correction system, comprising: a multi-line laser radar, a processor, a memory and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the navigation point correction method described in any of the above-mentioned embodiments.

[0033] In some embodiments, the multi-line laser radar is disposed on a photovoltaic cleaning device to collect point cloud data of a target photovoltaic panel;

[0034] The processor comprises:

[0035] A partitioning module is used to construct a ring point cloud map with the multi-line laser radar as the center, and partition the ring point cloud map; and obtain a representative point of each partition;

[0036] A screening module, used for connecting the representative points of the adjacent partitions to screen out the edge point cloud of the target photovoltaic panel;

[0037] A fitting module, used for performing straight line fitting on the screened edge point cloud to obtain an edge straight line of the target photovoltaic panel;

[0038] The calculation module is used to use the fitted edge straight line to calculate in real time the position of the latest target navigation point in the coordinate system centered on the multi-line laser radar.

[0039] In a third aspect, the present application further discloses a photovoltaic cleaning device, which at least includes the navigation point correction system described in any one of the above-mentioned embodiments.

[0040] Compared with the prior art, the present invention has at least one of the following beneficial effects:

[0041] 1. The coordinates of the target navigation point in this application can be adjusted in real time. Compared with the fixed navigation points used in the prior art, the technical content of this application can significantly improve the working stability and adaptability of the photovoltaic cleaning device. Traditional methods are difficult to deal with the angle changes of the photovoltaic panel caused by the tracking system. The present invention ensures that the photovoltaic cleaning device always maintains a fixed working distance from the photovoltaic panel through real-time monitoring and adjustment, thereby preventing poor cleaning effects or equipment damage due to improper distance. This improves cleaning efficiency, shortens cleaning cycles, and reduces maintenance costs.

[0042] 2. Traditional edge detection methods usually face the problem of laser point cloud sparsity, which makes it difficult to accurately detect edges in complex environments. The present invention constructs a ring point cloud map and uses point cloud partitioning technology to extract two-dimensional point cloud features from three-dimensional point cloud features for edge detection. By setting the size of the partition, the accuracy of the fitted edge line can be adjusted, overcoming the limitations caused by sparsity, making edge detection more reliable, and being able to effectively identify the edges of photovoltaic panels, maintaining high efficiency regardless of different lighting conditions or environmental complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The preferred implementation scheme will be described below in a clear and understandable manner with reference to the accompanying drawings to further illustrate the above-mentioned characteristics, technical features, advantages and implementation methods of the present application.

[0044] Figure 1 A flowchart showing a step of an embodiment of a navigation point correction method of the present application is shown;

[0045] Figure 2 A schematic diagram showing the effect of point cloud partitioning in one embodiment of the present application is shown;

[0046] Figure 3 A schematic diagram showing the effect of partition height in one embodiment of the present application is shown;

[0047] Figure 4A schematic diagram showing the effect of the connection lines formed by point cloud partitioning in one embodiment of the present application is shown;

[0048] Figure 5 A schematic diagram showing the effect of correcting a target navigation point in one embodiment of the present application is shown;

[0049] Figure 6 A structural block diagram showing an embodiment of a navigation point correction system of the present application

[0050] Figure 7 A schematic diagram of a specific scenario of a photovoltaic cleaning device cleaning a photovoltaic panel in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0051] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0052] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections.

[0053] In order to simplify the drawings, only the parts related to the invention are schematically shown in each figure, and they do not represent the actual structure of the product. In addition, in order to simplify the drawings and facilitate understanding, in some figures, only one of the parts with the same structure or function is schematically drawn or marked. In this article, "one" not only means "only one", but also means "more than one".

[0054] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0055] In this document, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0056] In a specific implementation, the terminal device described in the embodiments of the present application includes, but is not limited to, other portable devices such as mobile phones, laptop computers, tutoring machines, or tablet computers with touch-sensitive surfaces (e.g., touch screen displays and / or touch pads). It should also be understood that in some embodiments, the terminal device is not a portable communication device, but a desktop computer with a touch-sensitive surface (e.g., touch screen displays and / or touch pads).

[0057] In addition, in the description of the present application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0058] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the specific implementation methods of the present application will be described below with reference to the accompanying drawings. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings and other implementation methods can be obtained based on these drawings without creative work.

[0059] At present, the common photovoltaic cleaning devices on the market include split photovoltaic cleaning devices and integrated photovoltaic cleaning devices. Regardless of the type of photovoltaic cleaning device, when performing the cleaning task of the photovoltaic module, it is necessary to obtain a clear positioning point, that is, a navigation point.

[0060] In the prior art, photovoltaic cleaning devices usually rely on manually preset fixed navigation points. The integrated photovoltaic cleaning device runs to the fixed navigation point of each photovoltaic array before starting the cleaning work. Or, in the split photovoltaic cleaning device, the transfer vehicle runs to the fixed navigation point and starts the transfer cleaning robot. The fixed navigation point is generally set in front of the photovoltaic panel, at a position point that is kept a certain distance from the lower edge of the photovoltaic panel.

[0061] However, with the continuous development of photovoltaic technology, the photovoltaic tracking system monitors the position of the sun through sensors, and then drives the photovoltaic panel to rotate through the control system to keep the sunlight directly irradiating the photovoltaic array. This tracking system can continuously adjust the angle of the photovoltaic panel to track the trajectory of the sun and improve the utilization rate of solar energy and power generation efficiency. However, due to the change in the angle of the photovoltaic panel due to the tracking system, the lower edge of the photovoltaic panel is constantly changing, and the fixed navigation point coordinates cannot effectively adapt to this dynamic change. This causes the photovoltaic cleaning device to be unable to reach the exact position when performing the task, and thus cannot be successfully transferred to the next task point or cannot effectively perform the cleaning task. In reality, the wrong target navigation point will bring great harm. If the navigation point is close to the photovoltaic panel due to the rotation of the bracket, if the photovoltaic cleaning device fails to detect the real-time position of the photovoltaic panel in time during the movement, it may also collide with the photovoltaic panel. The photovoltaic cleaning device may hit the photovoltaic panel or cause obstacle avoidance to stop running. In addition, if it is far away, it will not achieve a good cleaning effect, which will not only reduce the cleaning efficiency, but also cause equipment damage and increase maintenance costs.

[0062] The present application proposes a navigation point correction method, system and photovoltaic cleaning device. The present application uses point cloud partitioning technology to screen the point cloud of the edge of the photovoltaic panel and fit it into a straight line to obtain the edge straight line of the photovoltaic panel. By real-time detection of the distance and direction to the edge straight line of the photovoltaic panel, the coordinate position of the new target navigation point in the coordinate system of the photovoltaic cleaning device is generated in real time to ensure that no matter how the inclination angle of the photovoltaic panel changes, the photovoltaic cleaning device can always maintain a fixed cleaning distance from the photovoltaic panel, thereby improving cleaning efficiency and safety.

[0063] Reference Manual Attached Figure 1 An embodiment of a navigation point correction method provided by the present application includes the following steps:

[0064] S100, using a multi-line laser radar arranged on the photovoltaic cleaning device to collect point cloud data of the target photovoltaic panel.

[0065] S200: construct a ring point cloud map with the multi-line laser radar as the center, partition the ring point cloud map, and obtain a representative point of each partition.

[0066] S300, connecting the representative points of the adjacent partitions to filter out the edge point cloud of the target photovoltaic panel.

[0067] S400, performing straight line fitting on the screened edge point cloud to obtain an edge straight line of the target photovoltaic panel.

[0068] S500, using the fitted edge straight line, calculate in real time the position of the latest target navigation point in the coordinate system centered on the multi-line laser radar.

[0069] Specifically, compared with ordinary single-line laser radar, multi-line laser radar has high resolution and data volume, and common ones are 16 / 32 / 64-line laser radar. Multi-line laser radar can emit and receive multiple laser beams at the same time, so as to obtain more data points in the same time, which enables them to provide higher resolution and more detailed environmental models. Since multi-line laser radars can perform multi-line scanning at the same time, they can cover a larger area per unit time, so the scanning speed is faster. In general, multi-line laser radars have better spatial perception. Multi-line laser radar can provide 360-degree environmental perception and collect all-round environmental point cloud data. It plays a key role in the step of constructing a ring point cloud map in this application.

[0070] In another embodiment of the present application, based on the above embodiment, step S200 specifically includes the following sub-steps:

[0071] S210, performing an all-round scan of the environment by means of the multi-line laser radar, acquiring all-round point cloud data, and generating a ring point cloud map.

[0072] S220, establishing a three-dimensional coordinate system with the multi-line laser radar as the center, and dividing the annular point cloud image into a plurality of partitions according to a set angle.

[0073] For details, please refer to the attached manual. Figure 2 As shown, Figure 2 A schematic diagram showing the effect of point cloud partitioning in one embodiment of the present application is shown.

[0074] In a three-dimensional coordinate system, the XYZ coordinate axes usually represent three mutually perpendicular axes, which correspond to different spatial directions. The X-axis represents the horizontal direction, for example, it can be regarded as the direction from south to north. The Y-axis also represents the horizontal direction, which is perpendicular to the X-axis, for example, it can be regarded as the direction from west to east. The specific azimuth angles of the X-axis and the Y-axis are not limited in this application. The Z-axis represents the direction perpendicular to the ground or horizontal plane. The X-axis and the Y-axis define a horizontal plane, then the Z-axis is perpendicular to this plane and can be regarded as the direction from the ground upward.

[0075] In this application, partitioning the annular point cloud image means dividing the 360° annular shape into several angle zones with the center point of the annular point cloud image as the origin, and the size of each zone is determined by the angle of the division. Figure 2 In the figure, the annular point cloud is divided into 12 zones, that is, 30° / zone. According to the sampling density and working environment of the multi-line laser radar, the annular point cloud can be divided into different numbers of zones.

[0076] Figure 2The size of the point cloud partition shown is relatively large and is only for illustration. Optionally, the angle of the partition is set by the relevant technicians according to the requirements of use. In this embodiment, the angle of the partition is not further limited.

[0077] The step S200 further includes the following sub-steps:

[0078] S230, obtaining a representative point of each partition.

[0079] In one implementation of the present embodiment, step S230 specifically includes: S231, in the xy plane of the three-dimensional coordinate system with the multi-line laser radar as the center, dividing the Euclidean distance of the xy plane into blocks with a preset resolution for each partition, obtaining the z value of all point clouds in each block, and taking the maximum z value of the point cloud in the block as the height of the block.

[0080] S232, depicting points corresponding to the heights of the various blocks in a two-dimensional graph according to the distance from the multi-line laser radar.

[0081] S233, sequentially connecting the points corresponding to the height values ​​of the respective blocks, and determining a representative point of each of the partitions from the points.

[0082] In this embodiment, in this application, the Euclidean distance of the xy plane is divided into blocks at a preset resolution for each partition, and size division is performed on the basis of a partition. The size division is based on the division of the X-axis or Y-axis on the plane, and no division is performed on the Z-axis.

[0083] For example, Figure 2 Each partition is divided into 3 blocks. Each partition is divided into several blocks according to the Euclidean distance in the xy plane. The specific number of blocks is determined by the point cloud sampling density and the working environment. For example, the blocks are divided according to different resolutions such as 3cm, 5cm, and 10cm.

[0084] Figure 2 The size of the point cloud partitions shown is relatively large and is only for illustration. Optionally, the size of the partitions and blocks can be set by relevant technicians according to the requirements of use. In this embodiment, the number and size of the partitions are not further limited.

[0085] Optionally, each block includes at least one point cloud, and the maximum value of the z value of the point cloud in each block is used as the height value of the block. Figure 3 As shown, Figure 3 This application is shown as Figure 2A schematic diagram of the three-dimensional effect of the height of each block in the block example shown. The height value of each block represents the maximum height of the point cloud in the block. Optionally, the smaller the size and angle of the point cloud partition, the denser the partitions of the annular point cloud map obtained by the partition, and the higher the accuracy of the edge point cloud filtered. On the contrary, the larger the size and angle of the partition, the sparser the partitions of the annular point cloud map obtained by the partition, and the lower the accuracy of the edge point cloud filtered.

[0086] In another implementation of this embodiment, step S231 specifically includes: obtaining the height value of each point cloud in each of the blocks, and taking the maximum height of the point cloud in the block as the height of the block.

[0087] As in the above embodiment, the smaller the size and angle of the point cloud partitioning are, the denser the partitions of the annular point cloud image obtained by the partitioning are, and the higher the precision of the edge point cloud screened is. On the contrary, the larger the size and angle of the partitioning are, the sparser the partitions of the annular point cloud image obtained by the partitioning are, and the lower the precision of the edge point cloud screened is.

[0088] In one implementation of this embodiment, step S232 depicts points corresponding to the heights of the various blocks in a two-dimensional graph according to the distance from the multi-line laser radar.

[0089] Step S233, sequentially connecting the points corresponding to the height values ​​of the various blocks, and determining the representative point of each of the partitions from the points.

[0090] Specifically, as attached Figure 4 As shown, the distance between each block in each partition and the multi-line laser radar is different, and the two-dimensional graph is drawn with the distance as the X value and the height of different blocks as the Y value.

[0091] Connect as shown in the attached Figure 4 The representative points of each partition can be selected from the points depicted in . Each partition can be drawn in a two-dimensional graph in the above manner to determine the representative points of each partition.

[0092] In another embodiment of the present application, based on the above embodiment, step S300 specifically includes the following sub-steps:

[0093] S310, calculating the gradient change value between the points connected in sequence, and when the gradient change value is greater than a first gradient threshold, taking the points corresponding to both ends of the gradient change value connection line as candidate representative points of the partition.

[0094] S320: Select, from among the candidate representative points, the candidate representative points whose heights are within a second height threshold as representative points of the partition.

[0095] Specifically, connect each divided point, such as Figure 4 As shown, the gradient change value of each point is calculated. When the gradient change value exceeds a certain threshold, the current point can be considered as a candidate representative point, and the starting point of the connecting line is also used as a candidate representative point. Specifically, Figure 4 The connecting lines formed by the point cloud partitions in one direction starting from the origin of the coordinate system in the annular point cloud diagram are shown. The gradient change between any point and its adjacent point is first calculated. The relationship between the gradient between two points and the height of the two points can be understood by simple trigonometric functions, namely: gradient = vertical height change / horizontal distance change. The change value of the gradient can be regarded as the change in height at a certain horizontal distance. If the gradient change value is greater than the first gradient threshold, it is considered that the point partition may be at the edge of the photovoltaic panel, and the corresponding point is used as an alternative representative point.

[0096] Among the candidate representative points, further screening is performed. Since the height of the photovoltaic panels in each site varies within a range, a second height threshold, that is, the z-axis threshold, is set. When the z-coordinate value of the candidate representative point is within the second height threshold, the candidate representative point is determined to be the representative point of the partition.

[0097] Specifically, since each photovoltaic panel includes at least its upper edge, lower edge, or also includes a side edge, in the present application, when searching for the target navigation point, only the lower edge of the photovoltaic panel needs to be identified. Therefore, in the present embodiment, the second height threshold of the facility is the height range where the lower edge of the photovoltaic panel is usually located.

[0098] Optionally, in this embodiment, the values ​​of the first gradient threshold and the second height threshold are set by personnel in this field according to specific circumstances and are not specifically limited in this embodiment.

[0099] Another embodiment of a navigation point correction method of the present application, based on an embodiment of the above method, step S400 includes the following steps:

[0100] S410: Perform outlier filtering on the edge point cloud.

[0101] S420: Clustering the filtered edge point cloud.

[0102] S430, performing straight line detection after clustering to obtain an expression equation of the edge straight line.

[0103] Specifically, outlier filtering is performed on the screened edge points, straight line detection is performed on the filtered point cloud, and equation parameters of the detected edge straight lines are recorded.

[0104] The method used for line detection is to cluster the filtered point cloud and distribute several different classes, then set the length and width thresholds, and filter the point cloud classes relative to the approximate position of the vehicle body, retain the class that meets the restrictions, and perform line detection on this class.

[0105] In some implementations of this embodiment, the above steps S410-S430 specifically include the following details:

[0106] 1. Calculate the centroid of a set of edge point cloud data and perform centralization on the edge point cloud. The centroid of the point cloud data refers to the geometric center point of the point cloud data in space, which represents the average position of all points.

[0107] 2. Translate the edge point cloud to a coordinate system with the centroid as the origin, and calculate the covariance matrix of the centralized edge point cloud.

[0108] 3. Perform eigenvalue decomposition on the covariance matrix, and select the eigenvector corresponding to the maximum eigenvalue in the covariance matrix as the direction vector of the edge line.

[0109] 4. Determine an expression equation of the edge straight line based on the direction vector and the centroid.

[0110] Specifically, assuming that the point cloud data consists of a set of three-dimensional points Composition, where each point It can be expressed as .

[0111] Calculate the centroid of the point cloud And the point cloud is centrally processed.

[0112] ;

[0113] Translate all points to a coordinate system with the centroid as the origin, that is, perform the following operations on each point:

[0114] ;

[0115] in It is the centralized point cloud data.

[0116] Calculate the covariance matrix of the centralized point cloud data .

[0117] ;

[0118] The specific form is:

[0119] ;

[0120] in, , the same goes for other items.

[0121] Covariance matrix Perform eigenvalue decomposition:

[0122] ;

[0123] in, is the eigenvalue of the covariance matrix, is the corresponding feature vector.

[0124] Select the covariance matrix The maximum eigenvalue of The corresponding eigenvector As the direction vector of the line :

[0125] ;

[0126] The direction of the line is now determined by the vector Given.

[0127] Direction vector of the line and centroid The equation of the line can be determined. The parametric equation of the line is:

[0128] ;

[0129] in, is a real number parameter. The explicit form of the straight line can be expressed as:

[0130] ;

[0131] in are the centroid coordinates, are the components of the direction vector.

[0132] Another embodiment of a navigation point correction method of the present application, based on an embodiment of the above method, step S500 includes the following steps:

[0133] S510, identifying the position coordinates of an initially set fixed navigation point, and identifying the position coordinates of a vertical projection point of the fixed navigation point on the edge straight line.

[0134] S520, calculating the direction vector of the line between the fixed navigation point and the vertical projection point.

[0135] S530: Taking the vertical projection point as the starting point, extending the standard working distance outward along the direction vector, and taking the new point position obtained after the extension as the target navigation point.

[0136] Specifically, the straight line equation between the originally set fixed navigation point and its projection point on the fitted edge straight line is calculated. Through the straight line equation, the coordinate value of the point on the straight line equation that is the standard working distance d from the projection point is calculated. This point is the new target navigation point. Figure 5 As shown, Figure 5 A schematic diagram showing the effect of target navigation point correction is shown.

[0137] Known: Fixed navigation points: ;

[0138] Equation of a Line: Through a Point and direction vector The straight line represented by

[0139] Standard working distance: ;

[0140] Calculate fixed navigation points Projection point on the line .

[0141] Projection Point The coordinate calculation steps are as follows:

[0142] Calculate the fixed navigation point to the starting point of the straight line A vector of:

[0143] ;

[0144] Calculating vectors In the direction vector Projection on:

[0145] ;

[0146] Get the projection point Coordinates:

[0147] ;

[0148] Calculate the standard working distance New target navigation point at ;

[0149] From the projection point Along the straight line Extend to the point with a distance of d, which is the new target navigation point :

[0150] Calculate new target navigation point Coordinates:

[0151] ;

[0152] The magnitude of the direction vector is for:

[0153] ;

[0154] Substituting the coordinates we get:

[0155] ;

[0156] Through the above steps, we calculated the new target navigation point The coordinates of the projection point are on the straight line and the distance from the projection point is the standard working distance. .

[0157] In another embodiment of the present application, the navigation point correction method further includes: S600, generating a navigation path from the current position of the photovoltaic cleaning device to the target navigation point. Specifically, based on the annular point cloud map, a path planning algorithm, such as a Dijkstra algorithm, an RRT algorithm, etc., is used to avoid obstacles on the path and plan a path.

[0158] The coordinates of the target navigation point in this application can be adjusted in real time. Compared with the fixed navigation points used in the prior art, the technical content of this application can significantly improve the working stability and adaptability of the photovoltaic cleaning device. Traditional methods are difficult to deal with the angle changes of the photovoltaic panel caused by the tracking system. The present invention ensures that the photovoltaic cleaning device always maintains a fixed working distance from the photovoltaic panel through real-time monitoring and adjustment, thereby preventing poor cleaning effects or equipment damage due to improper distance. This improves cleaning efficiency, shortens cleaning cycles, and reduces maintenance costs.

[0159] Different from the densely arranged fixed pixels of the camera, the point cloud data of the traditional LiDAR shows a certain sparsity. The distribution in three-dimensional space is very uneven. As a result, it is difficult for the edge detection algorithm to accurately detect the edge in a complex environment. The present invention constructs a ring point cloud map and uses point cloud partitioning technology to extract two-dimensional point cloud features from three-dimensional point cloud features for edge detection. The accuracy of the fitted edge line can be adjusted by setting the size of the partition. It overcomes the limitations brought by sparsity, making edge detection more reliable, and can effectively identify the edges of photovoltaic panels, maintaining high efficiency regardless of different lighting conditions or environmental complexity.

[0160] Based on the same technical concept, the present application also discloses a navigation point correction system, including a multi-line laser radar, a memory, a processor, etc. The system can be used to implement the steps of a navigation point correction method described in any of the above embodiments.

[0161] Specifically, a navigation point correction system embodiment of the present application is as shown in the attached specification. Figure 6As shown, the multi-line laser radar is arranged on a photovoltaic cleaning device to collect point cloud data of a target photovoltaic panel.

[0162] The processor includes: a partitioning module, which is used to construct a ring point cloud map with the multi-line laser radar as the center, and partition the ring point cloud map; and obtain edge points of each partition.

[0163] The screening module is used to connect the edge points of the adjacent partitions to screen out the edge point cloud of the target photovoltaic panel.

[0164] The fitting module is used to perform straight line fitting on the screened edge point cloud to obtain the edge straight line of the target photovoltaic panel.

[0165] The calculation module is used to use the fitted edge straight line to calculate in real time the position of the latest target navigation point in the coordinate system centered on the multi-line laser radar.

[0166] Specifically, multi-line laser radars have higher resolution and more data than ordinary single-line laser radars. Common ones are 16 / 32 / 64-line laser radars. Multi-line laser radars can emit and receive multiple laser beams at the same time, thereby acquiring more data points in the same time, which enables them to provide higher resolution and more detailed environmental models. Since multi-line laser radars can perform multi-line scanning at the same time, they can cover a larger area per unit time, so the scanning speed is faster. Generally speaking, multi-line laser radars have better spatial perception. Multi-line laser radars can provide 360-degree environmental perception and collect all-round environmental point cloud data.

[0167] The partitioning module is specifically used to: perform an all-round scan of the environment through the multi-line laser radar, obtain all-round point cloud data, and generate a ring point cloud map. A three-dimensional coordinate system is established with the multi-line laser radar as the center; and the ring point cloud map is divided according to the set angle to form multiple partitions.

[0168] The screening module is specifically used to: connect adjacent partitions, calculate the gradient change value between the heights of adjacent partitions connected by height; when the gradient change value is greater than a first gradient threshold, use the point cloud in the corresponding partition as a candidate edge point; and among the candidate edge points, screen the point cloud whose point cloud height is within a second height threshold as the edge point cloud.

[0169] The fitting module is used to: perform outlier filtering on the edge point cloud; cluster the filtered edge point cloud; perform straight line detection after clustering to obtain the expression equation of the edge straight line.

[0170] In other embodiments, the fitting module is specifically used to: calculate the centroid of a set of edge point cloud data and centralize the edge point cloud; translate the edge point cloud to a coordinate system with the centroid as the origin, and calculate the covariance matrix of the centralized edge point cloud; perform eigenvalue decomposition on the covariance matrix, and select the eigenvector corresponding to the maximum eigenvalue in the covariance matrix as the direction vector of the edge straight line; determine the expression equation of the edge straight line based on the direction vector and the centroid.

[0171] The calculation module is specifically used to: identify the position coordinates of the initially set fixed navigation point; identify the position coordinates of the vertical projection point of the fixed navigation point on the edge straight line; calculate the direction vector of the line between the fixed navigation point and the vertical projection point; take the vertical projection point as the starting point, and extend the standard working distance outward along the direction vector; and use the new point position obtained after the extension as the target navigation point.

[0172] Based on the same concept, the present application also discloses a photovoltaic cleaning device, which at least includes the navigation point correction system described in any one of the above embodiments.

[0173] refer to Figure 7 , Figure 7 A schematic diagram of a specific scenario of the photovoltaic cleaning device cleaning a photovoltaic panel in this embodiment is shown. In other embodiments of the present application, the photovoltaic cleaning device may also include:

[0174] Walking system: responsible for driving the cleaning device to move along the photovoltaic panel. The walking system may include different forms such as wheeled, crawler or self-propelled to adapt to different photovoltaic panel installation environments and terrain conditions.

[0175] Cleaning system: usually includes brushes, scrapers, vacuum devices, etc., used to remove dust, dirt and other foreign matter from photovoltaic panels.

[0176] Navigation and positioning system: including GPS, IMU (inertial measurement unit), LiDAR (laser rangefinder) and other technologies, used to achieve precise positioning and navigation of the cleaning device within the photovoltaic array.

[0177] Path planning system: Based on the collected environmental information, a path planning algorithm is used to determine the best cleaning path to ensure coverage of the entire photovoltaic array surface.

[0178] Energy supply system: Photovoltaic cleaning devices are usually equipped with batteries or obtain energy through solar panels to keep the robot running.

[0179] Intelligent control system: Integrates sensor data, path planning, environmental perception and other information, and enables the robot to perform cleaning tasks efficiently through real-time decision-making and adjustments.

[0180] Dust detection device: used to detect the degree of contamination on the surface of photovoltaic panels so that the robot can adjust the cleaning strategy to improve cleaning efficiency and effectiveness; etc.

[0181] These components work together to ensure that the photovoltaic cleaning device can efficiently and intelligently complete the cleaning task of the photovoltaic panels, thereby improving the power generation efficiency and economic benefits of the photovoltaic power station.

[0182] A navigation point correction method, system and photovoltaic cleaning device of the present application have the same technical concept, and the technical details of the three embodiments are applicable to each other. In order to reduce repetition, they will not be repeated here.

[0183] Based on the same concept, the present application also needs to disclose a computer-readable storage medium on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the steps of a navigation point correction method described in any of the above embodiments are implemented.

[0184] The present application also needs to disclose a computer program product including a computer program / instruction, which, when executed by a processor, implements the steps of a navigation point correction method described in any one of the above embodiments.

[0185] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned program modules is used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different program modules as needed, that is, the internal structure of the device can be divided into different program units or modules to complete all or part of the functions described above. The program modules in the embodiment can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into a processing unit, and the above-mentioned integrated unit can be implemented in the form of hardware or in the form of software program units. In addition, the specific names of the program modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application.

[0186] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0187] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A navigation point correction method for navigation and positioning of a photovoltaic cleaning device, characterized in that: The steps include: Using a multi-line laser radar disposed on the photovoltaic cleaning device to collect point cloud data of the target photovoltaic panel; Taking the multi-line laser radar as the center, constructing a ring point cloud map, and partitioning the ring point cloud map; obtaining a representative point of each partition; Connecting the representative points of the adjacent partitions to filter out the edge point cloud of the target photovoltaic panel; Performing straight line fitting on the screened edge point cloud to obtain the edge straight line of the target photovoltaic panel; using the fitted edge straight line, calculating in real time the position of the latest target navigation point in the coordinate system centered on the multi-line laser radar; The method of performing straight line fitting on the screened edge point cloud to obtain the edge straight line of the target photovoltaic panel comprises the following steps: Performing outlier filtering on the edge point cloud; Clustering the edge point cloud after filtering; performing straight line detection after clustering to obtain the expression equation of the edge straight line; The detection obtains the expression equation of the edge straight line; specifically includes: Calculating a centroid of a set of edge point cloud data and performing centralization processing on the edge point cloud; Translating the edge point cloud to a coordinate system with the centroid as the origin, and calculating the covariance matrix of the centralized edge point cloud; Performing eigenvalue decomposition on the covariance matrix, and selecting an eigenvector corresponding to the maximum eigenvalue in the covariance matrix as the direction vector of the edge line; An expression equation of the edge straight line is determined based on the direction vector and the centroid.

2. A navigation point correction method as claimed in claim 1, characterized in that: The method of constructing a ring point cloud map with the multi-line laser radar as the center and partitioning the ring point cloud map includes: The environment is scanned in all directions by the multi-line laser radar to obtain all-round point cloud data and generate a ring point cloud map; A three-dimensional coordinate system is established with the multi-line laser radar as the center; the annular point cloud image is divided according to a set angle to form a plurality of partitions.

3. A navigation point correction method as claimed in claim 2, characterized in that: The step of obtaining the representative point of each partition comprises the following steps: In the xy plane of the three-dimensional coordinate system centered on the multi-line laser radar, the Euclidean distance of the xy plane is divided into blocks at a preset resolution for each partition, the z value of all point clouds in each block is obtained, and the maximum value of the z value of the point cloud in the block is used as the height of the block; Depicting points corresponding to the heights of the blocks in a two-dimensional graph according to the distances from the multi-line laser radar; The points corresponding to the height values ​​of the blocks are connected in sequence, and the representative point of each partition is determined from the points.

4. A navigation point correction method as claimed in claim 3, characterized in that: The step of determining the representative point of each partition from the points comprises the following steps: Calculating the gradient change value between the points connected in sequence, and when the gradient change value is greater than a first gradient threshold, taking the points corresponding to both ends of the gradient change value connection line as candidate representative points of the partition; Among the candidate representative points, the candidate representative points whose heights are within a second height threshold are screened as representative points of the partition.

5. A navigation point correction method as claimed in claim 1, characterized in that: The method of using the fitted edge straight line to calculate in real time the position of the latest target navigation point in a coordinate system centered on the multi-line laser radar comprises the following steps: Identify the position coordinates of the initially set fixed navigation point; identify the position coordinates of the vertical projection point of the fixed navigation point on the edge straight line; Calculate the direction vector of the line between the fixed navigation point and the vertical projection point; Taking the vertical projection point as the starting point, the standard working distance is extended outward along the direction vector; and the new point position obtained after the extension is used as the target navigation point.

6. A navigation point correction system, characterized in that: include: A multi-line laser radar, a processor, a memory and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the navigation point correction method according to any one of claims 1 to 5.

7. A navigation point correction system as claimed in claim 6, characterized in that: The multi-line laser radar is arranged on the photovoltaic cleaning device to collect point cloud data of the target photovoltaic panel; The processor comprises: A partitioning module is used to construct a ring point cloud map with the multi-line laser radar as the center, and partition the ring point cloud map; and obtain a representative point of each partition; A screening module, used for connecting the representative points of the adjacent partitions to screen out the edge point cloud of the target photovoltaic panel; A fitting module, used for performing straight line fitting on the screened edge point cloud to obtain an edge straight line of the target photovoltaic panel; The calculation module is used to use the fitted edge straight line to calculate in real time the position of the latest target navigation point in the coordinate system centered on the multi-line laser radar.

8. A photovoltaic cleaning device, characterized in that: The cleaning device at least includes the navigation point correction system described in any one of claims 6-7.

Images