Method and system for warning engineering vehicle intrusion under high-voltage lines

Abstract

The present invention discloses a method and system for warning engineering vehicles intruding under high-voltage power lines. The method includes: obtaining point cloud data collected by a radio frequency radar and located under the high-voltage power lines and performing preprocessing; clustering the preprocessed point cloud data to detect intruding targets; matching the level of the three-dimensional warning zone to which the intruding targets belong based on their spatial coordinates; controlling a visible laser array component to track the intruding targets and emit a laser warning light curtain of a corresponding level based on the spatial coordinates of the intruding targets and the level of the three-dimensional warning zone to which they belong; and controlling a voice speaker component to issue a voice warning of a corresponding level or a customized voice warning based on the level of the three-dimensional warning zone to which the intruding targets belong. By detecting the intruding targets and their spatial positions in real time, when an intrusion into the three-dimensional warning zone is determined, controlling the visible laser array component to emit a laser warning light curtain of a corresponding level and to track the intruding targets for warning, and combining this with a voice warning, the warning effect is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission line safety, and in particular to a method and system for warning of engineering vehicle intrusion under a high-voltage line. Background Art

[0002] In recent years, transmission line tripping and power outages caused by external factors, such as illegal construction, have shown a significant upward trend. External damage has become a major factor impacting the safe and reliable operation of transmission lines. In particular, large construction machinery, such as cranes, approaching live wires or accidentally contacting them in blind spots below high-voltage transmission lines, causing short circuits, wire damage, widespread power outages, and even fatalities. This is especially true at night or during rainy, snowy, foggy, or hazy weather, where visibility is poor. Crane operators are more likely to approach or even contact high-voltage lines due to a lack of effective observation and long-range warning lights and sounds, potentially leading to electric shock accidents. Traditional, regular manual inspections are inefficient, time-consuming, and labor-intensive, often prone to negligence and lax oversight, making them ineffective for timely prevention and on-site control.

[0003] The existing light source warning schemes and their shortcomings are as follows: In the current light source warning device applications supporting various early warning system equipment, scattered source LEDs or tungsten filament lamps are used for on-site visual warnings. The visible diffuse scattered light source is only about 8-30 meters away depending on the weather. The warning distance of its visible light is extremely short. It does not have the ability to penetrate a large range of long-distance visible warnings at night, in rain, fog, and haze weather. It cannot play a role in on-site warnings beyond 30 meters, which can easily cause on-site accidents. The current light source warning device only has the function of switching on and off, and does not have the function of matrix-forming light curtain outputs with different warning colors according to different application scenarios. It cannot coordinate with different scene applications to automatically or remotely control the omnidirectional light beam to form a light curtain for target positioning, tracking, pointing, warning, strobe frequency adjustment, and power supply adjustment.

[0004] The existing sound warning scheme and its shortcomings are as follows: the current existing voice alarm warning device's broadcast control method cannot be remotely modified and recorded in real time. It basically adopts the form of solidifying the sound into the built-in voice chip. Like the warning light control method, it is triggered by a switch to drive the playback of a specified single content. During the period, it cannot be remotely controlled or changed. Its control is very inflexible. The same voice equipment cannot be applied to different scenarios. It needs to be pre-recorded and stored in the voice chip in advance for solidification. In the application of shouting, it does not have the remote speaking function with low data volume. The shouting and broadcasting functions of the existing voice warning device are separate and independent It does not support the remote voice function. If voice calls are required, an additional set of real-time voice acquisition and drive transmission devices must be added. Remote calls can only be made through traditional telephone call analog transmission. At the same time, remote real-time voice calls not only occupy a large amount of analog voice audio traffic bandwidth, but also have to take into account the intermittent and stuck signals caused by poor mobile base station coverage in remote mountainous areas, and the instability caused by non-standard pronunciation of different callers and errors. These problems greatly limit the scope of use of flexible early warning broadcasts, and cannot be broadcast in real time for different function broadcast commands issued by the remote background alarm control platform center.

[0005] The above-mentioned traditional sound and light warning devices have a very small effective area and can only provide early warning prompts to various types of workers and engineering machinery vehicles working under high-voltage transmission lines in a small range. They are unable to solve the problem of on-site sound and light early warning prompts between towers within 500 meters on the transmission line or even at a longer distance. Summary of the Invention

[0006] In view of the deficiencies in the prior art, the present invention provides a method and system for warning of engineering vehicle intrusion under high-voltage power lines to solve the problem that the existing sound and light warning methods are difficult to apply to warning of engineering vehicle intrusion under high-voltage power transmission lines.

[0007] In a first aspect, a method for warning of intrusion of engineering vehicles under high-voltage lines is provided, comprising the following steps:

[0008] S1: Obtain point cloud data collected by RF radar under high-voltage power lines and perform coordinate conversion and filtering preprocessing;

[0009] S2: Cluster the pre-processed point cloud data to detect intrusion targets;

[0010] S3: Match the three-dimensional warning zone level to which the intrusion target belongs based on its spatial coordinates. The three-dimensional warning zone is set at multiple levels based on the distance from the high-voltage cable.

[0011] S4: According to the spatial coordinates of the intruding target and the level of the three-dimensional warning zone to which it belongs, the visible laser array component is controlled to track the intruding target and emit a laser warning light curtain of the corresponding level to warn;

[0012] S5: According to the level of the three-dimensional warning zone to which the intrusion target belongs, the voice speaker component is controlled to issue a voice warning of the corresponding level or a customized voice warning.

[0013] Furthermore, after completing the detection of the intrusion target using the point cloud data collected by the RF radar, the position of the intrusion target is calibrated using the pan-tilt camera assembly, wherein the pan-tilt camera assembly is set adjacent to the RF radar, and the 0-degree direction of the camera in the pan-tilt camera assembly is parallel to the normal of the RF radar. The calibration process includes:

[0014] Control the pan / tilt camera assembly to aim at the intrusion target based on the location information of the intrusion target, and adjust the camera focus according to the distance between the intrusion target and the RF radar;

[0015] Capture an image of the intruding target and track the top of the intruding target through image recognition detection, control the pan-tilt camera assembly to place the top of the intruding target at the center of the image, and obtain the horizontal deflection angle α and vertical deflection angle β of the pan-tilt camera assembly at the current moment;

[0016] The x-axis coordinate of the intrusion target detected by the point cloud data collected by the radio frequency radar is used to approximate the current x-axis coordinate of the top of the intrusion target. The three-dimensional spatial coordinate of the top of the intrusion target is calculated by combining the horizontal deflection angle α and the vertical deflection angle β of the gimbal camera assembly at the current moment.

[0017] The three-dimensional spatial coordinates of the top of the intrusion target are used as the final three-dimensional spatial coordinates of the intrusion target.

[0018] Furthermore, the visible laser array assembly is arranged on the pan-tilt camera assembly, and the irradiation direction of the visible laser array assembly is the same as the camera direction.

[0019] Furthermore, the parameters of the laser warning light curtain emitted by the visible laser array assembly include the irradiated target position, and one or more of the number of beams, laser color, laser spot size, and laser strobe period.

[0020] Furthermore, in step S1, the coordinate conversion process is expressed as follows:

[0021] X=X'

[0022] Y=Y'

[0023] Z=Z'+(X*tgθ)+Z0

[0024] Where (X', Y', Z') represents the point cloud coordinates in the radar coordinate system, (X, Y, Z) represents the point cloud coordinates converted to the ground coordinate system, Z0 represents the installation height of the RF radar, and θ represents the calibration conversion angle.

[0025] Furthermore, in step S1, the point cloud filtering process includes:

[0026] S11: Estimate the two-dimensional plane of the point cloud data through RANSAC, and use the points inside the point as ground points and other points as non-ground points to achieve the segmentation of ground points and non-ground points;

[0027] S12: Filter out ground points and retain non-ground points;

[0028] S13: Perform straight-through filtering on the retained non-ground points (x, y, z, RCS) within a limited range. For non-ground points, only those within the multi-level three-dimensional warning zone and whose RCS values ​​are within the set range are selected; non-engineering vehicle point cloud data are filtered by pre-setting the engineering vehicle RCS value range.

[0029] Furthermore, after step S2, the following steps are further included:

[0030] For other targets within the preset offset range of the stable and effective intrusion target, determine whether their RCS values ​​are equal to the RCS value of the stable and effective intrusion target. If they are equal, they are considered to be the same intrusion target and recorded; if the RCS values ​​are not equal, they are considered to be new intrusion targets and recorded.

[0031] Furthermore, between steps S2 and S3, the following steps are also included:

[0032] S01: Match the detected intrusion target in the current frame with the intrusion target list; the intrusion target list records the location information and health value of each intrusion target and is initially empty;

[0033] S02: Calculate the distance between the intrusion target detected in the current frame and each intrusion target in the intrusion target list. If the distance between the two is less than the set distance threshold, they are considered to be the same intrusion target, and the location information of the intrusion target is updated, and its health value is restored to the initial value N. If the distance between the two is not less than the set distance threshold, it is considered to be a new intrusion target, and the intrusion target and its location information are added to the intrusion target list, and its health value is set to the initial value N.

[0034] S03: If there is an intrusion target in the intrusion target list that has not been matched to itself in the current frame, the health value of the intrusion target in the intrusion target list is reduced by 1; when the health value of an intrusion target in the intrusion target list reaches 0, it is deleted from the intrusion target list;

[0035] S04: After continuously collecting M frames and performing the above steps S01 to S03, the intrusion target and its position information finally retained are used as the intrusion target detection result; wherein M>N.

[0036] Furthermore, a three-dimensional warning zone is set directly below the high-voltage line. The entire three-dimensional warning zone is divided into multi-level three-dimensional warning zones by multiple planes with the same horizontal xy plane coordinates and different vertical z-axis coordinates; the closer the three-dimensional warning zone is to the high-voltage line, the higher its warning level.

[0037] In the second aspect, a high-voltage line engineering vehicle intrusion warning system is provided, including an MCU controller component module and a power supply and voltage stabilization module electrically connected thereto, a wireless communication component module, a pan-tilt motor linkage control drive component module, a TTS digital intelligent voice synthesis amplifier component module, a PWM-TTL laser array modulation drive output component module, a camera, and a radio frequency radar. It also includes a voice speaker component connected to the TTS digital intelligent voice synthesis amplifier component module, a visible laser array component connected to the PWM-TTL laser array modulation drive output component module, and a pan-tilt motor component connected to the pan-tilt motor linkage control drive component module; the camera is arranged on the pan-tilt motor component;

[0038] The MCU controller component module is configured to execute the above-mentioned method for warning of engineering vehicle intrusion under high-voltage lines.

[0039] Furthermore, when the MCU controller component module receives the custom voice warning instruction issued by the external warning host, it parses it, converts the format and outputs the hexadecimal data to the AI ​​chip built into the TTS digital intelligent voice synthesis amplifier component module for decoding, and finally intelligently converts the low-data-volume text form obtained by decoding into a natural voice stream of an analog signal, and drives the voice speaker component through the built-in audio power amplifier output of the TTS digital intelligent voice synthesis amplifier component module to broadcast or shout warnings.

[0040] The present invention proposes a method and system for warning of engineering vehicle intrusion under high-voltage power lines, which has the following beneficial effects:

[0041] (1) By detecting the intrusion target and its spatial position in real time, when it is determined that the intrusion target has invaded the three-dimensional warning zone, the visible laser array component is controlled to emit a laser warning light curtain of the corresponding level and track the intrusion target to issue an early warning. In addition, combined with voice warning, the early warning effect is greatly improved;

[0042] (2) By using a laser warning light curtain combined with voice for graded warning, the problem of short warning distance and non-remote control at night, in foggy and rainy weather is solved;

[0043] (3) When performing spatial positioning of an intrusion target, preliminary positioning is performed using point cloud data collected by the radio frequency radar, and then precise positioning of the top of the intrusion target is performed in combination with video images. This can solve the problem of poor positioning and intrusion judgment results caused by the loss of point cloud data on the top of the intrusion target under noise interference (especially large noise interference at long distances), thereby improving the accuracy of early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 This is a flow chart of a method for warning of engineering vehicle intrusion under a high-voltage line provided by an embodiment of the present invention;

[0046] Figure 2 is a schematic diagram of coordinate system conversion provided by an embodiment of the present invention;

[0047] Figure 3 This is a schematic diagram of the splitting of multi-level three-dimensional warning zones provided by an embodiment of the present invention;

[0048] Figure 4 Schematic diagram of an engineering vehicle intrusion warning system under high-voltage lines provided by an embodiment of the present invention;

[0049] Figure 5 It is a schematic structural diagram of the pan-tilt motor assembly provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] To make the objectives, technical solutions, and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other implementations obtained by those of ordinary skill in the art without inventive effort are within the scope of protection of the present invention.

[0051] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "center", "longitudinal", "lateral", "vertical", "horizontal", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they cannot be understood as limiting the present invention. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be a central element at the same time. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0052] like Figure 1 As shown, an embodiment of the present invention provides a method for warning of engineering vehicle intrusion under a high-voltage line, comprising the following steps:

[0053] S1: Obtain point cloud data under the high-voltage power line collected by the radio frequency radar and perform coordinate conversion and filtering preprocessing.

[0054] Radar coordinate system convention: Convert the radar coordinate system to the ground coordinate system. Specifically, the radar ground projection is used as the coordinate origin, the radar normal in the ground projection direction is the X-axis, the perpendicular to the X direction is the Y-axis, and the Z-axis is the plumb line of the X, Y plane.

[0055] Spatial coordinate data calibration and data conversion calculation implementation: The schematic diagram of calibration with two calibration objects is as follows Figure 2 As shown, the coordinate transformation process is expressed as follows:

[0056] X=X'

[0057] Y=Y'

[0058] Z=Z'+(X*tgθ)+Z0

[0059] Where (X', Y', Z') represents the point cloud coordinates in the radar coordinate system, (X, Y, Z) represents the point cloud coordinates converted to the ground coordinate system, Z0 represents the installation height of the RF radar detection sensor, and θ represents the calibration conversion angle.

[0060] Filtering preprocessing: The point cloud filtering process includes:

[0061] S11: Estimate the two-dimensional plane of the point cloud data through RANSAC, take the local points as ground points and other points as non-ground points to achieve the segmentation of ground points and non-ground points.

[0062] The RANSAC algorithm takes as input a set of observations, a parameterized model that can explain or adapt to the observations, and some credible parameters. It achieves this goal by repeatedly selecting a random subset of the data. The selected subsets are assumed to be inliers, and the specific process can be described as follows:

[0063] Step 1. Randomly select at least three points to fit the plane;

[0064] Step 2. Use the plane obtained in step 1 to test all other point cloud data. If the distance between a point and the plane is less than the threshold, it is considered to be an internal point, otherwise it is an external point.

[0065] Step 3. If enough points are classified as hypothetical inliers, then the estimated model is reasonable enough;

[0066] Step 4. Then, use all the hypothesized inliers to re-estimate the plane, since it has only been estimated by the initial hypothesized inliers.

[0067] Step 5. Repeat steps 1-4 for a preset number of times to select the plane with the most fitted points. The points in the fitted plane are considered ground points, and the remaining points are considered non-ground points.

[0068] S12: Filter out ground points and retain non-ground points. By filtering out ground points, the amount of data calculation in the later stage is reduced, and the intrusion target detection process is accelerated.

[0069] S13: Perform straight-through filtering on the retained non-ground points (x, y, z, RCS) within a limited range. For non-ground points, only those within the multi-level three-dimensional warning zone and whose RCS values ​​are within the set range are selected. Non-engineering vehicle point cloud data can be filtered by pre-setting the engineering vehicle RCS value range.

[0070] First, point cloud data is filtered based on the coordinates of the points, retaining only those within the 3D warning zone. Secondly, considering that different objects have different RCS values ​​(RCS refers to the ratio of the target's return power density to the incident power density of the radar wave), by setting the RCS value range for engineering vehicles, the point cloud data of some non-intrusive objects can be filtered out, thereby eliminating interference from people, trees, etc.

[0071] S2: Cluster the pre-processed point cloud data to detect intrusion targets (such as the top of the mechanical arm of an engineering vehicle, such as the top of the crane's boom).

[0072] This example uses a clustering method based on Euclidean distance. Its core concept is to group closely spaced points into one category, while further points are grouped into different categories. Each resulting point cloud cluster represents an intrusion target point cloud cluster, enabling intrusion target detection. This algorithm is commonly used for hierarchical clustering and can generate tree-like clustering results, where each node represents a cluster and child nodes represent subclusters of that node.

[0073] The steps of the Euclidean clustering algorithm in this embodiment mainly include:

[0074] S21: Initialization, each sample point is regarded as an independent cluster;

[0075] S22: Calculate the distance between adjacent clusters, usually using single linkage or complete linkage methods;

[0076] S23: Merge the two closest clusters to form a new cluster;

[0077] S24: Repeat the above steps until all points are clustered into one category or the preset number of clusters is reached.

[0078] In all practical applications of RF radar, due to varying surrounding environments, especially when there are multiple strong reflective objects in the distance, the target will be reflected by multiple echoes due to the characteristics of electromagnetic wave scattering, and there will be more or less instability in the appearance of some valid targets. The point cloud data output by the radar sensor will appear as offset jitter when displayed in 3D, which can easily cause the accurate identification of valid targets to be filtered out as noise signals. Therefore, the effective identification and filtering of the offset portion of the valid intrusion target is as follows: For other targets within the preset offset range of the stable and valid intrusion target, the RCS value is determined to be equal to the RCS value of the stable and valid intrusion target. If equal, the two are considered to be the same intrusion target and recorded. However, only the valid intrusion target is retained in subsequent judgment and visualization. If the RCS values ​​are not equal, it is considered a new intrusion target and recorded.

[0079] When the RF radar itself monitors engineering vehicles at a long distance (a crane is used as an example in this embodiment), the radar's feedback signal attenuation and noise will cause insufficient point cloud data and insufficient accuracy, resulting in a weak signal and target offset and flickering, which can easily be filtered out as noise. At the same time, the boom is detected but it is difficult to identify whether it is the top of the crane boom. When the radar itself detects the top of the crane boom at a long distance, it cannot be guaranteed from the radar data that the position monitored is the top of the crane boom, which affects the accuracy of the subsequent judgment of the intrusion of the three-dimensional warning zone. Therefore, in some preferred embodiments, after the intrusion target is detected using the point cloud data collected by the RF radar, the position of the intrusion target is calibrated using a pan-tilt camera assembly, wherein the pan-tilt camera assembly is set adjacent to the RF radar, and the 0-degree direction of the camera in the pan-tilt camera assembly is parallel to the normal of the RF radar. The calibration process includes:

[0080] A1: The gimbal camera assembly is controlled to aim at the intrusion target based on the target's location information, and the camera focus is adjusted based on the distance between the intrusion target and the RF radar.

[0081] The coordinates of the intrusion target in the radar coordinate system can be expressed as (x', y', z'). Therefore, the distance between the intrusion target and the RF radar is sqrt(x'^2 + y'^2 + z'^2). Therefore, the horizontal deflection angle is atan(y' / x'), and the vertical deflection angle is atan(z' / x'). The gimbal camera assembly is aligned with the intrusion target based on the horizontal and vertical deflection angles. The gimbal camera focal length is adjusted based on the distance between the intrusion target and the RF radar, referring to a distance-focal length comparison table.

[0082] A2: Capture an image of the intruding target and track the top of the intruding target through image recognition detection. Control the gimbal camera assembly to position the top of the intruding target at the center of the image. Obtain the horizontal deflection angle α and vertical deflection angle β of the gimbal camera assembly at the current moment.

[0083] Specifically, a large number of various boom top images (including photos from different angles and backgrounds) are collected in the early stage, and the boom top is labeled to form a sample set. The sample set is then used to train a boom top detection model based on a convolutional neural network model. The trained boom top detection model is then used to perform target detection on the images taken by the gimbal camera. The detected boom top is initialized as the region of interest, and the KCF tracking algorithm is then used to track the current boom. The gimbal is continuously controlled to place the boom top in the center of the picture, and the horizontal deflection angle α and vertical deflection angle β of the gimbal camera component at the current moment are obtained.

[0084] A3: The x-axis coordinate of the intrusion target detected by the point cloud data collected by the RF radar is used to approximate the current x-axis coordinate of the top of the intrusion target. Combined with the horizontal deflection angle α and vertical deflection angle β of the gimbal camera assembly at the current moment, the three-dimensional spatial coordinate of the top of the intrusion target is calculated.

[0085] The horizontal and vertical deflection angles of the gimbal correspond to the angles of the boom tip. The X-axis coordinate of the highest Z-axis point of the intrusion target in the RF radar point cloud (the point closest to the boom tip detected by the RF radar) is used to approximate the current X-axis coordinate of the intrusion target tip. Coordinate conversion is performed to obtain the three-dimensional spatial coordinates of the intrusion target tip. The coordinate conversion method is as follows:

[0086] x=x';

[0087] y=x'*tan(α);

[0088] z=x'*tan(β);

[0089] Although the x-axis coordinates are approximated using the x-coordinates of the original point cloud, the x-axis accuracy remains unchanged. However, the increased accuracy of the horizontal and vertical deflection angles improves the positioning accuracy of the y- and z-axis coordinates. In practical applications, the level of the 3D intrusion warning zone is determined by the distance from the high-voltage power line, and the y- and z-axis coordinates are the most important factors determining the distance from the high-voltage power line. Therefore, as long as the positioning accuracy of the y- and z-axis coordinates is improved, the positioning accuracy required in practical engineering applications can be improved. It should be noted that the 3D spatial coordinates of the top of the intrusion target are still in the radar coordinate system and need to be converted to the 3D spatial coordinates in the ground coordinate system according to the aforementioned method.

[0090] In some preferred embodiments, the visible laser array assembly is mounted on the pan-tilt camera assembly, and its illumination direction is aligned with the camera's. This arrangement ensures that when the pan-tilt camera assembly is controlled to center the top of an intruding object in the image, the visible laser array assembly is also aligned with the top of the intruding object, eliminating the need for secondary illumination direction control based on the position of the intruding object's top.

[0091] A4: The three-dimensional spatial coordinates of the top of the intrusion target are used as the final three-dimensional spatial coordinates of the intrusion target.

[0092] S3: Match the level of the three-dimensional warning zone to which the intrusion target belongs based on the spatial coordinates of the intrusion target. The three-dimensional warning zone is set to multiple levels based on the distance from the high-voltage cable.

[0093] The number of three-dimensional warning zones is set according to actual needs. In this embodiment, the three-level three-dimensional warning zone is set as an example. The three-level three-dimensional warning zone is divided into planes such as Figure 3 As shown in the figure, after the independent single layer is split, the point space between planes (A, B, C, D) and planes (A1, B1, C1, D1) is defined as a low-level 3D warning zone; the point space between planes (A1, B1, C1, D1) and planes (A2, B2, C2, D2) is defined as a medium-level 3D warning zone; and the point space between planes (A2, B2, C2, D2) and planes (A3, B3, C3, D3) is defined as a high-level 3D warning zone. Among them, planes (A1, B1, C1, D1) and planes (A2, B2, C2, D2) are superimposed shared surface points. A multi-level 3D warning zone is set directly below the high-voltage power lines. The entire warning zone is divided into three 3D warning zones by multiple planes with the same horizontal xy coordinates but different vertical z-axis coordinates. The closer the 3D warning zone is to the high-voltage power line, the higher the warning level.

[0094] The rules for determining the level of intrusion into the three-dimensional warning zone are as follows:

[0095] (1)X1(A,B,A1,B1,A2,B2,A3,B3) <X<X2(C、D、C1、D1、C2、D2、C3、D3)

[0096] (2)Y1(B,C,B1,C1,B2,C2,B3,C3) <Y<Y2(A、D、A1、D1、A2、D2、A3、D3)

[0097] (3) Z1(A, B, C, D) <Z<Z2(A1、B1、C1、D1)

[0098] (4) Z2(A1, B1, C1, D1) <Z<Z3(A2、B2、C2、D2)

[0099] (5) Z3(A2, B2, C2, D2) <Z<Z4(A3、B3、C3、D3)

[0100] According to whether the three-dimensional coordinates (X, Y, Z) of the intrusion target meet the above conditions, if (1), (2), and (3) are met, the intrusion target belongs to the low-level three-dimensional warning zone; if (1), (2), and (4) are met, the intrusion target belongs to the medium-level three-dimensional warning zone; if (1), (2), and (5) are met, the intrusion target belongs to the high-level three-dimensional warning zone.

[0101] S4: According to the spatial coordinates of the intruding target and the level of the three-dimensional warning zone to which it belongs, the visible laser array component is controlled to track the intruding target and emit a laser warning light curtain of the corresponding level and issue a warning.

[0102] The parameters of the laser warning light curtain emitted by the visible laser array assembly include the irradiated target position, as well as one or more of the number of beams, laser color, laser spot size and laser strobe cycle. Number of beams: 1 beam, multi-beam array, 1+N beam mixed array. Laser color: single color, dual color, mixed color. Directional tracking of target object: start the pan-tilt motor to match the position and adjust the drive motor to point to the predetermined position point. Laser spot size: start the lens to match the position distance to adjust the size and thickness of the spot. The larger the spot is adjusted, the smaller the energy dispersion distance, and the smaller the spot is, the larger the energy focusing distance. Laser strobe cycle: adjust the strobe cycle according to the pulse width amplitude, oscillation frequency and timing cycle signal. 0 means no strobe, 1 means strobe, 2 means mixed strobe timing interval cycle, and 3 means combined pattern.

[0103] S5: According to the level of the three-dimensional warning zone to which the intrusion target belongs, the voice speaker component is controlled to issue a voice warning of the corresponding level or a customized voice warning.

[0104] In some preferred embodiments, before calibrating the position of the intrusion target using the pan-tilt camera, the method further includes:

[0105] S01: Match the detected intrusion target in the current frame with the intrusion target list; the intrusion target list records the location information and health value of each intrusion target and is initially empty;

[0106] S02: Calculate the distance between the intrusion target detected in the current frame and each intrusion target in the intrusion target list. If the distance between the two is less than the set distance threshold, they are considered to be the same intrusion target, and the location information of the intrusion target is updated, and its health value is restored to the initial value N. If the distance between the two is not less than the set distance threshold, it is considered to be a new intrusion target, and the intrusion target and its location information are added to the intrusion target list, and its health value is set to the initial value N.

[0107] S03: If there is an intrusion target in the intrusion target list that has not been matched to itself in the current frame, the health value of the intrusion target in the intrusion target list is reduced by 1; when the health value of an intrusion target in the intrusion target list reaches 0, it is deleted from the intrusion target list;

[0108] S04: After continuously collecting M frames and performing the above steps S01 to S03, the intrusion target and its position information finally retained are used as the intrusion target detection result; wherein M>N.

[0109] In rainy, snowy, foggy, and hazy weather conditions, many intrusion target point clouds appear in only one or two frames before disappearing. By setting a life value, if an intrusion target is detected in the current frame, its life value is restored to its initial value, N. If the current frame does not detect an intrusion target that has appeared previously, its life value is reduced by 1. If there is no intrusion target for N consecutive frames, the life value is zero, and the intrusion target is considered to be interference noise and deleted from subsequent calculations and detection processes. This system can adapt to complex environments such as rainy, snowy, foggy, and hazy weather, preventing false alarms, missed alarms, and non-alarms.

[0110] The method for warning the intrusion of engineering vehicles under high-voltage power lines proposed in the above embodiment detects the intrusion target and its spatial position in real time. When it is determined that the intrusion target has invaded the three-dimensional warning area, the visible laser array component is controlled to emit a laser warning light curtain of corresponding level and track the intrusion target for warning. Combined with voice warning, the warning effect is greatly improved; by adopting the laser warning light curtain combined with voice for graded warning, the problem of short warning distance and non-remote control at night, in foggy and rainy weather is solved; when spatially locating the intrusion target, the point cloud data collected by the radio frequency radar is used for preliminary positioning, and then the top of the intrusion target is precisely positioned in combination with the video image, which can solve the problem of poor positioning and intrusion judgment results caused by the loss of point cloud data on the top of the intrusion target under noise interference (especially large long-distance noise interference), thereby improving the accuracy of warning.

[0111] The embodiment of the present invention also provides a high-voltage line engineering vehicle intrusion warning system, such as Figure 4 As shown, it includes an MCU controller component module and a power supply and voltage stabilization module electrically connected thereto, a wireless communication component module, a pan-tilt motor linkage control drive component module, a TTS digital intelligent voice synthesis amplifier component module, a PWM-TTL laser array modulation drive output component module, a camera, and a radio frequency radar. It also includes a voice speaker component connected to the TTS digital intelligent voice synthesis amplifier component module, a visible laser array component connected to the PWM-TTL laser array modulation drive output component module, and a pan-tilt motor component connected to the pan-tilt motor linkage control drive component module; the camera is arranged on the pan-tilt motor component;

[0112] The MCU controller component module is configured to execute the above-mentioned method for warning of engineering vehicle intrusion under high-voltage lines.

[0113] Specifically, the power supply and voltage stabilization module is responsible for the entire system's power supply. It receives power from an external solar cell or DC power supply, and connects to an internal voltage stabilization circuit to output stable DC12V and DC5V to power the various internal circuit modules. The power supply also includes a communication interface that connects to the MCU controller module for communication, data collection, and control of the power supply.

[0114] Wireless communication component module: responsible for communicating with the external early warning control host wireless network, using software communication protocol can be customized to modify the processing matching connection communication, with custom configuration networking and self-organizing network communication functions, can adapt to the communication baud rate.

[0115] The PTZ motor linkage control drive module is connected via the RS485 bus serial port and is responsible for communication and output control with the MCU controller module. It has MOS, relay, and PWM pulse width modulation hardware output ports. Upon detecting the control logic protocol issued by the MCU controller module, it executes different drive signals based on the protocol data type to output the PTZ motor linkage action.

[0116] TTS digital intelligent speech synthesis and amplifier module: TTS stands for Text To Speech, meaning "text to speech." It is a key component of human-computer dialogue systems, enabling machines to speak. Drawing on both linguistics and psychology, it communicates with the MCU controller module via serial ports such as TTL / UART / IO / RS485. The MCU controller module receives commands from the external warning host, parses them, converts them into hexadecimal data according to the corresponding GB2312, GBK, BIG5, and UNICODE formats, and transmits them to the built-in AI chip for decoding and synthesis. Finally, the low-volume text is intelligently converted into a natural speech stream of analog signals. The built-in audio power amplifier drives the voice speaker module to broadcast or issue warnings.

[0117] PWM-TTL laser array modulation drive output component module: responsible for the modulation drive output module of the light-emitting laser source, which contains the PWM signal pulse frequency, TTL long and short distance spot size mixed demodulation part and the power drive part output interface. Through the PWM-TTL interface, the pulse width modulation PWM-TTL port of the MCU controller component module is connected. The MCU controller component module sends the laser control instructions (such as: application scenario, warning level, number of beams of light, output color, intrusion target location and directional tracking, spot size, stroboscopic cycle range, etc.) to this module, and the corresponding mixed demodulation power amplification is sent to the corresponding laser, and finally the visible light curtain corresponding to different scene functions is output (such as application scenarios: under high-voltage lines, substations, maintenance alerts, etc.). Warning level: Multi-level warnings are configurable. What actions are performed for low-level warnings, what actions are performed for medium-level warnings, and what actions are performed for high-level warnings. What action is performed for the early warning, or what action is performed for the 1-N level early warning. Number of beams of light: 1 beam, multi-beam array, 1+N beam mixed array. Output color: single color, dual color, mixed color. Target position and directional tracking: Start the pan / tilt motor to match the position and adjust the drive motor to point to the predetermined position. Spot size: Start the lens to match the position distance to adjust the size and thickness of the spot. The larger the spot, the smaller the energy dispersion distance, and the smaller the spot, the larger the energy focusing distance. Strobe cycle range: Adjust the strobe cycle according to the pulse width amplitude, oscillation frequency and timing cycle signal. 0 means no strobe, 1 means strobe, 2 means mixed strobe timing interval cycle, and 3 means combined pattern).

[0118] The external warning detector interface module is also included: it can be independently connected to external warning detectors for warning and alerting purposes. When used independently without an alarm host, it functions as an MCU controller module to expand and match control applications for connected external detectors. It supports TTL, UART, RS485 bus serial ports, and wired input IO hardware interfaces connected to the MCU controller module, which collects, analyzes, and processes data.

[0119] Voice speaker component: voice broadcast and shouting execution device.

[0120] Visible laser array assembly: Visible laser light-emitting actuator.

[0121] The pan / tilt motor assembly consists of: up / down, left / right position positioning, and patrol execution devices.

[0122] like Figure 5As shown, in this embodiment, the pan-tilt motor assembly includes a pan-tilt motor 1, a lower bracket expandable device mounting base 2 installed on the output end of the pan-tilt motor 1, a metal waterproof heat dissipation integrated shell 3 installed on the lower bracket expandable device mounting base 2, an upper bracket expandable device mounting base 4, a camera 5 installed on the upper bracket expandable device mounting base 4, and a visible laser array assembly 6 and a voice speaker assembly 7 are all arranged on the metal waterproof heat dissipation integrated shell 3.

[0123] The above embodiment provides a warning system for intrusion of engineering vehicles under high-voltage lines, which has the following beneficial effects:

[0124] 1) The problem of structural technical control function of integrating array laser beam, voice matrix warning device and omnidirectional pan-tilt motor technology control structure to achieve real-time modulation control is solved.

[0125] 2) It solves the problem of short warning distance and non-remote control at night, in foggy, rainy and foggy weather.

[0126] 3) Solved the problem that the light source cannot form a matrix to output light curtains with different warning colors according to different application scenarios.

[0127] 4) Solved the functional technical problems of automatically and remotely triggering and adjusting different light beams to form different shapes of width indicator light curtains according to application scenarios, positioning and tracking the target, pointing to the width warning, strobe frequency adjustment and automatic adjustment of distance brightness.

[0128] 5) It solves the problem of adding multiple sets of voice real-time acquisition and transmission driver devices for local broadcast and remote shouting broadcast of integrated voice warnings, and solves the problem of large audio traffic occupied and unstable and intermittent signals caused by transmission when simulating real-time voice remotely.

[0129] 6) The solution uses a remote and local digital integrated TTL voice device, enabling the sending, conversion, local storage, and analysis output of local and remote audio in real-time digital, small, and lightweight protocol commands. This allows for real-time configuration and output of different voices, intonations, speech rates, and mixed and synthesized audio. What you send is what you get. Automatically and remotely adjust and trigger different fixed voices or send different real-time, low-data, lightweight voice commands to broadcast warnings to targets.

[0130] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0131] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for warning of intrusion of engineering vehicles under high-voltage lines, characterized in that: The steps include: S1: Obtain point cloud data collected by RF radar under high-voltage power lines and perform coordinate conversion and filtering preprocessing; S2: Cluster the pre-processed point cloud data to detect intrusion targets; S3: Matches the level of the three-dimensional warning zone to which the intrusion target belongs based on its spatial location. The three-dimensional warning zone is set to multiple levels based on the distance from the high-voltage cable. S4: According to the spatial position of the intruding target and the level of the three-dimensional warning zone to which it belongs, the visible laser array component is controlled to track the intruding target and emit a laser warning light curtain of the corresponding level to warn; S5: According to the level of the three-dimensional warning zone to which the intrusion target belongs, the voice speaker component is controlled to issue a corresponding level of voice warning or a customized voice warning; In step S1, the point cloud filtering process includes: S11: Estimate the two-dimensional plane of the point cloud data through RANSAC, and use the points inside the point as ground points and other points as non-ground points to achieve the segmentation of ground points and non-ground points; S12: Filter out ground points and retain non-ground points; S13: Performs a limited range of straight-through filtering on the retained non-ground points (x, y, z, RCS). Only non-ground points within the multi-level 3D warning zone and with RCS values ​​within the set range are selected. Non-engineering vehicle point cloud data is filtered using a pre-set engineering vehicle RCS value range. The RCS value is the ratio of the target's return power density to the incident radar wave to the incident power density.

2. The method for warning of intrusion of engineering vehicles under high-voltage lines according to claim 1, characterized in that: After detecting intrusion targets using the point cloud data collected by the RF radar, the position of the intrusion targets is calibrated using the pan-tilt camera assembly. The pan-tilt camera assembly is set adjacent to the RF radar, and the 0-degree direction of the camera in the pan-tilt camera assembly is parallel to the normal of the RF radar. The calibration process includes: Control the pan / tilt camera assembly to aim at the intrusion target based on the location information of the intrusion target, and adjust the camera focus according to the distance between the intrusion target and the RF radar; Capture the image of the intrusion target and track the top of the intrusion target through image recognition detection, control the pan-tilt camera component to place the top of the intrusion target in the center of the picture, and obtain the horizontal deflection angle of the pan-tilt camera component at the current moment and vertical deflection angle ; The x-axis coordinate of the intrusion target detected by the point cloud data collected by the radio frequency radar is approximated to the current x-axis coordinate of the top of the intrusion target, and combined with the horizontal deflection angle of the pan-tilt camera assembly at the current moment and vertical deflection angle , calculate and obtain the three-dimensional spatial coordinates of the top of the intrusion target; The three-dimensional spatial coordinates of the top of the intrusion target are used as the final three-dimensional spatial coordinates of the intrusion target.

3. The method for warning of intrusion of engineering vehicles under high-voltage lines according to claim 2, characterized in that: The visible laser array component is arranged on the pan-tilt camera component, and the irradiation direction of the visible laser array component is the same as the camera direction.

4. The method for warning of intrusion of engineering vehicles under high-voltage lines according to claim 1, characterized in that: Parameters of the laser warning light curtain emitted by the visible laser array assembly include the irradiated target position, and one or more of the number of beams, laser color, laser spot size, and laser strobe period.

5. The method for warning of intrusion of engineering vehicles under high-voltage lines according to claim 1, characterized in that: After step S2, the following steps are also included: For other targets within the preset offset range of the intrusion target, determine whether their RCS values ​​are equal to the RCS value of the intrusion target. If they are equal, they are considered to be the same intrusion target and recorded; if the RCS values ​​are not equal, they are considered to be new intrusion targets and recorded.

6. The method for warning of intrusion of engineering vehicles under high-voltage lines according to claim 1, characterized in that: The steps between step S2 and S3 also include: S01: Match the detected intrusion target in the current frame with the intrusion target list; the intrusion target list records the location information and health value of each intrusion target and is initially empty; S02: Calculate the distance between the intrusion target detected in the current frame and each intrusion target in the intrusion target list. If the distance between the two is less than the set distance threshold, they are considered to be the same intrusion target, and the location information of the intrusion target is updated, and its health value is restored to the initial value N. If the distance between the two is not less than the set distance threshold, it is considered to be a new intrusion target, and the intrusion target and its location information are added to the intrusion target list, and its health value is set to the initial value N. S03: If there is an intrusion target in the intrusion target list that has not been matched to itself in the current frame, the health value of the intrusion target in the intrusion target list is reduced by 1; when the health value of an intrusion target in the intrusion target list reaches 0, it is deleted from the intrusion target list; S04: After continuously collecting M frames and performing the above steps S01 to S03, the intrusion target and its position information finally retained are used as the intrusion target detection result; wherein M>N.

7. The method for warning of intrusion of engineering vehicles under high-voltage power lines according to claim 1, characterized in that: The three-dimensional warning zone is set directly below the high-voltage line. The entire three-dimensional warning zone is divided into multi-level three-dimensional warning zones by multiple planes with the same horizontal xy plane coordinates and different vertical z-axis coordinates. The closer the three-dimensional warning zone is to the high-voltage line, the higher its warning level.

8. A high-voltage line engineering vehicle intrusion warning system, characterized in that: It includes an MCU controller component module and a power supply and voltage stabilization module electrically connected thereto, a wireless communication component module, a pan-tilt motor linkage control drive component module, a TTS digital intelligent voice synthesis amplifier component module, a PWM-TTL laser array modulation drive output component module, a camera, and a radio frequency radar. It also includes a voice speaker component connected to the TTS digital intelligent voice synthesis amplifier component module, a visible laser array component connected to the PWM-TTL laser array modulation drive output component module, and a pan-tilt motor component connected to the pan-tilt motor linkage control drive component module; the camera is arranged on the pan-tilt motor component; The MCU controller component module is configured to execute the method for warning of intrusion of engineering vehicles under high-voltage power lines as described in any one of claims 1 to 7.

9. The high-voltage line engineering vehicle intrusion warning system according to claim 8, characterized in that: When the MCU controller component module receives the custom voice warning instruction issued by the external warning host, it parses it, converts the format and outputs the hexadecimal data to the AI ​​chip built into the TTS digital intelligent voice synthesis amplifier component module for decoding. Finally, the low-data-volume text form obtained by decoding is intelligently converted into a natural voice stream of an analog signal, and the built-in audio power amplifier output of the TTS digital intelligent voice synthesis amplifier component module drives the voice speaker component to broadcast or shout warnings.

Images