阅读说明：本技术 一种基于实景三维模型的变电站电子安全围栏及监控方法 (Transformer substation electronic security fence based on live-action three-dimensional model and monitoring method ) 是由 褚国伟 何佳熹 陈浩 马剑勋 张关应 束云豪 刘泽 王艺钢 蒋君 于乔乔 于 2021-07-13 设计创作，主要内容包括：本发明提出一种基于实景三维模型的变电站电子安全围栏,其结构包括若干云台摄像机,工作区域,图像分析服务器,交换机；若干云台摄像机布置在工作区域周围,若干云台摄像机通过交换机与图像分析服务器连接。一种基于实景三维模型的变电站电子安全围栏监控方法,该方法包括如下步骤：1、建立变电站三维模型；2、建立通讯网络将云台摄像机的摄像机视频或图像传输到图像分析服务器；3、根据实际变电站工作需要在变电站三维模型中设置电子围栏空间位置,结合变电站三维模型确定安全工作区域,并调用对应云台摄像机进行实时图像采集和分析,通过位置计算识别工作区域内的人员和设备,对离开安全工作区域的行为进行告警。(The invention provides a substation electronic security fence based on a live-action three-dimensional model, which structurally comprises a plurality of pan-tilt cameras, a working area, an image analysis server and a switch; the plurality of pan-tilt cameras are arranged around the working area and are connected with the image analysis server through the switch. A transformer substation electronic security fence monitoring method based on a real-scene three-dimensional model comprises the following steps: 1. establishing a three-dimensional model of the transformer substation; 2. establishing a communication network to transmit the video or image of the pan-tilt camera to an image analysis server; 3. setting an electronic fence space position in a three-dimensional model of the transformer substation according to the actual working requirement of the transformer substation, determining a safe working area by combining the three-dimensional model of the transformer substation, calling a corresponding pan-tilt camera to perform real-time image acquisition and analysis, identifying personnel and equipment in the working area through position calculation, and giving an alarm for the behavior of leaving the safe working area.)

1. A transformer substation electronic security fence based on a live-action three-dimensional model is characterized by comprising a plurality of pan-tilt cameras (1), a working area (2), an image analysis server (3) and a switch (4); the plurality of pan-tilt cameras (1) are arranged around the working area (2), and the plurality of pan-tilt cameras (1) are connected with the image analysis server (3) through the switch (4).

2. The substation electronic safety fence based on the live-action three-dimensional model as claimed in claim 1, wherein the number of the pan-tilt cameras (1) is more than or equal to 2.

3. The method for monitoring the substation electronic security fence based on the live-action three-dimensional model as claimed in claim 1, characterized by comprising the following steps:

1. establishing a three-dimensional model of the transformer substation;

2. establishing a communication network to transmit the video or image of the pan-tilt camera to an image analysis server;

3. setting an electronic fence space position in a three-dimensional model of the transformer substation according to the actual working requirement of the transformer substation, determining a safe working area by combining the three-dimensional model of the transformer substation, calling a corresponding pan-tilt camera to perform real-time image acquisition and analysis, identifying personnel or equipment in the working area through position calculation, and giving an alarm for the behavior of leaving the safe working area.

4. The method for monitoring the electronic safety fence of the transformer substation based on the live-action three-dimensional model as claimed in claim 3, wherein the personnel and equipment in the working area are identified through position calculation, and the behavior leaving the safe working area is alarmed, and the method specifically comprises the following steps:

1. determining two-dimensional space coordinates of the pan-tilt camera; the two-dimensional space coordinate of each pan-tilt camera is expressed as A_i(x_i，y_i，z_i) I is 1, 2, 3 … … n, n is more than or equal to 2, n represents the number of pan-tilt cameras used;

2. determining a monitoring area in the transformer substation through a three-dimensional model of the transformer substation;

3. determining a safety region (omega) by the spatial position of the charged body of the substation₀₁)；

4. Manually setting a safety fence according to the actual situation of the field work, and defining a working area (omega)₀₂)；

5. Calculating a safety region (Ω)₀₁) And working area (Ω)₀₂) Is a safe working area (omega)₀)；

6. Acquiring images of a monitoring area once at regular intervals by a pan-tilt camera arranged around the working area to obtain acquired images;

7. identifying a worker or a working device in the captured image as an identification target person (p);

8. the method comprises the steps of adjusting the orientation of the pan-tilt camera, placing a recognition target person (p) in the center of a picture, and obtaining the azimuth angle of the recognition target person (p) relative to the pan-tilt camera through the orientation parameters of the adjacent m pan-tilt camerasm is less than or equal to n; wherein, theta_iRepresents the included angle between the connecting line from the ith pan-tilt camera to the identified target person (p) and the z axis,representing the included angle between the projection of the connecting line from the ith pan-tilt camera to the identified target person (p) on the xOy plane and the x axis;

9. on an xOy plane, calculating (x, y) coordinates of the identification target person (p) according to the two-dimensional space coordinates of the pan-tilt camera and the azimuth angle of the identification target person (p) relative to the spherical coordinate of the ith pan-tilt camera;

10. on an xOz plane, calculating a z coordinate of the identification target person (p) according to the two-dimensional space coordinate of the pan-tilt camera and the azimuth angle of the identification target person (p) relative to the spherical coordinate of the ith pan-tilt camera;

11. compare-computing identifies spatial coordinates (x, y, z) of a target person (p) and a safe working area (Ω)₀) If the spatial coordinates (x, y, z) are included in the safe working area (Ω)₀) If the system is safe, the system does not send an alarm; if the spatial coordinates (x, y, z) are not contained in (Ω)₀) If the person is dangerous, the system gives an alarm that the person leaves the safety area.

5. The substation electronic security fence monitoring method based on the live-action three-dimensional model as claimed in claim 4, wherein the (x, y) coordinates of the identification target person (p) are calculated on the xOy plane according to the two-dimensional space coordinates of the pan-tilt camera and the azimuth angle of the identification target person (p) relative to the spherical coordinate of the ith pan-tilt camera, and the method specifically comprises the following steps:

1) dividing every two pan-tilt cameras into a group, combining every two cameras to form C_m ²Group C_m ²The number of combinations for extracting 2 from m samples; the two-dimensional space coordinates of the two pan-tilt cameras in each group in the xOy plane are respectively expressed as (x)_i1，y_i1) And (x)_i2，y_i2) I1 and i2 are any numbers which are not equal in the value range of i, i is 1, 2 and 3 … … n, n is not less than 2, i is the number of the adjacent m camera numbers, n represents the total number of the used pan-tilt cameras, and m is the total number of the adjacent cameras;

2) on the xOy plane, according to the two-dimensional space coordinates (x) of the two pan-tilt cameras in each group_i1，y_i1)(x_i2，y_i2) And identifying the azimuth angle of the target person (p) with respect to the two pan-tilt cameras in each groupCalculating (x) for identifying the target person (p)_i1，i2，y_i1，i2) Coordinates i1 and i2 are any number unequal in the value range of i;

3) identifying (x) of the target person (p) found by different groups of pan-tilt cameras_i1，i2，y_i1，i2) And combining the coordinates pairwise to obtain the final (x, y) coordinates of the recognition target person (p).

6. The method for monitoring the substation electronic security fence based on the live-action three-dimensional model as claimed in claim 5, wherein the two-dimensional space coordinates (x) of the two pan-tilt cameras in each group are used as reference_i1，y_i1)(x_i2，y_i2) And identifying the azimuth angle of the target person (p) with respect to the two pan-tilt cameras in each groupCalculating (x) for identifying the target person (p)_i1，i2，y_i1，i2) The coordinates are as follows:

7. the method for monitoring the substation electronic security fence based on the live-action three-dimensional model as claimed in claim 5, wherein the final (x, y) coordinates of the identified target person (p) are obtained by combining all pan-tilt cameras two by two to calculate (x)_i1，i2，y_i1，i2) The coordinates are averaged, i.e.:

C_m ²to extract the number of combinations of 2 out of m samples.

8. The substation electronic security fence monitoring method based on the live-action three-dimensional model as claimed in claim 4, wherein the z-coordinate of the identification target person (p) is calculated on the xOz plane according to the two-dimensional space coordinate of the pan-tilt camera and the azimuth angle of the ball coordinate of the identification target person (p) relative to the ith pan-tilt camera, and the method specifically comprises the following steps:

1) dividing every two pan-tilt cameras into a group, combining every two cameras to form C_m ²Group C_m ²The number of combinations for extracting 2 from m samples; the two-dimensional space coordinates of the two pan-tilt cameras in each group in the xOz plane are respectively expressed as (x)_i1，z_i1) And (x)_i2，z_i2) I1 and i2 are any number unequal within the value range of iI is 1, 2, 3 … … n, n is more than or equal to 2, i is the number of adjacent m camera numbers, n represents the total number of the used pan-tilt cameras, and m is the total number of the adjacent cameras;

2) on the xOz plane, according to the two-dimensional space coordinates (x) of the two pan-tilt cameras in each group_i1，z_i1)(x_i2，z_i2) And identifying the azimuth angle of the target person (p) with respect to the two pan-tilt cameras in each groupCalculating Z for identifying target person (p)_i1，i2Coordinates i1 and i2 are any number unequal in the value range of i;

3) z for identifying the target person (p) obtained by different groups of pan-tilt cameras_i1，i2And combining the coordinates pairwise to obtain the final Z coordinate of the recognition target person (p).

9. The method as claimed in claim 8, wherein the method is based on two-dimensional space coordinates (x) of two pan-tilt cameras in each group_i1，z_i1)(x_i2，z_i2) And identifying the azimuth angle of the target person (p) with respect to the two pan-tilt cameras in each groupCalculating Z for identifying target person (p)_i1，i2The coordinates are as follows:

in the formula, alpha_iIs composed ofThe included angle between the projection of the determined azimuth angle on the xOz plane and the x axis is specifically as follows:

10. the method for monitoring the electronic security fence of the transformer substation based on the live-action three-dimensional model as claimed in claim 8, wherein the final Z coordinate for identifying the target person (p) is obtained by combining all the pan-tilt cameras two by two to calculate the Z coordinate_i1，i2The coordinates are averaged, i.e.:

Technical Field

The invention relates to a substation electronic security fence based on a live-action three-dimensional model and a monitoring method, and belongs to the technical field of substation security.

Background

When part of power failure work exists in the transformer substation, in order to prevent electric shock of maintainers, the transformer substation operation and maintenance personnel generally need to define an operation and maintenance work area for the maintainers and arrange fences around the operation and maintenance personnel, and for the condition that live wires exist above the operation and maintenance area, the operation and maintenance personnel also need to perform relevant explanation on the maintainers to prompt the maintainers to keep a safe distance with a live object; however, in the actual work, the condition that a maintainer leaves the defined range of the fence and cannot keep a safe distance with an upper electrified body still exists, so that the transformer operation and maintenance personnel improve the fence so as to alarm the maintainer leaving the safe working area; electronic safety fences with warning functions adopted by the existing transformer substation are in a laser type, a radar type and the like, and the electronic safety fences detect people entering or leaving the fence boundary through a sensing element and send out a person entering or leaving warning; however, the electronic security fence adopted by the existing transformer substation cannot remotely monitor the actual situation of the site in real time; the laser or radar device needs to be moved by personnel, and the fence is heavy and is difficult to set and move.

At present, a plurality of transformer substations establish a real-scene three-dimensional model of the transformer substation through digital twin transformation, and if the real-scene three-dimensional model of the transformer substation can be used for defining an area in the model and monitoring the safety condition of workers in the actual transformer substation, the transformer substation has the advantages of real-time monitoring of the condition in the substation, convenience in setting and timely warning, so that the transformer substation is very important; therefore, the invention of the safety fence based on the real-scene three-dimensional model of the transformer substation has important significance for establishing a novel safety supervision means and setting a novel safety measure to improve the safety operation level of the transformer substation under the background of the transformation of the digital transformer substation.

Disclosure of Invention

The invention provides a substation electronic safety fence based on a real-scene three-dimensional model and a monitoring method, and aims to solve the problem that the actual condition of a substation field cannot be monitored remotely in real time by the existing substation electronic safety fence.

The technical solution of the invention is as follows: a transformer substation electronic security fence based on a live-action three-dimensional model structurally comprises a plurality of pan-tilt cameras 1, a working area 2, an image analysis server 3 and a switch 4; a plurality of pan-tilt cameras 1 are arranged around a working area 2, and the pan-tilt cameras 1 are connected with an image analysis server 3 through a switch 4.

A transformer substation electronic security fence monitoring method based on a real-scene three-dimensional model comprises the following steps:

1. establishing a three-dimensional model of the transformer substation;

2. establishing a communication network to transmit a camera video of the pan-tilt camera to an image analysis server;

3. setting an electronic fence space position in a three-dimensional model of the transformer substation according to the actual working requirement of the transformer substation, determining a safe working area by combining the three-dimensional model of the transformer substation, calling a corresponding pan-tilt camera to perform real-time image acquisition and analysis, identifying personnel and equipment in the working area through position calculation, and giving an alarm for the behavior of leaving the safe working area.

The invention has the advantages that:

1) the invention can monitor the actual working condition on site in real time and give an alarm in real time when entering a dangerous area;

2) can conveniently set up the security fence that satisfies different application scenarios, reduce and set up security fence work load.

Drawings

Fig. 1 is a schematic view of the installation position of the pan/tilt camera 1.

Fig. 2 is a schematic diagram of a connection structure of a substation electronic security fence based on a real-scene three-dimensional model.

Fig. 3 is a schematic workflow diagram of a substation electronic security fence monitoring method based on a real-scene three-dimensional model.

Fig. 4 is a schematic view of a safe area space.

Fig. 5 is a schematic view of identifying the orientation of a target person relative to a pan/tilt camera.

Fig. 6 is a schematic plan view of the recognition target person position xOy.

Fig. 7 is a schematic plan view of the recognition target person position xOz.

In the figure, 1 is a pan-tilt camera, 2 is a working area, 3 is an image analysis server, 4 is a switch, A₁、A₂Respectively two-dimensional space coordinates of two pan-tilt cameras in the embodiment, L is a live wire, and Q is₀₁Is a safe region, Ω₀₂Is the working area, Ω₀Is a safe working area, and p is an identification target person.

Detailed Description

A transformer substation electronic security fence based on a live-action three-dimensional model structurally comprises a plurality of pan-tilt cameras 1, a working area 2, an image analysis server 3 and a switch 4; a plurality of pan-tilt cameras 1 are arranged around a working area 2, and the pan-tilt cameras 1 are connected with an image analysis server 3 through a switch 4.

The number of the pan-tilt cameras 1 is more than or equal to 2.

A substation electronic safety fence based on a live-action three-dimensional model is suitable for remotely monitoring the site of the substation electronic safety fence in real time.

A transformer substation electronic security fence monitoring method based on a real-scene three-dimensional model comprises the following steps:

1. establishing a three-dimensional model of the transformer substation;

2. establishing a communication network to transmit the video or image of the pan-tilt camera to an image analysis server;

3. setting an electronic fence space position in a three-dimensional model of the transformer substation according to the actual working requirement of the transformer substation, determining a safe working area by combining the three-dimensional model of the transformer substation, calling a corresponding pan-tilt camera to perform real-time image acquisition and analysis, identifying personnel or equipment in the working area through position calculation, and giving an alarm for the behavior of leaving the safe working area.

The method for identifying the personnel and the equipment in the working area through position calculation and giving an alarm for the behavior of leaving the safe working area specifically comprises the following steps:

1. determining two-dimensional space coordinates of the pan-tilt camera; the two-dimensional space coordinate of each pan-tilt camera is expressed as A_f(x_f，y_f，z_f) I is 1, 2, 3 … … n, n is more than or equal to 2, n represents the number of pan-tilt cameras used;

2. determining a monitoring area in the transformer substation through a three-dimensional model of the transformer substation;

3. determining a safety region omega through the space position of a charged body of the transformer substation₀₁；

4. Root of herbaceous plantAccording to the actual situation of the field work, manually setting a safety fence to define the working area omega₀₂；

5. Calculating a safety region omega₀₁And working area omega₀₂The common part of (2) is a safe working area omega₀；

6. Acquiring images of a monitoring area once at regular intervals by a pan-tilt camera arranged around the working area to obtain acquired images;

7. identifying workers or working equipment in the collected image as an identification target person p;

8. the method comprises the steps of adjusting the orientation of the pan-tilt camera, placing a recognition target person p in the center of a picture, and obtaining the azimuth angle of the recognition target person p relative to the pan-tilt camera through the orientation parameters of the adjacent m pan-tilt camerasm is less than or equal to n; wherein, theta_iRepresenting the included angle between the connecting line from the ith pan-tilt camera to the identified target person p and the z axis,representing the included angle between the projection of the connecting line from the ith pan-tilt camera to the identified target person p on the xOy plane and the x axis;

9. on an xOy plane, calculating (x, y) coordinates of the recognition target person p according to the two-dimensional space coordinates of the pan-tilt camera and the azimuth angle of the recognition target person p relative to the spherical coordinate of the ith pan-tilt camera;

10. on an xOz plane, calculating a z coordinate of the identification target person p according to the two-dimensional space coordinate of the pan-tilt camera and the azimuth angle of the spherical coordinate of the identification target person p relative to the ith pan-tilt camera;

11. comparing, calculating and identifying space coordinates (x, y, z) of target person p and safe working area omega₀If the spatial coordinates (x, y, z) are included in the safe working area Ω₀If the system is safe, the system does not send an alarm; if the spatial coordinates (x, y, z) are not contained in Ω₀If the person is dangerous, the system gives an alarm that the person leaves the safety area.

On the xOy plane, calculating the (x, y) coordinate of the identified target person p according to the two-dimensional space coordinate of the pan-tilt camera and the azimuth angle of the identified target person p relative to the spherical coordinate of the ith pan-tilt camera, and specifically comprising the following steps:

1) dividing every two pan-tilt cameras into a group, combining every two cameras to form C_m ²Group C_m ²The number of combinations for extracting 2 from m samples; the two-dimensional space coordinates of the two pan-tilt cameras in each group in the xOy plane are respectively expressed as (x)_i1，y_i1) And (x)_i2，y_i2) I1 and i2 are any numbers which are not equal in the value range of i, i is 1, 2 and 3 … … n, n is not less than 2, i is the number of the adjacent m camera numbers, n represents the total number of the used pan-tilt cameras, and m is the total number of the adjacent cameras;

2) on the xOy plane, according to the two-dimensional space coordinates (x) of the two pan-tilt cameras in each group_i1，y_i1)(x_i2，y_i2) And identifying the azimuth angle of the target person p relative to the two pan-tilt cameras in each groupCalculating (x) of the recognition target person p_i1，i2，y_i1，i2) Coordinates i1 and i2 are any number unequal in the value range of i;

3) (x) of the recognition target person p to be found by the pan/tilt cameras of different groups_i1，i2，y_i1，i2) And combining the coordinates pairwise to obtain the final (x, y) coordinates of the recognition target person p.

The two-dimensional space coordinate (x) passing through two pan-tilt cameras in each group_i1，y_i1)(x_i2，y_i2) And identifying the azimuth angle of the target person p relative to the two pan-tilt cameras in each groupCalculating (x) of the recognition target person p_i1，i2，y_i1，i2) The coordinates are as follows:

the final (x, y) coordinates of the identified target person p are obtained by combining all the pan-tilt cameras in pairs_i1，i2，y_i1，i2) The coordinates are averaged, i.e.:

C_m ²to extract the number of combinations of 2 out of m samples.

On the xOz plane, calculating the z coordinate of the identified target person p according to the two-dimensional space coordinate of the pan-tilt camera and the azimuth angle of the identified target person p relative to the spherical coordinate of the ith pan-tilt camera, and specifically comprises the following steps:

1) dividing every two pan-tilt cameras into a group, combining every two cameras to form C_m ²Group C_m ²The number of combinations for extracting 2 from m samples; the two-dimensional space coordinates of the two pan-tilt cameras in each group in the xOz plane are respectively expressed as (x)_i1，z_i1) And (x)_i2，z_i2) I1 and i2 are any numbers which are not equal in the value range of i, i is 1, 2 and 3 … … n, n is not less than 2, i is the number of the adjacent m camera numbers, n represents the total number of the used pan-tilt cameras, and m is the total number of the adjacent cameras;

2) on the xOz plane, according to the two-dimensional space coordinates (x) of the two pan-tilt cameras in each group_i1，z_i1)(x_i2，z_i2) And identifying the target person p with respect to the two pan-tilt cameras in each groupAzimuth angleCalculating z for identifying the target person p_i1，i2Coordinates i1 and i2 are any number unequal in the value range of i;

3) z of the identified target person p obtained by different groups of pan-tilt cameras_i1，i2And combining the coordinates pairwise to obtain the final z coordinate of the recognition target person p.

The two-dimensional space coordinate (x) on the xOz plane passing through the two pan-tilt cameras in each group_i1，z_i1)(x_i2，z_i2) And identifying the azimuth angle of the target person p relative to the two pan-tilt cameras in each groupCalculating z for identifying the target person p_i1，i2The coordinates are as follows:

in the formula, alpha_iIs composed ofThe included angle between the projection of the determined azimuth angle on the xOz plane and the x axis is specifically as follows:

the step of solving the final z coordinate of the recognition target person p is concretely to calculate the z coordinate by combining all the cameras two by two_i1，i2The coordinates are averaged, i.e.:

the coordinates of the recognition target person p are determined to be (x, y, z) according to the equations (1), (2) and (3).

When the system works, the process can be repeated at intervals to realize the implementation monitoring of the operation of the transformer substation.

Example 1

A transformer substation electronic security fence monitoring method based on a real-scene three-dimensional model comprises the following steps:

1. scanning and modeling the transformer substation through a three-dimensional modeling technology, and determining a spatial coordinate of a pan-tilt camera; the spatial coordinates of each pan-tilt camera are denoted as A_iI is 1, 2, 3 … … n, n is more than or equal to 2; in this embodiment, two pan-tilt cameras are preferably used, for example, the two-dimensional space coordinates of the two pan-tilt cameras in fig. 4 to fig. 7 are respectively represented as: a. the₁，A₂；

2. Determining a safety region omega through the space position of a charged body of the transformer substation₀₁(ii) a Referring to FIG. 4, the live body of the substation is a live wire L, the safety distance is determined according to the substation safety rule, a certain distance around the live wire L is a dangerous area, and the area outside the dangerous area is a safety area omega₀₁；

3. Manually setting a safety fence according to the actual situation of the field work, and defining a working area omega₀₂；

4. Calculating a safety region omega₀₁And working area omega₀₂The common part of (2) is a safe working area omega₀；

5. Acquiring images of a monitoring area once at regular intervals by two pan-tilt cameras arranged around a working area to obtain acquired images;

6. identifying workers or working equipment in the collected image as an identification target person p;

7. the orientation of the pan-tilt camera is adjusted, the identified target person p is arranged in the center of the picture, and the azimuth angle of the identified target person p relative to the two pan-tilt cameras is obtained through the orientation parameters of the pan-tilt cameras

8. On the xOy plane, according to the two-dimensional space coordinate A of the pan-tilt camera₁，A₂And azimuth angleThe (x, y) coordinates of the recognition target person p can be calculated;

9. according to the azimuth angle of two opposite pan-tilt cameras of a person on the xOz plane The included angle alpha between the projection of the recognition target person p on the xOz plane and the x axis can be calculated₁，α₂(ii) a According to the two-dimensional space coordinate A of the pan-tilt camera₁，A₂And identifying the azimuth angle alpha of the target person p after projection in the xOz plane₁，α₂Calculating the (x, z) coordinates of the recognition target, and then determining the spatial coordinates (x, y, z) of the recognition target person p;

10. the space coordinates (x, y, z) of the target person p and the safe working area omega are compared and calculated₀If (x, y, z) is contained in Ω₀If the system is safe, the system does not send an alarm; if (x, y, z) is not contained in Ω₀If the person is dangerous, the system gives an alarm that the person leaves the safe area;

11. the process is repeated at intervals to realize the monitoring of the operation of the transformer substation.