A method for detecting invasion hotspot paths of invasive animals

Abstract

The present invention relates to the technical field of small animal path detection in substations, and more specifically, to a method for detecting hot spots of invading animals. A radar detection device is used to detect the path of an invading animal trajectory area, accumulate historical trajectory data of the invading animals, and convert it into a semantic trajectory, which is represented by an undirected weighted graph. Through weighted degree sorting analysis, the key nodes of the invading animals invading the substation are obtained, and the hot spots of the invading animals invading the substation are further obtained, which is conducive to realizing the high-efficiency, low-cost and targeted expulsion of invading animals from substations.

Description

Technical Field

[0001] The present invention relates to the technical field of small animal path detection in substations, and more specifically, to a method for detecting intrusion hotspot paths of intruding animals. Background Art

[0002] At present, with the increasing number of substations, the operation and maintenance of substations is very important. Especially with the change of seasons, various small animals frequently enter and exit substations. Small animals enter the substation through holes, and the insulation layer of wires and cables is easily damaged, causing the insulation effect to decrease. In severe cases, small animal short circuit accidents may occur, posing a great safety hazard to the safe and stable operation of the substation. In order to avoid small animal invasion incidents in the daily operation of substations, the prevention of small animals in substations is becoming more and more important.

[0003] The existing substation anti-small animal inspection work is mostly carried out through manual patrol measures, which periodically patrol the interior of the substation and place small animal repellent devices in places where obvious signs of small animal activities are found, so as to achieve the effect of repelling small animals, which consumes a lot of manpower and time. The unmanned working mode of substations is gradually becoming popular, and the number of patrol personnel has decreased. It mainly relies on the deployment of a large number of active small animal repellent devices for comprehensive coverage. However, the large-scale deployment of repellent devices is not conducive to cost savings.

[0004] The prior art discloses a DBSCAN life detection radar point trace condensation method and system based on KD-tree space search, including: obtaining the original echo image of the life detection radar and creating a KD-tree using the target pixel points of the original echo image; using the KD-tree to search for a subsample set within the Eps neighborhood radius of any point p in the tree; obtaining a cluster of point p according to the subsample set; repeating S2 to S3, traversing the unjudged points in the original echo image until all points are judged, and obtaining a clustered image of the original echo image; performing point trace condensation processing on the clustered image to obtain a point trace condensation result. Although this scheme can be used for small animal path detection in substations, it is impossible to obtain the hotspot path of small animal invasion, and it is necessary to drive away small animals in a large range of substations, which is inefficient and costly. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a method for detecting hot spot paths of invading animals invading substations. The method can obtain the hot spot paths of invading animals invading substations. When used for detecting hot spot paths of small animals invading substations, the method is conducive to achieving efficient, low-cost and targeted expulsion of small animals in substations.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A method for detecting invasion hotspot paths of invasive animals is provided, comprising the following steps:

[0008] S1: Install several radar detection devices based on the topological structure;

[0009] S2: pre-processing the tracks of the invading animals detected by the radar detection device, adding detection of the stopping points of the invading animals, and converting the tracks of the invading animals into semantic tracks;

[0010] S3: According to the semantic trajectory data obtained in step S2, an undirected weighted graph is used to represent it, and an adjacency matrix is ​​further created based on the undirected weighted graph, and then all adjacency matrices are accumulated to obtain a statistical adjacency matrix;

[0011] S4: Mining the key nodes of the statistical adjacency matrix based on weighted degree sorting, and then connecting the key nodes to each other to form invasion hotspot paths of invasive animals.

[0012] The invading animal invasion hotspot path detection method of the present invention is used for detecting the invading animal invasion hotspot path of a substation. The invading animal trajectory area path is detected by a radar detection device, and the invading animal historical trajectory data is accumulated and converted into semantic trajectories, which are represented by an undirected weighted graph. Through weighted degree sorting analysis, the key nodes of the invading animal invading the substation are obtained, and the hotspot path of the invading animal invading the substation is further obtained. The method can support targeted protection of key areas of invading animal invasion in the substation and effectively combat and drive away the invading animals, which is conducive to realizing efficient, low-cost and targeted driving away of invading animals in the substation.

[0013] Preferably, in step S1, the radar detection device includes a processing module, a power supply module, a detection module and a wireless module, the processing module, the detection module and the wireless module are respectively connected to the power supply module, and the detection module and the processing module are respectively connected to the wireless module, wherein the detection module is a Doppler radar, the processing module is a single-chip microcomputer, the wireless module is WIFI or zigBee, and the power supply module is a DC power supply.

[0014] Preferably, the detection area of ​​the Doppler radar is fan-shaped, and is laid out using a topological structure according to the environment. Each radar detection device is numbered, and each numbered radar detection device corresponds to an actual intrusion area.

[0015] Preferably, in step S2, the radar detection device detects invading animals in real time, and the single-chip computer sends location record information to the host computer through the wireless module at regular intervals. The location record information includes the invading animal number, the invading area number, the current time and the target radial relative speed.

[0016] Preferably, in step S2, the radial relative velocity of the invading animal relative to the radar detection device is obtained by Doppler radar measurement, and the radial relative velocity of the target is calculated based on the Doppler frequency shift:

[0017]

[0018] Where F is the original frequency of the signal; Fs is the difference between the received signal frequency and the transmitted signal frequency; c is the speed of light; and v is the radial relative velocity of the invading animal to the radar detection device.

[0019] Preferably, in step S2, the trajectory of the invading animal is divided into staying points and moving points, the moving point area where the invading animal is detected is used as the moving point, and the staying point area where the invading animal is detected is used as the staying point, and the moving points and the staying points are arranged in chronological order to obtain the semantic trajectory.

[0020] Preferably, in step S2, the step of stopping point detection is:

[0021] S21: pre-processing the trajectory location information of the invading animals in each area and setting a time threshold;

[0022] S22: Detect whether the intruding animal stays in the corresponding area for a time greater than a time threshold. If so, proceed to step S23. If not, mark the corresponding area as a moving point area, and end the detection of the corresponding intruding animal in the corresponding area.

[0023] S23: Setting speed threshold;

[0024] S24: Determine whether the average value of the sum of all speed features of the corresponding area trajectory is greater than the speed threshold. If so, mark the corresponding area as a moving point area. If not, mark the corresponding area as a staying point area, and end the detection of the corresponding invading animal in the corresponding area.

[0025] Preferably, in step S3, the moving points and the staying points in the undirected weighted graph are collectively referred to as nodes, and an adjacency matrix aFc[i][j] is created:

[0026]

[0027] In the formula, w ij Represents the weight between the i-th node and the j-th node;

[0028] The initial weight w between nodes ij is 0, and in an undirected graph w ij =w ji ,The weight depends on the number of repetitions of the invasive animal trajectory path, and the number of repetitions of the semantic trajectory path is equal to the weight between nodes.

[0029] Preferably, in step S4, the steps of forming the invasion hotspot path of invading animals are:

[0030] S41: Traverse the nodes of the statistical adjacency matrix and calculate the weighted degree of each node as d i , and calculate the average weighted degree d of the statistical adjacency matrix;

[0031] S42: Determine whether d i >d, if yes, mark the i-th node as a key node;

[0032] S43: sorting the key nodes in step S42 to obtain a node importance sorting table;

[0033] S44: Based on the node importance ranking table, key nodes are connected to each other on an undirected graph to form an invasion hotspot path of invasive animals.

[0034] Preferably, the calculation formula for counting the weighted degree of each node in the adjacency matrix and the average weighted degree of the adjacency matrix is:

[0035]

[0036]

[0037] In the formula, w ij represents the weight between the i-th node and the j-th node; d i represents the weighted degree of the i-th node; d represents the average weighted degree; n represents the number of nodes in the statistical adjacency matrix.

[0038] Compared with the background technology, the invading animal invasion hotspot path detection method of the present invention has the following beneficial effects:

[0039] The radar detection device is used to detect the path of the invading animal trajectory area, and the historical trajectory data of the invading animals is accumulated and represented by an undirected weighted graph. The key nodes are obtained through weighted degree sorting analysis, and then the key nodes are connected to each other to obtain the hotspot path of the invading animals invading the substation. When used for the detection of the hotspot path of invading animals in the substation, it can support the targeted protection of the key areas of the invading animals in the substation and effectively combat and drive away the invading animals, which is conducive to the efficient, low-cost and targeted expulsion of invading animals from the substation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a flow chart of a method for detecting hotspot paths of invading animals in an embodiment of the present invention;

[0041] Figure 2 A schematic diagram of the layout of a radar detection device in an embodiment of the present invention;

[0042] Figure 3This is a flow chart of the stay point detection in an embodiment of the present invention;

[0043] Figure 4 A schematic diagram of converting a semantic trajectory into an undirected weighted graph in an embodiment of the present invention;

[0044] Figure 5 The figure is a flow chart of forming an intrusion hotspot path in an embodiment of the present invention.

[0045] In the accompanying drawings: 1- radar detection device. DETAILED DESCRIPTION

[0046] The present invention will be further described below in conjunction with specific implementation modes.

[0047] Embodiment 1

[0048] A method for detecting hotspot paths of invasive animals, such as Figure 1 As shown, the following steps are included: S1: installing a plurality of radar detection devices 1 based on the topological structure;

[0049] S2: pre-processing the track of the invading animal detected by the radar detection device 1, adding the detection of the stopping point of the invading animal, and converting the track of the invading animal into a semantic track;

[0050] S3: According to the semantic trajectory obtained in step S2, an undirected weighted graph is used to represent it, and an adjacency matrix is ​​further created based on the undirected weighted graph. Then, all adjacency matrices are accumulated to obtain a statistical adjacency matrix.

[0051] S4: Mining the key nodes of the statistical adjacency matrix based on weighted degree sorting, and then connecting the key nodes to each other to form invasion hotspot paths of invasive animals.

[0052] The invading animal intrusion hotspot path detection method of the present embodiment is used for detecting the intrusion hotspot path of animals in a substation. The animal trajectory area path is detected by a radar detection device 1, and the animal historical trajectory data is accumulated and converted into a semantic trajectory, which is represented by an undirected weighted graph. Through weighted degree sorting analysis, the key nodes of animal invasion of the substation are obtained, and the hotspot paths of animal invasion of the substation are further obtained. This can support targeted protection of key areas of animal invasion in substations and effective crackdown and expulsion of animals, which is conducive to achieving efficient, low-cost and targeted expulsion of animals from substations.

[0053] In step S1, the radar detection device 1 includes a processing module, a power supply module, a detection module and a wireless module. The processing module, the detection module and the wireless module are respectively connected to the power supply module, and the detection module and the processing module are respectively connected to the wireless module. Among them, the detection module is a Doppler radar, the processing module is a single-chip microcomputer, the wireless module is WIFI, and the power supply module is a DC power supply.

[0054] The detection range of the Doppler radar is 2m to 10m, and the detection area is fan-shaped. The topological structure is used for layout according to the environment, and each radar detection device 1 is numbered. Each numbered radar detection device 1 corresponds to an actual intrusion area.

[0055] In step S2, the radar detection device 1 detects invading animals in real time, and the single chip sends location record information to the host computer through the wireless module at regular intervals. The location record information includes the invading animal number, the invading area number, the current time and the radial relative speed of the invading animal.

[0056] Embodiment 2

[0057] This embodiment is similar to the first embodiment, except that this embodiment is a method for detecting hot spots of intrusion of intruding animals in a rectangular substation environment with an area of ​​10m×20m. It is necessary to install a plurality of radar detection devices 1, and the radar detection devices 1 are numbered respectively. The detection module in the radar detection device 1 adopts HB100 Doppler radar. Since the detection range of the Doppler radar is 2m to 10m, the detection area is fan-shaped, and it is necessary to be based on the following example: Figure 2 The topological structure shown is equipped with 8 radar detection devices 1, and the detection distance of the Doppler radar is adjusted to 5m, and the detection angle is 90°. The 8 radar detection devices 1 correspond to actual areas ①~⑧ respectively. The processing module uses STM32F103C8T6 microcontroller, the wireless module uses ZigBee, and the power supply module uses a 10V DC power supply.

[0058] The signal detected by the radar detection device 1 is preprocessed. The radar detection device 1 sends the location record information of the invading animal to the host computer for processing at an interval of about 5 minutes. The location record information includes the invading animal number, the invading area number, the current time and the radial relative speed of the invading animal. The radial relative speed of the invading animal is obtained by Doppler radar measurement, and the radial relative speed of the target is calculated based on the Doppler frequency shift:

[0059]

[0060] Where F is the original frequency of the signal; F S is the difference between the receiving signal frequency and the sending signal frequency; c is the speed of light; v is the radial relative velocity of the invading animal to the radar detection device.

[0061] Embodiment 3

[0062] This embodiment is similar to the second embodiment, except that the radar detection device 1 is used for detection to obtain location record information of several invading animals, wherein the location record information of the invading animal numbered 1 is shown in Table 1:

[0063] Table 1 Location record information of invasive animals numbered 1

[0064] Invasive Animal Number Invasion Area Number Current time Radial relative velocity of invasive animals 1 ① 2021 / 6 / 8 22:30:18 0.2m / s 1 ② 2021 / 6 / 8 22:35:08 0.5m / s 1 ② 2021 / 6 / 8 22:40:20 0.6m / s 1 ④ 2021 / 6 / 8 22:45:25 0.8m / s 1 ④ 2021 / 6 / 8 22:50:20 0.1m / s 1 ④ 2021 / 6 / 8 22:56:50 0.08m / s 1 ④ 2021 / 6 / 8 23:01:18 0.03m / s 1 ④ 2021 / 6 / 8 23:06:20 0.01m / s

[0065] like Figure 3 As shown, the trajectory location information of the invading animals in each area is preprocessed to remove abnormal data and individual stop points. The abnormal data includes trajectory data with too fast speed and too short stay time, and a time threshold is set;

[0066] The residence time of the invasive animals in each area is further obtained: the time when the invasive animals first appear in an area is recorded as t1, and the time when the corresponding invasive animal signal disappears in the same area and appears in other areas is recorded as t2, then the residence time of the invasive animals in the corresponding area is (t2-t1);

[0067] Detect whether the invading animal stays in each area for longer than the time threshold t = 20min. If not, mark the area as a moving point area. If so, set the speed threshold v m =0.1m / s, to determine whether the average speed of the trajectory points in the area is greater than the set speed threshold v m , if yes, mark the area as a moving point area, if no, mark the area as a staying point area;

[0068] The non-stop point areas where the invading animals are detected are used as the moving points of the trajectory, and the stop point areas where the invading animals are detected are used as the stop points of the trajectory. These moving points and stop points are arranged in chronological order to obtain the semantic trajectory. According to Table 1, the trajectory of the invading animal numbered 1 is: ①→②→④, where ① and ② are moving points and ④ is a stop point. Similarly, the semantic trajectory records of some invading animals can be obtained, as shown in Table 2:

[0069] Table 2 Semantic track records of some intrusion objects

[0070] Invasive Animal Number Semantic Trajectory 1 ①→②→④ 2 ①→③→④ 3 ②→①→③ 4 ①→②→④→② 5 ②→④ 6 ②→①→③ 7 ①→③→④ 8 ③→④→⑤

[0071] Embodiment 4

[0072] This embodiment is similar to the third embodiment, except that the radar detection device 1 is used to detect invading animals, and semantic trajectories for one month are collected, and a total of 50 semantic trajectory data are obtained. The semantic trajectory data are clustered. Specifically, each semantic trajectory is first converted into an undirected weighted graph structure. Specifically, the moving points and the staying points of the semantics are represented by the nodes of the graph, and the trajectory direction is removed and further converted into an undirected graph structure. The weights between the nodes represent the number of round trips of the invading animals between the staying points and the moving points. One semantic trajectory is ①→②→③→④→②→①, and its conversion diagram is shown as follows: Figure 4As shown in Figure 2, after converting all semantic trajectories into an undirected weighted graph structure, all semantic trajectories are represented by an adjacency matrix, and the adjacency matrix arc[i][j] is:

[0073]

[0074] In the formula, w ij Represents the weight between the i-th node and the j-th node, and the initial weight w between the nodes ij is 0, and in an undirected graph w ij =w ji , the weight is equal to the number of trajectory path repetitions, and then the adjacency matrix is ​​accumulated to obtain the statistical adjacency matrix M:

[0075]

[0076] The statistical adjacency matrix has 8 nodes, and the weighted degree of each node is d i :

[0077]

[0078] The eight nodes are marked with ① to ⑧ respectively, and the node weighted degree table is obtained, as shown in Table 3.

[0079] Table 3 Node weighted degree table

[0080] node ① ② ③ ④ ⑤ ⑥ ⑦ ⑧ Weighted 38 10 56 48 25 6 3 4

[0081] The average weighted degree of the statistical adjacency matrix is ​​d:

[0082]

[0083] like Figure 5 As shown in Table 3, nodes ①, ③, ④ and ⑤ are key nodes. The larger the weighted degree of the node, the greater the importance. The node importance ranking results are ③④①⑤. The key nodes are connected to each other to obtain one or more hot node paths, that is, the hot path of invading animals. One of the hot paths is ①→③→④→⑤. According to the hot path, the staff can carry out targeted protection of the key areas of invading animals in the substation and effectively combat and drive away the invading animals. When the invading animals are animals, it is conducive to the efficient, low-cost and targeted expulsion of substation animals.

[0084] In the specific contents of the above-mentioned specific implementation methods, the various technical features can be combined in any non-contradictory manner. In order to make the description concise, not all possible combinations of the above-mentioned technical features are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0085] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for detecting hotspot paths of invading animals, characterized in that: The following steps are involved: S1: installing a number of radar detection devices (1) based on the topological structure; S2: pre-processing the track of the invading animal detected by the radar detection device (1), adding the detection of the stopping point of the invading animal, and converting the track of the invading animal into a semantic track; S3: According to the semantic trajectory obtained in step S2, an undirected weighted graph is used to represent it, and an adjacency matrix is ​​further created based on the undirected weighted graph. Then, all adjacency matrices are accumulated to obtain a statistical adjacency matrix. S4: Mining the key nodes of the statistical adjacency matrix based on weighted degree sorting, and then connecting the key nodes to each other to form invasion hotspot paths of invasive animals.

2. The method for detecting hotspot paths of invading animals according to claim 1, characterized in that: In step S1, the radar detection device (1) comprises a processing module, a power supply module, a detection module and a wireless module, wherein the processing module, the detection module and the wireless module are respectively connected to the power supply module, and the detection module and the processing module are respectively connected to the wireless module, wherein the detection module is a Doppler radar, the processing module is a single-chip microcomputer, the wireless module is WIFI or ZigBee, and the power supply module is a DC power supply.

3. The method for detecting hotspot paths of invading animals according to claim 2, characterized in that: The detection area of ​​the Doppler radar is fan-shaped and is laid out using a topological structure according to the environment. Each radar detection device (1) is numbered, and each numbered radar detection device (1) corresponds to an actual intrusion area.

4. The method for detecting hotspot paths of invading animals according to claim 3, characterized in that: In step S2, the radar detection device (1) detects intruding animals in real time, and the single-chip microcomputer sends location record information to the host computer through the wireless module at regular intervals. The location record information includes the intruding animal number, the intrusion area number, the current time and the radial relative speed of the intruding animal.

5. The method for detecting hotspot paths of invading animals according to claim 4, characterized in that: In step S2, the radial relative velocity of the invading animal relative to the radar detection device (1) is obtained by Doppler radar measurement, and the radial relative velocity of the target is calculated based on the Doppler frequency shift: Where F is the original frequency of the signal; F S is the difference between the receiving signal frequency and the sending signal frequency; c is the speed of light; and v is the radial relative velocity of the invading animal to the radar detection device (1).

6. The method for detecting hotspot paths of invading animals according to claim 3, characterized in that: In step S2, the trajectory of the invading animal is divided into a stay point and a moving point. The moving point area where the invading animal is detected is used as the moving point, and the staying point area where the invading animal is detected is used as the staying point. The moving points and the staying points are arranged in chronological order to obtain a semantic trajectory.

7. The method for detecting hotspot paths of invading animals according to claim 6, characterized in that: In step S2, the steps of staying point detection are: S21: pre-processing the trajectory location information of the invading animals in each area and setting a time threshold; S22: Detect whether the intruding animal stays in the corresponding area for a time greater than a time threshold. If so, proceed to step S23. If not, mark the corresponding area as a moving point area, and end the detection of the corresponding intruding animal in the corresponding area. S23: Setting speed threshold; S24: Determine whether the average value of the sum of all speed features of the corresponding area trajectory is greater than the speed threshold. If so, mark the corresponding area as a moving point area. If not, mark the corresponding area as a staying point area, and end the detection of the corresponding invading animal in the corresponding area.

8. The method for detecting hotspot paths of invading animals according to claim 6, characterized in that: In step S3, the moving points and the staying points in the undirected weighted graph are collectively referred to as nodes, and the adjacency matrix arc[i][j] is created: In the formula, w ij Represents the weight between the i-th node and the j-th node; The initial weight w between nodes ij is 0, and in an undirected graph w ij =w ji ,The weight depends on the number of repetitions of the invasive animal trajectory path, and the number of repetitions of the semantic trajectory path is equal to the weight between nodes.

9. The method for detecting hotspot paths of invasion of invasive animals according to any one of claims 1 to 8, characterized in that: In step S4, the steps for forming the invasion hotspot path of invasive animals are: S41: Traverse the nodes of the statistical adjacency matrix and calculate the weighted degree of each node as d i , and calculate the average weighted degree d of the statistical adjacency matrix; S42: Determine whether d i >d, if yes, mark the i-th node as a key node; S43: sorting the key nodes in step S42 to obtain a node importance sorting table; S44: Based on the node importance ranking table, key nodes are connected to each other on an undirected graph to form an invasion hotspot path of invasive animals.

10. The method for detecting hotspot paths of invading animals according to claim 9, characterized in that: The calculation formula for the weighted degree of each node in the statistical adjacency matrix and the average weighted degree of the statistical adjacency matrix is: In the formula, w ij represents the weight between the i-th node and the j-th node; d i represents the weighted degree of the i-th node; d represents the average weighted degree; n represents the number of nodes in the statistical adjacency matrix.

Images