A substation small animal intrusion monitoring and early warning method and system

Abstract

The invention relates to the technical field of small animal monitoring and early warning in substations, and discloses a small animal invasion monitoring and early warning method and system for substations. The system comprises an image preprocessing module considering micro-meteorological factors, an intrusion target detection module based on image recognition and target tracking, and a trajectory prediction module for a non-monitored area. The invention adopts an improved YOLOv5s algorithm as a benchmark detection model, and adopts an improved EIOU loss function to replace the defect of the loss function in the model. At the same time, the defect that ordinary convolution is difficult to extract irregular target features is solved by replacing the deformable convolution, and the recognition accuracy of the model is improved, so as to realize the detection of targets in the substation and judge whether they are intrusive small animals, effectively reduce the recognition time, and improve the accuracy, so as to meet the growing intelligent demand.

Description

Technical Field

[0001] The present invention relates to the technical field of small animal monitoring and early warning in transformer substations, and in particular to a small animal invasion monitoring and early warning method and system in transformer substations. Background Art

[0002] In recent years, with the rapid development of my country's economy, the demand for electricity has become increasingly higher, and the construction of power lines has also increased. As one of the core components of the power system, the substation undertakes core functions such as regulating voltage and current, receiving and distributing electric energy. With the development of the power industry, substations are also under continuous construction. For substations located in suburbs and mountainous areas, due to the lack of frequent human activities, the environmental pollution is relatively small and the natural conditions are excellent, and a large number of small animals such as birds, mice, snakes, and rabbits are active around them. As a result, outdoor substations have repeatedly experienced busbar phase short circuits and ground short circuits caused by small animals moving inside high-voltage electrical equipment or leaving nests, causing electrical equipment damage or even burning, seriously affecting the safe and stable operation of the power system, and further restricting the economic development of the region; in addition, the existing intrusion target early warning technology in substations has the defects of long recognition time and low accuracy, which cannot meet the growing demand for intelligence.

[0003] Therefore, it is necessary to further solve the above problems. The applicant has proposed a substation small animal intrusion monitoring and early warning method and system. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a substation small animal intrusion monitoring and early warning method and system, which makes up for the defects of the prior art proposed by the above background technology in the early warning process of intrusion targets in substations, such as long recognition time and low accuracy, and solves the problem that the prior art cannot meet the growing demand for intelligence.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A small animal intrusion monitoring and early warning system for a substation, the system comprising an image preprocessing module taking micro-meteorological factors into consideration, an intrusion target detection module based on image recognition and target tracking, and a trajectory prediction module for areas without monitoring;

[0007] The image preprocessing module considering micro-meteorological factors is used to monitor the meteorological factors in the current environment in real time, obtain meteorological judgment data, and perform image preprocessing on the camera image according to the current meteorological conditions;

[0008] The intrusion target detection module based on image recognition and target tracking is used to identify and track the intrusion targets in the substation in real time, determine whether they are small animals, and trigger an early warning prompt;

[0009] The trajectory prediction module of the non-monitored area is used to predict the movement trajectory of the intruding target in the blind area of ​​the camera to form a complete trajectory route.

[0010] A method for a small animal intrusion monitoring and early warning system for a substation, the method comprising the following steps:

[0011] S1. Use the micro-meteorological monitoring device to obtain the meteorological data of the current environment, including light intensity, temperature, humidity, ultraviolet index, and rainfall, to determine the meteorological state of the environment where the camera is currently located;

[0012] S2, based on the weather conditions determined by S1, image preprocessing is performed on low-brightness images in foggy and nighttime environments to improve image quality and clarity. A defogging algorithm based on mean filtering is used for foggy images, and an adaptive gamma correction algorithm is used for image enhancement of low-brightness images.

[0013] S3, identify the intrusion target in the substation, determine whether it is a small animal and track the movement trajectory of the target;

[0014] S4. Combine the three-dimensional model of the substation to predict the trajectory of the unmonitored area.

[0015] As a further solution of the present invention: the specific steps of step S1 include:

[0016] S11, firstly, the rainfall parameter in the monitoring data is judged, and if rain is detected, it is directly determined that a rainy day has occurred;

[0017] S12. If no rain is detected, the temperature, humidity and ultraviolet intensity are judged. If the foggy day condition is met at the same time, whether it is morning dew is further judged based on the time, and the morning dew formula is substituted for calculation. The formula is as follows:

[0018]

[0019] In the formula, is the dew point temperature, is the relative humidity at sunset, is the temperature at sunset, where the sunrise and sunset times are divided by the jump point of the ultraviolet intensity at 0 o'clock;

[0020] S13. For weather conditions that do not satisfy rainy and foggy conditions, cloudy and sunny days are distinguished based on light intensity. Finally, all monitoring data and meteorological judgments are sent to the meteorological curve for updating.

[0021] As a further solution of the present invention: the specific steps of step S3 include:

[0022] S31. The improved YOLOv5s algorithm is used as the benchmark detection model. Aiming at the defects of the loss function in the model, the improved EIOU loss function is used to replace it. At the same time, the defect that ordinary convolution is difficult to extract irregular target features is solved by replacing the deformable convolution, which improves the recognition accuracy of the model, detects the targets in the substation, and determines whether they are invading small animals.

[0023] S32. Combined with the improved ByteTrack multi-target tracking algorithm, the real-time tracking of invading small animals is carried out. In order to solve the state space description defect of non-rigid targets in the tracking algorithm, the aspect ratio of the bounding box is replaced by the width of the bounding box. At the same time, in order to solve the problem of temporary disappearance and discontinuity of the tracking trajectory caused by the occlusion of the target, the occluded target trajectory division is proposed, and the strategy of retaining frames is adopted for the occluded target trajectory, which effectively improves the continuity of the occluded target trajectory;

[0024] S33. Use monocular vision ranging technology to estimate the three-dimensional coordinate position of the invading small animal in the camera coordinate system, where the ranging formula for the ground target is as follows:

[0025]

[0026]

[0027] Among them, dh and dw are the vertical longitudinal distance and the horizontal lateral distance from the ranging point to the camera, respectively. H is the height between the camera and the ground. α is the angle between the optical axis of the camera and the horizontal plane, that is, the pitch angle. u and v are the coordinates of the midpoint of the image. u0 and v0 are the positions of the origin in the image coordinate system transformed to the pixel coordinate system. fx and fy are the focal lengths of the camera in the horizontal direction and the vertical direction, respectively.

[0028] As a further solution of the present invention: the specific steps of step S4 include:

[0029] S41. Combined with the three-dimensional model of the substation, the position of the intrusion target in the camera is integrated into the world coordinate system through the coordinate fusion formula, and the trajectory of the intrusion target is described from a global perspective. The formula is as follows:

[0030]

[0031] in, , is the target position in the world coordinate system, , is the target position in the camera, , is the position of the camera's origin in the world coordinate system, is the angle between the camera coordinate system and the horizontal axis of the world coordinate system;

[0032] S42. For discontinuous trajectories of intrusion targets in unmonitored areas, the PSO algorithm is used to predict the optimal trajectory.

[0033] Compared with the prior art, the present invention has the following beneficial effects: the present invention adopts an improved YOLOv5s algorithm as a benchmark detection model, and adopts an improved EIOU loss function to replace the defects of the loss function in the model. At the same time, by replacing the deformable convolution, the defect that ordinary convolution is difficult to extract irregular target features is solved, and the recognition accuracy of the model is improved, so as to detect the target in the substation and judge whether it is an invading small animal, effectively reducing the recognition time, while improving the accuracy, and meeting the growing demand for intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] 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 prior art descriptions. 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 the structures shown in these drawings without creative work. Among them:

[0035] Figure 1 It is a structural block diagram of the substation small animal intrusion monitoring and early warning system of the present invention;

[0036] Figure 2 It is a structural diagram of the micro-meteorological monitoring device of the present invention;

[0037] Figure 3 It is the target tracking flow chart of the present invention;

[0038] Figure 4 It is a schematic diagram of coordinate system fusion of the present invention;

[0039] Figure 5 It is a schematic diagram of the non-monitoring area of ​​the present invention. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Example

[0042] See also Figure 1 and Figure 2 As shown, the present invention provides a substation small animal intrusion monitoring and early warning system, the system includes an image preprocessing module considering micro-meteorological factors, an intrusion target detection module based on image recognition and target tracking, and a trajectory prediction module for non-monitored areas;

[0043] The image preprocessing module considering micro-meteorological factors is used to monitor the meteorological factors in the current environment in real time, obtain meteorological judgment data, and perform image preprocessing on the camera image according to the current meteorological conditions;

[0044] The intrusion target detection module based on image recognition and target tracking is used to identify and track the intrusion targets in the substation in real time, determine whether they are small animals, and trigger an early warning prompt;

[0045] The trajectory prediction module of the non-monitored area is used to predict the movement trajectory of the intruding target in the blind area of ​​the camera to form a complete trajectory route.

[0046] The present invention also provides a method for a small animal intrusion monitoring and early warning system for a substation, the method comprising the following steps:

[0047] Use the micro-meteorological monitoring device to obtain the current environmental meteorological data, including light intensity, temperature, humidity, UV index, and rainfall, to determine the meteorological state of the current camera environment. The specific steps are as follows:

[0048] First, the rainfall parameters in the monitoring data are judged. If rain is detected, it is directly determined that a rainy day has occurred;

[0049] If no rain is detected, the temperature, humidity and ultraviolet intensity are judged. If the foggy day conditions are met at the same time, it is further judged whether it is morning dew phenomenon based on time, and it is substituted into the morning dew formula for calculation. The formula is as follows:

[0050]

[0051] In the formula, is the dew point temperature, is the relative humidity at sunset, is the temperature at sunset, where the sunrise and sunset times are divided by the jump point of the ultraviolet intensity at 0 o'clock;

[0052] For weather conditions that do not meet the requirements of rainy and foggy days, cloudy and sunny days are distinguished based on light intensity. Ultimately, all monitoring data and meteorological judgments are sent to the meteorological curve for updating.

[0053] According to the meteorological conditions determined above, image preprocessing is performed on low-brightness images in foggy and nighttime environments to improve image quality and clarity. A defogging algorithm based on mean filtering is used for foggy images, and an adaptive gamma correction algorithm is used for image enhancement of low-brightness images.

[0054] Identify the intrusion target in the substation, determine whether it is a small animal, and track the target's movement trajectory. The specific steps are as follows:

[0055] The improved YOLOv5s algorithm is used as the benchmark detection model. Aiming at the defects of the loss function in the model, the improved EIOU loss function is used to replace it. At the same time, the defect that ordinary convolution is difficult to extract irregular target features is solved by replacing the deformable convolution, which improves the recognition accuracy of the model, detects the targets in the substation, and determines whether they are invading small animals.

[0056] like Figure 3 As shown in the figure, the real-time tracking of invading small animals is carried out by combining the improved ByteTrack multi-target tracking algorithm. In order to solve the state space description defect of non-rigid targets in the tracking algorithm, the aspect ratio of the bounding box is replaced by the width of the bounding box. At the same time, in order to solve the problem of temporary disappearance and discontinuity of the tracking trajectory caused by the occlusion of the target, the occluded target trajectory division is proposed, and the strategy of retaining frames is adopted for the occluded target trajectory, which effectively improves the continuity of the occluded target trajectory.

[0057] The monocular vision ranging technology is used to estimate the three-dimensional coordinate position of the intruding small animals in the camera coordinate system. The ranging formula for ground targets is as follows:

[0058]

[0059]

[0060] Among them, dh and dw are the vertical longitudinal distance and the horizontal lateral distance from the ranging point to the camera, respectively. H is the height between the camera and the ground. α is the angle between the optical axis of the camera and the horizontal plane, that is, the pitch angle. u and v are the coordinates of the midpoint of the image. u0 and v0 are the positions of the origin in the image coordinate system transformed to the pixel coordinate system. fx and fy are the focal lengths of the camera in the horizontal direction and the vertical direction, respectively.

[0061] Combined with the three-dimensional model of the substation, the trajectory of the unmonitored area is predicted. The specific steps are as follows:

[0062] like Figure 4As shown in the figure, combined with the three-dimensional model of the substation, the position of the intrusion target in the camera is integrated into the world coordinate system through the coordinate fusion formula, and the trajectory of the intrusion target is described from a global perspective. The formula is as follows:

[0063]

[0064] in, , is the target position in the world coordinate system, , is the target position in the camera, , is the position of the camera's origin in the world coordinate system, is the angle between the camera coordinate system and the horizontal axis of the world coordinate system;

[0065] like Figure 5 As shown in the figure, for the discontinuous trajectory of the intrusion target in the unmonitored area, the PSO algorithm is used to predict the optimal trajectory.

[0066] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A substation small animal intrusion monitoring and early warning system, characterized by: The system includes an image preprocessing module that considers micro-meteorological factors, an intrusion target detection module based on image recognition and target tracking, and a trajectory prediction module for unmonitored areas; The image preprocessing module considering micro-meteorological factors is used to obtain meteorological judgment data through real-time monitoring and perform image preprocessing on the camera images; The intrusion target detection module based on image recognition and target tracking: identifies and tracks intrusion targets in the substation in real time, triggering early warning prompts; the details are as follows: The improved YOLOv5s algorithm is used as the benchmark detection model, and the improved EIOU loss function is used to replace the loss function in the model. At the same time, the accuracy of model recognition is improved by replacing the deformable convolution method, and the targets in the substation are detected to determine whether they are invading small animals. Combined with the improved ByteTrack multi-target tracking algorithm for real-time tracking of invading small animals, the bounding box aspect ratio is replaced by the bounding box width. At the same time, the occluded target trajectory division is proposed, and the strategy of retaining frames is adopted for the occluded target trajectory, which effectively improves the continuity of the occluded target trajectory; The monocular vision ranging technology is used to estimate the three-dimensional coordinate position of the intruding small animals in the camera coordinate system. The ranging formula for ground targets is as follows: Among them, dh and dw are the vertical longitudinal distance and the horizontal lateral distance from the ranging point to the camera, respectively; H is the height between the camera and the ground; α is the angle between the optical axis of the camera and the horizontal plane, i.e. the pitch angle; u and v are the coordinates of the midpoint of the image; u0 and v0 are the positions of the origin in the image coordinate system transformed to the pixel coordinate system, and are the positions of the center point of the image; fx and fy are the horizontal focal length and the vertical focal length of the camera, respectively; The trajectory prediction module of the unmonitored area predicts the movement trajectory of the intrusion target in the camera blind spot to form a complete trajectory route; the details are as follows: Combined with the three-dimensional model of the substation, the position of the intrusion target in the camera is integrated into the world coordinate system through the coordinate fusion formula, and the trajectory of the intrusion target is described from a global perspective. The formula is as follows: in, , is the target position in the world coordinate system, , is the target position in the camera, , is the position of the camera's origin in the world coordinate system, is the angle between the camera coordinate system and the horizontal axis of the world coordinate system; For the discontinuous trajectory of the intrusion target in the unmonitored area, the PSO algorithm is used to predict the optimal trajectory.

2. A method for a substation small animal intrusion monitoring and early warning system as claimed in claim 1, characterized in that: The method comprises the following steps: S1. Using a micro-meteorological monitoring device to obtain meteorological data related to the current environment, and determine the meteorological state of the environment where the current camera is located; S2, based on the weather conditions determined by S1, image preprocessing is performed on low-brightness images in foggy days and nighttime environments; S3, identify the intrusion target in the substation, determine whether it is a small animal and track the movement trajectory of the target; S4. Combine the three-dimensional model of the substation to predict the trajectory of the unmonitored area.

3. A method for substation small animal intrusion monitoring and early warning system according to claim 2, characterized in that: The specific steps of step S1 include: S11, firstly, the rainfall parameter in the monitoring data is judged, and if rain is detected, it is directly determined that a rainy day has occurred; S12. If no rain is detected, the temperature, humidity and ultraviolet intensity are judged. If the foggy day condition is met at the same time, whether it is morning dew is further judged based on the time, and the morning dew formula is substituted for calculation. The formula is as follows: In the formula, is the dew point temperature, is the relative humidity at sunset, is the temperature at sunset, where the sunrise and sunset times are divided by the jump point of the ultraviolet intensity at 0 o'clock; S13. For weather conditions that do not satisfy rainy and foggy conditions, cloudy and sunny days are distinguished based on light intensity. Finally, all monitoring data and meteorological judgments are sent to the meteorological curve for updating.

Images