CN-121986775-A - Multi-mode omnibearing self-sensing bird-repellent device

CN121986775ACN 121986775 ACN121986775 ACN 121986775ACN-121986775-A

Abstract

The application relates to the technical field of electric power facility protection, and discloses a multi-mode omnibearing self-sensing bird repellent device. The bird threat fusion device comprises an acquisition module, a calculation module, a comparison module, an analysis module and a driving module, wherein the acquisition module is used for acquiring the electromagnetic field intensity of a power transmission line and setting an interference threshold value, the calculation module is used for acquiring signal quality scores of all sensors, the comparison module is used for calculating an electromagnetic interference grade index, inquiring history accuracy and calculating weight of all sensors, the analysis module is used for carrying out weighted fusion on detection probability and weight to obtain bird threat fusion confidence, and the driving module is used for selecting bird repelling intensity grade according to the fusion confidence and starting an actuator combination. The application solves the technical problems of high multi-sensor fusion false alarm rate, sensor performance degradation, low energy management efficiency and lack of self-adaptive feedback of the bird-repellent strategy of the existing bird-repellent device in the strong electromagnetic environment of the electric power facility, and improves the detection accuracy, long-term operation stability and energy utilization efficiency of the bird-repellent device in the complex electromagnetic environment.

Inventors

YIN HONGDE
LI WEIDONG
SHANG HUIHUI
Duan Jieling
CHEN YATAO
Dong Yingxiao
Xiao Liuwei

Assignees

国网河南省电力公司汝州市供电公司

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (10)

1. A multi-modal omnidirectional self-perceived bird repellent device, comprising: The acquisition module is used for acquiring the power frequency magnetic field intensity, the high-frequency electromagnetic radiation intensity and the electric field intensity of the power transmission line, establishing a three-dimensional electromagnetic field intensity baseline vector, and setting a low interference threshold value, a medium interference threshold value and a high interference threshold value according to the statistical median of the three-dimensional electromagnetic field intensity baseline vector; The computing module is used for respectively acquiring signal quality scores of the visual sensor, the infrared sensor, the acoustic sensor and the vibration sensor; The comparison unit is used for comparing the real-time electromagnetic field intensity with the low interference threshold value, the medium interference threshold value and the high interference threshold value, calculating an electromagnetic interference level index, inquiring the historical accuracy of each sensor in the current electromagnetic environment from a performance attenuation database according to the electromagnetic interference level index, and the weighting unit is used for carrying out weighted summation on the historical accuracy and the signal quality score to obtain a comprehensive credibility score of each sensor, and normalizing the comprehensive credibility score to obtain a visual weight, an infrared weight, an acoustic weight and a vibration weight; The analysis module is used for acquiring bird detection probability of each sensor, and carrying out weighted fusion on the bird detection probability and the visual weight, the infrared weight, the acoustic weight and the vibration weight respectively to obtain bird threat fusion confidence; The driving module comprises a recognition unit and a driving unit, wherein the recognition unit is used for comparing the bird threat fusion confidence coefficient with a preset decision threshold, when the bird threat fusion confidence coefficient exceeds the preset decision threshold, the bird repellent intensity level is selected according to the numerical interval of the bird threat fusion confidence coefficient, and the driving unit is used for starting a bird repellent actuator combination of the corresponding level.
2. The multi-modal omnidirectional self-perceived bird repellent device of claim 1, wherein the acquisition module is configured to: Acquiring the power frequency magnetic field intensity, the high frequency electromagnetic radiation intensity and the electric field intensity respectively through a power frequency magnetic field probe, a high frequency electromagnetic radiation probe and an electric field intensity probe in the triaxial electromagnetic field intensity sensor module to form a three-dimensional electromagnetic field intensity sampling sequence; Baseline data acquisition of preset time length is carried out on the three-dimensional electromagnetic field intensity sampling sequence, and a plurality of groups of three-dimensional electromagnetic field intensity vectors are obtained; Calculating the statistical median of the plurality of groups of three-dimensional electromagnetic field intensity vectors, comparing the statistical median with electromagnetic field intensity reference intervals of power transmission lines with different voltage levels prestored in a nonvolatile memory, and identifying the voltage level environment where the current device is positioned; And multiplying the statistical median by a first preset multiple, a second preset multiple and a third preset multiple respectively to obtain a low interference threshold, a medium interference threshold and a high interference threshold, wherein the first preset multiple is smaller than the second preset multiple, and the second preset multiple is smaller than the third preset multiple.
3. The multi-modal omnidirectional self-perceived bird repellent device of claim 1, wherein the computing module is configured to: Acquiring an image frame acquired by a visual image sensor, calculating the Laplace gradient square sum of the image frame as a definition index, calculating the gray histogram entropy of the image frame as a contrast index, normalizing the definition index and the contrast index, and then obtaining geometric average to obtain a visual quality score; Acquiring a thermal image frame acquired by an infrared thermal imaging sensor, calculating a temperature distribution standard deviation of the thermal image frame, and comparing the temperature distribution standard deviation with a preset temperature difference threshold value to obtain an infrared quality score; Acquiring an audio signal acquired by an acoustic sensor, performing power frequency notch filtering processing on the audio signal, calculating the signal-to-noise ratio of the filtered audio signal, and converting the signal-to-noise ratio into an acoustic quality score through a Sigmoid mapping function; and acquiring a triaxial acceleration signal acquired by the vibration sensor, carrying out wavelet decomposition on the triaxial acceleration signal, extracting the energy duty ratio of the high-frequency detail layer, and carrying out normalization processing on the energy duty ratio to obtain a vibration quality score.
4. The multi-modal omnidirectional self-perceived bird repellent device of claim 3, wherein the comparison unit is configured to: acquiring the power frequency magnetic field intensity, the high-frequency electromagnetic radiation intensity and the electric field intensity at the current moment, and calculating the ratio of the interference threshold value in the power frequency magnetic field, the interference threshold value in the high-frequency electromagnetic radiation and the interference threshold value in the electric field intensity, which correspond to the medium interference threshold value respectively; and carrying out weighted average on the ratio of the power frequency magnetic field intensity to the interference threshold value in the power frequency magnetic field, the ratio of the high-frequency electromagnetic radiation intensity to the interference threshold value in the high-frequency electromagnetic radiation and the ratio of the electric field intensity to the interference threshold value in the electric field intensity to obtain an electromagnetic interference grade index.
5. The multi-modal omnidirectional self-perceived bird repellent device of claim 4, wherein the query unit is configured to: And according to the electromagnetic interference level index, retrieving the vision sensor historical accuracy, the infrared sensor historical accuracy, the acoustic sensor historical accuracy and the vibration sensor historical accuracy from a performance attenuation database, wherein the performance attenuation database stores the mapping relation between each sensor accuracy and the electromagnetic interference level index.
6. The multi-modal omnidirectional self-perceived bird repellent device of claim 5, wherein the weighting unit is configured to: Carrying out weighted summation on the historical accuracy rate of the visual sensor and the visual quality score by a first weight coefficient and a second weight coefficient to obtain a comprehensive credibility score of the visual sensor; Carrying out weighted summation on the first weight coefficient and the second weight coefficient on the historical accuracy rate of the infrared sensor and the infrared quality score to obtain an integrated reliability score of the infrared sensor; Carrying out weighted summation on the first weight coefficient and the second weight coefficient on the historical accuracy of the acoustic sensor and the acoustic quality score to obtain an integrated reliability score of the acoustic sensor; carrying out weighted summation on the first weight coefficient and the second weight coefficient on the historical accuracy rate of the vibration sensor and the vibration quality score to obtain a comprehensive reliability score of the vibration sensor; and respectively carrying out exponential operation on the visual sensor comprehensive credibility score, the infrared sensor comprehensive credibility score, the acoustic sensor comprehensive credibility score and the vibration sensor comprehensive credibility score, and then carrying out summation normalization processing to obtain visual weight, infrared weight, acoustic weight and vibration weight.
7. The multi-modal omnidirectional self-perceived bird repellent device of claim 1, wherein the analysis module is configured to: The method comprises the steps of carrying out target detection on an image frame through a visual detection module, outputting visual detection probability, carrying out hot spot segmentation and feature classification on the hot image frame through an infrared detection module, outputting infrared detection probability, carrying out frequency spectrum analysis and voiceprint recognition on an audio signal through an acoustic detection module, outputting acoustic detection probability, carrying out feature extraction and pattern recognition on an acceleration signal through a vibration detection module, and outputting vibration detection probability; Multiplying the visual detection probability with the visual weight, multiplying the infrared detection probability with the infrared weight, multiplying the acoustic detection probability with the acoustic weight, multiplying the vibration detection probability with the vibration weight, and carrying out summation operation on four product results to obtain the bird threat fusion confidence.
8. The multi-modal omnidirectional self-perceived bird repellent device of claim 1, wherein the recognition unit is configured to: when the bird threat fusion confidence is greater than the preset decision threshold and less than a first threat threshold, determining that the bird repellent intensity level is a low threat level; When the bird threat fusion confidence is greater than or equal to the first threat threshold and less than a second threat threshold, determining that the bird repellent intensity level is a medium threat level; and when the bird threat fusion confidence is greater than or equal to the second threat threshold, determining that the bird repellent intensity level is a high threat level.
9. The multi-modal omnidirectional self-perceived bird repellent device of claim 8, wherein the drive unit is configured to: when the bird repellent intensity level is a low threat level, starting an ultrasonic transmitter for a first preset duration; When the bird repelling intensity level is a medium threat level, simultaneously starting the ultrasonic transmitter and the LED strong light lamp array, and enabling the LED strong light lamp array to perform color switching and flickering at a first preset frequency for a second preset duration; when the bird repelling intensity level is a high threat level, the ultrasonic transmitter, the LED strong light lamp array and the mechanical swing arm are started simultaneously, the LED strong light lamp array performs random color conversion at a second preset frequency, and the mechanical swing arm performs reciprocating swing at a preset swing frequency for a third preset duration.
10. The multi-modal omnidirectional self-perceived bird repellent device of claim 9, further comprising a decision module for: recalculating the bird threat fusion confidence coefficient at preset feedback intervals during the execution of bird repelling to obtain real-time fusion confidence coefficient; comparing the real-time fusion confidence coefficient with an initial fusion confidence coefficient at the start time of bird repellent, and when the real-time fusion confidence coefficient is smaller than a preset attenuation proportion of the initial fusion confidence coefficient for a plurality of times, judging that birds are repelled, and stopping the operation of a bird repellent actuator combination; And when the real-time fusion confidence level is continuously kept above a preset high confidence level threshold value in a first preset time period, a second preset time period or a third preset time period, the bird-repellent intensity level is raised to the next level, and a corresponding bird-repellent actuator combination is started.

Description

Multi-mode omnibearing self-sensing bird-repellent device Technical Field The application relates to the technical field of electric power facility protection, in particular to a multi-mode omnibearing self-sensing bird repellent device. Background The power facilities such as a power transmission line, a transformer substation and the like are exposed in an outdoor environment for a long time, birds nest, stay or excrete on the power facilities, so that faults such as equipment pollution, insulator flashover, line short circuit and the like can be caused, and the safe operation of a power grid is seriously threatened. The existing bird repellent device generally adopts a multi-mode sensor fusion technology, monitors bird activities in a cooperative manner by deploying a plurality of sensors such as vision, infrared, acoustic, vibration and the like, and combines with bird repellent means such as ultrasonic waves, strong light flickering, mechanical swing and the like for protection. The devices generally adopt a multi-sensor data fusion method of equal weight fusion or manual preset fixed weight, the detection results of all the sensors are voted or weighted average, and when the fusion result exceeds a set threshold value, bird repelling action is triggered. However, there is a strong electromagnetic environment around the power facility, and the degree of influence of the power frequency magnetic field generated by the power transmission line, the high-frequency electromagnetic radiation generated by the switching operation, and the strong electric field around the high-voltage line on different types of sensors is significantly different. The micro-bolometer of the infrared thermal imaging sensor is easy to generate reading drift in a strong electromagnetic field, electromagnetic interference in an audio acquisition circuit of the acoustic sensor is coupled into noise to reduce the signal-to-noise ratio, and the visual sensor and the vibration sensor are relatively less affected by the electromagnetic. The existing fixed weight fusion method does not consider the differential influence of electromagnetic environment on the performance of each sensor, under the strong electromagnetic interference environment, the sensors with seriously attenuated performance still participate in fusion decisions by fixed weight, so that the false alarm rate is as high as 25-40%, frequent false triggering not only wastes energy, but also can cause birds to generate adaptability due to excessive stimulation, and the long-term bird repelling effect is reduced. Disclosure of Invention The application provides a multi-mode omnibearing self-sensing bird-scaring device, which is used for solving the technical problems of high fusion false alarm rate of multiple sensors, sensor performance degradation, low energy management efficiency and lack of self-adaptive feedback of bird-scaring strategies of the traditional bird-scaring device in a strong electromagnetic environment of an electric power facility, and improving the detection accuracy, long-term operation stability and energy utilization efficiency of the bird-scaring device in a complex electromagnetic environment. The application provides a multimode omnibearing self-sensing bird repellent device, which comprises: The acquisition module is used for acquiring the power frequency magnetic field intensity, the high-frequency electromagnetic radiation intensity and the electric field intensity of the power transmission line, establishing a three-dimensional electromagnetic field intensity baseline vector, and setting a low interference threshold value, a medium interference threshold value and a high interference threshold value according to the statistical median of the three-dimensional electromagnetic field intensity baseline vector; The computing module is used for respectively acquiring signal quality scores of the visual sensor, the infrared sensor, the acoustic sensor and the vibration sensor; The comparison unit is used for comparing the real-time electromagnetic field intensity with the low interference threshold value, the medium interference threshold value and the high interference threshold value, calculating an electromagnetic interference level index, inquiring the historical accuracy of each sensor in the current electromagnetic environment from a performance attenuation database according to the electromagnetic interference level index, and the weighting unit is used for carrying out weighted summation on the historical accuracy and the signal quality score to obtain a comprehensive credibility score of each sensor, and normalizing the comprehensive credibility score to obtain a visual weight, an infrared weight, an acoustic weight and a vibration weight; The analysis module is used for acquiring bird detection probability of each sensor, and carrying out weighted fusion on the bird detection probability and the visual weight, the infrared weight, the acousti