Intelligent bird repelling device for electric power transmission and distribution facilities and repelling strategy generation method

Abstract

The invention discloses an intelligent bird repelling device for electric power transmission and distribution facilities and a repelling strategy generation method. The invention relates to the field of bird repelling devices, and provides a method for acquiring feature data of birds, including body type, flight speed, flight height and sound frequency; converting the feature data into feature vectors; calculating the probability P (C | X) of occurrence of each bird according to the relationship between the feature vector X and the bird type C; calculating an estimated value of P (C | X) according to the model parameters; determining the type of the birds according to the estimated value of the formula P (C | X); and executing a corresponding driving strategy according to the identification result of the bird type. Therefore, bird repelling strategies can be flexibly selected according to the types of the birds, and the optimal bird repelling effect is achieved.

Description

Technical Field

[0001] The present invention relates to the field of power transmission and distribution, and in particular to an intelligent bird-repelling device and a method for generating a bird-repelling strategy. Background Art

[0002] Power transmission and distribution facilities are the backbone of energy supply in modern society, and their safe and stable operation is of vital importance. However, the threat of bird activities to power facilities has become increasingly prominent, becoming one of the important factors affecting the safety of the power system. Birds' nesting and roosting behaviors on key facilities such as transmission line towers, distribution line poles, and substations may not only threaten the stable operation of power equipment, but may also cause accidents such as short circuits and flashovers in transmission lines, and in severe cases may even cause fires, posing great hidden dangers to the safety and stability of the power system.

[0003] Traditional bird-repelling measures mainly include physical isolation, sound repelling, and visual interference. The physical isolation method prevents birds from approaching or nesting by installing obstacles on power facilities, but this method is difficult to construct, costly, and may affect the normal operation and maintenance of power facilities. The sound repelling method drives away birds by playing specific sounds, but this method can only temporarily repel birds, and there is a problem that birds are highly adaptable to sounds and the repelling effect is difficult to last. The visual interference method scares away birds by setting up dummies, reflective objects, etc., but it also has problems such as limited effect and easy adaptation.

[0004] A major shortcoming of existing bird-repelling technology is the lack of targeting and differentiation. There are many species of birds, and different species of birds have different adaptability to the environment, threat perception, and behavioral habits. For example, some birds are more sensitive to sound, while others may be more sensitive to visual stimulation. In addition, the activity time, habitat habits, migration patterns, etc. of different birds are also different. Therefore, it is difficult to adapt to the characteristics of different birds and achieve the ideal bird-repelling effect by using a single bird-repelling method.

[0005] Using differentiated bird-repelling methods for different birds is the key to improving the efficiency and effectiveness of bird-repelling. Differentiated bird-repelling methods can be targeted to select or combine different bird-repelling technologies according to the species, habits, and activity patterns of birds. For example, for birds that are sensitive to sound, sounds of a specific frequency can be used to repel them; for birds that are sensitive to vision, visual stimulation methods such as lasers and flashes can be used; for birds that are active at night, special light sources such as infrared or ultraviolet rays can be used to repel them. In addition, the start time and frequency of bird-repelling measures can be adjusted in combination with the habitat habits and activity patterns of birds to achieve better bird-repelling effects.

[0006] However, there are still many challenges in achieving differentiated bird repellent with existing technologies. First, there are many types of birds, and the identification and classification of the characteristics of different birds requires a large amount of data support and advanced algorithms, and different bird repellent strategies are also required for different birds. Secondly, the implementation of differentiated bird repellent means requires a high degree of flexibility and adjustability to adapt to different scenarios and environments. In addition, differentiated bird repellent also needs to take into account the requirements of environmental protection and ecological protection to avoid adverse effects and accidental damage to birds and the ecological environment. Summary of the invention

[0007] The present invention is aimed at the deficiencies of the prior art and proposes a new type of intelligent bird-repelling device and a bird-repelling method. The device can intelligently identify and select differentiated bird-repelling means according to the species, habits, activity patterns, etc. of the birds, so as to achieve accurate bird-repelling. By integrating advanced detection technology, intelligent recognition algorithms and a variety of bird-repelling technologies, the present invention can monitor bird activities in real time, automatically start the most suitable bird-repelling measures, and effectively improve the efficiency and effect of bird-repelling. At the same time, the present invention also fully considers the requirements of environmental protection and ecological protection, and adopts harmless bird-repelling technology to ensure that the bird-repelling process is friendly to birds and the environment.

[0008] Through the implementation of the present invention, the problems of insufficient targeting of bird species and low efficiency in existing bird-repelling technologies can be effectively solved, and a more intelligent, efficient and environmentally friendly bird-repelling solution can be provided for power transmission and distribution facilities.

[0009] In one embodiment of the present application, a method for identifying bird species in an intelligent bird repellent device for power transmission and distribution facilities is provided, comprising:

[0010] Obtain characteristic data of birds, including body size, flight speed, flight altitude, and sound frequency;

[0011] Convert the feature data into a feature vector X = {x1, x2, ..., x n};

[0012] According to the relationship between the feature vector X and the bird species C, the probability of each bird appearing P(C|X) is calculated as:

[0013]

[0014] Where K is the number of known bird species, α, β, and γ are model parameters, and ||X|| is the norm of the feature vector X;

[0015] The negative log-likelihood function L(α, β, γ) is calculated by the following formula to estimate the model parameters α, β, and γ:

[0016]

[0017] Calculate an estimate of P(C|X) based on the model parameters α, β and γ;

[0018] The type of bird is determined based on the estimated value of the formula P(C|X).

[0019] Optionally, in some embodiments, the method further comprises the step of: calculating the probability P(C|X) of each bird appearing according to the relationship between the feature vector X and the bird species C as follows:

[0020]

[0021] Where K is the number of known bird species, α, β, and γ are model parameters, and ||X|| is the norm of the feature vector X;

[0022] The negative log-likelihood function L(α, β, γ) is calculated by the following formula to estimate the model parameters α, β, and γ:

[0023]

[0024] Calculate an estimate of P(C|X) based on the model parameters α, β and γ;

[0025] The type of bird is determined based on the estimated value of the formula P(C|X).

[0026] In order to construct a mathematical model including logarithmic and exponential calculations to describe the mixed shaping of ultrasonic and acoustic signals, and to adjust the components of ultrasonic and acoustic signals according to different bird species, the present application can design a scenario in which the mixed signal S(t)S(t) is a mixture of acoustic signal A(t)A(t) and ultrasonic signal U(t)U(t), and the mixing ratio varies with time tt and bird species kk. The present application will construct a mathematical model including logarithmic and exponential operations to define the mixing method of the two signals.

[0027] Optionally, the method for identifying bird species further comprises the steps of:

[0028] Generate a mixed signal of the acoustic wave signal A(t) and the ultrasonic signal U(t), and the mixed signal S(t,k) is defined as:

[0029]

[0030] where δ k and γ k is a parameter adjusted according to the bird species k, A0 and U0 are the initial amplitudes of the acoustic and ultrasonic signals, α and β are the attenuation coefficients of the acoustic and ultrasonic signals, respectively, and f a and f u are the frequencies of the acoustic and ultrasonic signals, respectively,a and φ u are the phases of the acoustic and ultrasonic signals, log δk is δ k The logarithmic function with base is , and e is the natural exponential function.

[0031] Optionally, the method for generating a driving strategy further comprises the steps of:

[0032] Generate a mixed signal of the acoustic wave signal A(t) and the ultrasonic signal U(t), and the mixed signal S(t,k) is defined as:

[0033]

[0034] where δ k and γ k is a parameter adjusted according to the bird species k, A0 and U0 are the initial amplitudes of the acoustic and ultrasonic signals, α and β are the attenuation coefficients of the acoustic and ultrasonic signals, respectively, and f a and f u are the frequencies of the acoustic and ultrasonic signals, respectively, a and φ u are the phases of the acoustic and ultrasonic signals, log δk is δ k The logarithmic function with base is , and e is the natural exponential function.

[0035] Optionally, in some embodiments, the method for identifying bird species further comprises the steps of:

[0036] Generate a mixed signal of the acoustic signal A(t) and the ultrasonic signal U(t). The complete form of the mixed signal S(t,k) can be defined by the following formula:

[0037]

[0038] where δ k and γ k is a parameter adjusted according to the bird species k, A0 and U0 are the initial amplitudes of the acoustic and ultrasonic signals, α and β are the attenuation coefficients of the acoustic and ultrasonic signals, respectively, and f a and f u are the frequencies of the acoustic and ultrasonic signals, respectively, a and φ u are the phases of the acoustic and ultrasonic signals, log δk is δ k The logarithmic function with base is , and e is the natural exponential function.

[0039] Optionally, the method for identifying bird species is characterized by:

[0040] The ultrasonic signal U(t)U(t) is:

[0041] U(t)=U0·e -βt ·sin(2πf u t+φ u )

[0042] The mixed signal S(t) is defined as:

[0043] S(t,k)=λ(t,k)·A(t)+(1-λ(t,k))·U(t)

[0044] in,

[0045] U0 is the initial amplitude of the ultrasonic signal;

[0046] β is the attenuation coefficient of the ultrasonic signal;

[0047] f u is the frequency of the ultrasonic signal;

[0048] φ u is the phase of the ultrasonic signal;

[0049] λ(t,k) is the weight of the acoustic signal in the mixed signal, 0≤λ(t,k)≤1, which depends on the time tt and the bird species k.

[0050] According to the bird species identification result, the driving control submodule performs beamforming according to the corresponding mixed signal S(t, k) and transmits it to the monitoring area.

[0051] Optionally, the method for generating a driving strategy is characterized by further comprising the steps of:

[0052] Generate a mixed signal of the acoustic wave signal A(t) and the ultrasonic signal U(t), and the mixed signal S(t,k) is defined as:

[0053] where δ k and γ k is a parameter adjusted according to the bird species k, A0 and U0 are the initial amplitudes of the acoustic and ultrasonic signals, α and β are the attenuation coefficients of the acoustic and ultrasonic signals, respectively, and f a and f u are the frequencies of the acoustic and ultrasonic signals, respectively, a and φ u

[0054] are the phases of the acoustic and ultrasonic signals, log δk is δ kis a logarithmic function with base , and e is a natural exponential function. Optionally, the method for generating a driving strategy in some embodiments further includes the steps of:

[0055] Generate a mixed signal of the acoustic wave signal A(t) and the ultrasonic signal U(t), and the mixed signal S(t,k) is defined as:

[0056]

[0057] where δ k and γ k is a parameter adjusted according to the bird species k, A0 and U0 are the initial amplitudes of the acoustic and ultrasonic signals, α and β are the attenuation coefficients of the acoustic and ultrasonic signals, respectively, and f a and f u are the frequencies of the acoustic and ultrasonic signals, respectively, a and φ u are the phases of the acoustic and ultrasonic signals, log δk is δ k The logarithmic function with base is , and e is the natural exponential function.

[0058] The methods in some embodiments of the present application, by establishing a mathematical model, can adaptively adjust the trunk strategy for different bird species, and have higher flexibility, higher accuracy and robustness than the existing technology.

[0059] In practical applications, these mathematical models can be further optimized to improve recognition accuracy, and model parameters can be adjusted through more training data to obtain better performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0061] Figure 1 The figure is a schematic diagram of signal flow in a vehicle decision-making method based on a PFC model according to an embodiment of the present invention.

[0062] Figure 2 Schematic diagram of the agent architecture of the PFC model;

[0063] Figure 3 A schematic diagram showing the structure of a computing device according to an embodiment of the present application is shown;

[0064] Figure 4 The system main interface of the host computer of the intelligent bird-repelling device of the embodiment of the present application is shown;

[0065] Figure 5A , Figure 5B The front view and the side view of the intelligent bird-repelling device of the embodiment of the present application are respectively shown;

[0066] Figure 6 The figure shows an axial view of the appearance of the intelligent bird-repelling device according to an embodiment of the present application. DETAILED DESCRIPTION

[0067] The intelligent bird-repelling device in some embodiments of this application is an intelligent bird-repelling device designed by Nanjing Shoufeng specifically for power transmission and distribution scenarios, aiming to effectively solve the potential safety hazards caused by bird nesting and roosting on key facilities such as transmission line towers, distribution line poles, and substations. By integrating advanced detection technology, intelligent recognition algorithms, and environmentally friendly bird-repelling methods, the technical solution of this application can monitor bird activities in real time, automatically initiate non-harmful bird-repelling measures, effectively reduce the incidence of accidents such as short circuits and flashovers in power facilities caused by bird activities, and ensure the safe and stable operation of transmission lines.

[0068] The technical solution of the present application adopts a variety of advanced bird-repelling means, which simulates threat signals in the natural environment to make birds feel uncomfortable, thereby effectively achieving the effect of bird-repelling. The technical solution of the present application emits specific sound waves, various sounds that scare birds, and changing laser beams to make birds feel threatened and stay away from power equipment. At the same time, the technical solution of the present application reduces the adaptability of birds through the coordination and changing combination of multiple repelling means.

[0069] The technical solution of this application is equipped with an intelligent expulsion strategy adjustment system, which allows users to adjust the expulsion strategy based on the bird activity monitored in real time. The system has multiple preset modes, and the appropriate expulsion strategy can be selected according to different scenarios and needs.

[0070] The technical solution of this application uses harmless bird-repelling technology and will not cause harm to birds. The frequency and intensity of sound and light have been scientifically adjusted and will not cause interference and harm to humans. The product materials meet environmental protection standards, are non-toxic and harmless, and will not pollute the environment.

[0071] The technical solution of this application is compact in design and easy to install, suitable for different types of power equipment and scenarios. The device is easy to operate and equipped with a user-friendly interface, which can be set and monitored remotely. The product has stable and reliable performance and can keep the transmission line away from the threat of bird damage for a long time.

[0072] Regarding the external structure of the intelligent bird repellent device A90, Figure 5A , Figure 5B , Figure 6 As shown in FIG. 1 , the internal structure of the intelligent bird repellent device A90 is as follows: Figure 2 As shown, including:

[0073] A camera submodule A99 operable to capture images of birds;

[0074] Radar monitoring module A96, operable to monitor bird activity in real time and collect bird data;

[0075] The sound expelling module A91 can be operated to play various pre-stored audios;

[0076] A specific sound wave repelling module A92, operable to emit sound waves of a specific frequency;

[0077] Laser Repellent Module A93, operable to emit a laser beam;

[0078] Wherein, the radar monitoring module A96 further comprises:

[0079] a transmitter submodule A961 operable to transmit radar wave signals;

[0080] The antenna submodule A963 is operable to receive the radar wave signal of the transmitter submodule A961 and radiate it to the monitoring area, and at the same time receive the reflected radar wave signal;

[0081] The receiver submodule A962 is operable to receive the reflected radar wave signal from the antenna submodule A963 and convert it into an electrical signal; and,

[0082] The intelligent bird-repelling device also includes:

[0083] The signal processing and intelligent identification submodule A95 analyzes the received radar wave signal and analyzes the data to identify the type, quantity and behavior pattern of the bird;

[0084] The driving control submodule A94 is operable to: generate a mixed signal of a corresponding sound wave signal and an ultrasonic signal according to the identification result of the bird species by the signal processing and intelligent identification submodule A95; and control the specific sound wave driving module A92 and the sound driving module A91 to execute a corresponding driving strategy;

[0085] The controlling the specific sound wave driving module A92 and the sound driving module A91 to execute the corresponding driving strategy further includes: beamforming the mixed signal, and transmitting the sound wave and / or ultrasonic wave of the mixed signal to the monitoring area.

[0086] Among them, the connection relationship between modules and sub-modules and the signal flow

[0087] Radar monitoring module A96

[0088] Transmitter submodule A961: responsible for transmitting radar wave signals.

[0089] ●Antenna submodule A963: Receives radar wave signals from transmitter submodule A961 and radiates them to the monitoring area; at the same time, receives the reflected radar wave signals.

[0090] Receiver submodule A962: Receives the reflected radar wave signal from the antenna submodule A963 and converts it into an electrical signal.

[0091] Signal processing and intelligent identification submodule A95

[0092] ●Receive the electrical signal (ie radar echo signal) from the receiver submodule A962.

[0093] Analyze the received data to identify the species, numbers and behavior patterns of birds.

[0094] Drive control submodule A94

[0095] Generate a mixed signal of the corresponding sound wave signal and ultrasonic wave signal according to the information provided by the signal processing and intelligent identification submodule A95.

[0096] Control the specific sound wave expelling module A92 and the sound expelling module A91 to execute the corresponding expelling strategy.

[0097] Beamforming the mixed signal and transmitting acoustic and / or ultrasonic waves to the monitoring area.

[0098] Sound drive module A91

[0099] Play various pre-stored audios according to the instructions of the driving control submodule A94.

[0100] Specific sonic repellent module A92

[0101] ·Emit sound waves of specific frequency according to the instructions of the expulsion control submodule A94.

[0102] Laser drive module A93

[0103] Operable to emit a laser beam.

[0104] Application

[0105] In some embodiments of the present application, the intelligent bird-repellent device technology is applicable to power transmission and distribution scenarios, including transmission line towers, distribution line poles, substations and other scenarios.

[0106] Transmission line towers: Transmission line towers are an important part of power transmission. Birds may nest, roost or forage on the towers. These birds may threaten the safe operation of the transmission lines, causing accidents such as short circuits or grounding. The technical solution of this application is installed on the tower, which can form a perfect protection for the cross arm area of ​​the tower and ensure the normal operation of the transmission line.

[0107] Distribution line towers: Distribution line towers are usually located in urban or rural areas, and birds may nest or roost on the towers. These birds may have an impact on the switches, transformers and other equipment on the distribution lines and towers, such as blocking the equipment vents, contaminating the equipment surface or causing electric shock accidents. The technical solution of this application is installed on the tower, which can effectively protect the safe operation of the distribution line.

[0108] Substation scenario: Substation is a key link in the power transmission and distribution system. Birds entering the substation may cause equipment failure, electric shock or even fire. This intelligent bird repellent device can be used in the surrounding area or specific location of the substation to ensure the normal operation and safety of the substation equipment by repelling birds.

[0109] In some embodiments of the present application, the intelligent bird-repellent device technology plays an important role in the above scenarios, effectively reducing the interference of birds on power equipment and ensuring the stable operation of the power transmission and distribution system.

[0110] Function Configuration

[0111] Bird activity monitoring function

[0112] Use microwave radar detection technology and camera equipment to detect and record bird activities;

[0113] By monitoring and counting bird activities, we can obtain information on the species, number, behavior, etc. of invasive birds.

[0114] Sound frightening and repelling function

[0115] Play various pre-stored audios to simulate warning sounds in nature that can threaten birds, so as to scare them away;

[0116] The intelligent bird-repelling device has a variety of bird-repelling audios pre-stored in it. Users can replace the pre-stored audios from the audio library to reduce the adaptability of birds.

[0117] Specific sound wave repelling function

[0118] Emitting various specific sound waves and their combinations can make birds feel uncomfortable and drive them away;

[0119] Specific sound wave types and patterns can be adjusted according to the bird population conditions in different installation areas to enhance the repellent effect.

[0120] Laser drive function

[0121] Use lasers to emit laser beams to irritate and disturb birds, causing them to feel uncomfortable and fly away;

[0122] Image forensics function

[0123] Equipped with video equipment to record images of bird activities for the purpose of evaluating the effectiveness of bird repelling and bird damage.

[0124] Repelling effect evaluation function

[0125] Analyze and evaluate the effectiveness of repelling measures based on statistics such as the success rate of bird repelling, the frequency of bird invasions, and bird images;

[0126] According to the evaluation results, the expulsion strategy is adjusted and optimized to achieve better expulsion effect.

[0127] Eviction strategy maintenance function

[0128] Develop more targeted expulsion strategies based on the habits and activity patterns of birds in the product installation area;

[0129] The expulsion strategy can be adjusted dynamically based on real-time monitoring data and expulsion effect evaluation results to solve bird adaptability problems.

[0130] Bird situation and bird damage assessment function

[0131] Evaluate the potential damage caused by birds to transmission lines or power facilities by monitoring and analyzing bird activities and behaviors;

[0132] Based on the assessment results, formulate corresponding protection measures and repelling strategies to reduce the impact and damage of birds on power equipment.

[0133] Technical Effects

[0134] Some technical solutions of this application have the following technical effects:

[0135] Highly sensitive bird activity detection: real-time monitoring of bird invasion behavior to provide support for precise expulsion;

[0136] Multi-mode sound repelling technology: simulates the warning sounds of natural enemies and various frightening sounds to effectively scare birds away;

[0137] Random sound wave combination technology uses specific sound waves to create interference within the hearing range of birds, with random patterns, forcing birds to leave the protected area;

[0138] Safe laser repellent: using variable laser beams to stimulate and disturb birds, three-dimensional area protection to ensure no harm during the repelling process;

[0139] Intelligent expulsion strategy optimization: dynamically adjust the expulsion method based on statistical monitoring data to enhance the expulsion effect;

[0140] Video forensics: Recording videos and images of bird activities to provide a basis for subsequent analysis and evaluation;

[0141] Evaluation and analysis of expulsion effect: optimize expulsion strategy and improve effect through data statistics and evaluation;

[0142] Intelligent adjustment of driving mode: flexibly adjust the driving mode according to the surrounding environment;

[0143] High-efficiency and energy-saving design: adopts energy-saving technology and optimized power management technology to ensure long-term reliable operation;

[0144] Cloud-based unified management: Distributed installation and unified management improve efficiency and facilitate monitoring and maintenance;

[0145] User-friendly interface: Simple and intuitive operation interface, easy to set up and manage the functions and parameters of the intelligent bird repellent device.

[0146] Sound repellent function

[0147] The principle of repelling birds by playing audio such as the calls of natural enemies and gunshots is based on the response of birds to potential threats. Playing audio such as the calls of natural enemies and gunshots can simulate potential threat signals, making birds nervous and alert. These sounds may imitate the calls of predators or other potentially dangerous sounds, causing birds to have a self-protection instinct, making them feel dangerous and stay away from the area where the sound is played. Birds will gradually adapt to and ignore common sounds, but for sounds that simulate threats, they may remain sensitive and continue to react to them. Playing audio such as the calls of natural enemies and gunshots can break the adaptability of birds and make them unable to ignore the sounds, thereby achieving the purpose of repelling birds. Playing sounds associated with danger or threats, such as gunshots, may trigger fear memories in birds. Birds will form negative memories associated with specific sounds when facing danger, and when they hear similar sounds again, they will produce fear and avoidance behaviors.

[0148] Specific sound wave repelling function

[0149] The principle of specific sound wave bird repellent is based on the sensitivity of birds to sound and their response to specific sound waves. Different types of birds have different perception ranges of sound frequencies. By emitting sound waves of specific frequencies, birds' normal auditory perception can be disturbed, causing them to feel uncomfortable or confused. This interference can give birds an unpleasant experience and force them to stay away from the source of the sound waves. Certain sound waves of specific frequencies can simulate birds' natural enemies or other threat signals, triggering birds' defensive instincts. These sound waves may imitate the calls of predators or other warning signals of other birds, alerting birds and prompting them to leave potentially dangerous areas. For birds that are exposed to a specific frequency sound wave environment for a long time, they may gradually develop habitual avoidance behavior of the sound wave. Through continuous sound wave repellent, birds may learn to associate the area with unpleasant sound waves and naturally avoid entering.

[0150] Laser drive function

[0151] The impact of light waves on birds

[0152] Eyes are the most sensitive and most dependent organs of birds. The proportion of bird eyes to the entire head is much higher than that of other vertebrates, usually accounting for half of the head or even more. Bird eyes are distributed on both sides of the head and have an extremely wide field of vision. Bird eyes have large pupils, which allows birds' eyes to receive more light under the same lighting conditions, especially at night, and enhances the ability of birds to move at night. Light waves of different wavelengths have different stimulations on cone cells. Studies have shown that birds are most sensitive to green, followed by red. The most sensitive wavelength range for most birds is 500-570nm. This shows that green and red light in this wavelength range can be used to stimulate bird vision to achieve the effect of bird repelling. Light sources of different colors have a certain impact on bird migration routes. Studies have shown that white and red light sources have a significant impact on bird migration, leading to bird disorientation, and birds are likely to lose their migration direction; blue light sources have no effect at all; green light sources have a very weak effect. From an ecological perspective, intelligent bird repellent devices should be bird-friendly, so red light is not suitable for bird repelling, and green light is the best choice.

[0153] Laser bird repellent

[0154] Laser bird repellent is a method of using laser beams to repel birds. The principle is based on the visual sensitivity of birds to laser beams and their adaptability to light. The bright spots and moving light produced by laser beams can cause visual sensitivity in birds. Birds will be alert to the sudden appearance of spots and lights because they may regard them as potential threats or danger signals. This visual disturbance can make birds feel uncomfortable and force them to leave the area illuminated by the laser beam. The laser beam can simulate the movement of predators by moving or sweeping quickly. Birds are generally sensitive to movement and will regard fast-moving objects as potential threats. By manipulating the movement of the laser beam, the behavior of predators can be simulated, causing birds to be alert and prompting them to stay away from the area illuminated by the laser beam. Birds are highly adaptable to common light and light in the environment, but they may not be able to adapt to and ignore the strong and concentrated light such as laser beams. The bright and focused nature of the laser beam can attract the attention of birds and limit their stay in the area illuminated by the laser beam.

[0155] Deportation completion judgment time

[0156] When the intelligent bird repellent device turns on the expulsion module (such as sound, laser, etc.) and executes an expulsion time, it will determine whether the repelled birds have left the target monitoring area. The expulsion completion judgment time is used to control the strictness of the judgment. After the expulsion module is executed, the microwave radar does not detect signs of bird activity within the expulsion completion judgment time, which means that the birds have been completely driven away and the expulsion is effective. If the microwave radar detects signs of bird activity again within the time, it means that the bird has not left or has returned quickly after leaving, which means that the expulsion is invalid and the expulsion module needs to be turned on again for expulsion. When the expulsion module is turned on three times in a row, the microwave radar can still detect bird activity within the expulsion completion judgment time, which means that the current expulsion strategy can no longer drive away birds and the bird repellent fails. The longer the expulsion completion judgment time, the stricter the intelligent bird repellent device's judgment on the expulsion completion. Setting a reasonable expulsion completion judgment time can ensure the effect of a single bird repellent. The default time for judging whether the eviction is complete is 10 seconds, and the configurable range is 5 to 60 seconds, as shown in Table 1-1.

[0157] Table 1-1 Device time parameter configuration

[0158]

[0159] Eviction mode time period configuration

[0160] The intelligent bird repellent device has three bird repellent modes: laser, sonic wave, and sound. Different modes have different degrees of impact on the surrounding environment. Users can adjust the repellent mode of the intelligent bird repellent device at different time periods according to factors such as the severity of bird damage near the installation location of the intelligent bird repellent device and the human activities nearby. The current system supports the configuration of up to five time periods. In each time period, users can choose whether to turn on laser bird repellent, sonic wave bird repellent, or sound bird repellent. The system default time period configuration and its repellent mode configuration are shown in Table 1-2. For the laser bird repellent mode, if there is a building diagonally above the installation location of the intelligent bird repellent device, it is recommended not to turn it on to prevent it from damaging the human vision in the building. For the sound bird repellent mode, if there are residents living near the installation location of the intelligent bird repellent device, it is recommended not to turn it on during the residents' rest time because some audio has a large sound intensity and penetration. For the sound wave bird repellent mode, its sound intensity, penetration, and degree of fright are weaker than those of the sound bird repellent mode, so it can be turned on during the rest time of nearby residents.

[0161] Table 1-2 Time period configuration of eviction method

[0162] Time period Laser bird repellent Sound bird repellent Sonic bird repellent 00:00:00～07:00:00 Enable Disable Enable 07:00:00～12:30:00 Enable Enable Disable 12:30:00～14:30:00 Enable Disable Enable 14:30:00～18:00:00 Enable Enable Disable 18:00:00～23:59:59 Enable Disable Enable

[0163] Pre-stored audio file configuration

[0164] The smart bird repellent device can pre-store up to 10 audio files that can scare birds. When the smart bird repellent device starts the sound bird repellent mode, it will randomly select one of the pre-stored audio files to play. The cloud platform has a bird repellent audio file library, which stores audio files that can scare birds. Users can select 1 to 10 audio files as pre-stored audio files for the smart bird repellent device. When the device updates parameters, the modified pre-stored audio files will be downloaded from the cloud audio file library to the smart bird repellent device. The smart bird repellent device has 10 pre-stored audio files when it leaves the factory. The information is shown in Table 1-3, including alarms, explosions, eagle calls, gunshots, thunder and other sounds that can alert target birds.

[0165] Table 1-3 Pre-stored audio file configuration of intelligent bird repellent device

[0166] Sound Type Audio file name File size MD5 value Alarm alert_1.wav 133.58KB f20c01c2c7d2e682db599be4998d3ff3 Alarm alert_2.wav 295.51KB 9cea4762f8c798cd2f4ffdd46c3002db Alarm alert_3.wav 222.38KB c762008c42a925a1fd5b4de9f183e2ae Explosion boom_1.wav 273.73KB 0a1906f2d113d81bfa3a4e22bd2ba758 Eagle cry eagle_1.wav 137.54KB e0f5e4a9acbc6a7d4359da4400a842e8 Eagle cry eagle_2.wav 164.08KB 1dc3867415337c119d2fd9be5f53ea6b Gunshots gun_1.wav 204.93KB 9625992ae22ab70282741e9be967efd1 Screech scream_1.wav 297.01KB 803dc0a230d53f5937936832ebf9bf51 Thunder thunder_1.wav 283.20KB b03a4ad0604c18aa0d1ec3d0b4416d99 Train whistle trian_1.wav 254.60KB b076be03004cda3831994e4ff8b0a7f3

[0167] Detection sensitivity

[0168] The intelligent bird repellent device uses microwave radar to achieve real-time monitoring of invading birds. By configuring the sensitivity of microwave radar to detect invading birds, the sensitivity of the intelligent bird repellent device to bird invasion behavior can be controlled. The higher the detection sensitivity of the configured microwave radar, the more sensitive it is to bird invasion activities, and the more susceptible it is to environmental influences such as trees and vibrations. The specific sensitivity setting should take into account the specific installation location of the device and its surrounding conditions. If there are interferences such as trees near the installation location of the device, the sensitivity should be appropriately reduced to avoid false detection. If the area near the installation location is relatively open, the sensitivity can be appropriately increased to improve the detection capability of invading birds. The system default monitoring sensitivity is 10, and the setting range is 3 to 31. The smaller the value, the higher the sensitivity.

[0169] In order to illustrate the application effects of some technical solutions of the present application and compare them with the prior art, several specific numerical calculation cases are constructed here. A set of fictitious bird image parameters are used, and these parameters are combined with the steps and formulas of some embodiments of the present application to demonstrate the actual application effects of these steps and methods.

[0170] The main system interface of the host computer of the intelligent bird-repelling device is mainly divided into three parts: Figure 4 As shown in the figure, the left side is the page navigation bar, the middle main part is the overview of the smart bird repellent system information, and the right side is the calendar and bird damage events. The main interface of the system is shown in the figure.

[0171] The main function of the page navigation bar is to switch to the management interface of lines, poles and equipment for fine-grained management.

[0172] After clicking on Line Management, the line list will drop down. After clicking on the line list, you will enter the line management interface.

[0173] After clicking on pole management, the pole list will drop down. After clicking on the pole list, you will enter the pole management interface.

[0174] After clicking on Device Management, the device list, expulsion record, expulsion strategy, and device configuration list will be pulled down. Clicking on the corresponding option will enter the corresponding configuration interface.

[0175] The smart bird repellent system information overview includes the number of lines, indicating how many lines are currently equipped with smart bird repellent devices; the number of poles, indicating how many poles are currently equipped with smart bird repellent devices; the number of devices, indicating how many smart bird repellent devices are currently installed in the power system; a line graph of the repellent records for the past seven days, indicating the number of bird invasions per day in the past seven days; a circular graph of the repellent success rate, indicating the proportion of successful repellents (blue) and the proportion of failed repellents (green) in all repellent times; and the proportion of repellent time, indicating the percentage of bird invasions in different time periods.

[0176] The calendar and bird damage events include the current calendar in the upper right corner and the real-time events in the lower right corner. The real-time events scroll to display the specific year, month, day, and on which line and tower the bird invasion occurred.

[0177] Specific application scenarios

[0178] Assume that this application is using an intelligent bird-repellent device to monitor and identify bird species. The device is installed on a power transmission line tower and needs to identify two types of birds: pigeons and sparrows.

[0179] like Figure 1 As shown, in the driving away strategy generation method of the present application, a group of bird image data and feature data therein are collected through step S1, including body size, flight speed, flight altitude, sound frequency, etc. These data will be used to calculate the probability of occurrence of each bird P(C|X) and minimize the negative log-likelihood function L(α, β, γ).

[0180] Assumed bird image parameters

[0181] Assume that the present application has collected a set of characteristic data of birds, and in step S2, these characteristic data are converted into a characteristic vector X:

[0182] 1. Eigenvector X:

[0183] οBody size s=15cm

[0184] οFlight speed v = 10m / s

[0185] οFlight altitude h=5m

[0186] οSound frequency f=1500Hz

[0187] 2. Model parameters:

[0188] οα=0.5

[0189] οβ=0.2

[0190] ογ=0.1

[0191] 3. Number of known bird species:

[0192] οK=2(pigeons and sparrows)

[0193] Norm of eigenvector

[0194] First, this application needs to calculate the norm of the feature vector X ||X||:

[0195]

[0196] The probability of each bird appearing

[0197] Next, in step S3, the probability of each bird appearing P(C|X) is calculated. Assume that there are two known bird species K=2 in this application.

[0198]

[0199] Since exp(-150.025) is very close to zero, this application can simplify the above calculation:

[0200]

[0201] This means that the probability of each bird appearing is about 50%. In practical applications, the parameters α, β, and γ may be different for each bird, so the probability of each bird appearing will also be different.

[0202] In step S4, the maximum likelihood estimation method is used to estimate the model parameters α, β and γ. Assume that there is a set of training data: D = {(X1, C1), (X2, C2), ..., (X m ,C m )}, where Xi is the feature vector of the i-th sample and Ci is the corresponding bird species.

[0203] The goal of maximum likelihood estimation is to minimize the negative log-likelihood function L(α,β,γ):

[0204]

[0205] In step S5, according to the identification result of the bird species, the corresponding driving strategy is executed. Of course, in the device, there are pre-stored multiple driving strategies corresponding to the bird species. For example, for medium and large-sized pigeons or some birds with strong anti-scare ability, two or more mixed signals of specific sound waves, sound signals, laser signals, etc., or multiple signal polling methods can be used to drive the birds away. For small-sized birds such as sparrows, a single signal can be used to drive them away.

[0206] In the prior art, the comparison is based on a simple threshold. For example, if the flying speed is greater than a certain threshold, it is considered to be a specific bird species. The accuracy of this recognition is low and is easily affected by external factors.

[0207] Specific calculation process:

[0208] Assume that this application has a simple threshold comparison model for identifying pigeons and sparrows.

[0209] If the flying speed v>12v>12m / s, it is considered to be a pigeon; otherwise it is considered to be a sparrow.

[0210] In this case, since the flying speed is 10m / s, according to the simple threshold comparison model, the present application will identify the bird as a sparrow.

[0211] However, this simple approach ignores the importance of other characteristics, such as body size, flight altitude, and sound frequency.

[0212] Optionally, in another embodiment of the present application, the following mathematical model may be used to perform pattern recognition of bird categories:

[0213]

[0214] Where I represents the input bird image, K is the number of feature categories, J is the number of feature subcategories, w is the weight vector, b is the model parameter vector, and R(I) and G(I) are the specific feature sets extracted from image I.

[0215] Further optionally, R(I) and G(I) include a combination of color histograms, texture features, and shape descriptors.

[0216] Further optionally, the model parameter b is optimized by maximum likelihood estimation.

[0217] Further optionally, the weight vector w is adjusted by a gradient ascent method.

[0218] Optionally, a normalization step is further included:

[0219]

[0220] Where μ(I) and σ(I) are the mean and standard deviation of image I, respectively.

[0221] Further optionally, the torso strategy generation method of this embodiment also includes a feature space transformation step: T(I)=I·A+B, wherein A and B are a transformation matrix and a vector.

[0222] Further optionally, the model also includes a regularization term to prevent overfitting: L(w,b)=-log(F)+λ·R(w,b)

[0223] Among them, λ is the regularization parameter and R(w,b) is the regularization function.

[0224] Further optionally, the regularization function R(w,b) is an L2 norm or an L1 norm.

[0225] Further optionally, the model parameter b is optimized by the following iterative process:

[0226]

[0227] Where η is the learning rate, is the loss function L with respect to parameter b k The partial derivative of .

[0228] Further optionally, the loss function L includes cross entropy loss and / or mean square error loss.

[0229] Further optionally, the step of performing pattern recognition of bird categories based on the mathematical model F(I, w, b) further includes the following sub-steps:

[0230] Step 1: Image acquisition and preprocessing

[0231] Use camera equipment to capture images of birds at electric power transmission and distribution facilities.

[0232] Preprocess the captured images, including denoising, adjusting contrast, cropping unnecessary edges, and normalizing the image size.

[0233] Step 2: Feature Extraction

[0234] Extract R(I) and G(I) feature sets from the preprocessed image I, which may include color histograms, texture features, shape descriptors, etc.

[0235] Step 3: Feature enhancement and selection

[0236] Apply an exponential function to the feature set R(I) to enhance important features and suppress unimportant features.

[0237] Transform another feature set G(I) using a logarithmic function to select the most representative features.

[0238] Step 4: Weight adjustment and model training

[0239] Using the training dataset, adjust the weight vector w via gradient ascent to maximize the model’s prediction accuracy.

[0240] Tune the model parameters b using maximum likelihood estimation or other optimization algorithms.

[0241] Step 5: Regularization

[0242] To prevent the model from overfitting, a regularization term R(w,b) is introduced into the loss function L, where λ is the regularization parameter.

[0243] Step 6: Loss function calculation and optimization

[0244] Calculate the loss function L(w,b), which may include cross entropy loss and mean squared error loss.

[0245] Iteratively update the model parameters b and weights w using an optimization algorithm (e.g. gradient descent) to minimize the loss function.

[0246] Step 7: Model Evaluation and Validation

[0247] Use the validation set to evaluate the performance of the model and adjust the model parameters to optimize the recognition accuracy.

[0248] Perform cross validation to ensure the generalization ability of the model.

[0249] Step 8: Model deployment and real-time recognition

[0250] Deploy the trained model to the smart bird repellent device.

[0251] Realize real-time image capture, processing and automatic identification of bird species.

[0252] Step 9: Feedback mechanism and model iteration

[0253] Collect feedback information based on the repelling effect of the intelligent bird repellent device.

[0254] Use feedback information to iteratively optimize the model, adjusting feature extraction, weights, and parameters to improve recognition accuracy.

[0255] Step 10: Environmental Factor Integration

[0256] Integrate environmental factors such as weather, time of day, etc. that may affect bird behavior and image characteristics.

[0257] Adjust the model to account for these environmental factors and improve the robustness of the model under different environmental conditions.

[0258] In order to specifically demonstrate the technical effect of mathematical models such as "F(I, w, b)" in the pattern recognition step of bird categories, numerical calculations and comparisons are performed here through a hypothetical application scenario. In this scenario, this embodiment assumes that images of two types of birds have been captured: sparrows and eagles. This embodiment will define some specific parameters for these images and show how to substitute them into the model calculation formula.

[0259] Assumed bird image parameters

[0260] Sparrow image parameters:

[0261] οArea Asparrow=80 pixels^2

[0262] οLongest dimension Lsparrow=45 pixels

[0263] ο Texture feature vector Tsparrow = [10, 20, 30, 40, 50] (hypothetical value)

[0264] οColor feature vector Csparrow = [0.3, 0.4, 0.3] (normalized value in RGB color space)

[0265] Eagle image parameters:

[0266] οArea Aeagle=600 pixels^2

[0267] οLongest size League=120 pixels

[0268] οTexture feature vector Teagle = [70, 80, 90, 100, 110] (hypothetical value)

[0269] οColor feature vector Ceagle = [0.2, 0.3, 0.5] (normalized value in RGB color space)

[0270] Calculation process of existing technology:

[0271] Existing techniques might use simple area and longest dimension classification:

[0272] · Sparrow classification conditions: A < 100 pixels^2 and L < 50 pixels

[0273] Eagle classification conditions: A>500 pixels^2 and L>100 pixels

[0274] Based on the above conditions, the images of sparrows and eagles will be classified as follows:

[0275] The sparrow image is correctly classified as a sparrow.

[0276] Eagle images are correctly classified as eagles.

[0277] New technology calculation process:

[0278] The new technique uses the “F(I,w,b)” model for classification, and the steps are as follows:

[0279] 1. Feature extraction: Extract texture and color features from images.

[0280] 2. Model parameters: Assume that the weight w and parameter b are obtained through training, for example:

[0281] οw=[1,1,1,1,1](texture and color features have the same weight)

[0282] οb = [0.1, 0.1, 0.1, 0.1, 0.1] (same bias for texture and color features)

[0283] 3. Model calculation:

[0284] Model output for sparrow:

[0285]

[0286] Model output from Eagle:

[0287]

[0288] 4. Classification decision: Based on the value output by the model, use a classification algorithm to determine the type of bird.

[0289] Numerical calculation examples

[0290] Assume that the following simplified calculation is performed (for the sake of example, here, only one feature is calculated):

[0291] Model output for Sparrow (considering only one feature):

[0292] F sparrow =e (0.1.10) ·log(e (0 . 1.0.3) +e (0.1.0.4) +e (0.1.0.3) )

[0293] Model output of Eagle (considering only one feature):

[0294] F eagle =e (0.1·70) ·log(e (0.1·0.2) +e (0.1·0.3) +e (0.1·0.5) )

[0295] Comparison results

[0296] Accuracy: By combining texture and color features, the new technology can more accurately distinguish between birds, especially when morphological characteristics are similar.

[0297] Adaptability to complex environments: New technologies can better adapt to complex environments, such as different lighting conditions or background interference.

[0298] Differentiated processing: The new technology can adjust model parameters according to the characteristics of different birds to achieve differentiated bird-repelling strategies.

[0299] For a specific application scenario

[0300] Assume that this embodiment is using an intelligent bird repellent device to monitor and identify bird species. The device is installed on a power transmission line tower and needs to identify two types of birds: pigeons and sparrows. This embodiment will use a set of fictitious bird image parameters and substitute these parameters into the previously constructed mathematical model to demonstrate the actual application effect of the mathematical model expressed by the formula.

[0301] Assumed bird parameters

[0302] Assume that this embodiment collects characteristic data of a group of birds:

[0303] 1. Eigenvector X:

[0304] οBody size s=15cm

[0305] οFlight speed v = 10m / s

[0306] οFlight altitude h=5m

[0307] οSound frequency f=1500Hz

[0308] 2. Sound wave signal parameters:

[0309] ο Initial amplitude A0 = 1

[0310] οAttenuation coefficient α=0.1

[0311] Frequency fa = 1000 Hz

[0312] οPhase φa=π / 4

[0313] 3. Ultrasonic signal parameters:

[0314] ο Initial amplitude U0 = 1

[0315] Attenuation coefficient β = 0.2

[0316] Frequency f u =40000Hz

[0317] οPhase φ u =π / 4

[0318] 4. Mixed signal parameters:

[0319] οTime t=0.1s

[0320] οBird species k = 1 (assuming it is a pigeon)

[0321] οParameter δ k =2 and γ k =1

[0322] Specific numerical calculation of mixed signal S(t,k)

[0323] Now, this embodiment will use the above parameters to calculate the specific value of the mixed signal S(t,k).

[0324] Acoustic signal A(t)

[0325]

[0326] Ultrasonic signal U(t)

[0327]

[0328] The weight function λ(t,k) of the mixed signal S(t,k)

[0329]

[0330] Mixed signal S(t,k)

[0331] S(t,k)=λ(t,k)·A(t)+(1-λ(t,k))·U(t)=0.145·0.700133903379443+(1-0.145)·0.693178999279526·0.70 0133903379443+0.855·0.693178999279526≈0.101523141928713+0.592649824383097=0.694172966311810

[0332] In this example, the value of the mixed signal S(t,k) is calculated at a specific time point t=0.1s and bird species k=1 (assuming it is a pigeon).

[0333] In contrast, in the prior art, fixed sound and ultrasonic signals are used and cannot be adjusted according to the bird species. It is not possible to perform adaptive signal adjustment for different bird species. In the calculation process of the prior art, the mixed signal is usually fixed and does not change with time and bird species. Therefore, in the prior art, the value of the mixed signal is fixed throughout the time period.

[0334] In summary, through the above calculations, it can be seen that the present technology can be adaptively adjusted for different bird species through the mathematical model in this embodiment, and has higher flexibility than the prior art. In practical applications, these mathematical models can be further optimized to improve recognition accuracy, and model parameters can be adjusted through more training data to obtain better performance.

[0335] Comparison between some embodiments of the present application and the prior art

[0336] In the embodiments of the present application, a complex mathematical model is used to comprehensively evaluate the feature vector in combination with logarithmic and exponential calculations, taking into account the probability distribution of each bird species, thereby improving the accuracy and robustness of recognition.

[0337] In the specific calculation process, through the above complex mathematical model, the present application can more accurately calculate the probability of each bird appearing. In this case, the probability of each bird appearing is about 50%, which shows that the model of the present application can more comprehensively consider all features and can more accurately identify bird species. By using the maximum likelihood estimation method, the present application can further optimize the model parameters to improve the recognition accuracy.

[0338] Through the above calculations, it can be seen that this technology can more accurately identify bird species through complex mathematical models, and has higher accuracy and robustness than existing technologies. In practical applications, these mathematical models can be further optimized to improve recognition accuracy, and model parameters can be adjusted through more training data to obtain better performance.

[0339] In another embodiment, a mathematical model can also be constructed to describe the beamforming of ultrasonic and acoustic signals. The following is an exemplary mathematical model:

[0340] Mathematical model of hybrid beamforming

[0341] Assume that there are two types of sound wave signals in this embodiment: Ultrasonic Waves (UC) and Acoustic Waves (AW). This embodiment hopes to mix and shape these two signals to optimize the bird-repelling effect.

[0342] U(t): The instantaneous amplitude of the ultrasonic signal at time tt.

[0343] A(t): The instantaneous amplitude of the acoustic signal at time tt.

[0344] fu: frequency of ultrasonic signal.

[0345] fa: frequency of the sound wave signal.

[0346] ·θ: beam shaping angle.

[0347] d: beamforming spread distance.

[0348] ·ku and ka: the wave numbers of ultrasonic wave and sound wave respectively (k=2πλk=λ2π).

[0349] ·α and β: are the attenuation coefficients of ultrasonic and sound waves respectively.

[0350] g(θ): beamforming function, which describes the distribution of the beam with angle.

[0351] Beamforming function

[0352]

[0353] Mixed signal amplitude

[0354] M(t)=U(t)·exp(-α·k u ·d)+A(t)·exp(-β·k a d)

[0355] Mixed Signal Beamforming

[0356] B(t,θ)=M(t)·g(θ)·cos(2πf u t+φ u )+M(t)·(1-g(θ))·cos(2πf a t+φ a )

[0357] in:

[0358] ·φu and φa are the phases of ultrasonic and acoustic wave signals respectively.

[0359] d0 is the reference distance used to normalize the diffusion distance d.

[0360] The formula of mathematical model is:

[0361]

[0362] The formula combines the amplitude, frequency, attenuation, beamforming angle and diffusion distance of ultrasonic and acoustic signals to form a beamforming hybrid model, which can be used to predict the intensity of the acoustic wave field at a specific time and angle.

[0363] The meanings of the variables are as follows:

[0364] B(t,θ): The amplitude of the beamformed mixed signal at time t and angle θ.

[0365] U(t): The instantaneous amplitude of the ultrasonic signal at time t.

[0366] A(t): The instantaneous amplitude of the acoustic signal at time t.

[0367] exp: exponential function, indicating growth or decay.

[0368] α: The attenuation coefficient of the ultrasonic signal, which describes the attenuation of the signal with distance.

[0369] ku: The wave number of ultrasound, inversely proportional to the wavelength.

[0370] d: The beamforming diffusion distance, which affects the beam width and intensity distribution.

[0371] β: The attenuation coefficient of the sound wave signal, which describes the attenuation of the signal with distance.

[0372] ka: The wave number of the sound wave, which is inversely proportional to the wavelength.

[0373] fu: frequency of ultrasonic signal.

[0374] φu: The phase of the ultrasonic signal, describing the starting point of the signal.

[0375] fa: frequency of the sound wave signal.

[0376] φa: The phase of the sound wave signal, describing the starting point of the signal.

[0377] g(θ): beamforming function, describing the distribution of the beam with angle θθ.

[0378] d0: Reference distance, used to normalize the diffusion distance d, usually used in beamforming calculations.

[0379] sin(θ) and cos(θ): Sine and cosine functions used to describe the variation of beamforming with angle.

[0380] · The denominator of the beamforming function, used to adjust the beam width.

[0381] This formula is a complex mathematical model that describes the mixing and shaping effects of ultrasonic and acoustic signals at specific times and angles. It takes into account factors such as signal amplitude, frequency, attenuation, beamforming, and diffusion distance to predict the intensity distribution of the acoustic wave field.

[0382] In order to build a hybrid shaping mathematical model that can automatically adjust the ultrasonic and acoustic signal components, we need to consider the differences in sensitivity of different birds to sound frequencies. The following is an example of a complex mathematical model that combines logarithmic and exponential functions and can automatically adjust the signal components according to the sensitivity of different birds:

[0383] Hybrid Shape Mathematical Model

[0384] Let S(t) be the mixed sound wave signal at time t, U(t) be the ultrasonic signal, and A(t) be the sound wave signal. u,t and S a,t The ultrasonic and acoustic signal components are adjusted for specific birds. u,b and R a,b is the sensitivity response function of different bird species to ultrasonic and sound waves.

[0385] Sensitivity response function

[0386] For ultrasonic and sound waves, the sensitivity response function Ru,b and R a,b Can be expressed as:

[0387]

[0388] in:

[0389] γ is the control parameter of the logarithmic sensitivity response.

[0390] ·f u u is the frequency of the ultrasonic wave.

[0391] ·f u,ref is the reference ultrasonic frequency.

[0392] ·f a is the frequency of the sound wave.

[0393] ·f a,opt,b It is the sound wave frequency that birds are most sensitive to.

[0394] ·σ fa,b is the standard deviation of the distribution of sound frequency sensitivity for bird b.

[0395] Adjusted signal components

[0396] S u，t =U(t)·R u，b (f u )

[0397] S a，t =A(t)·R a，b (f a )

[0398] Mixed Signal

[0399] S(t)=w u，b ·S u,t +w a，b ·S a，t

[0400] in:

[0401] ·w u,b and w a,b is the weight coefficient automatically adjusted according to the sensitivity response of bird b.

[0402] Automatically adjust weight coefficients

[0403]

[0404] Single formula expression

[0405] The single formula expression of the mixed signal S(t) can be written as:

[0406] S(t)=ω u,b ·S u,t +ω a,b ·S a,t

[0407] Further substituting the weight coefficient and sensitivity response function, we get:

[0408]

[0409] This formula combines logarithmic and exponential functions to automatically adjust the components of ultrasonic and sonic signals according to the sensitivity of different birds. The model can dynamically adjust the weight of the signal according to the specific sensitivity curve of the bird species bb to achieve the best bird repellent effect.

[0410] In one possible design, Figure 3 The data annotation apparatus of the illustrated embodiment may be implemented as a computing device, which may include a storage component 71 and a processing component 72 as shown in the figure;

[0411] The storage component 71 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 72 .

[0412] The processing component 72 is used to: obtain a query object provided by a user; based on the query object, query multiple data to be labeled that match the query object in a pre-established data set to be labeled, and generate multiple initial segmentation results corresponding to the data to be labeled; perform fine segmentation on the initial segmentation results corresponding to the multiple data to be labeled to generate multiple target segmentation results of the data to be labeled; use a pre-created labeling tool to perform labeling operations on the target segmentation results of the selected data to be labeled, and apply the labeling operations in batches to the remaining multiple target segmentation results of the data to be labeled.

[0413] The processing component 72 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0414] The storage component 71 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0415] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0416] The display component 73 can be an electroluminescent (EL) element, a liquid crystal display or a micro display with a similar structure, or a retinal direct display or a similar laser scanning display.

[0417] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0418] The communication components are configured to facilitate wired or wireless communications between the computing device and other devices.

[0419] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0420] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The data labeling method of the illustrated embodiment.

[0421] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0422] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0423] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

Claims

1. An intelligent bird-repelling device for power transmission and distribution facilities, characterized in that: include: A radar monitoring module that can be operated to monitor bird activities in real time and collect bird data; The sound expulsion module can be operated to play various pre-stored audios; A specific sound wave repelling module operable to emit sound waves of a specific frequency; Wherein, the radar monitoring module further comprises: a transmitter submodule operable to transmit a radar wave signal; and, The intelligent bird-repelling device also includes: The signal processing and intelligent identification submodule analyzes the received radar wave signals and analyzes the data to identify the types, numbers and behavior patterns of birds; The driving control submodule is operable to: generate a mixed signal of a corresponding sound wave signal and an ultrasonic signal according to the identification result of the bird species by the signal processing and intelligent identification submodule; and control the specific sound wave driving module and the sound driving module to execute a corresponding driving strategy; The controlling the specific sound wave driving module and the sound driving module to execute the corresponding driving strategy further includes: beamforming the mixed signal, and transmitting the sound wave and / or ultrasonic wave of the mixed signal to the monitoring area.

2. The intelligent bird-repelling device according to claim 1, characterized in that: Also includes: A camera module operable to capture image feature data of birds; a laser repellent module operable to emit a laser beam; an antenna submodule operable to receive the radar wave signal from the transmitter submodule and radiate it to the monitoring area, and simultaneously receive the reflected radar wave signal; a receiver submodule operable to receive the reflected radar wave signal transmitted from the antenna submodule and convert it into an electrical signal; The signal processing and intelligent identification submodule can be operated to perform the following steps: Obtain characteristic data of birds, including body size, flight speed, flight altitude, and sound frequency; Convert the feature data into a feature vector X = {x1, x2, ..., x n }; According to the relationship between the feature vector X and the bird species C, the probability P(C|X) of each bird appearing is calculated; Calculate the negative log-likelihood function L(α,β,γ) to estimate the model parameters α, β, and γ: Calculate an estimate of P(C|X) based on the model parameters α, β and γ; The type of bird is determined based on the estimated value of the formula P(C|X).

3. The intelligent bird-repelling device according to claim 2, characterized in that: The driving control submodule is operable to: perform beamforming according to the mixed signal S(t, k) corresponding to the type of the bird, and transmit the sound waves of the mixed signal to the monitoring area through the sound driving module and the specific sound wave driving module.

4. The intelligent bird-repelling device according to claim 3, wherein: The driving control submodule is operable to generate a mixed signal of the acoustic wave signal A(t) and the ultrasonic wave signal U(t), and the mixed signal S(t, k) is defined as: Among them, δ k and γ k is a parameter adjusted according to the bird species k, A0 and U0 are the initial amplitudes of the acoustic and ultrasonic signals, α and β are the attenuation coefficients of the acoustic and ultrasonic signals, respectively, and f a and f u are the frequencies of the acoustic and ultrasonic signals, respectively, a and φ u are the phases of the acoustic and ultrasonic signals, log δk is δ k The logarithmic function with base is , and e is the natural exponential function.

5. A method for generating a driving-away strategy for an intelligent bird-repelling device for power transmission and distribution facilities, comprising: Obtain characteristic data of birds, including body size, flight speed, flight altitude, and sound frequency; Convert the feature data into a feature vector X = {x1, x2, ..., x n }; According to the relationship between the feature vector X and the bird species C, the probability of each bird appearing P(C|X) is calculated. Calculating a negative log-likelihood function L(α, β, γ) to estimate the model parameters α, β and γ; Calculate an estimate of P(C|X) based on the model parameters α, β and γ; Determine the type of bird based on the estimated value of the formula P(C|X); According to the identification result of the bird species, a corresponding driving strategy is executed.

6. The method for generating a driving strategy according to claim 5, characterized in that: Also includes the steps: Generate a mixed signal of the acoustic wave signal A(t) and the ultrasonic signal U(t), and the mixed signal S(t,k) is defined as: where δ k and γ k is a parameter adjusted according to the bird species k, A0 and U0 are the initial amplitudes of the acoustic and ultrasonic signals, α and β are the attenuation coefficients of the acoustic and ultrasonic signals, respectively, and f a and f u are the frequencies of the acoustic and ultrasonic signals, respectively, a and φ u are the phases of the acoustic and ultrasonic signals, log δk is δ k The logarithmic function with base is , and e is the natural exponential function.

7. The method for generating a driving strategy according to claim 6, characterized in that: The ultrasonic signal U(t) is: U(t)=U0·e -βt ·sin(2πf u t+φ u ) The mixed signal is defined as: S(t,k)=λ(t,k)·A(t)+(1-λ(t,k))·U(t) in, U0 is the initial amplitude of the ultrasonic signal; β is the attenuation coefficient of the ultrasonic signal; f u is the frequency of the ultrasonic signal; φ u is the phase of the ultrasonic signal; λ(t,k) is the weight of the acoustic signal in the mixed signal, 0≤λ(t,k)≤1, depending on the time t and the bird species k; Beamforming is performed on the mixed signal corresponding to the type of the bird, and the mixed signal after beamforming is transmitted to a monitoring area.

8. The method for generating a driving strategy according to claim 6, characterized in that: The step of executing a corresponding driving strategy according to the identification result of the bird species by the signal processing and intelligent identification submodule further includes: Provides a series of predetermined center frequencies f1, f2, ..., f N ; Generate a frequency hopping sequence H(t) according to time t, where the frequency hopping sequence is defined as: The instantaneous frequency of the carrier signal is adjusted according to the frequency hopping sequence H(t) and the instantaneous frequency offset Δf(t), where the instantaneous frequency offset is defined as: Δf=A f ·e -γt ·sin(ωt+φ) 9. The method for generating a driving strategy according to claim 8, characterized in that: The method further comprises the step of calculating the instantaneous frequency f(t) of the microwave signal by the following formula: Among them, α and β are the parameters for adjusting the frequency hopping sequence, A f is the amplitude of the instantaneous frequency offset, γ is the attenuation coefficient of the frequency offset, ω is the angular frequency of the frequency offset, φ is the initial phase of the frequency offset, It means round down, logβ is the logarithmic function with β as the base, and e is the natural exponential function; as well as According to the frequency hopping sequence and / or the carrier signal, beamforming is performed on the mixed signal corresponding to the type of the bird, and the sound wave and / or ultrasonic wave of the mixed signal is transmitted to the monitoring area.

10. An electronic device, characterized in that: It includes a processor, which is operable to execute the method for generating a repelling strategy in an intelligent bird repelling device for electric power transmission and distribution facilities as described in any one of claims 5-9.

Images