A Bird Intelligent Matching and Repelling System and Method Based on Voiceprint Database

Abstract

This invention discloses a bird intelligent matching and repelling system and method based on a voiceprint database, belonging to the field of intelligent bird repelling technology. The invention includes: S10: predicting the sensitivity of target birds to the calls of various predators at different frequencies; S20: predicting the sensitivity of target birds to various predators; S30: generating a repelling strategy for the birds to be repelled; S40: controlling a call transmitter to repel the birds according to the repelling strategy. This invention segments the birds to be repelled based on their sensitivity to different predators at different frequencies and their flight status, and then combines this with bird behavior to achieve targeted repelling, further improving the effectiveness of bird repelling. The bird repelling strategy generated by this invention has high reliability and will not result in birds not being repelled.

Description

Technical Field

[0001] This invention relates to the field of intelligent bird matching and deterrence technology, specifically to an intelligent bird matching and deterrence system and method based on a voiceprint database. Background Technology

[0002] Birds have different sensitivities and fear responses to different sounds. Therefore, specific sound signals can be played to drive away or control bird activity, and specific voiceprint signals can influence bird behavior, thereby achieving eco-friendly bird removal.

[0003] When driving away birds with high activity density, multiple driving methods need to be used. Currently, the order of use of driving methods is determined according to the driving distance, which cannot guarantee that the driving will achieve the desired effect each time. This requires multiple adjustments to the driving strategy, and it is also impossible to make fine-grained adjustments to the strategy according to the different bird habits. Summary of the Invention

[0004] The purpose of this invention is to provide a bird intelligent matching and repelling system and method based on a voiceprint database to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a bird intelligent matching and repelling method based on a voiceprint database, the method comprising:

[0006] S10: Simulate the calls of various target bird predators based on the voiceprint database, and play the simulated calls of various target bird predators at different times at a certain distance from the target bird. By continuously changing the playback frequency of the simulated calls of various target bird predators and continuously shortening the playback distance of the simulated calls of various target bird predators, data on the response of the target bird to different calls of various target bird predators are collected. Based on the collection results, the sensitivity of the target bird to the calls of various target bird predators at different frequencies is predicted.

[0007] S20: Predict the sensitivity of target birds to various natural enemies of the target birds;

[0008] S30: The monitoring device in the target area determines the birds to be driven away in the target area. Based on the positional relationship of each bird to be driven away relative to the driving position, the type, frequency, and time period of the driving away sound emitted by the sound transmitter are analyzed. Based on the analysis results, a driving away strategy for the birds to be driven away is generated.

[0009] S40: Control the sound transmitter to drive away each bird to be driven away according to the driving strategy.

[0010] Furthermore, S10 includes:

[0011] S101: Simulate the call of the target bird's natural enemy i based on the voiceprint database. Let the simulated call of the target bird's natural enemy i be denoted as call A. Play call A at a distance of X meters from the target bird in time intervals, with only one type of target bird's natural enemy's call played in each time interval. By shortening the playback distance of call A, the alert distance V for the target bird at the frequency of call A at f is adjusted. fA 、Surprising distance Y fA And escape distance Z fA The data collection is performed, and the playback distance is shortened by x meters each time. Here, X and x are constants and both X and x are greater than 0. i = 1, 2, ..., n represents the number of the natural enemy of each target bird, and n represents the total number of natural enemies of the target bird.

[0012] S102: Regarding the warning distance V... fA With weighting coefficient a and flight distance Y fA With weighting coefficient b, and escape distance Z fA The product between the product and the weight coefficient c is calculated, and all products are summed. The summation result is denoted as G. fA With base e, -G fA Constructing an exponential model W for the exponential fA W fA =[1-exp(-G fA )]×100%, predicting the sensitivity of target birds to the call A of frequency f, where exp() represents an exponential function with base e and e=2.73;

[0013] Iterate through all simulated calls of the target bird's natural enemies and determine the target bird's sensitivity W to the call of the target bird's natural enemy i at frequency j. ji Make predictions, where j=1,2,…,m, representing the number corresponding to each playback frequency of each sound, and m represents the number of times the playback frequency of the sound changes.

[0014] Furthermore, the specific method for S20 to predict the sensitivity of the target bird to various natural enemies of the target bird is as follows:

[0015] For W ji The corresponding maximum value maxW ji Search for maxW ji The corresponding warning distance V i 、Surprising distance Y i And escape distance Z i To determine the warning distance V i 、Surprising distance Y i Escape distance Z i The sum R of these three parameters i Perform calculations;

[0016] According to maxW ji The duration L of the target bird's sustained vigilance behavior is obtained;

[0017] According to H i =[exp(-L)×(1 / R i The sensitivity of the target bird to its natural enemy i is predicted by 100% ( )]×100%.

[0018] Incorporating the duration of bird alertness during the prediction process can improve prediction accuracy to some extent.

[0019] Furthermore, S30 includes:

[0020] S301: Identify the birds to be driven away within the target area using monitoring devices within the target area, construct a three-dimensional spatial coordinate system using any point within the target area as the origin, and determine the position coordinates (x, y) of the bird to be driven away (p) and the driving position at time t. pt ,y pt ,z pt ), (x rt ,y rt ,z rt Data is collected, and the distance D from the bird p to be driven away to the driving position at time t is calculated according to the three-dimensional spatial distance formula. pt Calculations are performed based on the formula for calculating the angle between vectors in three-dimensional space, specifically the azimuth angle U of the bird p to be driven away at time t relative to its driving position. pt The calculation is performed, where p=1,2,…,q, representing the number of each bird to be driven away, and q represents the total number of birds to be driven away in the target area;

[0021] S302: Randomly select a direction angle U pt For those with the same direction angle U pt The birds to be driven away are identified, and it is determined whether there is overlap in the natural enemies of the identified birds. If there is overlap, the birds are classified into different stages based on their distance from the driving position, and the sound emitter is positioned at a direction angle of U. pt The frequency, type, and time period of the sound emitted from the location are determined, and the phases are divided into the alert phase, the flight phase, and the silence phase.

[0022] S303: If there is no overlap, then according to the determined distance values ​​of the birds to be driven away from the driving position in ascending order, the sound transmitters are directed at a directional angle of U. pt The frequency, type, and time period of the sound emitted from the location are determined.

[0023] S304: If there is partial overlap, priority shall be given to driving away birds with overlapping natural enemies, using the same method as in S302. Then, the remaining birds shall be driven away using the same method as in S303.

[0024] S305: Generate a deterrent strategy for birds to be deterred based on the type, frequency, and duration of the deterrent calls emitted by the sound transmitter in each directional angle.

[0025] Furthermore, S302 also includes:

[0026] The identified birds to be driven away are numbered in ascending order of distance. The numbering result is: β=1,2,…,τ; τ represents the total number of birds to be driven away.

[0027] The minimum number corresponding to the birds to be driven away in the silent stage is determined and denoted as ψ, where ψ = 1, 2, ..., τ. If the bird to be driven away ψ-1 is in the alert stage, then the distance D of the bird to be driven away ψ-1 from the driving position at time t is determined. (ψ-1)t As a warning distance ψ-1 for birds to be driven away, and D (ψ-1)t The startled flight distance and escape distance of the birds to be driven away (ψ-1) under the calls of predators at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird to be driven away, ψ-1, to the calls of predators at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call emitter at a direction angle of U. pt The frequency of the call emitted from the location, the call type is the call of the overlapping predator simulated based on the voiceprint database, and the emission time period of the call is [t, D]. 1t / δ1), where D 1t δ1 represents the distance of the determined bird 1 to be driven away from the driving position at time t, and δ1 represents the average flight speed of the determined bird 1 to be driven away within the target area.

[0028] If all the identified birds to be driven away are in the alert or flight-avoidance phase, then D τt The startled flight distance and escape distance of the birds to be driven away τ under the calls of their natural enemies at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird species to be driven away (τ) to the calls of predators at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call transmitter at a direction angle of U. pt The frequency of the call emitted from the location, the call type is the call of the overlapping predator simulated based on the voiceprint database, and the emission time period of the call is [t, D]. 1t / δ1).

[0029] Furthermore, S303 also includes:

[0030] The sensitivity of the bird β to various predators was collected. The predator corresponding to the maximum sensitivity value was taken as the target predator of the bird β to be driven away. The call of the target predator simulated according to the voiceprint database was taken as the type of call emitted by the call transmitter.

[0031] D βt The startled flight distance and escape distance of the birds to be driven away (β) under the calls of target predators at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird species β to the calls of its predator at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call transmitter at a direction angle of U. pt The frequency of the sound emitted at the location, and the emission time period of the sound is [t, D]. βt / δ β ), where D βt δ represents the distance of a specific bird β to be driven away from the driving location at time t. β This represents the average flight speed of the identified bird β within the target area.

[0032] Iterate through all the distance values ​​of the identified birds to be driven away from the driving location, and apply the sound emission to the bird at a direction angle of U. pt The location of the sound is determined in real time.

[0033] Furthermore, the specific method for generating the deportation strategy for each bird to be deported in step S305 is as follows:

[0034] The sound emission time periods of the sound transmitter at each directional angle are acquired. Redundancy processing is performed on the acquired sound emission time periods. The end times of each sound emission after redundancy processing are sorted according to the time sequence. Each sound emission end time is bound to the corresponding directional angle. The end times of each sound emission are numbered according to the sorting order. The numbering result is: ξ=1,2,…,φ; φ represents the total number of sound emission end times.

[0035] The reliability of the ordering between the sound emission end time ξ+1 and the sound emission end time ξ is analyzed;

[0036] When the sorting reliability is 1, the rotation speed of the audible transmitter is not adjusted between the audible emission end time ξ and the audible emission end time ξ+1.

[0037] When the sorting reliability is 0, the rotational speed p of the audible transmitter between the audible emission end time ξ and the audible emission end time ξ+1 is... ξ→ξ+1 Make adjustments;

[0038] The generated strategy for driving away birds is as follows: the movement trajectory of the sound transmitter is determined according to the sorting order. When the sorting reliability is 0, the rotation speed of the sound transmitter is adjusted at the starting position of the corresponding movement trajectory segment. The sound transmitter emits a sound type and frequency that matches the birds to be driven away at each movement trajectory stop point.

[0039] Furthermore, the specific method for S305 to analyze the reliability of the ordering between the sound emission end time ξ+1 and the sound emission end time ξ is as follows:

[0040] The time difference TV between the sound emission end time ξ+1 and the sound emission end time ξ ξ→ξ+1 And the absolute value a of the difference between the azimuth angle corresponding to the end time ξ+1 of the sound emission and the azimuth angle corresponding to the end time ξ of the sound emission. ξ→ξ+1 Perform calculations for a ξ→ξ+1 The ratio a' between the standard time required for the sound emitter to rotate 1° and the standard time required for the sound emitter to rotate 1° ξ→ξ+1 Perform calculations on TV ξ→ξ+1 With P 1-ξ ×ER ξ→ξ+1 The difference between ER´ ξ→ξ+1 Perform calculations on a´ ξ→ξ+1 With ER´ ξ→ξ+1 The difference θ between ξ→ξ+1 Calculations are performed to obtain the reliability judgment value of the order between the two selected sound emission end times. If θ ξ→ξ+1 If θ > 0, then the reliability of the ordering between the sound emission end time ξ and the sound emission end time ξ+1 is considered to be 0. ξ→ξ+1 If ≤0, then the reliability of the ordering between the sound emission end time ξ and the sound emission end time ξ+1 is considered to be 1, ER ξ→ξ+1 P represents the average dwell time of the audible emitter at the audible emission location. 1-ξ =1 or P 1-ξ =0, when 1-ξ=0, P 1-ξ =0, when 1-ξ<0, P 1-ξ =1;

[0041] Rotational speed p ξ→ξ+1 The adjustment formula is: p ξ→ξ+1 =a ξ→ξ+1 / (TV ξ→ξ+1 -P 1-ξ ×ER ξ→ξ+1 ).

[0042] A bird intelligent matching and repelling system based on a voiceprint database, the system comprising a first sensitivity prediction module, a second sensitivity prediction module, a repelling strategy generation module, and a repelling strategy driving module;

[0043] The first sensitivity prediction module is used to predict the sensitivity of the target bird to the calls of its natural enemies at different frequencies;

[0044] The second sensitivity prediction module is used to predict the sensitivity of the target bird to various natural enemies of the target bird;

[0045] The repulsion strategy generation module is used to identify birds to be repelled in the target area through monitoring devices in the target area, and to analyze the type, frequency, and time period of the repulsion calls emitted by the sound transmitter based on the positional relationship of each bird to be repelled relative to the repulsion location. Based on the analysis results, a repulsion strategy for the birds to be repelled is generated.

[0046] The drive-off strategy module is used to control the sound transmitter to drive away each bird to be driven away according to the drive-off strategy.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] 1. This invention simulates the calls of various bird predators based on a voiceprint database. By changing the playback frequency and distance of the calls, it predicts the sensitivity of target birds to the calls of various target bird predators at different frequencies, as well as the sensitivity of target birds to various target bird predators. Combined with the flight status of the birds to be driven away, the birds to be driven away are segmented and processed. Then, combined with the bird habits, the birds to be driven away are driven away in a targeted manner, which further improves the effect of driving away birds.

[0049] 2. When implementing targeted repulsion of birds, this invention classifies bird groups by the angle of sound emission, ensuring an effective repulsion effect in a short time.

[0050] 3. The bird deterrence strategy generated by this invention has high reliability and will not result in birds not being deterred. Attached Figure Description

[0051] Figure 1 This is a schematic diagram illustrating the workflow of a bird intelligent matching and repelling method based on a voiceprint database according to the present invention. Detailed Implementation

[0052] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] like Figure 1 As shown, this invention provides a technical solution for a bird intelligent matching and repelling system and method based on a voiceprint database. The method includes:

[0054] S10: Simulate the calls of various target bird predators based on the voiceprint database, and play the simulated calls of various target bird predators at different times at a certain distance from the target bird. By continuously changing the playback frequency of the simulated calls of various target bird predators and continuously shortening the playback distance of the simulated calls of various target bird predators, data on the response of the target bird to different calls of various target bird predators are collected. Based on the collection results, the sensitivity of the target bird to the calls of various target bird predators at different frequencies is predicted.

[0055] S10 includes:

[0056] S101: Simulate the call of the target bird's natural enemy i based on the voiceprint database. Let the simulated call of the target bird's natural enemy i be denoted as call A. Play call A at a distance of X meters from the target bird in time intervals, with only one type of target bird's natural enemy's call played in each time interval. By shortening the playback distance of call A, the alert distance V for the target bird at the frequency of call A at f is adjusted. fA (Warning distance refers to the distance between the target bird and the predator when the target bird first shows wariness behavior towards the approaching predator; the longer the warning distance, the higher the sensitivity of the target bird to the predator, and the earlier it can detect the threat.) Startle distance Y fA (The startled flight distance refers to the distance between the target bird and its predator when it flees; the longer the startled flight distance, the farther the target bird can fly away when facing a predator, demonstrating a stronger escape ability.) and flight distance Z. fA (The escape distance refers to the distance between the escape location and the initial landing location of the target bird; the longer the escape distance, the farther the target bird can stay after escaping, further reflecting its sensitivity to predators.) The data is collected, and the playback distance is shortened by x meters each time, where X and x are constants and both X and x are greater than 0, i=1,2,…,n, representing the number of each target bird's predator, and n represents the total number of predators of the target bird.

[0057] S102: Regarding the warning distance V... fA With weighting coefficient a and flight distance Y fA With weighting coefficient b, and escape distance Z fA The product between the product and the weight coefficient c is calculated, and all products are summed. The summation result is denoted as G. fA With base e, -G fA Constructing an exponential model W for the exponential fA WfA =[1-exp(-G fA The value is calculated as 100%, which predicts the sensitivity of target birds to the call A of frequency f, where exp() represents an exponential function with base e and e = 2.73, and W fA This indicates the sensitivity of the target bird species to a song A with frequency f;

[0058] Iterate through all simulated calls of the target bird's natural enemies and determine the target bird's sensitivity W to the call of the target bird's natural enemy i at frequency j. ji Make predictions, where j=1,2,…,m, representing the number corresponding to each playback frequency of each sound, and m represents the number of times the playback frequency of the sound changes;

[0059] S20: Predict the sensitivity of target birds to various natural enemies of the target birds;

[0060] The specific method for S20 to predict the sensitivity of target birds to various natural enemies of the target birds is as follows:

[0061] For W ji The corresponding maximum value maxW ji Search for maxW ji The corresponding warning distance V i 、Surprising distance Y i And escape distance Z i To determine the warning distance V i 、Surprising distance Y i Escape distance Z i The sum R of these three parameters i Perform calculations;

[0062] According to maxW ji The duration L of the target bird's sustained vigilance behavior is obtained;

[0063] According to H i =[exp(-L)×(1 / R i The 100% accuracy rate is used to predict the sensitivity of the target bird species to its natural enemy i.

[0064] S30: The monitoring device in the target area determines the birds to be driven away in the target area. Based on the positional relationship of each bird to be driven away relative to the driving position, the type, frequency, and time period of the driving away sound emitted by the sound transmitter are analyzed. Based on the analysis results, a driving away strategy for the birds to be driven away is generated.

[0065] S30 includes:

[0066] S301: Identify birds to be driven away within the target area using monitoring devices, including cameras. Construct a three-dimensional coordinate system using any point within the target area as the origin, and determine the position coordinates (x, y) of the bird to be driven away (p) and the driving position at time t. pt ,y pt ,z pt ), (x rt ,y rt ,z rt Data is collected, and the distance D from the bird p to be driven away to the driving position at time t is calculated according to the three-dimensional spatial distance formula. pt Calculations are performed based on the formula for calculating the angle between vectors in three-dimensional space, specifically the azimuth angle U of the bird p to be driven away at time t relative to its driving position. pt Perform calculations, U pt =arccos((s·h) / (|s||h|)), where s·h represents the dot product of vectors s and h, and |s| and |h| are the magnitudes of vectors s and h, respectively. Vector s = (x...) pt -x rt ,y pt -y rt ,z pt -z rt ), vector h = (w, y) rt ,z rt ), w represents a constant and w>0, where p=1,2,…,q, represents the number corresponding to each bird to be driven away, and q represents the total number of birds to be driven away in the target area;

[0067] S302: Randomly select a direction angle U pt For those with the same direction angle U pt The birds to be driven away are identified, and it is determined whether there is overlap in the natural enemies of the identified birds. If there is overlap, the birds are classified into different stages based on their distance from the driving position, and the sound emitter is positioned at a direction angle of U. pt The frequency, type, and time period of the calls emitted from the location are determined. The specific method is as follows: the identified birds to be driven away are numbered in ascending order of distance. The numbering result is: β=1,2,…,τ; τ represents the total number of birds to be driven away.

[0068] The minimum number corresponding to the birds to be driven away in the silent stage is determined and denoted as ψ, where ψ = 1, 2, ..., τ. If the bird to be driven away ψ-1 is in the alert stage, then the distance D of the bird to be driven away ψ-1 from the driving position at time t is determined. (ψ-1)t As a warning distance ψ-1 for birds to be driven away, and D (ψ-1)tThe startled flight distance and escape distance of the birds to be driven away (ψ-1) under the calls of predators at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird to be driven away, ψ-1, to the calls of predators at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call emitter at a direction angle of U. pt The frequency of the call emitted from the location, the call type is the call of the overlapping predator simulated based on the voiceprint database, and the emission time period of the call is [t, D]. 1t / δ1), where D 1t δ1 represents the distance of the determined bird 1 to be driven away from the driving position at time t, and δ1 represents the average flight speed of the determined bird 1 to be driven away within the target area.

[0069] If all the identified birds to be driven away are in the alert or flight-avoidance phase, then D τt The startled flight distance and escape distance of the birds to be driven away τ under the calls of their natural enemies at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird species to be driven away (τ) to the calls of predators at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call transmitter at a direction angle of U. pt The frequency of the call emitted from the location, the call type is the call of the overlapping predator simulated based on the voiceprint database, and the emission time period of the call is [t, D]. 1t / δ1);

[0070] The phases are divided into three stages: the alert stage, the flight stage, and the silence stage. The alert stage refers to the distance between the bird to be driven away and the driving location being within the range of [γ, l]. γ represents the distance between the bird to be driven away and the overlapping predator when the bird last exhibited alert behavior, and l represents the distance between the bird to be driven away and the overlapping predator when the bird first exhibited alert behavior. The flight stage refers to the distance between the bird to be driven away and the driving location being within the range of [0, γ). The silence stage refers to the distance between the bird to be driven away and the driving location being within the range of (l, +∞).

[0071] S303: If there is no overlap, then according to the determined distance values ​​of the birds to be driven away from the driving position in ascending order, the sound transmitters are directed at a directional angle of U. pt The frequency, type, and duration of the sound emitted at the selected location are determined. pt If there is only one bird to be driven away, denoted as bird L, then based on the distance of bird L from the driving position, the sound emitter is directed at a direction angle U. ptThe frequency, type, and duration of the call emitted from the location are determined, with the duration being [t, T), where T represents the distance of the bird L to be driven away from the driving location and the average flight speed of the bird L in the target area. The average flight speed does not include the grounded state of the bird L in the calculation process. The specific method is as follows: the sensitivity of the bird β to various predators is collected, and the predator corresponding to the maximum sensitivity value is taken as the target predator of the bird β. The call of the target predator simulated according to the sound pattern database is taken as the type of call emitted by the call transmitter.

[0072] D βt The startled flight distance and escape distance of the birds to be driven away (β) under the calls of target predators at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird species β to the calls of its predator at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call transmitter at a direction angle of U. pt The frequency of the sound emitted at the location, and the emission time period of the sound is [t, D]. βt / δ β ), where D βt δ represents the distance of a specific bird β to be driven away from the driving location at time t. β This represents the average flight speed of the identified bird β within the target area.

[0073] Iterate through all the distance values ​​of the identified birds to be driven away from the driving location, and apply the sound emission to the bird at a direction angle of U. pt The location of the sound emitted in real time is determined;

[0074] S304: If there is partial overlap, priority shall be given to driving away birds with overlapping natural enemies, using the same method as in S302. Then, the remaining birds shall be driven away using the same method as in S303.

[0075] S305: Based on the type, frequency, and duration of the deterrent calls emitted by the sound transmitter at various directional angles, a deterrent strategy for the birds to be deterred is generated. The specific method is as follows:

[0076] The sound emission time periods of the sound transmitter at each directional angle are acquired. Redundancy processing is performed on the acquired sound emission time periods. The end times of each sound emission after redundancy processing are sorted according to the time sequence. Each sound emission end time is bound to the corresponding directional angle. The end times of each sound emission are numbered according to the sorting order. The numbering result is: ξ=1,2,…,φ; φ represents the total number of sound emission end times.

[0077] The time difference TV between the sound emission end time ξ+1 and the sound emission end time ξ ξ→ξ+1 And the absolute value a of the difference between the azimuth angle corresponding to the end time ξ+1 of the sound emission and the azimuth angle corresponding to the end time ξ of the sound emission. ξ→ξ+1 Perform calculations for a ξ→ξ+1 The ratio a' between the standard time required for the sound emitter to rotate 1° and the standard time required for the sound emitter to rotate 1° ξ→ξ+1 Perform calculations on TV ξ→ξ+1 With P 1-ξ ×ER ξ→ξ+1 The difference between ER´ ξ→ξ+1 Perform calculations on a´ ξ→ξ+1 With ER´ ξ→ξ+1 The difference θ between ξ→ξ+1 Calculations are performed to obtain the reliability judgment value of the order between the two selected sound emission end times. If θ ξ→ξ+1 If θ > 0, then the reliability of the ordering between the sound emission end time ξ and the sound emission end time ξ+1 is considered to be 0. ξ→ξ+1 If ≤0, then the reliability of the ordering between the sound emission end time ξ and the sound emission end time ξ+1 is considered to be 1, ER ξ→ξ+1 P represents the average dwell time of the audible emitter at the audible emission location. 1-ξ =1 or P 1-ξ =0, when 1-ξ=0, P 1-ξ =0, when 1-ξ<0, P 1-ξ =1; Standard time refers to the time required for the audible transmitter to rotate 1° at its default speed.

[0078] When the sorting reliability is 1, the rotation speed of the audible transmitter is not adjusted between the audible emission end time ξ and the audible emission end time ξ+1.

[0079] When the sorting reliability is 0, the rotational speed p of the audible transmitter between the audible emission end time ξ and the audible emission end time ξ+1 is... ξ→ξ+1 Adjustments are made, and the specific adjustment formula is: p ξ→ξ+1 =a ξ→ξ+1 / (TV ξ→ξ+1 -P 1-ξ ×ER ξ→ξ+1 );

[0080] The generated strategy for driving away birds is as follows: the movement trajectory of the sound transmitter is determined according to the sorting order. When the sorting reliability is 0, the rotation speed of the sound transmitter is adjusted at the starting position of the corresponding movement trajectory segment. The sound transmitter emits a sound type and frequency that matches the birds to be driven away at each movement trajectory stop point.

[0081] S40: Control the sound transmitter to drive away each bird to be driven away according to the driving strategy.

[0082] A bird intelligent matching and repelling system based on a voiceprint database, the system includes a first sensitivity prediction module, a second sensitivity prediction module, a repelling strategy generation module, and a repelling strategy driving module;

[0083] The first sensitivity prediction module is used to predict the sensitivity of target birds to the calls of their natural enemies at different frequencies.

[0084] The second sensitivity prediction module is used to predict the sensitivity of target birds to various natural enemies of the target birds;

[0085] The bird removal strategy generation module is used to identify birds to be removed in the target area through monitoring devices in the target area. Based on the positional relationship of each bird to be removed relative to the removal location, the module analyzes the type, frequency, and time period of the removal calls emitted by the sound transmitter. Based on the analysis results, a bird removal strategy is generated.

[0086] The deterrence strategy driver module is used to control the sound transmitter to deter each bird according to the deterrence strategy.

[0087] Example 1: Suppose that the sound emission time periods corresponding to the sound transmitter at a directional angle of 30° are [t, t+2) and [t, t+20), the sound emission time periods corresponding to a directional angle of 45° are [t, t+11), and the sound emission time periods corresponding to a directional angle of 20° are [t, t+26]. The units of 2, 8, 6, and 11 are all seconds. The end times of each sound emission are sorted in chronological order, and each sound emission end time is bound to the corresponding directional angle to obtain the sorted sequence Q1: (t+2). 30° →(t+11) 45° →(t+20) 30° →(t+26) 20° ;

[0088] Assume the standard time required for the audible transmitter to rotate 1° is 4 seconds, and the average dwell time of the audible transmitter at the audible emission position is 5 seconds, then:

[0089] θ 1→2 =(|45-30| / 4)-[(t+11)-(t+2)-0×5]=-5.25;

[0090] θ 2→3 =(|30-45| / 4)-[(t+20)-(t+11)-1×5]=-0.25;

[0091] θ 3→4=(|30-20| / 4)-[(t+26)-(t+20)-1×5]=1.5;

[0092] It can be seen that the reliability of the order between the sound emission end time t+2 and the sound emission end time t+11 is 1, the reliability of the order between the sound emission end time t+13 and the sound emission end time t+11 is 1, and the reliability of the order between the sound emission end time t+22 and the sound emission end time t+13 is 0.

[0093] The rotational speed p of the sound transmitter between the sound emission end time t+26 and the sound emission end time t+20 3→4 Make adjustments, p 3→4 =|30-20| / {[(t+26)-(t+20)]-5}=10, meaning that the time required for the sound transmitter to rotate 1° is 0.1 seconds.

[0094] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A bird intelligent matching and repelling method based on a voiceprint database, characterized in that: The method includes: S10: Simulate the calls of various target bird predators based on the voiceprint database, and play the simulated calls of various target bird predators at different times at a certain distance from the target bird. By continuously changing the playback frequency of the simulated calls of various target bird predators and continuously shortening the playback distance of the simulated calls of various target bird predators, data on the response of the target bird to different calls of various target bird predators are collected. Based on the collection results, the sensitivity of the target bird to the calls of various target bird predators at different frequencies is predicted. S10 includes: S101: Simulate the call of the target bird's natural enemy i based on the voiceprint database. Let the simulated call of the target bird's natural enemy i be denoted as call A. Play call A at a distance of X meters from the target bird in time intervals, with only one type of target bird's natural enemy's call played in each time interval. By shortening the playback distance of call A, the alert distance V for the target bird at the frequency of call A at f is adjusted. fA 、Surprising distance Y fA And escape distance Z fA The data collection is performed, and the playback distance is shortened by x meters each time. Here, X and x are constants and both X and x are greater than 0. i = 1, 2, ..., n represents the number of the natural enemy of each target bird and n represents the total number of natural enemies of the target bird. S102: Regarding the warning distance V... fA With weighting coefficient a and flight distance Y fA With weighting coefficient b, and escape distance Z fA The product between the product and the weight coefficient c is calculated, and all products are summed. The summation result is denoted as G. fA With base e, -G fA Constructing an exponential model W for the exponential fA W fA =[1-exp(-G fA )]×100%, predicting the sensitivity of target birds to the call A of frequency f, where exp() represents an exponential function with base e and e=2.73; Iterate through all simulated calls of the target bird's natural enemies and determine the target bird's sensitivity W to the call of the target bird's natural enemy i at frequency j. ji Make predictions, where j=1,2,…,m, representing the number corresponding to each playback frequency of each sound, and m represents the number of times the playback frequency of the sound changes; S20: Predict the sensitivity of target birds to various natural enemies of the target birds. The specific method is as follows: For W ji The corresponding maximum value maxW ji Search for maxW ji The corresponding warning distance V i 、Surprising distance Y i And escape distance Z i To determine the warning distance V i 、Surprising distance Y i Escape distance Z i The sum R of these three parameters i Perform calculations; According to maxW ji The duration L of the target bird's sustained vigilance behavior is obtained; According to H i =[exp(-L)×(1 / R i The 100% accuracy rate is used to predict the sensitivity of the target bird species to its natural enemy i. S30: The monitoring device in the target area determines the birds to be driven away in the target area. Based on the positional relationship of each bird to be driven away relative to the driving position, the type, frequency, and time period of the driving away sound emitted by the sound transmitter are analyzed. Based on the analysis results, a driving away strategy for the birds to be driven away is generated. S30 includes: S301: Identify the birds to be driven away within the target area using monitoring devices within the target area, construct a three-dimensional spatial coordinate system using any point within the target area as the origin, and determine the position coordinates (x, y) of the bird to be driven away (p) and the driving position at time t. pt ,y pt ,z pt ), (x rt ,y rt ,z rt Data is collected, and the distance D from the bird p to be driven away to the driving position at time t is calculated according to the three-dimensional spatial distance formula. pt Calculations are performed based on the formula for calculating the angle between vectors in three-dimensional space, specifically the azimuth angle U of the bird p to be driven away at time t relative to its driving position. pt The calculation is performed, where p=1,2,…,q, representing the number of each bird to be driven away, and q represents the total number of birds to be driven away in the target area; S302: Randomly select a direction angle U pt For those with the same direction angle U pt The birds to be driven away are identified, and it is determined whether there is overlap in the natural enemies of the identified birds. If there is overlap, the birds are classified into different stages based on their distance from the driving position, and the sound emitter is positioned at a direction angle of U. pt The frequency, type, and time period of the sound emitted from the location are determined, and the phases are divided into the alert phase, the flight phase, and the silence phase. S303: If there is no overlap, then according to the determined distance values ​​of the birds to be driven away from the driving position in ascending order, the sound transmitters are directed at a directional angle of U. pt The frequency, type, and time period of the sound emitted from the location are determined. S304: If there is partial overlap, priority shall be given to driving away birds with overlapping natural enemies, using the same method as in S302. Then, the remaining birds shall be driven away using the same method as in S303. S305: Generate a deterrent strategy for birds to be deterred based on the type, frequency, and duration of the deterrent calls emitted by the sound transmitter in each directional angle. S40: Control the sound transmitter to drive away each bird to be driven away according to the driving strategy.

2. The bird intelligent matching and repelling method based on a voiceprint database according to claim 1, characterized in that: S302 further includes: The identified birds to be driven away are numbered in ascending order of distance. The numbering result is: β=1,2,…,τ; τ represents the total number of birds to be driven away. The minimum number corresponding to the birds to be driven away in the silent stage is determined and denoted as ψ, where ψ = 1, 2, ..., τ. If the bird to be driven away ψ-1 is in the alert stage, then the distance D of the bird to be driven away ψ-1 from the driving position at time t is determined. (ψ-1)t As a warning distance ψ-1 for birds to be driven away, and D (ψ-1)t The startled flight distance and escape distance of the birds to be driven away (ψ-1) under the calls of predators at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird to be driven away, ψ-1, to the calls of predators at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call emitter at a direction angle of U. pt The frequency of the call emitted from the location, the call type is the call of the overlapping predator simulated based on the voiceprint database, and the emission time period of the call is [t, D]. 1t / δ1), where D 1t δ1 represents the distance of the determined bird 1 to be driven away from the driving position at time t, and δ1 represents the average flight speed of the determined bird 1 to be driven away within the target area. If all the identified birds to be driven away are in the alert or flight-avoidance phase, then D τt The startled flight distance and escape distance of the birds to be driven away τ under the calls of their natural enemies at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird species to be driven away (τ) to the calls of predators at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call transmitter at a direction angle of U. pt The frequency of the call emitted from the location, the call type is the call of the overlapping predator simulated based on the voiceprint database, and the emission time period of the call is [t, D]. 1t / δ1).

3. The bird intelligent matching and repelling method based on a voiceprint database according to claim 2, characterized in that: The S303 further includes: The sensitivity of the bird β to various natural enemies was collected. The natural enemy corresponding to the maximum sensitivity value was taken as the target natural enemy of the bird β to be driven away. The call of the target natural enemy simulated according to the voiceprint database was taken as the type of call emitted by the call transmitter. D βt The startled flight distance and escape distance of the birds to be driven away (β) under the calls of target predators at different frequencies are respectively input into the exponential model W. fA The sensitivity of the bird species β to the calls of its predator at different frequencies was obtained. The frequency corresponding to the maximum sensitivity was used as the frequency of the call transmitter at a direction angle of U. pt The frequency of the sound emitted at the location, and the emission time period of the sound is [t, D]. βt / δ β ), where D βt δ represents the distance of a specific bird β to be driven away from the driving location at time t. β This represents the average flight speed of the identified bird β within the target area. Iterate through all the distance values ​​of the identified birds to be driven away from the driving location, and apply the sound emission to the bird at a direction angle of U. pt The location of the sound is determined in real time.

4. The bird intelligent matching and repelling method based on a voiceprint database according to claim 3, characterized in that: The specific method for S305 to generate the deportation strategy for each bird to be deported is as follows: The sound emission time periods of the sound transmitter at each directional angle are acquired. Redundancy processing is performed on the acquired sound emission time periods. The end times of each sound emission after redundancy processing are sorted according to the time sequence. Each sound emission end time is bound to the corresponding directional angle. The end times of each sound emission are numbered according to the sorting order. The numbering result is: ξ=1,2,…,φ; φ represents the total number of sound emission end times. The reliability of the ordering between the sound emission end time ξ+1 and the sound emission end time ξ is analyzed; When the sorting reliability is 1, the rotation speed of the audible transmitter is not adjusted between the audible emission end time ξ and the audible emission end time ξ+1. When the sorting reliability is 0, the rotational speed p of the audible transmitter between the audible emission end time ξ and the audible emission end time ξ+1 is... ξ→ξ+1 Make adjustments; The generated strategy for driving away birds is as follows: the movement trajectory of the sound transmitter is determined according to the sorting order. When the sorting reliability is 0, the rotation speed of the sound transmitter is adjusted at the starting position of the corresponding movement trajectory segment. The sound transmitter emits a sound type and frequency that matches the birds to be driven away at each movement trajectory stop point.

5. The bird intelligent matching and repelling method based on a voiceprint database according to claim 4, characterized in that: The specific method for S305 to analyze the reliability of the sorting between the sound emission end time ξ+1 and the sound emission end time ξ is as follows: The time difference TV between the sound emission end time ξ+1 and the sound emission end time ξ ξ→ξ+1 And the absolute value a of the difference between the azimuth angle corresponding to the end time ξ+1 of the sound emission and the azimuth angle corresponding to the end time ξ of the sound emission. ξ→ξ+1 Perform calculations for a ξ→ξ+1 The ratio a' between the standard time required for the sound emitter to rotate 1° and the standard time required for the sound emitter to rotate 1° ξ→ξ+1 Perform calculations on TV ξ→ξ+1 With P 1-ξ ×ER ξ→ξ+1 The difference between ER´ ξ→ξ+1 Perform calculations on a´ ξ→ξ+1 With ER´ ξ→ξ+1 The difference θ between ξ→ξ+1 Calculations are performed to obtain the reliability judgment value of the order between the two selected sound emission end times. If θ ξ→ξ+1 If θ > 0, then the reliability of the ordering between the sound emission end time ξ and the sound emission end time ξ+1 is considered to be 0. ξ→ξ+1 If ≤0, then the reliability of the ordering between the sound emission end time ξ and the sound emission end time ξ+1 is considered to be 1, ER ξ→ξ+1 P represents the average dwell time of the audible emitter at the audible emission location. 1-ξ =1 or P 1-ξ =0, when 1-ξ=0, P 1-ξ =0, when 1-ξ<0, P 1-ξ =1; Rotational speed p ξ→ξ+1 The adjustment formula is: p ξ→ξ+1 =a ξ→ξ+1 / (TV ξ→ξ+1 -P 1-ξ ×ER ξ→ξ+1 ).

6. A bird intelligent matching and deterrence system based on a voiceprint database, applied to the bird intelligent matching and deterrence method based on a voiceprint database as described in any one of claims 1-5, characterized in that: The system includes a first sensitivity prediction module, a second sensitivity prediction module, a deportation strategy generation module, and a deportation strategy driving module; The first sensitivity prediction module is used to predict the sensitivity of the target bird to the calls of its natural enemies at different frequencies; The second sensitivity prediction module is used to predict the sensitivity of the target bird to various natural enemies of the target bird; The repulsion strategy generation module is used to identify birds to be repelled in the target area through monitoring devices in the target area, and to analyze the type, frequency, and time period of the repulsion calls emitted by the sound transmitter based on the positional relationship of each bird to be repelled relative to the repulsion location. Based on the analysis results, a repulsion strategy for the birds to be repelled is generated. The drive-off strategy module is used to control the sound transmitter to drive away each bird to be driven away according to the drive-off strategy.

Images