Intelligent aiming and positioning method based on high-altitude bird repelling bomb filling

Abstract

The invention relates to the technical field of high-altitude bird repelling, and particularly discloses an intelligent aiming and positioning method based on high-altitude bird repelling bomb filling, which comprises the following steps: utilizing multiple groups of different types of sensors to cooperatively work to acquire data, and fusing the acquired data; processing the fused data by using a deep learning model to obtain feature vectors of the birds, and predicting a future motion state sequence of the birds according to a historical motion data sequence by using a long-short term memory network model; selecting a proper bird repelling bomb according to the feature vector of the target bird and the predicted motion state sequence; calculating an angle vector of the transmitting device; in combination with environmental factors, the emission parameters are corrected by using the correction function; after emission, the bird repelling effect is monitored through a feedback system. According to the invention, a reliable data basis is provided for accurate aiming and positioning, and the adaptability to different environments and bird behaviors is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of high-altitude bird repelling, and particularly relates to an intelligent aiming and positioning method based on the loading of high-altitude bird repelling bombs. Background Art

[0002] In modern society, the activities of birds have caused serious impacts on many fields. Taking airports as an example, collisions between birds and aircraft occur from time to time, seriously threatening aviation safety. According to statistics, the losses of aviation accidents caused by bird strikes globally amount to hundreds of millions of US dollars every year. In the power system, birds building nests and perching on poles and towers will cause faults such as line short circuits and tripping, affecting the stability of power supply. In the agricultural field, birds pecking at crops result in a large reduction in production, bringing economic losses to farmers. At present, high-altitude bird repelling bombs are one of the commonly used bird repelling means, but there are significant defects in the existing loading and aiming and positioning technologies of bird repelling bombs. Manual loading of bird repelling bombs is not only inefficient, but also prone to problems such as misloading and missing loading in case of emergency, delaying the bird repelling opportunity. Manual aiming and positioning is limited by human eye vision, reaction speed and operation experience, and it is difficult to accurately lock on to fast-moving or distant bird targets, resulting in poor bird repelling effects and being unable to effectively guarantee the safety of related facilities and industries. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent aiming and positioning method based on the loading of high-altitude bird repelling bombs to solve the deficiencies of manual loading and aiming and positioning in the prior art and achieve efficient and accurate bird repelling operations.

[0004] The purpose of the present invention can be realized by the following technical solutions:

[0005] An intelligent aiming and positioning method based on the loading of high-altitude bird repelling bombs includes the following steps:

[0006] S1: Use multiple different types of sensors to work together to collect data and fuse the collected data;

[0007] S2: Use a deep learning model to process the fused data to obtain the feature vectors of birds, and use a long short-term memory network model to predict the future motion state sequence of birds according to the historical motion data sequence;

[0008] S3: Store multiple bird repelling bombs with different attributes in an intelligent ammunition depot, select a suitable bird repelling bomb through a selection function according to the feature vectors of the target birds and the predicted motion state sequence, and an automated robotic arm performs a loading operation under the control of an intelligent control system and monitors the loading process in real time;

[0009] S4: Calculate the angular vector of the launching device through the aiming function based on the real-time position, predicted position of the target bird, and the ballistic characteristics of the bird repellent projectile; correct the launching parameters using the correction function in combination with environmental factors; monitor the bird repelling effect through the feedback system after launching.

[0010] As a further solution of the present invention: In the said S1, it specifically includes the following steps:

[0011] Let the data collected by the i-th sensor at time t be D i (t), where i = 1, 2,..., n, and n is the number of sensors;

[0012] Fuse the data collected by each sensor using the Kalman filtering algorithm. The prediction equation of the Kalman filter is:

[0013]

[0014] P(t|t - 1) = F(t)P(t - 1|t - 1)F T (t) + Q(t);

[0015] Where, is the predicted state at time t, F(t) is the state transition matrix, B(t) is the control input matrix, U(t) is the control input vector, P(t∣t - 1) is the predicted error covariance matrix, and Q(t) is the process noise covariance matrix;

[0016] The update equation is:

[0017] K(t) = P(t|t - 1)H T (t)[H(t)P(t|t - 1)H T (t) + R(t)] -1 ;

[0018]

[0019] P(t|t) = [I - K(t)H(t)]P(t|t - 1);

[0020] Where, K(t) is the Kalman gain, H(t) is the observation matrix, R(t) is the observation noise covariance matrix, is the updated state at time t, and P(t∣t) is the updated error covariance matrix;

[0021] Obtain the fused data D fused (t).

[0022] As a further solution of the present invention: In the said S2, it specifically includes the following steps:

[0023] Analyze the preprocessed fusion data D fused (t) using deep learning algorithms. After passing through the deep learning model M feature obtain the feature vector V feature of the bird, that is:

[0024] V feature = M feature (D fused (t));

[0025] Adopt a long short-term memory network model. According to the input historical motion data sequence S history = [V1, V2,..., V m , obtain the predicted motion state sequence at the next k moments

[0026] As a further solution of the present invention: In the said S3, it specifically includes the following steps:

[0027] Construct an intelligent ammunition depot, which is built-in with various types of bird repellent bombs. The attribute vector of the j-th type of bird repellent bomb is A j = [a j1 , a j2 ,..., a jl ;

[0028] According to the feature vector V feature of the target bird and the predicted motion state sequence select the appropriate bird repellent bomb number j select through the selection function f selected , that is s is the number of types of bird repellent bombs;

[0029] Realize the loading operation through an automated robotic arm and an intelligent control system. The position vector of the robotic arm is P arm (t), and the target loading position vector is P target . Establish the constraint:

[0030]

[0031] And monitor the loading process in real time.

[0032] As a further solution of the present invention: In the said S4, it specifically includes the following steps:

[0033] According to the real-time position P bird (t) of the target bird, the predicted position and the ballistic characteristic parameter vector B ballistic of the bird repellent bomb;

[0034] Through the aiming function f aim Calculate the angle vector θ = [θ1, θ2] of the launching device, that is

[0035] Combine environmental factors to correct the original launching parameters

[0036] After launching, the bird repelling effect index E is monitored in real time through the feedback system. If E does not reach the expected threshold E threshold , then re - perform data acquisition, analysis, loading and aiming and launching operations

[0037] As a further solution of the present invention: The multiple groups of different types of sensors include high - definition thermal imaging cameras, millimeter - wave radars and ultrasonic sensors

[0038] As a further solution of the present invention: The environmental factors include wind speed, wind direction and air pressure

[0039] Advantages of the present invention

[0040] The multi - source data acquisition and fusion technology of the present invention can comprehensively and accurately obtain bird information, without being overly interfered by environmental factors, providing a reliable data basis for precise aiming and positioning, and greatly improving the adaptability to different environments and bird behaviors

[0041] The intelligent loading system realizes the rapid and accurate selection and loading of bird - repelling projectiles, avoids manual operation errors, improves the efficiency and safety of bird - repelling operations, and ensures rapid response in emergency situations

[0042] The precise aiming and positioning model and launching control algorithm fully consider the movement characteristics of birds and environmental factors, significantly improve the hit rate of bird - repelling projectiles, effectively enhance the bird - repelling effect, and provide more reliable protection for related facilities and industries Description of the drawings

[0043] The present invention will be further described below with reference to the accompanying drawings

[0044] Figure 1 is a schematic flow chart of an intelligent aiming and positioning method based on high - altitude bird - repelling projectile loading of the present invention Detailed implementation manners

[0045] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention

[0046] Please refer toFigure 1 As shown in the figure, the present invention is an intelligent aiming and positioning method based on the loading of high-altitude bird repelling bombs, including the following steps:

[0047] Multi-source data acquisition and preprocessing:

[0048] Data acquisition: Multiple groups of different types of sensors work together, including high-definition thermal imaging cameras, millimeter-wave radars, ultrasonic sensors, etc. Let the data collected by the i-th sensor at time t be D i (t), where i = 1, 2,..., n, and n is the number of sensors.

[0049] Data fusion: The Kalman filtering algorithm is used to fuse the data collected by each sensor. Let the system state vector be X(t), the observation vector be Z(t), and the prediction equation of the Kalman filter is:

[0050]

[0051] P(t|t - 1) = F(t)P(t - 1|t - 1)F T (t) + Q(t);

[0052] Among them, is the predicted state at time t, F(t) is the state transition matrix, B(t) is the control input matrix, U(t) is the control input vector, P(t∣t - 1) is the predicted error covariance matrix, and Q(t) is the process noise covariance matrix.

[0053] The update equation is:

[0054] K(t) = P(t|t - 1)H T (t)[H(t)P(t|t - 1)H T (t) + R(t)] -1 ;

[0055]

[0056] P(t|t) = [I - K(t)H(t)]P(t|t - 1);

[0057] Among them, K(t) is the Kalman gain, H(t) is the observation matrix, R(t) is the observation noise covariance matrix, is the updated state at time t, and P(t∣t) is the updated error covariance matrix. Through the above Kalman filtering algorithm, the fused data D fused (t) is obtained. By comprehensively analyzing the data of different sensors, data noise and errors are eliminated, and the accuracy and reliability of the data are improved, providing an accurate basis for subsequent aiming and positioning.

[0058] Target Feature Extraction and Behavior Prediction:

[0059] Target Feature Extraction: Using deep learning algorithms, analyze the preprocessed images and data to extract the feature information of birds. Let the input fused data be D fused (t), and after passing through the deep learning model M feature obtain the feature vector V feature of birds, that is:

[0060] V feature = M feature (D fused (t)).

[0061] Behavior Prediction: By establishing a bird behavior prediction model, based on the historical movement trajectories and current movement states of birds, predict their positions and movement trends in the next period of time. Adopt the Long Short-Term Memory (LSTM) model. Let the input historical movement data sequence be S history = [V1, V2,..., V m , where V j is the movement state vector at the j-th moment. After passing through the LSTM model MLSTM, obtain the predicted movement state sequence for the next k moments That is

[0062] Intelligent Loading System Design:

[0063] Bird Repellent Selection: Build an intelligent ammunition depot with various types of bird repellent ammunition. Let the attribute vector of the j-th type of bird repellent ammunition be A j = [a j1 , a j2 ,..., a jl , where a jl represents the l-th attribute of the j-th type of bird repellent ammunition (such as explosion power, effective dispersal range, etc.). According to the feature vector V feature of the target bird and the predicted movement state sequence select the appropriate bird repellent ammunition number j select through the selection function f selected , that is where s is the number of types of bird repellent ammunition.

[0064] Loading Operation: Achieve fast and accurate loading operations through an automated robotic arm and an intelligent control system. Let the position vector of the robotic arm be P arm (t), the target loading position vector be P target , and the motion control algorithm of the robotic arm calculates the motion trajectory and control parameters based on the current position and the target position, so that At the same time, monitor the loading process in real time. Once an abnormality is detected, stop immediately and perform fault diagnosis and repair.

[0065] Precise aiming and positioning and launch control:

[0066] Aiming and positioning model: Based on the real-time position P bird (t), predicted position and the ballistic characteristics of the bird repellent projectile, an accurate aiming and positioning model is established. Let the angle vector of the launching device be θ = [θ1, θ2] (representing the horizontal and vertical angles respectively), and the appropriate launch angle is calculated through the aiming function f aim , that is:

[0067] where B ballistic is the ballistic characteristic parameter vector of the bird repellent projectile.

[0068] Launch parameter correction: During the launch process, in combination with environmental factors such as wind speed v wind , wind direction α wind , air pressure p, etc., the launch parameters are corrected in real time.

[0069] Let the original launch parameter vector be P original = [v0, θ0] (v0 is the initial launch velocity, θ0 is the launch angle), and the corrected launch parameter vector P correct is obtained after passing through the correction function f corrected , that is, P corrected = f correct (P original , v wind , α wind , p).

[0070] Launch and feedback: After the launch, the bird repellent effect is monitored in real time through the feedback system. Let the monitored bird repellent effect index be E. If E does not reach the expected threshold E threshold , then according to the newly obtained bird information, a series of operations such as data collection, analysis, loading, and aiming and launching are automatically re-performed until E ≥ E threshold .

[0071] The above has described a detailed description of an embodiment of the present invention, but the described content is only a preferred embodiment of the present invention and cannot be considered as being used to limit the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the present invention application should still fall within the scope covered by the patent of the present invention.

Claims

1. An intelligent aiming and positioning method based on the loading of high-altitude bird repellent bombs, characterized in that, It includes the following steps: S1: Use multiple different types of sensors to work together to collect data, and fuse the collected data; S2: Use a deep learning model to process the fused data to obtain the feature vectors of birds, and use a long short-term memory network model to predict the future movement state sequence of birds based on the historical movement data sequence; S3: Store various types of bird repellent projectiles in an intelligent ammunition depot. According to the feature vectors of the target birds and the predicted movement state sequence, select a suitable bird repellent projectile through a selection function. The automated robotic arm performs loading operations under the control of an intelligent control system, and monitors the loading process in real time; S4: Calculate the angle vector of the launching device through an aiming function according to the real-time position, predicted position of the target bird, and the ballistic characteristics of the bird repellent projectile; Combine environmental factors and use a correction function to correct the launching parameters; Monitor the bird repellent effect through a feedback system after launching.

2. The intelligent aiming and positioning method based on the loading of high-altitude bird repellent bombs according to claim 1, wherein In the S1, it specifically includes the following steps: Let the data collected by the \(i\)-th sensor at time \(t\) be \(D i (t)\), where \(i = 1, 2, \cdots, n\) and \(n\) is the number of sensors; Use the Kalman filtering algorithm to fuse the data collected by each sensor. The prediction equation of the Kalman filter is: P(t|t - 1) = F(t)P(t - 1|t - 1)F T (t) + Q(t); wherein, is the predicted state at time t, F(t) is the state transition matrix, B(t) is the control input matrix, U(t) is the control input vector, P(t|t - 1) is the predicted error covariance matrix, and Q(t) is the process noise covariance matrix; The update equation is: K(t) = P(t|t - 1)H T (t)[H(t)P(t|t - 1)H T (t) + R(t)] -1 ; P(t|t) = [I - K(t)H(t)]P(t|t - 1); where K(t) is the Kalman gain, H(t) is the observation matrix, R(t) is the observation noise covariance matrix, is the updated state at time t, and P(t|t) is the updated error covariance matrix; Obtain the fused data D fused (t).

3. The intelligent aiming and positioning method based on the loading of high-altitude bird repellent bombs according to claim 2, wherein, In the S2, it specifically includes the following steps: Analyze the preprocessed fusion data D fused (t) using deep learning algorithms, and pass it through the deep learning model M feature to obtain the feature vector V feature of the birds, that is: V feature = M feature (D fused (t)); Using a long short-term memory network model, based on the input historical motion data sequence S history =[V1, V2,..., V m , the motion state sequence for the predicted next k time instants is obtained through the long short-term memory network model 4. The intelligent aiming and positioning method based on the loading of high-altitude bird repelling bombs according to claim 1, wherein In the S3, it specifically includes the following steps: Build an intelligent ammunition depot, which is equipped with various types of bird repellent ammunition. The attribute vector of the j-th type of bird repellent ammunition is A j =[a j1 ,a j2 ,...,a jl ; According to the feature vector V of the target bird feature and the predicted motion state sequence By using the selection function f select select the appropriate bird repellent ammunition number j selected , that is, j selected = f select (V feature , A1,A2,...,A s ), where s is the number of types of bird repellent ammunition; The loading operation is realized through an automated robotic arm and an intelligent control system. The position vector of the robotic arm is P arm (t), and the target loading position vector is P target , and the constraint is established as follows: And monitor the loading process in real time.

5. An intelligent aiming and positioning method based on the loading of high-altitude bird repellent bombs according to claim 1, characterized in that, In the S4, it specifically includes the following steps: According to the real-time position P bird (t) of the target bird, the predicted position and the ballistic characteristic parameter vector B of the bird repellent projectile ballistic ; Through the aiming function f aim Calculate the angle vector θ = [θ1, θ2] of the launching device, i.e., θ = f aim (P bird (t), Combine environmental factors to correct the direction of the original launching parameters; After launch, the anti-bird effect index E is monitored in real time through the feedback system. If E does not reach the expected threshold E threshold , the operations of data collection, analysis, loading, and aiming and launching are carried out again.

6. The intelligent aiming and positioning method based on the loading of high-altitude bird repelling bombs according to claim 1, wherein The multiple different types of sensors include a high-definition thermal imaging camera, a millimeter-wave radar, and an ultrasonic sensor.

7. An intelligent aiming and positioning method based on the loading of high-altitude bird repellent bombs according to claim 5, characterized in that The environmental factors include wind speed, wind direction, and air pressure.

Images