Pseudosciaena crocea parent selective fishing method and system based on otolith resonance and artificial intelligence

Abstract

The invention discloses a large yellow croaker parent selective fishing method and system based on otolith resonance and artificial intelligence, and the method comprises the steps: measuring the resonance frequency of otolith of large yellow croakers with different body lengths through an experiment, and building a body length-resonance frequency mapping model; designing frequency-adjustable sound wave signals by analyzing resonant frequency differences of otolith of large yellow croakers with different body lengths; inducing target parent fish according to the frequency-adjustable sound wave signal, and monitoring fish school dynamics and environmental noise in real time through an underwater acoustic sensor, an optical sensor and environmental parameter detection; using a deep learning model to analyze fish school feedback signals in real time, identifying fish body length and distribution, and optimizing sound wave parameters; by combining sound wave induction and light induction technologies, parent fishes are guided to be concentrated to a target area, then mechanical net drawing is started, and fishing is completed. According to the method, efficient and selective fishing of the larimichthys crocea parents is achieved, the mistaken fishing rate of juvenile fishes is remarkably reduced, and the fishing efficiency and the survival rate of the parent fishes are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of aquaculture fishing, and in particular relates to a method and system for selectively fishing large yellow croaker parent fish based on otolith resonance and artificial intelligence. Background Art

[0002] Traditional large yellow croaker fishing (such as the knocking method) causes fish schools to faint through acoustic resonance, but lacks selectivity, resulting in the accidental capture of juvenile fish and ecological damage. In addition, existing acoustic attraction technologies (such as fixed-frequency sound waves) are difficult to dynamically adapt to the otolith resonance characteristics of fish of different body lengths, and lack real-time feedback on the fishing effect. Therefore, it is of great significance to develop a technology that combines acoustic characteristics and intelligent control to accurately induce parent fish of a specific body length, reduce the accidental capture of juvenile fish, and improve fishing efficiency and ecological friendliness. Summary of the Invention

[0003] To solve the above technical problems, the present invention provides a method and system for selectively fishing large yellow croaker parent fish based on otolith resonance and artificial intelligence. Among them, a method for selectively fishing large yellow croaker parent fish based on otolith resonance and artificial intelligence includes:

[0004] Measuring the resonance frequencies of otoliths of large yellow croakers of different body lengths through experiments, and establishing a "body length - resonance frequency" mapping model;

[0005] Designing an adjustable-frequency acoustic signal by analyzing the differences in otolith resonance frequencies of large yellow croakers of different body lengths;

[0006] Inducing target parent fish according to the adjustable-frequency acoustic signal, and real-time monitoring the dynamics of the fish school and environmental noise through underwater acoustic sensors, optical sensors and environmental parameter detection;

[0007] Using a deep learning model to real-time analyze the fish school feedback signal, identify the fish body length and distribution, and optimize the acoustic parameters;

[0008] Combining acoustic induction and light attraction technology, guiding the parent fish to concentrate in the target area and then starting the mechanical netting to complete the fishing.

[0009] Preferably, the process of measuring the resonance frequencies of otoliths of large yellow croakers of different body lengths through experiments includes:

[0010] Collecting samples of large yellow croakers of different body lengths, performing vibration mode analysis on the otoliths through a laser Doppler vibrometer, recording the resonance frequencies, and determining the otolith resonance frequency range of large yellow croakers of different body lengths through numerical mode analysis; among them, the resonance frequency of adult fish with a body length greater than or equal to 30 cm is 200 - 500 Hz, and the resonance frequency of juvenile fish is 800 - 1200 Hz.

[0011] Preferably, the formula expression of the resonance frequency is:

[0012]

[0013] Among them, E is the elastic modulus of the otolith; ρ is the density of the otolith; V is the Poisson's ratio; L and w are the length and width of the otolith respectively.

[0014] Preferably, the process of inducing target broodstock according to the tunable frequency acoustic wave signal includes:

[0015] Using a broadband underwater acoustic transducer to emit a composite acoustic wave that superimposes an orthogonal chirp signal and a single-frequency resonance signal, covering the target frequency range; and using an embedded adaptive algorithm to adjust the acoustic wave intensity and frequency band in real time according to the ambient noise.

[0016] Preferably, the process of real-time monitoring of fish school dynamics and ambient noise through underwater acoustic sensors, optical sensors and environmental parameter detection includes:

[0017] Obtaining the fish school density, body length distribution and movement trajectory through an underwater camera and sonar. After real-time collecting the fish school response data, using a convolutional neural network to identify the fish body length in real time and calculate the ratio of broodstock to juveniles.

[0018] Preferably, the process of using a deep learning model to analyze the fish school feedback signal in real time, identify the fish body length and distribution, and optimize the acoustic wave parameters includes:

[0019] Using a convolutional neural network combined with reinforcement learning to optimize the acoustic wave parameters and dynamically adjust the acoustic wave signal;

[0020] When the proportion of juveniles exceeds the preset threshold, automatically switch to a higher frequency band or reduce the sound intensity.

[0021] Preferably, the method further includes: designing a double-layer trawl net; wherein, the upper layer of the double-layer trawl net is a 10 cm × 10 cm mesh that allows juveniles to escape, and the lower layer is a 25 cm × 25 cm mesh adapted to the size of broodstock.

[0022] Preferably, the process of combining acoustic wave induction and light attraction technology to guide broodstock to concentrate in the target area and then starting mechanical netting for fishing includes:

[0023] Inducing broodstock through acoustic wave induction combined with light attraction technology of green LED with a wavelength of 520 nm to simulate the twinkling of plankton. When the aggregation degree of broodstock reaches more than 80%, start the double-layer trawl net, and the juveniles escape through the upper mesh, and the broodstock enter the codend.

[0024] The present invention also provides a large yellow croaker parent selective fishing system based on otolith resonance and artificial intelligence, including:

[0025] A resonance frequency measurement module for experimentally measuring the resonance frequencies of otoliths of large yellow croakers with different body lengths and establishing a "body length - resonance frequency" mapping model;

[0026] A signal design module, which is used to design an adjustable-frequency acoustic wave signal by analyzing the resonance frequency differences of otoliths of large yellow croakers with different body lengths;

[0027] An induction monitoring module, which is used to induce target broodstock according to the adjustable-frequency acoustic wave signal, and to monitor the dynamics of the fish school and environmental noise in real time through underwater acoustic sensors, optical sensors and environmental parameter detection;

[0028] A feedback regulation module, which is used to analyze the fish school feedback signal in real time by using a deep learning model, identify the fish body length and distribution, and optimize the acoustic wave parameters;

[0029] A fishing module, which is used to combine acoustic wave induction and light trapping technology, guide the broodstock to concentrate in the target area and then start mechanical netting to complete the fishing.

[0030] Compared with the prior art, the present invention has the following advantages and technical effects:

[0031] By combining the otolith resonance characteristics and artificial intelligence technology, the present invention realizes the precise fishing of large yellow croaker broodstock, significantly reduces the misfishing rate of juvenile fish, improves the fishing efficiency and the survival rate of broodstock, and at the same time has strong environmental adaptability, providing an innovative solution for sustainable aquatic fishing. Brief Description of the Drawings

[0032] The drawings constituting a part of this application are used to provide a further understanding of this application. The schematic embodiments and descriptions thereof of this application are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0033] Figure 1 It is a schematic flowchart of the method of the embodiment of the present invention. Detailed Embodiments

[0034] It should be noted that, without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The following will refer to the drawings and combine the embodiments to detail this application.

[0035] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0036] Embodiment 1

[0037] As Figure 1 shown, in this embodiment, a method for selective fishing of large yellow croaker broodstock based on otolith resonance and artificial intelligence is provided, including:

[0038] Measure the resonance frequencies of the otoliths of large yellow croakers with different body lengths through experiments, and establish a "body length - resonance frequency" mapping model; the "body length - resonance frequency" mapping model is used as the basis for acoustic wave regulation.

[0039] By analyzing the differences in the resonance frequencies of the otoliths of large yellow croakers with different body lengths, design an adjustable - frequency acoustic wave signal to induce only the target broodstock and avoid accidental capture of juvenile fish;

[0040] Induce the target broodstock according to the adjustable - frequency acoustic wave signal, and through underwater acoustic sensors, optical sensors and environmental parameter detection, monitor the fish school dynamics and environmental noise in real - time;

[0041] Use a deep - learning model to analyze the fish school feedback signal in real - time, identify the fish body length and distribution, and optimize the acoustic wave parameters;

[0042] Combine acoustic wave induction and light - trapping technology, guide the broodstock to concentrate in the target area and then start mechanical netting to complete the fishing.

[0043] Furthermore, the process of measuring the resonance frequencies of the otoliths of large yellow croakers with different body lengths through experiments includes:

[0044] Collect samples of large yellow croakers with different body lengths, conduct vibration mode analysis on the otoliths using a laser Doppler vibrometer, record the resonance frequencies, and determine the otolith resonance frequency range of large yellow croakers with different body lengths through numerical mode analysis; among them, the resonance frequency of adult fish with a body length of ≥30 cm is 200 - 500 Hz, and the resonance frequency of juvenile fish is 800 - 1200 Hz.

[0045] Furthermore, the formula expression of the resonance frequency is:

[0046]

[0047] Among them, E is the elastic modulus of the otolith; ρ is the density of the otolith; V is the Poisson's ratio; L and w are the length and width of the otolith respectively.

[0048] Even further, sample collection and frequency measurement:

[0049] Sample size: Collect samples of large yellow croakers with different body lengths (15 - 45 cm) (n = 200), anatomically extract the otoliths, and record the body length (L) and the three - dimensional dimensions of the otoliths (length l, width w, thickness t).

[0050] Specifically, collect 100 adult wild large yellow croakers (body length ≥30 cm) and 100 juvenile fish (body length < 15 cm), and anatomically extract otolith samples.

[0051] Experimental method: Use a laser Doppler vibrometer (LDV) to apply a swept - frequency acoustic wave of 0 - 2000 Hz to the otoliths, conduct vibration mode analysis, and record the main resonance peak frequency (fres).

[0052] The data results are shown in Table 1:

[0053] Table 1

[0054] Body length range (cm) Average otolith resonance frequency (Hz) Standard deviation (Hz) 15-20 800-1200 ±50 25-30 500-800 ±40 ≥30 200-500 ±30

[0055] From this, we can see that the otolith resonance frequency is negatively correlated with body length; the formula is:

[0056] f=k / L 2

[0057] Where k is the material constant, k≈4.3×10 4 Hz·cm 2 .

[0058] Furthermore, the acoustic wave induced threshold was verified:

[0059] Experimental design: Different frequency sound waves (200-1200Hz) were emitted in the tank, and the responses of fish schools (aggregation rate, escape rate) were observed.

[0060] Key parameters: Effective induction frequency: The peak aggregation rate of adult fish reaches 85% at 300Hz, and the peak escape rate of juvenile fish reaches 90% at 1000Hz.

[0061] Sound pressure level range: The optimal induced sound pressure is 140-160dB (no response below the threshold, too high will cause stress).

[0062] Furthermore, the process of inducing target broodstock according to the frequency-adjustable sound wave signal includes:

[0063] A wide-band underwater acoustic transducer is used to emit composite sound waves that superimpose orthogonal chirp signals and single-frequency resonance signals, covering the target frequency range of 200-1200Hz; and an embedded adaptive algorithm is used to adjust the sound wave intensity and frequency band in real time according to environmental noise (such as water flow and ship interference).

[0064] Furthermore, the composite acoustic wave design includes:

[0065] Orthogonal chirp signal: frequency range 200-1200 Hz, duration 2 s, interval 1 s.

[0066] Single-frequency resonance signal: center frequency 300Hz (adult fish), bandwidth ±50Hz, sound pressure level 155dB.

[0067] Superposition method: time domain orthogonal modulation to avoid signal interference.

[0068] Furthermore, the generation of orthogonal Chirp signals:

[0069] The orthogonal Chirp signal is composed of two linear frequency modulation signals (LFM), with a phase difference of 90°, corresponding to the real part and the imaginary part respectively, forming a complex signal:

[0070]

[0071] Where: A is the signal amplitude; f0 is the starting frequency (such as 200 Hz); k is the frequency modulation slope, k = B / T, B is the bandwidth (such as 1000 Hz), and T is the signal duration (such as 2 seconds).

[0072] Implementation steps:

[0073] Set the signal parameters f0 = 200 Hz, B = 1000 Hz, T = 2 s;

[0074] Generate the time series t ∈ [0, T];

[0075] Generate the real part (I channel) and the imaginary part (Q channel) signals respectively:

[0076]

[0077] Furthermore, the generation of the single-frequency resonance signal:

[0078] The single-frequency resonance signal is a sine wave with a fixed frequency, and the frequency corresponds to the resonance frequency of the target fish otolith (such as 300 Hz):

[0079] S res (t) = A res ·sin(2πf res t)

[0080] Where, f res is determined by the "body length - resonance frequency" model (such as f res = 850·L-1.2).

[0081] Implementation steps:

[0082] Calculate f res according to the target fish body length L (for example, when L = 30 cm, f res = 300 Hz);

[0083] Generate the time-synchronized sine signal s res (t).

[0084] Furthermore, signal superposition and modulation:

[0085] Superpose the orthogonal Chirp signal and the single-frequency resonance signal in the time domain to form a composite signal:

[0086] S composite (t) = S chirp (t) + sres (t)

[0087] Expand to:

[0088]

[0089] Furthermore, the adaptive noise suppression includes:

[0090] Ambient noise library: Pre-store common noise spectra (ship engines, wave impacts).

[0091] Noise reduction algorithm: Adopt the LMS (Least Mean Square) adaptive filter, and the signal-to-noise ratio is increased by ≥15 dB.

[0092] Further, the process of real-time monitoring of fish school dynamics and ambient noise through underwater acoustic sensors, optical sensors, and environmental parameter detection includes:

[0093] Obtain the fish school density, body length distribution, and movement trajectory through an underwater camera and sonar. After real-time collecting the fish school response data, use a convolutional neural network to real-time identify the fish body length and calculate the ratio of adult fish to juvenile fish.

[0094] Further, the process of using a deep learning model to real-time analyze the fish school feedback signal, identify the fish body length and distribution, and optimize the acoustic wave parameters includes:

[0095] Use a convolutional neural network combined with reinforcement learning to optimize the acoustic wave parameters (such as frequency, pulse interval), and dynamically adjust the acoustic wave signal;

[0096] When the proportion of juvenile fish exceeds a preset threshold (such as 5%), automatically switch to a higher frequency band or reduce the sound intensity to reduce misfishing.

[0097] Even further, the model training includes:

[0098] 1. Dataset construction;

[0099] Data source: 10,000 videos of large yellow croaker swimming taken by an underwater camera (including different body lengths and lighting conditions).

[0100] Annotation method: Manually annotate the fish body length (error ±1 cm), and classify it into "adult fish" (≥30 cm) and "juvenile fish" (<15 cm) according to the body length.

[0101] 2. Convolutional neural network (CNN) parameters;

[0102] Model architecture: ResNet-50 pre-trained model, fine-tune the last fully connected layer.

[0103] The training process of the relevant parameters is shown in Table 2:

[0104] Table 2

[0105]

[0106]

[0107] 3. Reinforcement learning control strategy;

[0108] State space: fish school density, proportion of juvenile fish, environmental noise intensity.

[0109] Action space: acoustic wave frequency adjustment (±50 Hz), sound pressure level adjustment (±5 dB).

[0110] Reward function:

[0111] R = 0.7 (adult fish aggregation rate) - 0.3 (juvenile fish bycatch rate) - 0.1 (energy consumption)

[0112] Training result: After 200 times of simulation training, the control success rate of the model in turbid waters has increased from the initial 60% to 92%.

[0113] Furthermore, the method further includes: designing a double-layer trawl net; wherein, the upper layer of the double-layer trawl net has a mesh size of 10 cm × 10 cm allowing juvenile fish to escape, and the lower layer has a mesh size of 25 cm × 25 cm adapted to the size of adult fish.

[0114] Even further, the hierarchical netting design includes:

[0115] Upper layer mesh: 10 cm × 10 cm (allowing juvenile fish to escape).

[0116] Lower layer mesh: 25 cm × 25 cm (capturing adult fish).

[0117] Material: ultra-high molecular weight polyethylene (UHMWPE), breaking strength ≥ 200 MPa.

[0118] Furthermore, the process of combining acoustic wave induction and light attraction technology to guide adult fish to concentrate in the target area and then starting mechanical netting for fishing includes:

[0119] By combining acoustic wave induction with green LED, a light attraction technology with a wavelength of 520 nm is used to simulate the twinkling of plankton for adult fish induction. When the aggregation degree of adult fish reaches more than 80%, the double-layer trawl net is started, and juvenile fish escape through the upper layer mesh, while adult fish enter the codend.

[0120] Even further, the light attraction auxiliary system:

[0121] Wavelength: 520 nm (green LED, the light attraction peak value of large yellow croaker).

[0122] Illumination intensity: 2000 - 3000 lux, pulse frequency 2 Hz (simulating the twinkling of plankton).

[0123] For further optimized solutions, targeting conventional water area fishing;

[0124] Deploy a sound wave emission array and underwater monitoring equipment in the target sea area, and initialize the AI model parameters.

[0125] According to the otolith resonance frequency database, select the initial sound wave frequency as 300 Hz, and emit a composite sound wave that is the superposition of an orthogonal chirp signal and a single-frequency resonance signal.

[0126] Obtain the fish school density, body length distribution and movement trajectory in real time through underwater cameras and sonars. Use the CNN model to identify the fish body length in real time and calculate the ratio of adult fish to juvenile fish.

[0127] If the proportion of juvenile fish exceeds 5%, the AI model automatically adjusts the sound wave frequency to 400 Hz and reduces the sound pressure level by 10 dB.

[0128] When the aggregation degree of adult fish reaches more than 80%, start the double-layer trawl net. The juvenile fish escape through the upper mesh, and the adult fish enter the codend.

[0129] The fishing effect of this embodiment can reduce the miscapture rate of juvenile fish to 2.8%, increase the fishing efficiency of adult fish to 93%, and the survival rate of adult fish > 97%.

[0130] For further optimized solutions, targeting turbid water area fishing:

[0131] In turbid water areas, enable sonar to replace optical monitoring, and enhance the recognition accuracy by combining otolith resonance sound characteristics (such as the feeding sound spectrum).

[0132] Noise suppression uses an LMS adaptive filter to suppress environmental noises such as ship engines and wave slapping, and the signal-to-noise ratio is increased by ≥ 15 dB.

[0133] According to the fish school response, the AI model dynamically adjusts the sound wave parameters to ensure the induction efficiency.

[0134] In strong current areas, the adaptive algorithm increases the sound pressure level to 165 dB, and the induction efficiency only drops by 8%.

[0135] The fishing effect of this embodiment can reduce the miscapture rate of juvenile fish to below 3%, which is significantly better than traditional methods.

[0136] The fishing effect of this embodiment can increase the fishing efficiency by 30%, and the survival rate of adult fish > 95%.

[0137] This embodiment can be extended to other Sciaenidae fish (such as Bahaba taipingensis) through AI dynamic learning.

[0138] In the sonar-assisted mode in turbid waters of this embodiment, the accuracy rate of body length recognition remains at 91%, while it drops to 75% in the optical mode. In a strong current area, the adaptive algorithm increases the sound pressure level to 165 dB, and the induction efficiency only decreases by 8%.

[0139] Embodiment 2

[0140] Based on the same inventive concept, this embodiment also provides a selective capture system for large yellow croaker parent fish based on otolith resonance and artificial intelligence, including:

[0141] A resonance frequency measurement module, used to measure the resonance frequencies of otoliths of large yellow croakers with different body lengths through experiments and establish a "body length - resonance frequency" mapping model;

[0142] A signal design module, used to design an adjustable-frequency sound wave signal by analyzing the resonance frequency differences of otoliths of large yellow croakers with different body lengths;

[0143] An induction monitoring module, used to induce target parent fish according to the adjustable-frequency sound wave signal and real-time monitor the fish school dynamics and environmental noise through underwater acoustic sensors, optical sensors and environmental parameter detection;

[0144] A feedback regulation module, used to use a deep learning model to analyze the fish school feedback signal in real time, identify the fish body length and distribution, and optimize the sound wave parameters;

[0145] A capture module, used to combine sound wave induction and light trapping technology, guide the parent fish to concentrate in the target area and then start mechanical netting to complete the capture.

[0146] The selective capture system for large yellow croaker parent fish based on otolith resonance and artificial intelligence provided by this embodiment has all the advantages of the selective capture method for large yellow croaker parent fish based on otolith resonance and artificial intelligence provided by Embodiment 1.

[0147] Embodiment 3

[0148] This embodiment also discloses a computer device, including a memory, a processor and a computer program stored on the memory, and the processor executes the computer program to implement the steps of the method described in Embodiment 1.

[0149] Embodiment 4

[0150] This embodiment also discloses a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in Embodiment 1 are implemented.

[0151] Embodiment 5

[0152] This embodiment also discloses a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the method described in Embodiment 1 are implemented.

[0153] The above are only the preferred specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for selectively fishing large yellow croaker broodstock based on otolith resonance and artificial intelligence, characterized in that Including: Measuring the resonance frequencies of otoliths of large yellow croakers with different body lengths through experiments, and establishing a "body length - resonance frequency" mapping model; Designing an adjustable - frequency acoustic wave signal by analyzing the differences in the resonance frequencies of otoliths of large yellow croakers with different body lengths; Inducing target broodstock according to the adjustable - frequency acoustic wave signal, and real - time monitoring the fish school dynamics and environmental noise through underwater acoustic sensors, optical sensors and environmental parameter detection; Using a deep - learning model to analyze the fish school feedback signal in real time, identifying the fish body length and distribution, and optimizing the acoustic wave parameters; Combining acoustic wave induction and light - trapping technology, guiding the broodstock to concentrate in the target area and then starting mechanical netting to complete the fishing.

2. The method according to claim 1, wherein The process of measuring the resonance frequencies of otoliths of large yellow croakers with different body lengths through experiments includes: Collecting samples of large yellow croakers with different body lengths, conducting vibration mode analysis on the otoliths with a laser Doppler vibrometer, recording the resonance frequencies, and determining the otolith resonance frequency ranges of large yellow croakers with different body lengths through numerical mode analysis; among them, the resonance frequency of adult fish with a body length greater than or equal to 30 cm is 200 - 500 Hz, and the resonance frequency of juvenile fish is 800 - 1200 Hz.

3. The method according to claim 1, wherein The formula expression of the resonance frequency is: where E is the elastic modulus of the otolith; ρ is the density of the otolith; V is the Poisson's ratio; L and w are the length and width of the otolith respectively.

4. The method according to claim 1, wherein The process of inducing target broodstock according to the adjustable - frequency acoustic wave signal includes: Using a broadband underwater acoustic transducer to emit a composite acoustic wave that superimposes an orthogonal chirp signal and a single - frequency resonance signal, covering the target frequency range; and using an embedded adaptive algorithm to adjust the acoustic wave intensity and frequency band in real time according to the environmental noise.

5. The method according to claim 1, wherein The process of real - time monitoring the fish school dynamics and environmental noise through underwater acoustic sensors, optical sensors and environmental parameter detection includes: Obtaining the fish school density, body length distribution and movement trajectory through an underwater camera and sonar, collecting fish school response data in real time, and then using a convolutional neural network to identify the fish body length in real time and calculate the ratio of broodstock to juvenile fish.

6. The method according to claim 1, wherein The process of using a deep - learning model to analyze the fish school feedback signal in real time, identifying the fish body length and distribution, and optimizing the acoustic wave parameters includes: Using a convolutional neural network combined with reinforcement learning to optimize the acoustic wave parameters and dynamically adjust the acoustic wave signal; When the proportion of juvenile fish exceeds a preset threshold, automatically switch to a higher frequency band or reduce the sound intensity.

7. The method according to claim 1, wherein The method further includes: designing a double - layer trawl net; wherein, the upper layer of the double - layer trawl net is a 10 cm×10 cm mesh that allows juvenile fish to escape, and the lower layer is a 25 cm×25 cm mesh adapted to the size of broodstock.

8. The method according to claim 1, wherein The process of combining acoustic wave induction and light - trapping technology, guiding the broodstock to concentrate in the target area and then starting mechanical netting for fishing includes: By combining acoustic wave induction with green LEDs, a light trapping technique with a wavelength of 520 nm is used to simulate the flickering of plankton for inducing broodstock. When the aggregation degree of broodstock reaches over 80%, a double-layer trawl net is started, and the juvenile fish escape through the upper mesh while the broodstock enter the codend net.

9. A larimichthys crocea parent selective fishing system based on otolith resonance and artificial intelligence, characterized in that, It includes: A resonance frequency measurement module for experimentally measuring the resonance frequencies of the otoliths of large yellow croakers with different body lengths and establishing a "body length - resonance frequency" mapping model; A signal design module for designing an adjustable frequency acoustic wave signal by analyzing the resonance frequency differences of the otoliths of large yellow croakers with different body lengths; An induction monitoring module for inducing target broodstock according to the adjustable frequency acoustic wave signal and real-time monitoring of the fish school dynamics and environmental noise through underwater acoustic sensors, optical sensors and environmental parameter detection; A feedback regulation module for using a deep learning model to analyze the fish school feedback signal in real time, identify the fish body length and distribution, and optimize the acoustic wave parameters; A fishing module for combining acoustic wave induction with light trapping technology, guiding the broodstock to concentrate in the target area and then starting mechanical netting to complete the fishing.

Images