A light purse seine intelligent control method and system based on multi-fishing boat data analysis

Abstract

The application discloses a kind of based on multi-fishing boat data analysis's light purse seine intelligent control method and system.Method includes: constructing multi-fishing boat data acquisition network, setting different light purse seine scheme to carry out fish detection;Light, water environment and fish detection parameter are obtained, variable set is constructed and is generated light control-fish relationship model by multiple linear regression;Introduce bayesian optimization algorithm, with relationship equation as base function, combined with constraint condition search optimal parameter, generate optimized light control and target environmental parameter;Screening preferred fishing boat, apply optimized light control scheme and set purse seine scheme.The application integrates multi-fishing boat data, establishes accurate relationship model, realizes light purse seine intelligent control, improves fishing efficiency and resource utilization, solves the problem that prior art relies on artificial experience and low intelligent degree, applicable to various light purse seine fishing scenes.

Description

Technical Field

[0001] This invention relates to the field of intelligent fisheries, and more specifically, to an intelligent control method and system for lighted purse seine nets based on multi-vessel data analysis. Background Technology

[0002] Light-based purse seine fishing, an important method in distant-water fishing, operates on the principle of attracting fish to light by creating a specific lighting environment, which is then used to catch them. However, with the increasing scale of distant-water fishing, traditional light-based purse seine fishing has revealed numerous technical bottlenecks. Currently, most fishing vessels still rely on crew experience to manually adjust parameters such as brightness, spectrum, and flashing frequency, lacking precise adaptation to the marine environment, fish behavior, and operational scenarios. This extensive control method not only leads to excessive energy consumption and accelerated wear and tear on fishing gear but also results in low fish aggregation efficiency and unstable catch yields, severely restricting the economic benefits and sustainability of fishing operations.

[0003] In addition, existing technologies often lack comprehensive collaborative data analysis capabilities for multiple fishing vessels, lack comprehensive efficiency analysis and dynamic control of lighting based on the fish aggregation characteristics of multiple fishing vessels, and lack effective fusion analysis mechanisms, which prevents the full exploitation of data value and further restricts the intelligent and efficient development of light-controlled seine nets. Summary of the Invention

[0004] This invention overcomes the shortcomings of existing technologies and proposes an intelligent control method and system for light-filled purse seine nets based on multi-vessel data analysis.

[0005] The first aspect of this invention provides an intelligent control method for lighted purse seine nets based on multi-vessel data analysis, comprising:

[0006] S1: In the target waters, a data collection network is built based on multiple fishing vessel terminals. In multiple preset cycles, different light-filled netting schemes are set for multiple fishing vessels to detect fish schools.

[0007] S2: Acquire lighting parameters, aquatic environment parameters, and fish detection parameters of multiple fishing boats through a data acquisition network. Construct a first set of variables based on the lighting parameters and aquatic environment parameters, and construct a second set of variables based on the fish detection parameters. Fit the linear mapping relationship between the first and second variables through multiple linear regression, and generate a lighting control-fish school relationship model for each fishing boat.

[0008] S3: Using multiple relational models as objective basis functions, a Bayesian optimization algorithm is introduced. The goal is to maximize the mean of the objective basis functions, with the first variable as the decision variable and the second variable as the objective variable. The minimum range of fishing boat lighting control is used as the constraint to search for optimal parameters. Based on the optimal parameter results, optimized lighting control parameters and target environment parameters are generated.

[0009] S4: Analyze the difference between the aquatic environmental parameters and the target environmental parameters of each fishing vessel, and select the preferred fishing vessels. In multiple purse seine operation cycles, generate a lighting control scheme based on the optimized lighting control parameters and apply it to the preferred fishing vessels. Set the purse seine scheme through the preferred fishing vessels.

[0010] In this solution, S1 and S2 specifically refer to:

[0011] Multiple fishing boats are deployed based on the size of the target water area, and a data collection network is built for the terminals of multiple fishing boats through the Internet of Things.

[0012] For each fishing vessel, a monitoring unit is set up to acquire parameters. The monitoring unit includes an underwater monitoring device and a light control device.

[0013] Aquatic environmental parameters are acquired through underwater monitoring devices, including temperature, dissolved oxygen, pH, and salinity.

[0014] In this solution, S1 and S2 specifically include:

[0015] The lighting configuration of each fishing boat on and under the water is controlled by a lighting control device, and the lighting parameters are obtained.

[0016] Lighting parameters include color temperature, brightness ratio, duration, flicker value, and light coverage ratio;

[0017] Image data is acquired by an underwater camera device, and the image data is preprocessed by noise reduction and grayscale conversion. Image contour features are extracted using LBP mode, and YOLOv5 is introduced for target recognition. The number and density of fish are counted based on the recognition results, and the number and density of fish are used as fish detection parameters.

[0018] In this solution, S2 specifically includes:

[0019] The lighting parameters, aquatic environment parameters, and fish detection parameters of multiple fishing vessels are exchanged with the system terminal through a data acquisition network;

[0020] In each fishing vessel, the corresponding aquatic environmental parameters are compared with the environmental adaptation parameters, and the environmental adaptability is calculated.

[0021] In each fishing boat, the lighting parameters and environmental adaptability collected over multiple periods are integrated, and the data is cleaned and outliers are removed. The data is then sorted by time to construct the first set of variables.

[0022] In the fish school detection parameters, the number of fish schools identified in each preset period is weighted and averaged with the fish school density to obtain the fish school aggregation coefficient.

[0023] In each fishing vessel, the fish aggregation coefficient obtained from multiple periods is used as the second set of variables.

[0024] In this solution, S2 further includes:

[0025] A multiple linear regression equation is constructed based on the first and second variables;

[0026] The linear mapping relationship between the first and second variables was fitted by multiple linear regression. The regression coefficients were solved by the least squares method, and the first and second variable sets were imported for fitting evaluation.

[0027] The fitting equations were subjected to a significance test, and the fitting equations for each fishing boat were determined as the light control-fish school relationship model after passing the test.

[0028] In this solution, S3 includes:

[0029] Multiple relational models are used as objective basis functions, and the maximization of the mean of these objective basis functions is taken as the objective function. ;

[0030] Obtain the control range of the lighting parameters for each fishing vessel and set constraint condition 1. Filter out the minimum and maximum values ​​based on the environmental adaptability of each fishing vessel and set constraint condition 2.

[0031] Introducing the Bayesian optimization algorithm, using the first variable as the decision variable and the second variable as the target variable, and using constraints 1 and 2 as the constraint range of the optimal decision variable to search for the optimal parameters;

[0032] The root search result yields the optimal decision variable and the optimal target variable;

[0033] Data analysis is performed based on the optimal decision variables to obtain the optimized lighting control parameters and their adaptability to the target environment.

[0034] In this solution, S4 specifically refers to:

[0035] Analyze the difference between the environmental adaptability of each fishing vessel and the target environmental parameters, sort them based on the difference, and select the top K fishing vessels as the core selection;

[0036] After multiple preset cycles, multiple seine operation cycles are set, and multiple optimized light control parameters are sent to the core fishing vessel and the seine scheme is set.

[0037] The fencing solution includes lighting layout, activation timing, power allocation, and resource capture.

[0038] The resource capture scheme prioritizes purse seine fishing resources based on the selection of fishing vessels.

[0039] In this solution, S4 includes:

[0040] The optimized lighting control parameters, target environmental parameters, and netting scheme for each analysis cycle are sent to the fishing vessel terminal via the Internet of Things.

[0041] Fishing vessel terminals include computer terminals and mobile terminals.

[0042] A second aspect of the present invention also provides an intelligent control system for light-based purse seine nets based on multi-vessel data analysis. The system includes a memory, a processor, and a data interface. The memory includes an intelligent control program for light-based purse seine nets based on multi-vessel data analysis. When executed by the processor, the intelligent control program for light-based purse seine nets based on multi-vessel data analysis performs the following steps:

[0043] S1: In the target waters, a data collection network is built based on multiple fishing vessel terminals. In multiple preset cycles, different light-filled netting schemes are set for multiple fishing vessels to detect fish schools.

[0044] S2: Acquire lighting parameters, aquatic environment parameters, and fish detection parameters of multiple fishing boats through a data acquisition network. Construct a first set of variables based on the lighting parameters and aquatic environment parameters, and construct a second set of variables based on the fish detection parameters. Fit the linear mapping relationship between the first and second variables through multiple linear regression, and generate a lighting control-fish school relationship model for each fishing boat.

[0045] S3: Using multiple relational models as objective basis functions, a Bayesian optimization algorithm is introduced. The goal is to maximize the mean of the objective basis functions, with the first variable as the decision variable and the second variable as the objective variable. The minimum range of fishing boat lighting control is used as the constraint to search for optimal parameters. Based on the optimal parameter results, optimized lighting control parameters and target environment parameters are generated.

[0046] S4: Analyze the difference between the aquatic environmental parameters and the target environmental parameters of each fishing vessel, and select the preferred fishing vessels. In multiple purse seine operation cycles, generate a lighting control scheme based on the optimized lighting control parameters and apply it to the preferred fishing vessels. Set the purse seine scheme through the preferred fishing vessels.

[0047] A third aspect of the present invention also provides a computer-readable storage medium comprising a smart control program for a light-based purse seine based on multi-vessel data analysis. When executed by a processor, the smart control program for a light-based purse seine based on multi-vessel data analysis implements the steps of the smart control method for a light-based purse seine based on multi-vessel data analysis as described in any of the preceding claims.

[0048] This invention discloses an intelligent control method and system for light-based purse seine fishing based on multi-vessel data analysis. The method includes: constructing a multi-vessel data acquisition network; setting different light-based purse seine schemes for fish detection; acquiring parameters related to lighting, water environment, and fish detection; constructing a variable set and generating a lighting control-fish school relationship model through multiple linear regression; introducing a Bayesian optimization algorithm, using the relationship equation as the basic function, and combining constraints to search for optimal parameters, generating optimized lighting control and target environment parameters; selecting the best fishing vessels, applying the optimized lighting control scheme, and setting the purse seine scheme. This invention integrates multi-vessel data, establishes a precise relationship model, and achieves intelligent control of light-based purse seine fishing, improving fishing efficiency and resource utilization. It solves the problems of existing technologies relying on manual experience and having low levels of intelligence, and is applicable to various light-based purse seine fishing scenarios. Attached Figure Description

[0049] Figure 1 A flowchart of a smart control method for lighted seine nets based on multi-vessel data analysis according to the present invention is shown;

[0050] Figure 2 The flowchart for obtaining the variable set of this invention is shown;

[0051] Figure 3 A block diagram of a light-based purse seine intelligent control system based on multi-vessel data analysis is shown. Detailed Implementation

[0052] To better understand the above-mentioned objects, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It is understood that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0053] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0054] Figure 1 The flowchart of a smart control method for lighted seine nets based on multi-vessel data analysis according to the present invention is shown.

[0055] like Figure 1As shown, the first aspect of the present invention provides an intelligent control method for lighted purse seine nets based on multi-vessel data analysis, comprising:

[0056] S1: In the target waters, a data collection network is built based on multiple fishing vessel terminals. In multiple preset cycles, different light-filled netting schemes are set for multiple fishing vessels to detect fish schools.

[0057] S2: Acquire lighting parameters, aquatic environment parameters, and fish detection parameters of multiple fishing boats through a data acquisition network. Construct a first set of variables based on the lighting parameters and aquatic environment parameters, and construct a second set of variables based on the fish detection parameters. Fit the linear mapping relationship between the first and second variables through multiple linear regression, and generate a lighting control-fish school relationship model for each fishing boat.

[0058] S3: Using multiple relational models as objective basis functions, a Bayesian optimization algorithm is introduced. The goal is to maximize the mean of the objective basis functions, with the first variable as the decision variable and the second variable as the objective variable. The minimum range of fishing boat lighting control is used as the constraint to search for optimal parameters. Based on the optimal parameter results, optimized lighting control parameters and target environment parameters are generated.

[0059] S4: Analyze the difference between the aquatic environmental parameters and the target environmental parameters of each fishing vessel, and select the preferred fishing vessels. In multiple purse seine operation cycles, generate a lighting control scheme based on the optimized lighting control parameters and apply it to the preferred fishing vessels. Set the purse seine scheme through the preferred fishing vessels.

[0060] It is understood here that, in the embodiments, different light-controlled netting schemes can be set for multiple fishing boats to detect fish schools. The specific scheme is based on the environmental conditions and preset scheme settings, which are generally the default schemes. Based on the consistency of the environment, the same light control scheme can also be set for monitoring and evaluation at preset cycles.

[0061] According to an embodiment of the present invention, S1 and S2 specifically include:

[0062] Multiple fishing boats are deployed based on the size of the target water area, and a data collection network is built for the terminals of multiple fishing boats through the Internet of Things.

[0063] For each fishing vessel, a monitoring unit is set up to acquire parameters. The monitoring unit includes an underwater monitoring device and a light control device.

[0064] Aquatic environmental parameters are acquired through underwater monitoring devices, including temperature, dissolved oxygen, pH, and salinity.

[0065] It can be understood here that underwater monitoring devices include water area sensors and underwater camera devices.

[0066] According to an embodiment of the present invention, S1 and S2 specifically include:

[0067] The lighting configuration of each fishing boat on and under the water is controlled by a lighting control device, and the lighting parameters are obtained.

[0068] Lighting parameters include color temperature, brightness ratio, duration, flicker value, and light coverage ratio;

[0069] Image data is acquired by an underwater camera device, and the image data is preprocessed by noise reduction and grayscale conversion. Image contour features are extracted using LBP mode, and YOLOv5 is introduced for target recognition. The number and density of fish are counted based on the recognition results, and the number and density of fish are used as fish detection parameters.

[0070] Here, it can be understood that the color temperature can be set between 5000-6500K, with the specific parameter value expressed as a corresponding numerical value. For fishing boats, the flicker value for the surface and underwater areas is generally set within 1-5Hz. The brightness ratio is the ratio of the surface and underwater light brightness settings, which can be set from 0-100%. The light coverage ratio is the ratio of the surface and underwater light coverage areas, with underwater coverage set to 50-100 meters and surface coverage set to 100 meters. The specific ratios are calculated and then used to evaluate the relationship between light parameter control and fish-gathering effects. Furthermore, the light parameters are based on the actual water environment and preset schemes for multiple fishing boats; the surface and underwater areas can use the same light configuration values. The duration can also be set differently for the surface and underwater areas. For example, the lights can be turned on for a period of time on the surface first, and then the underwater lights can be turned on to achieve efficient fish gathering. In this case, the duration ratio can be used as a light control parameter, and the activation sequence scheme can be set using this parameter.

[0071] Before introducing YOLOv5 for target recognition, standardized image data of the target fish group can be used for model training, and a scale distribution adapted to underwater targets can be set. The training process can use a 2:8 ratio to set the test set and training set. LBP stands for Local Binary Pattern.

[0072] Figure 2 A flowchart illustrating the variable set acquisition process of this invention is shown.

[0073] According to an embodiment of the present invention, step S2 specifically includes:

[0074] The lighting parameters, aquatic environment parameters, and fish detection parameters of multiple fishing vessels are exchanged with the system terminal through a data acquisition network;

[0075] In each fishing vessel, the corresponding aquatic environmental parameters are compared with the environmental adaptation parameters, and the environmental adaptability is calculated.

[0076] In each fishing boat, the lighting parameters and environmental adaptability collected over multiple periods are integrated, and the data is cleaned and outliers are removed. The data is then sorted by time to construct the first set of variables.

[0077] In the fish school detection parameters, the number of fish schools identified in each preset period is weighted and averaged with the fish school density to obtain the fish school aggregation coefficient.

[0078] In each fishing vessel, the fish aggregation coefficient obtained from multiple periods is used as the second set of variables.

[0079] It can be understood here that the environmental adaptation parameters are the optimal environmental parameters suitable for the activity of the target fish school. These parameters are used to calculate environmental factors and are incorporated into the lighting control analysis process to increase the dimension of analysis of the influence of environmental factors. The higher the environmental adaptability, the closer each of the aquatic environmental parameters is to the environmental adaptation parameters. In the weighted average, the weight value of fish density is generally greater than that of fish number.

[0080] According to an embodiment of the present invention, step S2 further includes:

[0081] A multiple linear regression equation is constructed based on the first and second variables;

[0082] The linear mapping relationship between the first and second variables was fitted by multiple linear regression. The regression coefficients were solved by the least squares method, and the first and second variable sets were imported for fitting evaluation.

[0083] The fitting equations were subjected to a significance test, and the fitting equations for each fishing boat were determined as the light control-fish school relationship model after passing the test.

[0084] The multiple linear regression equation is as follows:

[0085] ;

[0086] Here, K1-Kn are the regression coefficients. -xn is the first variable, and there are multiple variables, such as lighting parameters and environmental adaptability, while Y is the fish aggregation coefficient.

[0087] According to an embodiment of the present invention, S3 includes:

[0088] Multiple relational models are used as objective basis functions, and the maximization of the mean of these objective basis functions is taken as the objective function. ;

[0089] Obtain the control range of the lighting parameters for each fishing vessel and set constraint condition 1. Filter out the minimum and maximum values ​​based on the environmental adaptability of each fishing vessel and set constraint condition 2.

[0090] Introducing the Bayesian optimization algorithm, using the first variable as the decision variable and the second variable as the target variable, and using constraints 1 and 2 as the constraint range of the optimal decision variable to search for the optimal parameters;

[0091] The root search result yields the optimal decision variable and the optimal target variable;

[0092] Data analysis is performed based on the optimal decision variables to obtain the optimized lighting control parameters and their adaptability to the target environment.

[0093] This can be understood as obtaining the control range of the lighting parameters for each fishing vessel and setting constraint condition 1, using the intersection of the control ranges as the constraint range, and setting constraint conditions. The main purpose of setting the search range of the decision variables here is to limit the search to the hardware limitations of the fishing vessels.

[0094] The objective function is as follows:

[0095] ;

[0096] in, The objective function value, For the i-th relational model, the corresponding numerical value is used to calculate the mean, where N is the total number of relational models (i.e., the total number of fishing boats). Here, in calculating the objective function value, each... Maximize the search by setting the same set of decision variables.

[0097] The first variable includes multiple independent variables x, and the objective function value is the mean of the second variable.

[0098] According to an embodiment of the present invention, S4 specifically includes:

[0099] Analyze the difference between the environmental adaptability of each fishing vessel and the target environmental parameters, sort them based on the difference, and select the top K fishing vessels as the core selection;

[0100] After multiple preset cycles, multiple seine operation cycles are set, and multiple optimized light control parameters are sent to the core fishing vessel and the seine scheme is set.

[0101] The fencing solution includes lighting layout, activation timing, power allocation, and resource capture.

[0102] The resource capture scheme prioritizes purse seine fishing resources based on the selection of fishing vessels.

[0103] Here, the target environmental parameter is understood to be environmental adaptability. The sorting is in descending order. The purpose of selecting preferred fishing vessels is to identify areas that have a positive impact on the water environment where the vessels are located, and to optimize the lighting control effect and fish density distribution of these preferred vessels. For non-preferred core vessels, the default lighting control scheme is used, achieving dynamic optimization and control of the lighting effect. In the priority settings, preferred vessels are given higher priority, and non-preferred vessels are given lower priority to ensure maximum efficiency of purse seine fishing.

[0104] The lighting layout can be set according to brightness, flicker value, etc., the activation sequence can be set according to duration, and the power distribution can be dynamically allocated according to brightness ratio, duration, color temperature, etc.

[0105] According to an embodiment of the present invention, S4 includes:

[0106] The optimized lighting control parameters, target environmental parameters, and netting scheme for each analysis cycle are sent to the fishing vessel terminal via the Internet of Things.

[0107] Fishing vessel terminals include computer terminals and mobile terminals.

[0108] It should be noted that this invention can effectively uncover the migration and distribution patterns of fish schools in various complex aquatic environments, assess the potential adaptability and tendency of fish schools to different lighting effects, and analyze the linear mapping between lighting state factors, environmental factors and fish density distribution based on the periodic lighting scheme of multiple fishing boats. Furthermore, it can search for optimal lighting parameters based on optimization algorithms. Here, dynamic analysis of environmental factors is introduced to identify fishing boat areas that respond positively to the default lighting scheme and fish aggregation, and core fishing boats are set, along with corresponding optimization schemes.

[0109] This invention can effectively solve the problems of low efficiency, high energy consumption, and blind decision-making in traditional operation modes, and promote the development of fisheries in multiple scenarios towards high efficiency, energy saving, and precision.

[0110] According to an embodiment of the present invention, the preferred fishing vessel further includes:

[0111] Select a target fishing vessel;

[0112] In multiple preset cycles, the average deviation ratio of fish detection parameters between the target fishing vessel and neighboring fishing vessels is analyzed.

[0113] The average deviation ratio is analyzed by calculating the deviation rate between the number of fish and the density of fish.

[0114] Calculate the average deviation ratio of all fishing vessels, mark the fishing vessels with an average deviation ratio higher than the preset ratio and consistent environmental adaptability, and obtain the preferred fishing vessels.

[0115] The environmental adaptability consistency is judged by the deviation rate of the environmental adaptability values ​​between the target fishing vessel and the neighboring fishing vessels being within 10%.

[0116] Here, in addition to selecting fishing vessels through environmental factor deviation analysis, based on waters with high environmental consistency, we can also evaluate fishing vessels with a certain positive impact on fish density and those with high fluctuations in fish school influence by assessing the correlation between the density effect of fishing vessels and neighboring fishing vessels, and select fishing vessels with high fluctuations in fish school influence as preferred fishing vessels. Fishing vessels with high fluctuations in fish school influence can achieve better fish school density effect through optimization of certain lighting control parameters.

[0117] Neighboring fishing vessels are those that are geographically adjacent or within a certain preset range, such as a certain radius.

[0118] Figure 3 A block diagram of a light-based purse seine intelligent control system based on multi-vessel data analysis is shown.

[0119] A second aspect of the present invention also provides an intelligent control system for lighted purse seine nets based on multi-vessel data analysis. The system includes a memory, a processor, and a data interface. The data interface is connected to a system terminal and a fishing vessel terminal, and a data transmission link is established through the Internet of Things (IoT). The memory includes an intelligent control program for lighted purse seine nets based on multi-vessel data analysis. When the processor executes the intelligent control program for lighted purse seine nets based on multi-vessel data analysis, it performs the following steps:

[0120] S1: In the target waters, a data collection network is built based on multiple fishing vessel terminals. In multiple preset cycles, different light-filled netting schemes are set for multiple fishing vessels to detect fish schools.

[0121] S2: Acquire lighting parameters, aquatic environment parameters, and fish detection parameters of multiple fishing boats through a data acquisition network. Construct a first set of variables based on the lighting parameters and aquatic environment parameters, and construct a second set of variables based on the fish detection parameters. Fit the linear mapping relationship between the first and second variables through multiple linear regression, and generate a lighting control-fish school relationship model for each fishing boat.

[0122] S3: Using multiple relational models as objective basis functions, a Bayesian optimization algorithm is introduced. The goal is to maximize the mean of the objective basis functions, with the first variable as the decision variable and the second variable as the objective variable. The minimum range of fishing boat lighting control is used as the constraint to search for optimal parameters. Based on the optimal parameter results, optimized lighting control parameters and target environment parameters are generated.

[0123] S4: Analyze the difference between the aquatic environmental parameters and the target environmental parameters of each fishing vessel, and select the preferred fishing vessels. In multiple purse seine operation cycles, generate a lighting control scheme based on the optimized lighting control parameters and apply it to the preferred fishing vessels. Set the purse seine scheme through the preferred fishing vessels.

[0124] When the system is running, it can perform one or more steps of the above-described intelligent control method for light-filled embankments based on multi-vessel data analysis.

[0125] A third aspect of the present invention also provides a computer-readable storage medium comprising a smart control program for a light-based purse seine based on multi-vessel data analysis. When executed by a processor, the smart control program for a light-based purse seine based on multi-vessel data analysis implements the steps of the smart control method for a light-based purse seine based on multi-vessel data analysis as described in any of the preceding claims.

[0126] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this application can be generated, in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic cable, data subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital universal optical disc), or a semiconductor medium (e.g., solid-state drive). In the various embodiments of this application, unless otherwise specified or logically conflicting, the terminology and / or descriptions between different embodiments are consistent and can be mutually referenced. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0127] In the embodiments of this application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone, where A and B can be singular or plural. In the textual description of the embodiments of this application, the character " / " generally indicates that the preceding and following associated objects have an "or" relationship. In this application, "first," "second," and various numerical designations are merely for descriptive convenience and are not used to limit the scope of the embodiments of this application. For example, they are used to distinguish different messages, rather than to describe a specific order or sequence.

[0128] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application. The order of the process numbers does not imply the order of execution; the execution order of each process should be determined by its function and internal logic.

[0129] Finally, it should be noted that the above description is only a specific implementation of this application, but the protection scope of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the protection scope of this application.

Claims

1. A light purse seine intelligent control method based on multi-fishing boat data analysis, characterized in that, The method comprises the following steps: S1: In the target water area, a data collection network is constructed based on the terminals of multiple fishing vessels. In multiple preset periods, different light purse seine schemes are set for the multiple fishing vessels to detect fish schools; S2: The light parameters, water environment parameters, and fish school detection parameters of the multiple fishing vessels are obtained through the collection network. The first variable set is constructed based on the light parameters and water environment parameters, and the second variable set is constructed based on the fish school detection parameters. The linear mapping relationship between the first variable and the second variable is fitted through multiple linear regression, and the light control-fish school relationship model of each fishing vessel is generated; S3: The multiple relationship models are taken as target basic functions, the Bayesian optimization algorithm is introduced, the mean maximization of the target basic function is taken as the objective function, the first variable is taken as the decision variable, the second variable is taken as the target variable, and the optimal parameter search is performed with the minimum range of the fishing vessel light control as the constraint condition. The optimal light control parameters and target environment parameters are generated according to the optimal parameter result; S4: The difference between the water environment parameters of each fishing vessel and the target environment parameters is analyzed, and the preferred fishing vessel is selected. In multiple purse seine operation periods, the light control scheme based on the optimized light control parameters is generated and applied to the preferred fishing vessel, and the purse seine scheme is set through the preferred fishing vessel; In S1 and S2, the specific steps are as follows: According to the size of the target water area, multiple fishing vessels are set up, and a data collection network is constructed through the Internet of Things. For each fishing vessel, a monitoring unit is set up to obtain parameters. The monitoring unit includes an underwater monitoring device and a light control device. The water environment parameters are obtained through the underwater monitoring device. The water environment parameters include temperature, dissolved oxygen content, pH, and salinity. In S1 and S2, the specific steps include: The light configuration of each fishing vessel on the water surface and underwater is controlled through the light control device, and the light parameters are obtained. The light parameters include color temperature, brightness ratio, duration, frequency flash value, and light coverage range ratio. Image data is collected through the underwater camera, and the image data is preprocessed by noise reduction and grayscale. The image contour features are extracted using the LBP mode, and the yolov5 is introduced for target recognition. The number and density of fish schools are calculated based on the recognition results, and the number and density of fish schools are taken as the fish school detection parameters. In S2, the specific steps include: The light parameters, water environment parameters, and fish school detection parameters of the multiple fishing vessels are exchanged through the data collection network and the system terminal. In each fishing vessel, the corresponding water environment parameters and adaptive environment parameters are compared and the environmental adaptability is calculated. In each fishing vessel, the light parameters and environmental adaptability obtained in multiple periods are integrated, cleaned, and the numerical outliers are removed. The data is sorted by time dimension to construct the first variable set. In the fish school detection parameters, the number and density of fish schools identified in each preset period are weighted and averaged to obtain the fish school aggregation coefficient. In each fishing vessel, the fish school aggregation coefficient obtained in multiple periods is taken as the second variable set.

2. The method according to claim 1, wherein, S2 also includes: A multiple linear regression equation is constructed based on the first variable and the second variable. The linear mapping relationship between the first variable and the second variable is fitted by multiple linear regression, the regression coefficient is solved by least square method in the fitting process, and the first variable set and the second variable set are introduced for fitting evaluation; Significance test is performed on the fitted equation, and the fitted equation of each fishing boat is determined as the light control-fish school relationship model after passing the test.

3. The method according to claim 2, wherein, The S3 comprises: maximizing the mean of the target base functions as an objective function ; The light control parameter control range of each fishing boat is obtained and constraint condition 1 is set, the minimum and maximum values are screened according to the environmental fitness of each fishing boat, and constraint condition 2 is set; The Bayesian optimization algorithm is introduced, the first variable is used as the decision variable, the second variable is used as the target variable, and the constraint conditions 1 and 2 are used as the constraint range of the optimal decision variable for optimal parameter search; The optimal decision variable and the optimal target variable are obtained according to the search result; The optimal light control parameter and the target environmental fitness are obtained by data analysis according to the optimal decision variable.

4. The method according to claim 3, wherein, The S4 specifically comprises: The difference between the environmental fitness and the target environmental parameter of each fishing boat is analyzed, the difference is sorted, and the top K fishing boats are selected as the preferred fishing boats; After a plurality of preset periods, a plurality of purse seine operation periods are set, a plurality of optimized light control parameters are sent to the preferred fishing boats, and a purse seine scheme is set; The purse seine scheme includes a light arrangement scheme, an opening timing scheme, a power distribution scheme, and a captured resource scheme; The captured resource scheme sets the priority of the purse seine operation resources based on the preferred fishing boats.

5. The method of claim 4, wherein the method is characterized by, The S4 comprises: The optimized light control parameter, the target environmental parameter and the purse seine scheme of each analysis period are sent to the fishing boat terminal through the Internet of Things; The fishing boat terminal comprises a computer terminal and a mobile terminal.

6. A light purse seine intelligent control system based on multi-fishing boat data analysis, characterized in that, The system comprises a memory, a processor and a data interface, the memory comprises a light purse seine intelligent control program based on multi-fishing boat data analysis, and the light purse seine intelligent control program based on multi-fishing boat data analysis is executed by the processor to realize the steps of the light purse seine intelligent control method based on multi-fishing boat data analysis in claim 1.

7. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a light purse seine intelligent control program based on multi-fishing boat data analysis, and the light purse seine intelligent control program based on multi-fishing boat data analysis is executed by the processor to realize the steps of the light purse seine intelligent control method based on multi-fishing boat data analysis in any one of claims 1 to 5.

Images