Method and system for desalting lateolabrax japonicus offspring seeds based on characteristic data analysis

Abstract

The invention relates to the technical field of aquaculture data analysis, in particular to a lateolabrax japonicus fry desalination method and system based on feature data analysis. Importing a real-time working parameter data set of the desalination pool into the water body salinity dynamic prediction model for prediction, and predicting to obtain predicted salinity dynamic change data of the desalination pool in a preset time period; performing salinity dynamic analysis on the desalination pool according to the predicted salinity dynamic change data of the desalination pool in the preset time period, and judging whether the salinity of the water body of the desalination pool needs to be adjusted or not according to an analysis result; and if the salinity of the water body of the desalination pool needs to be adjusted, generating a corresponding regulation and control scheme, and adjusting the salinity of the water body of the desalination pool according to the generated regulation and control scheme. The salinity of the desalination pond can be accurately controlled, a suitable salinity environment is provided for lateolabrax japonicus seedlings, and the survival rate and the growth quality of the seedlings are improved.

Description

Technical Field

[0001] The invention relates to the technical field of aquaculture data analysis, and in particular to a method and system for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis. Background Art

[0002] The culture of striped seabass has unique salinity requirements: high-salinity seawater is required for broodstock and fertilized eggs, while desalination culture is more suitable for the seedling stage. Moreover, desalination culture can significantly increase the growth rate and survival rate of striped seabass seedlings and reduce the incidence of diseases. In addition, considering the market demand and the salinity of local aquaculture waters, desalinated seedlings have more advantages. Traditional methods of desalination of striped seabass seedlings mainly rely on manual experience and simple monitoring methods. It is difficult to accurately control the changes in salinity during the desalination process, resulting in problems such as slow growth of seedlings, low survival rate, and susceptibility to disease.

[0003] In recent years, with the rapid development of technologies such as artificial intelligence, big data, and the Internet of Things, new ideas and methods have been provided for the desalination technology of japonica seabass seedlings. By real-time monitoring and analysis of parameters such as salinity, temperature, and dissolved oxygen in the desalination pond, and combining machine learning algorithms to establish a prediction model, intelligent control of the desalination process can be achieved. However, most of the existing japonica seabass seedling desalination technologies lack accurate prediction and intelligent regulation of the dynamic changes in the salinity of the desalination pond water, making it difficult to meet the differentiated management of the salinity requirements of japonica seabass seedlings at different growth stages. Summary of the invention

[0004] The present invention overcomes the shortcomings of the prior art and provides a method and system for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis.

[0005] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0006] The invention discloses a method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis, comprising the following steps:

[0007] Obtain the dynamic change data of salinity of the desalination pool under various working parameter conditions, and build a dynamic prediction model of water salinity based on the dynamic change data of salinity of the desalination pool under various working parameter conditions;

[0008] Acquire the real-time working parameter data of the desalination pool, and perform denoising on the collected real-time working parameter data to obtain a real-time working parameter data set of the desalination pool;

[0009] Importing the real-time working parameter data set of the desalination pool into the water body salinity dynamic prediction model for prediction, and predicting the dynamic change data of the salinity of the desalination pool within a preset time period;

[0010] Perform salinity dynamic analysis on the desalination pool according to the predicted salinity dynamic change data of the desalination pool within a preset time period, and determine whether it is necessary to adjust the salinity of the water in the desalination pool according to the analysis results;

[0011] If it is necessary to adjust the salinity of the water in the desalination pool, a corresponding adjustment plan is generated, and the salinity of the water in the desalination pool is adjusted according to the generated adjustment plan.

[0012] Preferably, the dynamic change data of the salinity of the desalination pool under various working parameter conditions are obtained, and a dynamic prediction model of water body salinity is constructed according to the dynamic change data of the salinity of the desalination pool under various working parameter conditions, specifically:

[0013] Obtain the dynamic change data of salinity in desalination pools under various working parameter conditions through the big data network;

[0014] Construct a Markov random field, use each working parameter as a conditional node of the Markov random field, and use the corresponding salinity dynamic change data as a variable node of the Markov random field;

[0015] Directed connections are made between each variable node and condition node to determine all possible value combinations of each variable node and condition node;

[0016] For each value combination, the potential function value between each value combination is calculated according to the Markov random field, and a potential function matrix is ​​constructed according to the calculated potential function value;

[0017] Constructing a water body salinity dynamic prediction model based on a graph neural network, and embedding the potential function matrix into the water body salinity dynamic prediction model for training;

[0018] When the prediction accuracy of the water body salinity dynamic prediction model is higher than a preset accuracy threshold, the final training parameters of the water body salinity dynamic prediction model are saved and a training end instruction is output.

[0019] Preferably, the collected real-time working parameter data is subjected to denoising to obtain a real-time working parameter data set of the desalination pool, specifically:

[0020] Preset a wavelet basis function and a decomposition layer number, perform wavelet decomposition on each real-time working parameter data according to the wavelet basis function and the decomposition layer number, and obtain wavelet coefficients of each real-time working parameter data at different scales, including low-frequency approximate coefficients and high-frequency detail coefficients;

[0021] Compare the wavelet coefficients of each real-time working parameter data at different scales with the preset threshold; directly set the wavelet coefficients smaller than the preset threshold to zero, and keep the wavelet coefficients larger than the preset threshold unchanged;

[0022] The wavelet coefficients after threshold processing are subjected to inverse wavelet transform, and the coefficients of different scales are gradually recombined to reconstruct the real-time working parameter data set of the denoised fading pool.

[0023] Preferably, the salinity dynamic analysis of the desalination pool is performed according to the predicted salinity dynamic change data of the desalination pool within a preset time period, and it is determined whether the salinity of the water in the desalination pool needs to be adjusted according to the analysis results, specifically:

[0024] Obtaining a preset desalination scheme, and obtaining the dynamic change data of the standard salinity of the desalination pool within a preset time period according to the preset desalination scheme;

[0025] The Spearman rank correlation coefficient algorithm is introduced to calculate the Spearman rank correlation coefficient between the standard salinity dynamic change data and the predicted salinity dynamic change data of the desalination pool within a preset time period based on the Spearman rank correlation coefficient algorithm;

[0026] comparing the Spearman rank correlation coefficient with a preset coefficient threshold;

[0027] If the Spearman rank correlation coefficient is greater than the preset coefficient threshold, it means that the salinity of the water in the desalination pool meets the preset requirements within the preset time period, and the salinity of the water is not regulated.

[0028] Preferably, if the salinity of the water in the desalination pool needs to be adjusted, a corresponding control scheme is generated, and the salinity of the water in the desalination pool is adjusted according to the generated control scheme, specifically:

[0029] If the Spearman rank correlation coefficient is not greater than a preset coefficient threshold, the salinity data difference between the standard salinity dynamic change data and the predicted salinity dynamic change data at each time node within the preset time period is calculated;

[0030] A salinity data deviation value threshold is preset, and the salinity data difference between the standard salinity dynamic change data and the predicted salinity dynamic change data at each time node is compared with the preset salinity data deviation value threshold;

[0031] The time node corresponding to the salinity data difference being greater than the preset salinity data deviation value threshold is defined as the salinity drift time node.

[0032] Preferably, if the salinity of the water in the desalination pool needs to be adjusted, a corresponding adjustment scheme is generated, and the salinity of the water in the desalination pool is adjusted according to the generated adjustment scheme, and the following steps are also included:

[0033] Obtaining a preset desalination scheme, obtaining a preset salinity requirement range value of the water body at the salinity drift time node in the preset desalination scheme; and obtaining preset working parameters of the desalination pool at the salinity drift time node in the preset desalination scheme;

[0034] Acquire structural parameter information of each control device in the desalination pool, and construct a simulation control system of the desalination pool according to the structural parameter information;

[0035] Introducing a genetic algorithm and setting a genetic algebra; regulating the preset working parameters of the desalination pool at the salinity drift time node in the simulation control system according to the genetic algebra, generating several control schemes, and obtaining predicted salinity data of the water body in the desalination pool after each control scheme is regulated;

[0036] Only the control scheme corresponding to the predicted salinity data within the preset salinity requirement range is extracted, and the control energy consumption value of the marked control scheme is obtained;

[0037] Constructing a sorting table, importing the control energy consumption values ​​of the marked control schemes into the sorting table, and sorting them from low to high according to the control energy consumption values ​​to generate a sorting result;

[0038] According to the sorting result, the marked control schemes are sent in sequence to the control terminal of the desalination pool, so that the salinity of the desalination pool is regulated in order according to each marked control scheme until the actual salinity data of the water body in the desalination pool is within the preset salinity requirement range.

[0039] The present invention also discloses a system for desalination of striped seabass seedlings based on characteristic data analysis, the system for desalination of striped seabass seedlings based on characteristic data analysis comprises a memory and a processor, the memory stores a program for desalination method of striped seabass seedlings based on characteristic data analysis, when the program for desalination method of striped seabass seedlings based on characteristic data analysis is executed by the processor, any step of the method for desalination of striped seabass seedlings based on characteristic data analysis is implemented.

[0040] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: by constructing a water body salinity dynamic prediction model, processing real-time working parameter data, predicting the dynamic change of salinity, analyzing and deciding whether to adjust the salinity, and generating and executing a control scheme, the salinity of the desalination pond can be accurately controlled. This helps to provide a suitable salinity environment for the japonica sea bass seedlings, improve the survival rate and growth quality of the seedlings, and reduce the stress response of the seedlings caused by salinity fluctuations or inappropriateness. The salinity can be flexibly adjusted according to the actual situation, realizing the scientific and intelligent management of the japonica sea bass seedling desalination process. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0042] Figure 1 The overall method flow chart of the desalination method;

[0043] Figure 2 A partial method flow chart of this desalination method;

[0044] Figure 3 System block diagram of the desalination system. DETAILED DESCRIPTION

[0045] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0046] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0047] like Figure 1 As shown, the present invention discloses a method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis, comprising the following steps:

[0048] S102: obtaining salinity dynamic change data of the desalination pool under various working parameter conditions, and constructing a water body salinity dynamic prediction model according to the salinity dynamic change data of the desalination pool under various working parameter conditions;

[0049] S104: acquiring real-time operating parameter data of the desalination pool, and performing denoising on the acquired real-time operating parameter data to obtain a real-time operating parameter data set of the desalination pool;

[0050] S106: importing the real-time working parameter data set of the desalination pool into the water body salinity dynamic prediction model for prediction, and predicting the dynamic change data of the salinity of the desalination pool within a preset time period;

[0051] S108: performing a salinity dynamic analysis on the desalination pool according to the predicted salinity dynamic change data of the desalination pool within a preset time period, and judging whether it is necessary to adjust the salinity of the water body in the desalination pool according to the analysis result;

[0052] S110: If the salinity of the water in the desalination pool needs to be adjusted, a corresponding adjustment scheme is generated, and the salinity of the water in the desalination pool is adjusted according to the generated adjustment scheme.

[0053] Among them, the working parameters include the injection speed and flow rate of fresh water, the amount of anti-stress drugs added, the water temperature, and the type and amount of bait added.

[0054] It should be noted that the salinity of the desalination pond can be accurately controlled through a series of steps such as building a dynamic prediction model for water salinity, processing real-time working parameter data, predicting dynamic changes in salinity, analyzing and deciding whether to adjust salinity, and generating and executing control plans. This helps to provide a suitable salinity environment for sea bass seedlings, improve the survival rate and growth quality of seedlings, and reduce problems such as seedling stress response caused by salinity fluctuations or inappropriateness. The salinity can also be flexibly adjusted according to actual conditions, realizing scientific and intelligent management of the sea bass seedling desalination process.

[0055] Preferably, the dynamic change data of the salinity of the desalination pool under various working parameter conditions are obtained, and a dynamic prediction model of water body salinity is constructed according to the dynamic change data of the salinity of the desalination pool under various working parameter conditions, specifically:

[0056] Obtain the dynamic change data of salinity in desalination pools under various working parameter conditions through the big data network;

[0057] Construct a Markov random field, use each working parameter as a conditional node of the Markov random field, and use the corresponding salinity dynamic change data as a variable node of the Markov random field;

[0058] Directed connections are made between each variable node and condition node to determine all possible value combinations of each variable node and condition node;

[0059] For each value combination, the potential function value between each value combination is calculated according to the Markov random field, and a potential function matrix is ​​constructed according to the calculated potential function value;

[0060] Constructing a water body salinity dynamic prediction model based on a graph neural network, and embedding the potential function matrix into the water body salinity dynamic prediction model for training;

[0061] When the prediction accuracy of the water body salinity dynamic prediction model is higher than a preset accuracy threshold, the final training parameters of the water body salinity dynamic prediction model are saved and a training end instruction is output.

[0062] Among them, the method steps for calculating the potential function value between each value combination based on the Markov random field are: first determine the dependency relationship between the variable node and the condition node according to the structure of the Markov random field, and then based on the acquired water body salinity dynamic change data and characteristic environmental data, combined with the probability distribution model (such as Gibbs distribution, etc.), in accordance with the principle of maximizing the overall probability or minimizing the energy, determine the potential function value corresponding to each value combination through data statistical analysis, parameter estimation and other means.

[0063] It should be noted that, firstly, the data of dynamic changes in salinity of the desalination pool under various working parameters (including injection speed and flow rate of fresh water, amount of anti-stress drugs, water temperature, type and amount of bait, etc.) are obtained through the big data network. Then, the Markov random field is constructed, with the working parameters as conditional nodes and the dynamic change data of salinity as variable nodes. The significance of this step is to establish the association framework between working parameters and salinity changes. The Markov random field can well describe this system with random characteristics and certain dependencies. After the variable nodes and conditional nodes are connected in a directed manner, all possible value combinations are determined. For each value combination, the potential function value is calculated according to the Markov random field. The potential function value reflects a kind of "energy" or "probability" state of the system under different value combinations, which is the basis for constructing the potential function matrix. The potential function matrix is ​​constructed according to the calculated potential function value, and then a water body salinity dynamic prediction model is constructed based on the graph neural network, and the potential function matrix is ​​embedded in the model for training. The graph neural network can effectively handle this data structure with complex node and edge relationships, and improve the prediction accuracy by continuously adjusting the model parameters. When the prediction accuracy of the water body salinity dynamic prediction model is higher than the preset accuracy threshold, the last training parameters of the model are saved and the training end instruction is output. This ensures that the model has sufficient accuracy and can be used for actual salinity dynamic prediction. The water body salinity dynamic prediction model constructed by this step can accurately predict the dynamic changes in salinity according to the working parameters of the desalination tank, so that during the desalination process of the striped sea bass seedlings, the changing trend of salinity can be known in advance, thereby providing a basis for accurately regulating the salinity of the water body in the desalination tank, helping to improve the success rate of the desalination of the striped sea bass seedlings, ensuring that the seedlings grow in a suitable salinity environment, reducing the stress response of the seedlings, slow growth or death caused by salinity fluctuations, and improving the quality and breeding efficiency of the striped sea bass seedlings.

[0064] Preferably, the collected real-time working parameter data is subjected to denoising to obtain a real-time working parameter data set of the desalination pool, specifically:

[0065] Preset a wavelet basis function and a decomposition layer number, perform wavelet decomposition on each real-time working parameter data according to the wavelet basis function and the decomposition layer number, and obtain wavelet coefficients of each real-time working parameter data at different scales, including low-frequency approximate coefficients and high-frequency detail coefficients;

[0066] Compare the wavelet coefficients of each real-time working parameter data at different scales with the preset threshold; directly set the wavelet coefficients smaller than the preset threshold to zero, and keep the wavelet coefficients larger than the preset threshold unchanged;

[0067] The wavelet coefficients after threshold processing are subjected to inverse wavelet transform, and the coefficients of different scales are gradually recombined to reconstruct the real-time working parameter data set of the denoised fading pool.

[0068] It should be noted that the wavelet basis function and the number of decomposition layers are first preset, which is the basic setting of wavelet decomposition. Wavelet decomposition is performed on each real-time working parameter data according to the selected wavelet basis function and the number of decomposition layers. Wavelet decomposition can decompose the real-time working parameter data into wavelet coefficients at different scales, including low-frequency approximate coefficients and high-frequency detail coefficients. The low-frequency approximate coefficients mainly contain the general outline information of the signal, while the high-frequency detail coefficients contain detailed information such as local changes of the signal. The preset threshold is a critical value set according to the characteristics of the data and the denoising requirements. The wavelet coefficients less than the preset threshold are directly set to zero. The purpose of this step is to remove those wavelet coefficients that may be generated by noise, because noise usually manifests as small fluctuations, and its wavelet coefficients are relatively small. The wavelet coefficients greater than the preset threshold remain unchanged, and these coefficients are considered to contain effective information, such as the part reflecting the real change of the working parameters. The wavelet coefficients after threshold processing are subjected to inverse wavelet transform. Inverse wavelet transform is the inverse process of wavelet decomposition, which can gradually recombine coefficients of different scales to reconstruct the real-time working parameter data set of the denoised fading pool. In this way, the obtained data set has been freed of noise interference while retaining the original effective information. The denoised data can more truly reflect the actual working status of the desalination pond, reduce the misjudgment of the working parameters of the desalination pond due to noise interference, and provide more accurate data support for the precise regulation of the desalination process of the japonica sea bass (such as salinity regulation, etc.), which will help improve the effectiveness and stability of the entire japonica sea bass desalination technology and ensure that the japonica sea bass grows in a suitable environment.

[0069] Preferably, the salinity dynamic analysis of the desalination pool is performed based on the predicted salinity dynamic change data of the desalination pool within a preset time period, and it is determined whether the salinity of the water in the desalination pool needs to be adjusted based on the analysis results, such as Figure 2 As shown, specifically:

[0070] S202: Obtaining a preset desalination scheme, and obtaining dynamic change data of standard salinity of a desalination pool within a preset time period according to the preset desalination scheme;

[0071] S204: introducing a Spearman rank correlation coefficient algorithm, and calculating the Spearman rank correlation coefficient between the standard salinity dynamic change data and the predicted salinity dynamic change data of the desalination pool within a preset time period based on the Spearman rank correlation coefficient algorithm;

[0072] S206: Compare the Spearman rank correlation coefficient with a preset coefficient threshold;

[0073] S208: If the Spearman rank correlation coefficient is greater than the preset coefficient threshold, it means that the salinity of the water in the desalination pool meets the preset requirements within the preset time period, and the salinity of the water is not regulated.

[0074] It should be noted that the preset desalination scheme is first obtained. This preset desalination scheme is a preset ideal scheme for desalination of sea bass seedlings. According to this preset desalination scheme, the standard salinity dynamic change data of the desalination pool within the preset time period can be obtained. This set of data represents the change trajectory that the salinity of the desalination pool should follow under the ideal desalination operation, and is the reference standard for subsequent judgment of whether the actual salinity meets the requirements. The Spearman rank correlation coefficient is a non-parametric statistic used to measure the monotonic relationship between two variables. Here, by calculating this coefficient, the similarity or correlation between the predicted salinity dynamic change data and the ideal standard salinity dynamic change data can be understood. The calculated Spearman rank correlation coefficient is compared with the preset coefficient threshold. This preset coefficient threshold is a judgment criterion for determining whether the relationship between the predicted salinity dynamic change data and the standard salinity dynamic change data is close enough. If the Spearman rank correlation coefficient is greater than the preset coefficient threshold, it means that the predicted salinity dynamic change data has a high similarity with the standard salinity dynamic change data, indicating that the salinity of the water body in the desalination pool meets the preset requirements within the preset time period, so there is no need to regulate the salinity of the water body. This decision-making process is based on quantitative analysis of data, rather than on experience or subjective judgment alone. Through the above series of operations, it is possible to accurately determine in a quantitative manner whether the water salinity of the desalination pool meets the preset requirements within the preset time period. The Spearman rank correlation coefficient algorithm is used to compare and analyze the predicted salinity dynamic change data and the standard salinity dynamic change data, avoiding the subjectivity and uncertainty in traditional judgment methods. This data-based analysis method makes the regulation of water salinity more scientific and accurate during the desalination process of sea bass seedlings, reduces unnecessary regulation operations, saves manpower and material resources, and ensures that sea bass seedlings are always in a suitable salinity environment, which is conducive to improving the growth quality and survival rate of seedlings.

[0075] Preferably, if the salinity of the water in the desalination pool needs to be adjusted, a corresponding control scheme is generated, and the salinity of the water in the desalination pool is adjusted according to the generated control scheme, specifically:

[0076] If the Spearman rank correlation coefficient is not greater than a preset coefficient threshold, the salinity data difference between the standard salinity dynamic change data and the predicted salinity dynamic change data at each time node within the preset time period is calculated;

[0077] A salinity data deviation value threshold is preset, and the salinity data difference between the standard salinity dynamic change data and the predicted salinity dynamic change data at each time node is compared with the preset salinity data deviation value threshold;

[0078] The time node corresponding to the salinity data difference being greater than the preset salinity data deviation value threshold is defined as the salinity drift time node.

[0079] It should be noted that when the Spearman rank correlation coefficient is not greater than the preset coefficient threshold, it means that the correlation between the predicted salinity dynamic change data and the standard salinity dynamic change data does not meet the requirements. At this time, it is necessary to calculate the salinity data difference between the standard salinity dynamic change data and the predicted salinity dynamic change data at each time node within the preset time period. This difference can intuitively reflect the degree of deviation between the predicted salinity and the ideal salinity at each time point. The preset salinity data deviation value threshold is used to determine whether the salinity deviation has reached the degree that needs to be adjusted. Comparing the salinity data difference at each time node with the preset salinity data deviation value threshold, this step is to screen out those time nodes with a large degree of deviation and need to adjust the salinity. The time node corresponding to the salinity data difference greater than the preset salinity data deviation value threshold is defined as the salinity drift time node. These time nodes are the moments when the salinity of the water body deviates from the preset requirements more seriously, and are the key basis for the subsequent generation of the control plan. By determining these time nodes, the salinity at these moments can be adjusted in a targeted manner, so that the salinity of the desalination pool is as close as possible to the preset ideal salinity dynamic change. Through the above operation, the time node at which the salinity of the water body deviates from the preset requirement within the preset time period, that is, the salinity drift time node, can be accurately found. This process enables the specific time point where the problem lies in the case where the salinity of the water body needs to be adjusted during the desalination process of the striped sea bass seedlings, thereby providing an accurate basis for generating targeted control plans. This helps to more accurately adjust the salinity of the water body in the desalination pool, ensure that the striped sea bass seedlings are in a suitable salinity environment, improve the growth quality of the seedlings and the success rate of the desalination process, and reduce seedling growth problems caused by inappropriate salinity.

[0080] Preferably, if the salinity of the water in the desalination pool needs to be adjusted, a corresponding adjustment scheme is generated, and the salinity of the water in the desalination pool is adjusted according to the generated adjustment scheme, and the following steps are also included:

[0081] Obtaining a preset desalination scheme, obtaining a preset salinity requirement range value of the water body at the salinity drift time node in the preset desalination scheme; and obtaining preset working parameters of the desalination pool at the salinity drift time node in the preset desalination scheme;

[0082] Acquire structural parameter information of each control device in the desalination pool, and construct a simulation control system of the desalination pool according to the structural parameter information;

[0083] Introducing a genetic algorithm and setting a genetic algebra; regulating the preset working parameters of the desalination pool at the salinity drift time node in the simulation control system according to the genetic algebra, generating several control schemes, and obtaining predicted salinity data of the water body in the desalination pool after each control scheme is regulated;

[0084] Only the control scheme corresponding to the predicted salinity data within the preset salinity requirement range is extracted, and the control energy consumption value of the marked control scheme is obtained;

[0085] Constructing a sorting table, importing the control energy consumption values ​​of the marked control schemes into the sorting table, and sorting them from low to high according to the control energy consumption values ​​to generate a sorting result;

[0086] According to the sorting result, the marked control schemes are sent in sequence to the control terminal of the desalination pool, so that the salinity of the desalination pool is regulated in order according to each marked control scheme until the actual salinity data of the water body in the desalination pool is within the preset salinity requirement range.

[0087] It should be noted that the preset desalination scheme is obtained, from which the preset salinity demand range value of the water body and the preset working parameters of the desalination pool at the salinity drift time node are obtained. These preset information provide targets and initial references for the subsequent generation of control schemes. For example, the preset salinity demand range value clarifies the ideal range that the salinity of the water body should reach at a specific time node, and the preset working parameters are the working state-related parameters that the desalination pool should have at this time node under ideal circumstances. The structural parameter information of each control device (such as freshwater injection equipment, salinity adjustment equipment, etc.) in the desalination pool is obtained, and a simulation control system for the desalination pool is constructed based on this information. This simulation system can simulate the state changes of the desalination pool under different control operations, providing a platform for testing the control scheme in a virtual environment, and avoiding a large number of trial and error operations directly in the actual desalination pool. Genetic algorithm is introduced and genetic algebra is set. Genetic algorithm is an optimization algorithm based on the principle of biological evolution. By simulating processes such as natural selection, crossover and mutation, the preset working parameters of the desalination pool at the salinity drift time node are regulated in the simulation control system. After multiple generations of evolution, several control schemes are generated, and the predicted salinity data of the water body in the desalination pool after each control scheme is regulated is obtained. This step uses the global search capability of the genetic algorithm to find multiple possible control schemes that meet the salinity requirements. Only the control schemes corresponding to the predicted salinity data within the preset salinity requirement range are extracted, and the control energy consumption values ​​of these marked control schemes are obtained. This step is to screen the many generated control schemes, retain only those that can make the salinity reach the ideal range, and consider the important factor of control energy consumption, because energy consumption is directly related to the breeding cost. Construct a sorting table, import the control energy consumption values ​​of the marked control schemes into the sorting table, and sort them from low to high according to the control energy consumption values ​​to generate a sorting result. In this way, the control scheme with low energy consumption can be given priority, and the cost can be optimized while meeting the salinity control requirements. According to the sorting results, the marked control schemes are sent to the control terminal of the desalination pool in order, so as to regulate the salinity of the desalination pool in an orderly manner according to each marked control scheme, until the actual salinity data of the water body in the desalination pool is within the preset salinity requirement range. This ensures that in actual operation, the control schemes are tried in the optimal order (from low energy consumption to high energy consumption) to keep the water salinity within the ideal range. Through the above steps, when the water salinity needs to be adjusted during the desalination process of the striped sea bass seedlings, a control scheme that meets the salinity control requirements and has low energy consumption can be systematically generated and screened out. Building a simulation control system and using genetic algorithms can quickly find a variety of feasible control schemes. Through screening, marking and sorting, the most economical and efficient control scheme can be selected and executed in an orderly manner. This helps to accurately control the salinity of the water in the desalination pool, so that the striped sea bass seedlings are in a suitable salinity environment, improve the growth quality of the seedlings, and at the same time reduce the cost of control and improve the efficiency and economy of the entire striped sea bass seedling desalination process.

[0088] In addition, the method further comprises the following steps:

[0089] Obtaining equipment operation log information of the desalination pool, obtaining fault events of each control device in the desalination pool according to the equipment operation log information, and obtaining fault feature data corresponding to each fault event;

[0090] The fuzzy C-means clustering algorithm is introduced to obtain the fuzzy matrix of the fault feature data of each fault event based on the fault feature data corresponding to each fault event.

[0091] A random condition field is introduced, and based on the random condition field, a feature extraction is performed on the fuzzy matrix of the fault characteristic data to obtain a conditional transition probability of each control device from a normal state to a fault state under each fault characteristic data condition;

[0092] A support vector machine model is constructed based on a fuzzy neural network, and the conditional transition probability of each control device from a normal state to a fault state under each fault characteristic data condition is imported into the support vector machine model for training and learning until the model parameters meet the requirements and the training is completed;

[0093] If the actual salinity data of the water in the desalination pool is still within the preset salinity requirement range after the salinity of the water in the desalination pool is adjusted according to the generated control scheme, the real-time operating parameter data of the control equipment in the desalination pool is obtained;

[0094] Importing the real-time operating parameter data of the control equipment in the desalination pool into the support vector machine model for prediction, and obtaining the conditional transfer probability of the control equipment in the desalination pool;

[0095] If the conditional transfer probability is greater than the preset probability, a control equipment failure warning information is generated.

[0096] It should be noted that the fuzzy C-means clustering and random conditional field algorithms are used to process fault feature data, and fault prediction is performed based on the fuzzy neural network and support vector machine models, so that the operating status of the control equipment can be monitored in real time, and an early warning can be issued in time when a possible equipment failure is detected. This can effectively improve the stability and reliability of the desalination pond system, reduce the risk of failures, ensure the salinity control effect of the desalination pond water body, optimize the maintenance and management of the equipment, and improve the survival rate of the desalination of the sea bass seedlings.

[0097] The present invention also discloses a system for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis, such as Figure 3As shown, the striped seabass seedling desalination system based on characteristic data analysis includes a memory 10 and a processor 20. The memory 10 stores a striped seabass seedling desalination method program based on characteristic data analysis. When the striped seabass seedling desalination method program based on characteristic data analysis is executed by the processor 20, any one of the steps of the striped seabass seedling desalination method based on characteristic data analysis is implemented.

[0098] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0099] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0100] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0101] Those skilled in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), disks or optical disks, and other media that can store program codes.

[0102] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0103] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis, characterized in that: The following steps are involved: Obtain the dynamic change data of salinity of the desalination pool under various working parameter conditions, and build a dynamic prediction model of water salinity based on the dynamic change data of salinity of the desalination pool under various working parameter conditions; Acquire the real-time working parameter data of the desalination pool, and perform denoising on the collected real-time working parameter data to obtain a real-time working parameter data set of the desalination pool; Importing the real-time working parameter data set of the desalination pool into the water body salinity dynamic prediction model for prediction, and predicting the dynamic change data of the salinity of the desalination pool within a preset time period; Perform salinity dynamic analysis on the desalination pool according to the predicted salinity dynamic change data of the desalination pool within a preset time period, and determine whether it is necessary to adjust the salinity of the water in the desalination pool according to the analysis results; If the salinity of the water in the desalination pool needs to be adjusted, a corresponding adjustment plan is generated, and the salinity of the water in the desalination pool is adjusted according to the generated adjustment plan.

2. The method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis according to claim 1, characterized in that: Obtain the dynamic change data of salinity of the desalination pool under various working parameter conditions, and build a dynamic prediction model of water salinity based on the dynamic change data of salinity of the desalination pool under various working parameter conditions, specifically: Obtain the dynamic change data of salinity in desalination pools under various working parameter conditions through the big data network; Construct a Markov random field, use each working parameter as a conditional node of the Markov random field, and use the corresponding salinity dynamic change data as a variable node of the Markov random field; Directed connections are made to each variable node and condition node to determine all possible value combinations of each variable node and condition node; For each value combination, the potential function value between each value combination is calculated according to the Markov random field, and a potential function matrix is ​​constructed according to the calculated potential function values; Building a water body salinity dynamic prediction model based on a graph neural network, and embedding the potential function matrix into the water body salinity dynamic prediction model for training; When the prediction accuracy of the water body salinity dynamic prediction model is higher than a preset accuracy threshold, the final training parameters of the water body salinity dynamic prediction model are saved and a training end instruction is output.

3. The method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis according to claim 1, characterized in that: The collected real-time working parameter data is subjected to denoising to obtain the real-time working parameter data set of the desalination pool, which is specifically: Preset a wavelet basis function and a decomposition layer number, perform wavelet decomposition on each real-time working parameter data according to the wavelet basis function and the decomposition layer number, and obtain wavelet coefficients of each real-time working parameter data at different scales, including low-frequency approximate coefficients and high-frequency detail coefficients; Compare the wavelet coefficients of each real-time working parameter data at different scales with a preset threshold value; The wavelet coefficients smaller than the preset threshold are directly set to zero, and the wavelet coefficients larger than the preset threshold are kept unchanged; The wavelet coefficients after threshold processing are subjected to inverse wavelet transform, and the coefficients of different scales are gradually recombined to reconstruct the real-time working parameter data set of the denoised fading pool.

4. The method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis according to claim 1, characterized in that: According to the predicted salinity dynamic change data of the desalination pool within the preset time period, the salinity dynamic analysis of the desalination pool is carried out, and according to the analysis results, it is determined whether the salinity of the water in the desalination pool needs to be adjusted. Specifically: Obtaining a preset desalination scheme, and obtaining the dynamic change data of the standard salinity of the desalination pool within a preset time period according to the preset desalination scheme; The Spearman rank correlation coefficient algorithm is introduced to calculate the Spearman rank correlation coefficient between the standard salinity dynamic change data and the predicted salinity dynamic change data of the desalination pool within a preset time period based on the Spearman rank correlation coefficient algorithm; comparing the Spearman rank correlation coefficient with a preset coefficient threshold; If the Spearman rank correlation coefficient is greater than the preset coefficient threshold, it means that the salinity of the water in the desalination pool meets the preset requirements within the preset time period, and the salinity of the water is not regulated.

5. The method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis according to claim 4, characterized in that: If the salinity of the desalination pool water needs to be adjusted, a corresponding control scheme is generated, and the salinity of the desalination pool water is adjusted according to the generated control scheme, specifically: If the Spearman rank correlation coefficient is not greater than a preset coefficient threshold, the salinity data difference between the standard salinity dynamic change data and the predicted salinity dynamic change data at each time node within the preset time period is calculated; A salinity data deviation value threshold is preset, and the salinity data difference between the standard salinity dynamic change data and the predicted salinity dynamic change data at each time node is compared with the preset salinity data deviation value threshold; The time node corresponding to the salinity data difference being greater than the preset salinity data deviation value threshold is defined as the salinity drift time node.

6. The method for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis according to claim 5, characterized in that: If the salinity of the water in the desalination pool needs to be adjusted, a corresponding adjustment scheme is generated, and the salinity of the water in the desalination pool is adjusted according to the generated adjustment scheme, and the following steps are also included: Obtaining a preset desalination scheme, obtaining a preset salinity requirement range value of the water body at the salinity drift time node in the preset desalination scheme; and obtaining preset working parameters of the desalination pool at the salinity drift time node in the preset desalination scheme; Acquire structural parameter information of each control device in the desalination pool, and construct a simulation control system of the desalination pool according to the structural parameter information; Introducing a genetic algorithm and setting a genetic algebra; regulating the preset working parameters of the desalination pool at the salinity drift time node in the simulation control system according to the genetic algebra, generating several control schemes, and obtaining predicted salinity data of the water body in the desalination pool after each control scheme is regulated; Only the control scheme corresponding to the predicted salinity data within the preset salinity requirement range is extracted, and the control energy consumption value of the marked control scheme is obtained; Constructing a sorting table, importing the control energy consumption values ​​of the marked control schemes into the sorting table, and sorting them from low to high according to the control energy consumption values ​​to generate a sorting result; According to the sorting result, the marked control schemes are sent in sequence to the control terminal of the desalination pool, so that the salinity of the desalination pool is regulated in order according to each marked control scheme until the actual salinity data of the water body in the desalination pool is within the preset salinity requirement range.

7. A system for desalination of Lateolabrax japonicus seedlings based on characteristic data analysis, characterized in that: The striped seabass seedling desalination system based on characteristic data analysis includes a memory and a processor, wherein the memory stores a striped seabass seedling desalination method program based on characteristic data analysis, and when the striped seabass seedling desalination method program based on characteristic data analysis is executed by the processor, the striped seabass seedling desalination method steps based on characteristic data analysis as described in any one of claims 1 to 6 are implemented.

Images