A new type of grasshopper marking method and system based on intelligent regulation

Abstract

This invention relates to the field of aquaculture technology, specifically a novel method and system for raising cricket larvae based on intelligent regulation. The method involves constructing a Bayesian feeding prediction model for cricket larvae; importing the raising feature information into the Bayesian feeding prediction model for prediction, obtaining predicted feeding data of cricket larvae in the raising device within a preset time period; combining the actual feeding data and predicted feeding data within the preset time period to analyze the feeding status of the cricket larvae; if the feeding status of the cricket larvae is abnormal, performing characteristic analysis on the dynamic set of various actual raising environmental parameters to obtain various raising environmental states; if a raising environment is marked as an unstable raising environment, adjusting the working parameters of the corresponding functional adjustment modules in the raising device, aiming to improve the efficiency and quality of cricket larva raising through intelligent means and achieve sustainable ecological raising goals.

Description

Technical Field

[0001] This invention relates to the field of aquaculture technology, and in particular to a novel method and system for raising rice worms based on intelligent control. Background Technology

[0002] There are two main existing methods for raising rice worms: one is to directly divide the breeding area with colored tarpaulin netting, and the other is to use raising devices to introduce seedlings. However, both methods have obvious drawbacks, such as uneven distribution of rice worm larvae, cumbersome operation, difficulty in environmental control, and unstable raising results. In recent years, intelligent control technology has been widely used in the agricultural field, providing new ideas for the innovation of rice worm raising technology. Using modern raising facilities, such as recirculating aquaculture systems and automatic feeding systems, can improve breeding efficiency and achieve efficient, environmentally friendly, and sustainable rice worm breeding. These modern raising facilities can effectively solve the drawbacks of traditional methods and bring new breakthroughs to the development of the rice worm breeding industry. However, the pain point of these modern raising facilities is still the poor control over the environment and the introduction of seedlings and feed. The raising results are still not very stable. It is difficult to use sensors and monitoring systems to monitor key parameters such as temperature, humidity, light, and water quality in the raising environment in real time, and to automatically adjust the raising environment with intelligent algorithms, making it difficult to ensure the optimal conditions for rice worm larvae growth. Summary of the Invention

[0003] This invention overcomes the shortcomings of the prior art and provides a novel method and system for the cultivation of rice worms based on intelligent control.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] The first aspect of this invention discloses a novel method for marking grains of rice based on intelligent control, comprising the following steps:

[0006] Based on big data networks, standard food intake datasets of larvae in each growth stage of the worm are obtained within a unit of time, and a Bayesian food intake prediction model for worm larvae is constructed based on these datasets.

[0007] Obtain the marking feature information of the larvae of the worm in the marking device, and import the marking feature information into the Bayesian food prediction model for prediction to obtain the predicted food amount data of the larvae of the worm in the marking device within a preset time period.

[0008] The tailwater of the standardization device is sampled and tested at a preset time point to obtain the concentration of remaining feed in the tailwater. The actual food consumption data of the larvae of the worms within the preset time period is calculated based on the concentration of remaining feed. The feeding status of the larvae of the worms is analyzed by combining the actual food consumption data and the predicted food consumption data within the preset time period.

[0009] If the feeding state of the larvae of the rice worm is abnormal, various actual scale environmental parameters are obtained at several preset time points to obtain a dynamic set of various actual scale environmental parameters; the characteristics of various actual scale environmental parameters are analyzed to obtain the status of various scale environmental conditions.

[0010] If a roughing environment is marked as an unstable roughing environment, the operating parameters of the corresponding functional adjustment module in the roughing device will be adjusted.

[0011] Furthermore, based on big data networks, standard food intake datasets of larvae at each stage of the growth stage are obtained within a unit of time. A Bayesian network prediction model is then constructed based on these standard food intake datasets. Specifically:

[0012] Obtain a preset grading scheme for the larvae to be graded, construct key search terms based on the preset grading scheme, and search the big data network resource library based on the key search terms to obtain the standard food intake dataset of larvae at each grading stage within a unit of time.

[0013] The Apriori algorithm is introduced, and correlation analysis is performed on the standard feeding data of various larval stages of the mulberry worm within a unit time to obtain the correlation strength between the feeding of any mulberry worm stage and the feeding of other mulberry worm stages.

[0014] The Bayesian network's food quantity nodes are set according to various standard feeding stages of the larvae of the worm, and the edges of the Bayesian network are set according to the corresponding standard feeding dataset; where each feeding quantity node in the Bayesian network represents a specific standard feeding stage, and each edge represents the feeding quantity relationship between two feeding quantity nodes.

[0015] If the correlation strength between the amount of food in a certain standard stage and another standard stage is greater than a preset strength threshold, then the corresponding food nodes and edges are connected in the Bayesian network to generate a topology network structure graph; the topology relationship of food between each standard stage is obtained according to the topology network structure graph.

[0016] Based on the topological relationship of food consumption between each stage and combined with the maximum probability estimation method, the conditional probability range of each food consumption node in the Bayesian network is estimated. Based on the conditional probability range of each food consumption node, a food consumption range attribute is added to each food consumption node to obtain the Bayesian food consumption prediction model for the larvae of the worm.

[0017] Furthermore, the feeding status of the larvae of the rice worm was analyzed by combining actual and predicted feeding data within a preset time period, specifically as follows:

[0018] The difference between the actual and predicted feeding data of the larvae of the rice worm in the standard feeding device within a preset time period is calculated to obtain the feeding data deviation value;

[0019] Determine whether the deviation value of the food intake data is greater than a preset food intake deviation threshold;

[0020] If the deviation value of the food intake data is greater than the preset food intake deviation threshold, the feeding status of the larvae of the rice worm is marked as normal.

[0021] If the deviation value of the food intake data is not greater than the preset food intake deviation threshold, the feeding status of the larvae of the rice worm is marked as an abnormal state.

[0022] Furthermore, characteristic analysis is performed on the dynamic sets of various actual coarse-graining environment parameters to obtain the various coarse-graining environment states, specifically:

[0023] Obtain a preset grading scheme for the coarse worm to be calibrated, and obtain a dynamic set of various standard grading environmental parameters within a preset grading time period based on the preset grading scheme;

[0024] Construct a first blank matrix and a second blank matrix, randomly access a standard bolding environment, and obtain the dynamic set of actual standard bolding environment parameters and the dynamic set of standard bolding environment parameters corresponding to that standard bolding environment;

[0025] The actual dynamic set of standard roughing environment parameters and the standard dynamic set of standard roughing environment parameters corresponding to these roughing environment parameters are filled into the first blank matrix and the second blank matrix respectively in a preset order to obtain the actual roughing environment parameter matrix and the standard roughing environment parameter matrix.

[0026] Calculate the Euclidean distance between elements at the same position in the actual and standard roughened environmental parameter matrices; and compare the Euclidean distance between each element at the same position in the two matrices with a preset distance threshold.

[0027] If the Euclidean distance between elements at the same position in two matrices is not greater than a preset distance threshold, it means that the actual bolding environmental parameter at the corresponding time node has not drifted or the drift amplitude is within the allowable range. Then, the actual bolding environmental parameter at that time node is marked as a normal parameter node.

[0028] If the Euclidean distance between elements at the same position in two matrices is greater than a preset distance threshold, it indicates that the actual bolding environmental parameter at the corresponding time node has drifted and the drift amplitude exceeds the allowable range. Then, the actual bolding environmental parameter at that time node is marked as a parameter drift node.

[0029] The total number of parameter drift nodes and the total number of parameter normal nodes in the currently accessed standard environment are counted. The ratio of the total number of parameter drift nodes to the total number of parameter normal nodes in the currently accessed standard environment is calculated to obtain the percentage of parameter drift nodes in the currently accessed standard environment.

[0030] If the percentage of drifting nodes in the currently accessed bolding environment is greater than the preset percentage, the currently accessed bolding environment is marked as an unstable bolding environment; if the percentage of drifting nodes in the currently accessed bolding environment is not greater than the preset percentage, the parameters of the currently accessed bolding environment are marked as a stable bolding environment.

[0031] Visit the next bolded environment that has not been visited before, and repeat the above steps until all bolded environments have been visited, and obtain the status of each bolded environment.

[0032] Furthermore, if a grading environment is marked as an unstable grading environment, the operating parameters of the corresponding functional adjustment module in the grading device will be adjusted accordingly, specifically as follows:

[0033] Obtain the structural drawing information of the roughening device, and construct a dynamic simulation system for the roughening device based on the structural drawing information and simulation software.

[0034] Obtain the functional specification of the grading device, determine the functional adjustment module corresponding to the grading environment marked as unstable based on the functional specification, and obtain the real-time working parameters of the functional adjustment module and the dynamic set of actual grading environment parameters marked as unstable based on the Internet of Things.

[0035] Import the real-time operating parameters of the function adjustment module and the dynamic set of parameters of the actual roughing environment marked as the unstable roughing environment into the dynamic simulation system;

[0036] A genetic algorithm is introduced, and the preset number of iterations and the preset iteration adjustment class range are set according to the genetic algorithm.

[0037] Based on the preset iterative adjustment level range, the real-time working parameters of the functional adjustment module in the dynamic simulation system are simulated and adjusted to obtain the dynamic set of simulated rough environment parameters marked as unstable rough environment after simulation adjustment; at the same time, the simulation adjustment parameters corresponding to the simulation adjustment are output, and a working parameter adjustment scheme is generated according to the simulation adjustment parameters.

[0038] Repeat the above steps until the preset number of iterations is reached. This will yield the dynamic set of simulation standard environment parameters and the working parameter control scheme corresponding to each simulation control. The dynamic set of simulation standard environment parameters and the working parameter control scheme corresponding to each simulation control will be bundled together to obtain several data packets.

[0039] It also includes the following steps:

[0040] Obtain a preset grading scheme for the coarse worm to be calibrated, and obtain a dynamic set of standard grading environment parameters marked as unstable grading environments within a preset simulation time period based on the preset grading scheme;

[0041] Calculate the cosine similarity between the dynamic set of simulated bolding environment parameters in each data packet and the dynamic set of standard bolding environment parameters marked as unstable bolding environment within a preset simulation time period to obtain several cosine similarities.

[0042] The imported cosine similarity values ​​are sorted to obtain the maximum cosine similarity, and the dynamic set of simulated bolding environment parameters corresponding to the maximum cosine similarity is obtained; the working parameter control scheme is extracted from the data package of the dynamic set of simulated bolding environment parameters corresponding to the maximum cosine similarity.

[0043] The extracted working parameter control scheme is sent to the control terminal so that the control terminal can adjust the real-time working parameters of the functional adjustment module corresponding to the unstable standard environment based on the extracted working parameter control scheme.

[0044] The standard environment includes temperature, salinity, pH, feeding speed, air pressure, pH value, dissolved oxygen concentration, light intensity, water depth, and water flow speed.

[0045] The growth characteristics include the number of rice worms, the growth stage, and the feeding time.

[0046] A second aspect of this invention discloses a novel grain-growing system for rice worms based on intelligent control, applicable to any of the novel grain-growing methods for rice worms based on intelligent control described in any one of the claims. The novel grain-growing system for rice worms includes:

[0047] The main body of the nursery is used to provide a pre-designed nursery environment for the larvae of the rice worm;

[0048] The function adjustment module is used to control various parameters of the standard environment, including water temperature, salinity, pH, dissolved oxygen, pH value, air pressure, light intensity, water flow speed and water depth, in order to simulate the natural growth environment of the larvae of the worm.

[0049] The feed dispensing module automatically feeds the larvae of the rice worm according to their growth stage and food requirements, and controls the feeding frequency and type of feed.

[0050] The data acquisition and analysis module collects various data during the nursery process, including environmental parameters, feed intake, and growth status, and performs data analysis.

[0051] The intelligent control module integrates data from various modules and automatically controls the operating parameters of each module according to a preset scaling scheme.

[0052] A third aspect of the present invention discloses a computer-readable storage medium comprising a novel method program for marking and thickening rice worms based on intelligent control. When the novel method program for marking and thickening rice worms based on intelligent control is processed and executed, it implements any of the steps of the novel method program for marking and thickening rice worms based on intelligent control as described in the present invention.

[0053] This invention addresses the technical deficiencies in the prior art and offers the following advantages: It constructs a Bayesian feeding prediction model for larvae of the trevally; imports the larval feeding prediction model with the larval feeding characteristics to obtain predicted feeding data for larvae within a preset time period in the larval feeding device; analyzes the feeding status of the larvae by combining the actual feeding data with the predicted feeding data within the preset time period; if the feeding status of the larvae is abnormal, various actual larval feeding environment parameters are acquired at several preset time points to obtain a dynamic set of various actual larval feeding environment parameters; the dynamic set of various actual larval feeding environment parameters is analyzed to obtain various larval feeding environment states; if a larval feeding environment is marked as unstable, the operating parameters of the corresponding functional adjustment module in the larval feeding device are adjusted. The entire process realizes a closed-loop management strategy from data collection, model construction, predictive analysis to environmental control, aiming to improve the efficiency and quality of larval feeding through intelligent means and achieve sustainable ecological feeding goals. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0055] Figure 1 This is a flowchart of the overall method for the novel method of coarsely culturing rice worms;

[0056] Figure 2 This is a partial flowchart of the novel method for coarsely measuring rice worms.

[0057] Figure 3 This is a simplified structural diagram of the novel rice worm labeling system. Detailed Implementation

[0058] To better understand the above-mentioned objectives, 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 should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0059] 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.

[0060] like Figure 1 As shown, the first aspect of this invention discloses a novel method for marking grains based on intelligent control, comprising the following steps:

[0061] S102: Obtain the standard food intake dataset of each stage of the seedling stage of the worm larvae within a unit time based on the big data network, and construct a Bayesian food intake prediction model for the worm larvae based on the standard food intake dataset of each stage of the seedling stage within a unit time.

[0062] S104: Obtain the marking feature information of the larvae of the worm in the marking device, and import the marking feature information into the Bayesian food prediction model for prediction to obtain the predicted food data of the larvae of the worm in the marking device within a preset time period.

[0063] S106: Sample and test the tailwater of the standardization device at a preset time node to obtain the concentration of remaining feed in the tailwater. Calculate the actual food consumption data of the larvae of the worm within the preset time period based on the concentration of remaining feed. Analyze the feeding status of the larvae of the worm by combining the actual food consumption data and the predicted food consumption data within the preset time period.

[0064] S108: If the feeding state of the larvae of the rice worm is abnormal, various actual scale environmental parameters are obtained at several preset time points to obtain a dynamic set of various actual scale environmental parameters; the characteristics of the dynamic set of various actual scale environmental parameters are analyzed to obtain the status of various scale environmental conditions.

[0065] S110: If a roughing environment is marked as an unstable roughing environment, the working parameters of the corresponding functional adjustment module in the roughing device shall be adjusted.

[0066] The nurturing environment includes temperature, salinity, pH, feeding speed, air pressure, pH value, dissolved oxygen concentration, light intensity, water depth, and water flow velocity. The nurturing characteristics include the number of worms, the nurturing stage, and the feeding time.

[0067] It's important to note that, firstly, standardized feed intake datasets of larvae at different growth stages were obtained through big data networks, forming the foundation for the subsequent prediction model. Next, using these standardized datasets, a Bayesian feed intake prediction model for larvae was established. This model aims to predict the feed intake of larvae over a future period based on current growth characteristics, thereby enabling precise resource management and aquaculture planning. Based on the prediction, wastewater samples were collected at preset time points to detect the concentration of remaining feed, thus inferring the actual feed intake of the larvae during that period. By comparing the actual feed intake data with the predicted data, the feeding status of the larvae can be monitored in real time, allowing for the timely detection of any anomalies. Once an anomaly in the larvae's feeding status is detected, various actual growth environment parameters during the abnormal time period are collected to form a dynamic set, which is then subjected to in-depth characteristic analysis to determine the specific cause of the anomaly. These environmental parameters include water temperature, pH value, dissolved oxygen concentration, and nutrient content, which collectively constitute the state of the growth environment. For breeding environments marked as "unstable"—meaning excessive fluctuations in environmental parameters that affect the normal growth and food intake of the larvae—the system will specifically adjust the operating parameters of the corresponding functional regulation modules. This could involve altering the water circulation rate, increasing the frequency of nutrient replenishment, or adjusting lighting conditions to restore the stability of the breeding environment and promote the healthy growth of the larvae. The entire process implements a closed-loop management strategy from data collection, model building, predictive analysis to environmental control, aiming to improve the efficiency and quality of larval breeding through intelligent means and achieve sustainable ecological breeding goals.

[0068] Furthermore, such as Figure 2 As shown, a standard food intake dataset of larvae at each stage of the growth stage of *Gnaphalium affine* is obtained within a unit of time based on a big data network. A Bayesian network prediction model is then constructed based on this dataset. Specifically:

[0069] S202: Obtain a preset grading scheme for the larvae to be graded, construct key search terms based on the preset grading scheme, and search the big data network resource library based on the key search terms to obtain a standard food intake dataset of larvae at each grading stage within a unit of time.

[0070] S204: Introducing the Apriori algorithm, and performing correlation analysis on the standard feeding data of various larval stages of the mulberry worm within a unit of time based on the Apriori algorithm, to obtain the correlation strength between the feeding of any larval stage and the feeding of other larval stages.

[0071] Specifically, the Apriori algorithm reveals the correlation between the feeding amounts of larvae at different growth stages of the mulberry worm by identifying frequently occurring itemsets in a dataset and generating association rules based on these frequent itemsets. It first identifies itemsets in the dataset that appear more frequently than a set threshold, termed frequent itemsets. Then, it uses these frequent itemsets to generate association rules, which represent the probability of another itemset appearing when another itemset appears. By analyzing these association rules, the strength of the correlation between the feeding amounts of larvae at different growth stages can be understood. For example, an increase in feeding amount at one growth stage may be associated with changes in feeding amount at another stage, thus helping to understand the patterns of feeding amount changes during the mulberry worm's growth.

[0072] S206: Set the feeding nodes of the Bayesian network according to the various standard feeding stages of the larvae of the worm, and set the edges of the Bayesian network according to the corresponding standard feeding dataset; wherein, each feeding node in the Bayesian network represents a specific standard feeding stage, and each edge represents the feeding relationship between two feeding nodes.

[0073] S208: If the correlation strength between the amount of food in a certain standard stage and another standard stage is greater than a preset strength threshold, then the corresponding food nodes and edges are connected in the Bayesian network to generate a topology network structure diagram; the topology relationship of food between each standard stage is obtained according to the topology network structure diagram.

[0074] S210: Based on the topological relationship of food consumption between each stage and combined with the maximum probability estimation method, estimate the conditional probability range of each food consumption node in the Bayesian network. Add the food consumption range attribute to each food consumption node according to the conditional probability range of each food consumption node to obtain the Bayesian food consumption prediction model for the larvae of the rice worm.

[0075] It should be noted that, firstly, a pre-defined breeding scheme for the larvae to be bred was obtained. This scheme was developed in advance by relevant technical personnel, and key search terms were constructed, such as keywords related to larvae, breeding stage, and food intake data. These were then searched using large-scale online resource databases (such as Google Scholar and Baidu Scholar) to collect standard food intake datasets for each breeding stage of larvae within a unit of time. This step ensured the comprehensiveness and accuracy of the data, laying a solid foundation for subsequent analysis. Next, the Apriori algorithm was introduced to perform correlation analysis on the collected data to identify the food intake relationships between different breeding stages. Through analysis, the correlation strength between the food intake of a specific breeding stage and the food intake of other stages can be quantified, providing a scientific basis for understanding the food intake variation patterns in the larvae's growth cycle. Based on this, a Bayesian network model was constructed. Each node represents a specific breeding stage, and the edges between nodes represent the food intake relationship between two stages. By connecting nodes and edges with significant correlation strengths, a topological network structure graph was formed, intuitively displaying the topological relationships of food intake between various breeding stages. Finally, the conditional probability range of each node in the Bayesian network was estimated using the maximum probability ingress estimation method, and a food intake range attribute was added to each node to construct a Bayesian food intake prediction model for larvae of the sandworm. This model not only predicts the food intake of larvae at different growth stages but also provides the uncertainty range of food intake changes, offering more precise guidance for aquaculture decisions. Overall, this process integrates data mining, machine learning, and Bayesian network theory, achieving full automation from data collection to model construction, and providing an efficient and accurate solution for the growth management and prediction of sandworm larvae.

[0076] Furthermore, the feeding status of the larvae of the rice worm was analyzed by combining actual and predicted feeding data within a preset time period, specifically as follows:

[0077] The difference between the actual and predicted feeding data of the larvae of the rice worm in the standard feeding device within a preset time period is calculated to obtain the feeding data deviation value;

[0078] Determine whether the deviation value of the food intake data is greater than a preset food intake deviation threshold;

[0079] If the deviation value of the food intake data is greater than the preset food intake deviation threshold, the feeding status of the larvae of the rice worm is marked as normal.

[0080] If the deviation value of the food intake data is not greater than the preset food intake deviation threshold, the feeding status of the larvae of the rice worm is marked as an abnormal state.

[0081] It should be noted that the difference between the actual and predicted food intake data within a preset time period is calculated. This difference reflects the deviation between the actual and predicted intake. The calculation of the food intake deviation directly reflects the degree of difference between the actual feeding behavior of the larvae and the behavior predicted by the model. Next, a preset food intake deviation threshold is set in advance as a standard for judging the feeding status. The setting of this threshold needs to consider various factors, including but not limited to the larvae's growth stage, food type, and environmental conditions, to ensure that under normal circumstances, the food intake deviation will not exceed this threshold. If the calculated food intake deviation is greater than the preset threshold, it means that there is a significant difference between the actual and predicted food intake. This is usually considered to be within the normal range of the larvae's feeding status. This is because a normal food intake deviation indicates a reasonable fluctuation between the model-predicted food intake and the actual intake, which may be due to food availability, changes in the larvae's appetite, or the influence of other environmental factors. Conversely, if the deviation in food intake data is not greater than the preset food intake deviation threshold, it indicates that the difference between the actual intake and the predicted intake is small, which may mean that the feeding status of the larvae of the rice worm is abnormal. By quantifying the difference between the actual and predicted food intake, an objective and quantitative method is provided for monitoring the feeding status of the rice worm larvae, which helps to promptly identify and address problems that may affect the growth efficiency and yield of the rice worm.

[0082] Furthermore, characteristic analysis is performed on the dynamic sets of various actual coarse-graining environment parameters to obtain the various coarse-graining environment states, specifically:

[0083] Obtain a preset grading scheme for the coarse worm to be calibrated, and obtain a dynamic set of various standard grading environmental parameters within a preset grading time period based on the preset grading scheme;

[0084] Construct a first blank matrix and a second blank matrix, randomly access a standard bolding environment, and obtain the dynamic set of actual standard bolding environment parameters and the dynamic set of standard bolding environment parameters corresponding to that standard bolding environment;

[0085] The actual dynamic set of standard roughing environment parameters and the standard dynamic set of standard roughing environment parameters corresponding to this type of roughing environment parameter are filled into the first blank matrix and the second blank matrix respectively in a preset order to obtain the actual roughing environment parameter matrix and the standard roughing environment parameter matrix.

[0086] Calculate the Euclidean distance between elements at the same position in the actual and standard roughened environmental parameter matrices; and compare the Euclidean distance between each element at the same position in the two matrices with a preset distance threshold.

[0087] If the Euclidean distance between elements at the same position in two matrices is not greater than a preset distance threshold, it means that the actual bolding environmental parameter at the corresponding time node has not drifted or the drift amplitude is within the allowable range. Then, the actual bolding environmental parameter at that time node is marked as a normal parameter node.

[0088] If the Euclidean distance between elements at the same position in two matrices is greater than a preset distance threshold, it indicates that the actual bolding environmental parameter at the corresponding time node has drifted and the drift amplitude exceeds the allowable range. Then, the actual bolding environmental parameter at that time node is marked as a parameter drift node.

[0089] The total number of parameter drift nodes and the total number of parameter normal nodes in the currently accessed standard environment are counted. The ratio of the total number of parameter drift nodes to the total number of parameter normal nodes in the currently accessed standard environment is calculated to obtain the percentage of parameter drift nodes in the currently accessed standard environment.

[0090] If the percentage of drifting nodes in the currently accessed bolding environment is greater than the preset percentage, the currently accessed bolding environment is marked as an unstable bolding environment; if the percentage of drifting nodes in the currently accessed bolding environment is not greater than the preset percentage, the parameters of the currently accessed bolding environment are marked as a stable bolding environment.

[0091] Visit the next bolded environment that has not been visited before, and repeat the above steps until all bolded environments have been visited, and obtain the status of each bolded environment.

[0092] It should be noted that, according to a pre-set standardization scheme, a dynamic set of standard standardization environmental parameters is obtained within a preset standardization period. These parameters include temperature, humidity, light intensity, water quality indicators, etc., which are key factors in maintaining the stability of the standardization environment. To compare the actual parameters with the standard parameters, two blank matrices are constructed: one to store the dynamic set of actual standardization environmental parameters, and the other to store the dynamic set of standard standardization environmental parameters. A standardization environment is randomly selected, and its corresponding dynamic set of actual parameters and dynamic set of standard parameters are obtained. Then, the dynamic sets of actual parameters and dynamic sets of standard parameters are filled into the two blank matrices in a preset order, forming the actual standardization environmental parameter matrix and the standard standard standardization environmental parameter matrix. The Euclidean distance between parameters at the same position in the two matrices is calculated. Euclidean distance is a way to measure the straight-line distance between two points, and here it is used to quantify the difference between the actual parameters and the standard parameters. The calculated Euclidean distance is compared with a preset distance threshold. If the distance is not greater than the threshold, it means that there is no significant difference between the actual parameters and the standard parameters, the environmental parameters are stable, and it is marked as a "normal parameter node". Conversely, if the distance exceeds the threshold, it indicates that environmental parameters have drifted, and the drift exceeds the allowable range, marking it as a "parameter drift node." The total number of "parameter drift nodes" and the total number of "normal parameter nodes" in the currently accessed standardization environment are counted. By calculating the ratio of these two types of nodes, the "parameter drift node ratio" of the current standardization environment is obtained. If the "parameter drift node ratio" is greater than a preset ratio, it indicates that the standardization environment has a significant degree of parameter drift and is marked as an "unstable standardization environment." Otherwise, the environment is marked as a "stable standardization environment." The above steps are repeated until all standardization environments are evaluated. Finally, the status of all standardization environments is obtained, including which environments are stable and which are unstable. This method, through quantitative analysis of the stability and drift of standardization environment parameters, provides a reliable foundation for subsequent optimization of standardization environment management and prevention of environmental parameter anomalies, enabling precise control of the standardization environment and helping to improve the quality and efficiency of rice worm standardization.

[0093] Furthermore, if a grading environment is marked as an unstable grading environment, the operating parameters of the corresponding functional adjustment module in the grading device will be adjusted accordingly, specifically as follows:

[0094] Obtain the structural drawing information of the roughening device, and construct a dynamic simulation system for the roughening device based on the structural drawing information and simulation software.

[0095] Obtain the functional specification of the grading device, determine the functional adjustment module corresponding to the grading environment marked as unstable based on the functional specification, and obtain the real-time working parameters of the functional adjustment module and the dynamic set of actual grading environment parameters marked as unstable based on the Internet of Things.

[0096] Import the real-time operating parameters of the function adjustment module and the dynamic set of parameters of the actual roughing environment marked as the unstable roughing environment into the dynamic simulation system;

[0097] A genetic algorithm is introduced, and the preset number of iterations and the preset iteration adjustment class range are set according to the genetic algorithm.

[0098] Based on the preset iterative adjustment level range, the real-time working parameters of the functional adjustment module in the dynamic simulation system are simulated and adjusted to obtain the dynamic set of simulated rough environment parameters marked as unstable rough environment after simulation adjustment; at the same time, the simulation adjustment parameters corresponding to the simulation adjustment are output, and a working parameter adjustment scheme is generated according to the simulation adjustment parameters.

[0099] Repeat the above steps until the preset number of iterations is reached. This will yield the dynamic set of simulation standard environment parameters and the working parameter control scheme corresponding to each simulation control. The dynamic set of simulation standard environment parameters and the working parameter control scheme corresponding to each simulation control will be bundled together to obtain several data packets.

[0100] The functional regulation module is a crucial component of the nursery system, responsible for adjusting and controlling various parameters of the nursery environment to meet the growth and development needs of specific organisms (such as sandworms). This module may include, but is not limited to, temperature control systems, humidity control systems, light control systems, water quality regulation systems, and nutrient supply systems. For example, the temperature control system regulates the water temperature, ensuring it remains within the suitable range for sandworm growth; a stable water temperature is typically required to promote healthy growth and metabolic activities. The light control system provides necessary light conditions, aiding sandworms in photosynthesis and other light reactions required by other organisms, while also influencing their circadian rhythm and growth pattern. The water quality regulation system monitors and regulates the chemical properties of the water, such as pH, dissolved oxygen concentration, ammonia nitrogen, and nitrite, ensuring the water quality meets the sanitary standards for sandworm growth and preventing disease. The functional regulation module typically integrates sensors, actuators, and controllers to achieve real-time monitoring and automatic control of environmental parameters.

[0101] It is important to note that obtaining detailed design drawings and related technical documents of the roughing-out device is crucial. This information, including the device's structure, dimensions, materials, and component layout, forms the foundation for building a dynamic simulation. Using the acquired structural drawings and simulation software (such as ANSYS and SolidWorks Simulation), a dynamic simulation system capable of simulating the operating state of the roughing-out device is constructed. The simulation system can predict the device's performance under different operating conditions, providing a basis for subsequent optimization. By reviewing the functional specifications of the roughing-out device, the function, parameter range, and expected performance indicators of each functional adjustment module are clarified. Using Internet of Things (IoT) technologies (such as RFID and sensor networks), the operating parameters of the functional adjustment modules are monitored and collected in real time, including but not limited to current, voltage, temperature, and pressure, as well as the dynamic set of real-time roughing-out environment parameters marked as unstable roughing-out environments, such as water quality, light intensity, and oxygen concentration. The real-time operating parameters of the functional adjustment modules and the dynamic parameter set of the unstable roughing-out environment are imported into the dynamic simulation system, enabling the system to simulate the device's operating state under real-world conditions. A genetic algorithm is introduced for parameter optimization. The genetic algorithm finds the optimal solution by simulating the natural selection process (selection, crossover, and mutation). By setting a preset number of iterations and an iterative adjustment range, the parameters of the functional adjustment module are gradually adjusted to optimize the stability of the standardization environment. Through simulation control, a dynamic set of parameters for the adjusted simulated standardization environment is obtained. Simultaneously, the parameters used in each round of control are output, and a working parameter control scheme is generated based on these parameters, which are then adjusted using the parameters in actual control. The dynamic set of parameters after each simulation control and the corresponding control scheme are bundled into a data package, forming a series of optimized schemes. These data packages can be used for subsequent analysis, comparison, and selection, helping to make better parameter adjustment decisions. The entire process is a closed-loop, iterative optimization process. Through continuous simulation, control, and evaluation, the ultimate goal is to improve the performance of the standardization device and optimize the standardization environment.

[0102] It also includes the following steps:

[0103] Obtain a preset grading scheme for the coarse worm to be calibrated, and obtain a dynamic set of standard grading environment parameters marked as unstable grading environments within a preset simulation time period based on the preset grading scheme;

[0104] Calculate the cosine similarity between the dynamic set of simulated bolding environment parameters in each data packet and the dynamic set of standard bolding environment parameters marked as unstable bolding environment within a preset simulation time period to obtain several cosine similarities.

[0105] The imported cosine similarity values ​​are sorted to obtain the maximum cosine similarity, and the dynamic set of simulated bolding environment parameters corresponding to the maximum cosine similarity is obtained; the working parameter control scheme is extracted from the data package of the dynamic set of simulated bolding environment parameters corresponding to the maximum cosine similarity.

[0106] The extracted working parameter control scheme is sent to the control terminal so that the control terminal can adjust the real-time working parameters of the functional adjustment module corresponding to the unstable standard environment based on the extracted working parameter control scheme.

[0107] It should be noted that, through simulation technology, a set of standard parameters for the standard growing environment, marked as "unstable" within a preset time period, is obtained. Here, "unstable" typically refers to a state where the parameters of the growing environment fluctuate greatly, which is detrimental to the growth of rice worms. Next, from multiple collected data packets, the simulated dynamic set of parameters for the standard growing environment is selected and compared with the dynamic set of parameters for the standard unstable environment. A set of values ​​is obtained by calculating the cosine similarity between these two sets. Cosine similarity is an index that measures the similarity of the directions of two sets of vectors; the closer the value is to 1, the more consistent the directions of the two sets of vectors, meaning the more similar the parameter dynamic sets are. All calculated cosine similarities are sorted, and the maximum value is found. The simulated dynamic set of parameters for the standard growing environment corresponding to the maximum cosine similarity is considered to be most similar to the dynamic set of parameters for the standard unstable environment. From the simulated dynamic set of parameters for the standard growing environment corresponding to the maximum cosine similarity, corresponding working parameter control schemes are extracted. These schemes typically include specific adjustment parameters of the growing environment, such as temperature, humidity, and light, to improve the growing environment. Finally, the extracted working parameter control schemes are sent to the control terminal. After receiving the instruction, the control terminal adjusts the real-time operating parameters of the functional adjustment modules corresponding to the unstable seedling environment according to the plan. The purpose of this is to reduce fluctuations and improve the seedling environment by adjusting environmental parameters, thereby promoting the healthy growth of rice worms. This optimizes and adjusts the seedling environment, improving the success rate and efficiency of rice worm seedling cultivation.

[0108] In practical applications, this invention may further include the following steps:

[0109] If no thickening environment is marked as an unstable thickening environment, then the video frame information of the larvae of the grass worm in the thickening device is obtained, and the video frame information is divided into several consecutive frame feature images.

[0110] The edge detection algorithm based on the Canny operator is used to extract features from each feature image to obtain the position nodes of the larvae in each feature image; the position nodes of the larvae in each feature image are integrated to obtain the actual activity trajectory of the larvae.

[0111] The actual activity trajectory of the larvae of the worm is imported into a pre-constructed pathological prediction model for prediction, and the pathological probability value of the larvae of the worm is obtained.

[0112] The pathological images of larvae infected with different pathogens are obtained through big data networks, a knowledge graph is constructed, and the knowledge graph is divided into several storage spaces. The pathological images of larvae infected with different pathogens are imported into the corresponding storage spaces, and each storage space is assigned a corresponding pathogen infection label according to the different pathogen infection types.

[0113] If the pathological probability value of the larvae of the sandworm is greater than the preset threshold, a feature image of the larvae of the sandworm is randomly obtained, and the Pearson correlation coefficient between the feature image and each pathological image is calculated; the pathological image corresponding to the maximum Pearson correlation coefficient is obtained.

[0114] Extract the pathogen infection label assigned to the storage space of the pathological image corresponding to the maximum Pearson correlation coefficient to obtain the pathogen infection type of the larvae of the sandworm in the labeling device; send the pathogen infection type of the larvae of the sandworm in the labeling device to the preset terminal for display.

[0115] It should be noted that if no marked environment is identified as an unstable marked environment, it indicates that the abnormal feeding state of the larvae is not caused by the marked environment. In this case, the larvae are highly likely to be infected by pathogens, leading to abnormal feeding. Further analysis is required. Specifically, firstly, video frame information of the larvae within the marked environment is acquired and segmented into a series of consecutive feature images. Then, the system uses the Canny operator edge detection algorithm to process the feature images of each frame. The Canny operator is a widely used edge detection algorithm that can effectively identify edges and contours in images. Through this process, the system can locate and record the position nodes of the larvae in each frame. Integrating the extracted position nodes, the activity trajectory of the larvae in the entire video sequence can be constructed. This trajectory reveals the movement patterns and behavioral habits of the larvae, providing an important basis for subsequent health status assessment. Next, the system's pathological prediction model predicts the pathological probability value of the larvae based on the activity trajectory. This model may be trained based on historical data and known pathological patterns, predicting the health risk of the larvae by comparing the similarity between the activity trajectory and known pathological patterns. The system collects pathological images of rice worm larvae under different pathogen infections through a big data network and stores them in a categorized manner. Each storage space is associated with a specific pathogen infection type, forming a knowledge graph. This structured data storage method facilitates subsequent querying and analysis. When the pathological probability value of a rice worm larva exceeds a preset threshold, the system randomly selects a feature image and calculates its Pearson correlation coefficient with the pathological images in the database. The Pearson correlation coefficient is a statistical indicator that measures the degree of linear correlation between two variables; here, it is used to assess the similarity between the feature image and known pathological images. Then, the system selects the pathological image corresponding to the largest Pearson correlation coefficient and further extracts the pathogen infection label corresponding to its storage space. This label reveals the most likely pathogen infection type of the rice worm larva. Finally, the diagnostic result (i.e., the pathogen infection type of the rice worm larva) is sent to a preset terminal for display, providing farmers with real-time health warning information so that appropriate prevention and control measures can be taken in a timely manner. This step, through automated monitoring and intelligent analysis, greatly improves the efficiency and accuracy of rice worm larva health monitoring, enabling timely implementation of appropriate sterilization measures and improving the efficiency and effectiveness of rice worm rearing.

[0116] More specifically, the method for constructing the pathological prediction model is as follows:

[0117] A large amount of activity trajectory data of larvae of the rice worm under healthy and infected states was obtained through big data networks;

[0118] The activity trajectory data of healthy larvae of the sandworm is used as the first graph node, and the activity trajectory data of larvae of the sandworm under pathogen infection is used as the second graph node. A graph convolutional network is constructed based on the first graph node and the second graph node.

[0119] The graph convolutional layer is used to extract the feature information of each graph node to obtain the topological feature information of the activity trajectory data of the larvae of the rice worm under the conditions of health and infection.

[0120] A topological feature matrix is ​​constructed based on the topological feature information of the activity trajectory data of juvenile worms in healthy and infected states.

[0121] A pathological prediction model is constructed based on a convolutional neural network. The topological feature matrix is ​​imported into the pathological prediction model for backpropagation training. After each training, the evaluation index of the model is obtained. The trained pathological prediction model is output after all evaluation indexes meet the preset requirements.

[0122] The evaluation metrics include accuracy, recall, and F1 score.

[0123] In practical applications, this invention may further include the following steps:

[0124] If the pathological probability value of the larvae of the rice worm is greater than the preset threshold, the sterilization log in the standard equipment is obtained, and the historical sterilization condition data at the preset sterilization time node is obtained according to the sterilization log.

[0125] Based on the historical sterilization condition data, a three-dimensional model of the historical sterilization condition is constructed, and an overlap analysis is performed between the three-dimensional model of the historical sterilization condition and the three-dimensional model of the preset sterilization condition to obtain the degree of overlap between the three-dimensional model of the historical sterilization condition and the three-dimensional model of the preset sterilization condition.

[0126] If the overlap between the historical sterilization condition 3D model and the preset sterilization condition 3D model is not greater than the preset overlap threshold, then the non-overlapping area between the historical sterilization condition 3D model and the preset sterilization condition 3D model is obtained, defined as a sterilization abnormal area, and the sterilization abnormal area is sent to the control terminal to control the sterilization mechanism to sterilize the sterilization abnormal area.

[0127] If the overlap between the historical sterilization condition 3D model and the preset sterilization condition 3D model is greater than a preset overlap threshold, then the historical sterilization parameters of the sterilization mechanism at the preset sterilization time node are obtained; the historical sterilization parameters are compared with the preset sterilization parameters to obtain the sterilization parameter drift value; when the sterilization mechanism performs the next sterilization process, the sterilization parameters of the sterilization mechanism are corrected based on the sterilization parameter drift value.

[0128] It should be noted that when the system detects that the pathological probability value of rice worm larvae exceeds a preset threshold, it first retrieves the sterilization log from the standard equipment and extracts historical sterilization condition data based on the log information. Then, the system constructs a 3D model of the historical sterilization condition data and compares it with a preset 3D model of the sterilization condition. If the overlap between the historical and preset sterilization conditions is lower than the preset threshold, it indicates an anomaly in the historical sterilization process. The system identifies the abnormal area and sends it to the control terminal. The control terminal, based on the abnormal area information, instructs the sterilization unit to perform targeted sterilization treatment on that area to eliminate potential pathogen threats. Conversely, if the overlap between the historical and preset sterilization conditions is higher than the preset threshold, it indicates that the historical sterilization process basically conforms to the preset plan, but the system further analyzes the sterilization parameters to identify parameter drift values. In the next sterilization process, the system fine-tunes the sterilization parameters of the sterilization unit based on the parameter drift values ​​to ensure the stability and reliability of the sterilization effect. This method effectively improves sterilization efficiency, reduces the risk of bacterial infection during the growth of rice worms, and provides strong protection for the healthy growth of rice worm larvae.

[0129] The second aspect of this invention discloses a novel grain-growing system for rice worms based on intelligent control, applicable to any of the novel grain-growing methods for rice worms based on intelligent control described in any one of the claims, such as... Figure 3 As shown, the novel rice worm labeling system includes:

[0130] The main body of the nursery pool 1011 is used to provide a pre-set nursery growth environment for the larvae of the rice worm;

[0131] The function adjustment module 1012 is used to control various parameters of the standard environment, including water temperature, salinity, pH, dissolved oxygen, pH value, air pressure, light intensity, water flow speed and water depth, so as to simulate the natural growth environment of the larvae of the worm.

[0132] The feed dispensing module 1013 automatically feeds the larvae of the rice worm according to their growth stage and food requirements, and controls the feeding frequency and type of feed.

[0133] The data acquisition and analysis module 1014 collects various data during the nursery process, including environmental parameters, feed intake, and growth status, and performs data analysis.

[0134] The intelligent control module 1015 integrates data from various modules and automatically controls the operating parameters of each module according to a preset standardization scheme.

[0135] A third aspect of the present invention discloses a computer-readable storage medium comprising a novel method program for marking and thickening rice worms based on intelligent control. When the novel method program for marking and thickening rice worms based on intelligent control is processed and executed, it implements any of the steps of the novel method program for marking and thickening rice worms based on intelligent control as described in the present invention.

[0136] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A novel method for profiling rice worms based on intelligent control, characterized in that, Includes the following steps: Based on big data networks, standard food intake datasets of larvae in each growth stage of the worm are obtained within a unit of time, and a Bayesian food intake prediction model for worm larvae is constructed based on these datasets. Obtain the marking feature information of the larvae of the worm in the marking device, and import the marking feature information into the Bayesian food prediction model for prediction to obtain the predicted food amount data of the larvae of the worm in the marking device within a preset time period. The tailwater of the standardization device is sampled and tested at a preset time point to obtain the concentration of remaining feed in the tailwater. The actual food consumption data of the larvae of the worms within the preset time period is calculated based on the concentration of remaining feed. The feeding status of the larvae of the worms is analyzed by combining the actual food consumption data and the predicted food consumption data within the preset time period. If the feeding state of the larvae of the rice worm is abnormal, various actual scale environmental parameters are obtained at several preset time points to obtain a dynamic set of various actual scale environmental parameters; the characteristics of various actual scale environmental parameters are analyzed to obtain the status of various scale environmental conditions. If a roughing environment is marked as an unstable roughing environment, the operating parameters of the corresponding functional adjustment module in the roughing device will be adjusted.

2. The novel method for profiling rice worms based on intelligent control according to claim 1, characterized in that, Based on big data networks, standard food intake datasets of larvae at each growth stage of the *Gnaphalium affine* were obtained within a unit of time. A Bayesian network prediction model was then constructed based on these datasets. Specifically: Obtain a preset grading scheme for the larvae to be graded, construct key search terms based on the preset grading scheme, and search the big data network resource library based on the key search terms to obtain the standard food intake dataset of larvae at each grading stage within a unit of time. The Apriori algorithm is introduced, and correlation analysis is performed on the standard feeding data of various larval stages of the mulberry worm within a unit time to obtain the correlation strength between the feeding of any mulberry worm stage and the feeding of other mulberry worm stages. The Bayesian network's food quantity nodes are set according to various standard feeding stages of the larvae of the worm, and the edges of the Bayesian network are set according to the corresponding standard feeding dataset; where each feeding quantity node in the Bayesian network represents a specific standard feeding stage, and each edge represents the feeding quantity relationship between two feeding quantity nodes. If the correlation strength between the amount of food in a certain standard stage and another standard stage is greater than a preset strength threshold, then the corresponding food nodes and edges are connected in the Bayesian network to generate a topology network structure graph; the topology relationship of food between each standard stage is obtained according to the topology network structure graph. Based on the topological relationship of food consumption between each stage and combined with the maximum probability estimation method, the conditional probability range of each food consumption node in the Bayesian network is estimated. Based on the conditional probability range of each food consumption node, a food consumption range attribute is added to each food consumption node to obtain the Bayesian food consumption prediction model for the larvae of the worm.

3. The novel method for profiling rice worms based on intelligent control according to claim 1, characterized in that, The feeding status of rice worm larvae was analyzed by combining actual and predicted feeding data within a preset time period. Specifically: The difference between the actual and predicted feeding data of the larvae of the rice worm in the standard feeding device within a preset time period is calculated to obtain the feeding data deviation value; Determine whether the deviation value of the food intake data is greater than a preset food intake deviation threshold; If the deviation value of the food intake data is greater than the preset food intake deviation threshold, the feeding status of the larvae of the rice worm is marked as normal. If the deviation value of the food intake data is not greater than the preset food intake deviation threshold, the feeding status of the larvae of the rice worm is marked as an abnormal state.

4. The novel method for profiling rice worms based on intelligent control according to claim 1, characterized in that, Characteristic analysis was performed on the dynamic sets of various actual coarse-graining environment parameters to obtain the various coarse-graining environment states, as follows: Obtain a preset grading scheme for the coarse worm to be calibrated, and obtain a dynamic set of various standard grading environmental parameters within a preset grading time period based on the preset grading scheme; Construct a first blank matrix and a second blank matrix, randomly access a standard bolding environment, and obtain the dynamic set of actual standard bolding environment parameters and the dynamic set of standard bolding environment parameters corresponding to that standard bolding environment; The actual dynamic set of standard roughing environment parameters and the standard dynamic set of standard roughing environment parameters corresponding to these roughing environment parameters are filled into the first blank matrix and the second blank matrix respectively in a preset order to obtain the actual roughing environment parameter matrix and the standard roughing environment parameter matrix. Calculate the Euclidean distance between elements at the same position in the actual and standard roughened environmental parameter matrices; and compare the Euclidean distance between each element at the same position in the two matrices with a preset distance threshold. If the Euclidean distance between elements at the same position in two matrices is not greater than a preset distance threshold, it means that the actual bolding environmental parameter at the corresponding time node has not drifted or the drift amplitude is within the allowable range. Then, the actual bolding environmental parameter at that time node is marked as a normal parameter node. If the Euclidean distance between elements at the same position in two matrices is greater than a preset distance threshold, it indicates that the actual bolding environmental parameter at the corresponding time node has drifted and the drift amplitude exceeds the allowable range. Then, the actual bolding environmental parameter at that time node is marked as a parameter drift node. The total number of parameter drift nodes and the total number of parameter normal nodes in the currently accessed standard environment are counted. The ratio of the total number of parameter drift nodes to the total number of parameter normal nodes in the currently accessed standard environment is calculated to obtain the percentage of parameter drift nodes in the currently accessed standard environment. If the percentage of drifting nodes in the currently accessed bolding environment is greater than the preset percentage, the currently accessed bolding environment is marked as an unstable bolding environment; if the percentage of drifting nodes in the currently accessed bolding environment is not greater than the preset percentage, the parameters of the currently accessed bolding environment are marked as a stable bolding environment. Visit the next bolded environment that has not been visited before, and repeat the above steps until all bolded environments have been visited, and obtain the status of each bolded environment.

5. The novel method for profiling rice worms based on intelligent control according to claim 1, characterized in that, If a grading environment is marked as an unstable grading environment, the operating parameters of the corresponding functional adjustment module in the grading device will be adjusted accordingly, specifically as follows: Obtain the structural drawing information of the roughening device, and construct a dynamic simulation system for the roughening device based on the structural drawing information and simulation software. Obtain the functional specification of the grading device, and determine the functional adjustment module corresponding to the grading environment marked as unstable based on the functional specification; The system acquires the real-time operating parameters of the function adjustment module and the dynamic set of actual coarseness environment parameters marked as unstable coarseness environment based on the Internet of Things. Import the real-time operating parameters of the function adjustment module and the dynamic set of parameters of the actual roughing environment marked as the unstable roughing environment into the dynamic simulation system; A genetic algorithm is introduced, and the preset number of iterations and the preset iteration adjustment class range are set according to the genetic algorithm. Based on the preset iterative adjustment level range, the real-time working parameters of the functional adjustment module in the dynamic simulation system are simulated and adjusted to obtain the dynamic set of simulated rough environment parameters marked as unstable rough environment after simulation adjustment; at the same time, the simulation adjustment parameters corresponding to the simulation adjustment are output, and a working parameter adjustment scheme is generated according to the simulation adjustment parameters. Repeat the above steps until the preset number of iterations is reached. This will yield the dynamic set of simulation standard environment parameters and the working parameter control scheme corresponding to each simulation control. The dynamic set of simulation standard environment parameters and the working parameter control scheme corresponding to each simulation control will be bundled together to obtain several data packets.

6. A novel method for profiling rice worms based on intelligent control according to claim 5, characterized in that, It also includes the following steps: Obtain a preset grading scheme for the coarse worm to be calibrated, and obtain a dynamic set of standard grading environment parameters marked as unstable grading environments within a preset simulation time period based on the preset grading scheme; Calculate the cosine similarity between the dynamic set of simulated bolding environment parameters in each data packet and the dynamic set of standard bolding environment parameters marked as unstable bolding environment within a preset simulation time period to obtain several cosine similarities. The imported cosine similarity values ​​are sorted to obtain the maximum cosine similarity, and the dynamic set of simulated bolding environment parameters corresponding to the maximum cosine similarity is obtained; the working parameter control scheme is extracted from the data package of the dynamic set of simulated bolding environment parameters corresponding to the maximum cosine similarity. The extracted working parameter control scheme is sent to the control terminal so that the control terminal can adjust the real-time working parameters of the functional adjustment module corresponding to the unstable standard environment based on the extracted working parameter control scheme.

7. The novel method for profiling rice worms based on intelligent control according to claim 1, characterized in that, The standard environment includes temperature, salinity, pH, feeding speed, air pressure, pH value, dissolved oxygen concentration, light intensity, water depth, and water flow speed.

8. The novel method for profiling rice worms based on intelligent control according to claim 1, characterized in that, The growth characteristics include the number of rice worms, the growth stage, and the feeding time.

9. A novel grain-growing system based on intelligent control, applied to the novel grain-growing method based on intelligent control as described in any one of claims 1 to 8, characterized in that, The novel rice worm labeling system includes: The main body of the nursery is used to provide a pre-designed nursery environment for the larvae of the rice worm; The function adjustment module is used to control various parameters of the standard environment, including water temperature, salinity, pH, dissolved oxygen, pH value, air pressure, light intensity, water flow speed and water depth, in order to simulate the natural growth environment of the larvae of the worm. The feed dispensing module automatically feeds the larvae of the rice worm according to their growth stage and food requirements, and controls the feeding frequency and type of feed. The data acquisition and analysis module collects various data during the nursery process, including environmental parameters, feed intake, and growth status, and performs data analysis. The intelligent control module integrates data from various modules and automatically controls the operating parameters of each module according to a preset scaling scheme.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a novel method program for the identification of grass worms based on intelligent control. When the novel method program for the identification of grass worms based on intelligent control is processed and executed, it implements the steps of the novel method for the identification of grass worms based on intelligent control as described in any one of claims 1 to 8.

Images