Aquaculture oxygenation equipment automatic adjusting method and system based on environmental parameters

Abstract

The invention provides an automatic adjustment method and system for aquaculture oxygenation equipment based on environmental parameters, and relates to the technical field of data prediction adjustment. Environmental parameter data in an aquaculture area are acquired, and data classification and preprocessing are performed based on the acquired environmental parameter data; an oxygen demand prediction model is constructed by adopting a deep learning algorithm, the oxygen demand prediction model is trained and optimized based on the preprocessed environmental parameter data, the oxygen demand in the water body in a period of time in the future is predicted based on the optimized oxygen demand prediction model, and an oxygenation mode is automatically selected and switched. According to the method, the working intensity of the oxygenation equipment is adjusted, the working intensity is not purely switched, and the working mode and intensity of the oxygenation equipment are dynamically adjusted by calculating the oxygen demand quantity in the water body in a period of time in the future, so that automatic switching of the working modes of the oxygenation equipment is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data prediction and regulation, and in particular to an automatic regulation method and system for aquaculture oxygenation equipment based on environmental parameters. Background Art

[0002] Dissolved oxygen, as an important parameter in the water environment, directly affects the growth and survival rate of aquatic organisms. The dissolved oxygen concentration in water is affected by many factors, including environmental parameters such as water temperature, pH value and ammonia nitrogen concentration. Traditional artificial oxygenation methods are inefficient and it is difficult to achieve precise control of oxygen demand. At the same time, different oxygenation equipment has a single working mode and it is difficult to flexibly adjust according to the real-time needs of the water body, resulting in energy waste and increased breeding costs.

[0003] Although the current aquaculture oxygen supply system or method has introduced some automation technology, it still has many shortcomings, specifically:

[0004] First, existing systems generally rely on fixed oxygenation strategies and lack real-time monitoring and intelligent adjustment functions of environmental parameters, and are unable to dynamically adjust the working intensity and mode of oxygenation equipment according to actual needs; second, the collection and processing of environmental parameters are relatively simple, and modern data analysis and machine learning technologies are not fully utilized to predict future oxygen demand; third, traditional systems can often only rely on a single oxygenation mode and cannot flexibly switch to the most suitable oxygenation mode under different oxygen demand scenarios. These problems limit the refined management of aquaculture and cannot ensure the best living environment for aquatic organisms.

[0005] Therefore, there is an urgent need to propose an automatic adjustment method and system for aquaculture oxygenation equipment based on environmental parameters. Summary of the invention

[0006] The present invention provides an automatic adjustment method and system for aquaculture oxygenation equipment based on environmental parameters, aiming to solve the technical problems in the related technology of lacking real-time monitoring and intelligent adjustment functions of environmental parameters, and being unable to dynamically adjust the working intensity and mode of the oxygenation equipment according to actual needs, avoiding the problem of not fully utilizing modern data analysis and machine learning technologies to predict future oxygen demand and use, and avoiding the subsequent problem of being unable to flexibly switch the most suitable oxygenation mode under different oxygen demand scenarios. By real-time calculation of the oxygen demand in the water body over a period of time in the future, the working mode and intensity of the oxygenation equipment are dynamically adjusted to realize automatic switching of the oxygenation mode, and fine-tuning is performed according to the oxygen demand of the water body, thereby improving the utilization efficiency of the oxygenation equipment, improving the stability of the water environment, and providing an automated solution for aquaculture.

[0007] In order to achieve the above purpose, the embodiments of the present application are implemented in the following manner:

[0008] In a first aspect, the present application provides an automatic adjustment method for aquaculture oxygenation equipment based on environmental parameters, including the following:

[0009] Step S100, obtaining environmental parameter data in the aquaculture area;

[0010] Step S200, performing data classification and preprocessing based on the environmental parameter data acquired in step S100;

[0011] Step S300: constructing an oxygen demand prediction model using a deep learning algorithm, and training and optimizing the oxygen demand prediction model based on the pre-processed environmental parameter data;

[0012] Step S400, predicting the oxygen demand in the water body within a future period of time based on the optimized oxygen demand prediction model obtained in step S300;

[0013] Step S500: Based on the prediction result obtained in step S400, the oxygen enrichment mode is automatically selected and switched to adjust the working intensity of the oxygen enrichment equipment.

[0014] In an embodiment of the present application, environmental parameter data in an aquaculture area is obtained, and data classification and preprocessing are performed based on the obtained environmental parameter data, an oxygen demand prediction model is constructed using a deep learning algorithm, and the oxygen demand prediction model is trained and optimized based on the preprocessed environmental parameter data. Based on the obtained optimized oxygen demand prediction model, the oxygen demand in the water body in the future period is predicted, the oxygenation mode is automatically selected and switched, and the working intensity of the oxygenation equipment is adjusted. This method does not simply switch the oxygenation mode, but automatically switches according to the prediction results and the actual environmental parameters. By real-time calculation of the oxygen demand in the water body in the future period, the working mode and intensity of the oxygenation equipment are dynamically adjusted to achieve automatic switching of the working mode of the oxygenation equipment, and fine-tuned regulation is performed according to the oxygen demand of the water body, thereby improving the utilization efficiency of the oxygenation equipment, reducing energy consumption, and improving the stability of the water environment, providing an automated solution for aquaculture.

[0015] In conjunction with the first aspect, in some specific implementations, obtaining environmental parameter data in an aquaculture area specifically includes:

[0016] Step S101, obtaining environmental parameter data, deploying environmental parameter data sensors to obtain environmental parameters of the aquaculture area, the environmental parameters including water temperature, dissolved oxygen concentration, pH value, ammonia nitrogen concentration, water flow velocity, atmospheric pressure and air temperature.

[0017] Among them, the environmental parameter data sensors include the following:

[0018] Water temperature sensor, used to monitor water temperature.

[0019] Dissolved oxygen sensor, used to monitor the concentration of dissolved oxygen in water.

[0020] pH sensor, used to measure the acidity or alkalinity of water.

[0021] Ammonia nitrogen sensor is used to measure the concentration of ammonia nitrogen in water. Ammonia nitrogen is an important water quality indicator in aquaculture.

[0022] Flow rate sensors are used to monitor the speed of water flow, which affects the oxygen transfer in the water.

[0023] Atmospheric pressure sensors are used to monitor the air pressure in the aquaculture area, as changes in air pressure can affect oxygen exchange in the water.

[0024] Air temperature sensor is used to monitor the external ambient temperature, which affects the water temperature and oxygen demand.

[0025] Step S102: Environmental parameter data transmission: The environmental parameter data sensor transmits the acquired environmental parameter data to a cloud database via a wireless communication network.

[0026] Among them, the database is a distributed time series database that supports data writing and query, as well as data redundancy and regular backup.

[0027] In combination with the first aspect, in some specific implementations, classification and preprocessing are performed based on the environmental parameter data acquired in step S100, specifically including:

[0028] Preprocessing includes data cleaning, outlier detection and dimensionality reduction.

[0029] Step S201 , data classification, based on the environmental parameter data obtained in step S100 , classify the data into different categories, and perform group processing based on the data of different categories.

[0030] Grouping processing includes classification by time, by space, and by data source, specifically:

[0031] Classify by time: classify on the time axis based on the time point or time interval of data collection to separate real-time data from historical data. The time intervals include seconds, minutes and hours.

[0032] According to spatial classification, the data from multiple sensors in the breeding area are spatially classified based on the physical locations of the sensors, which include different depths and different water areas.

[0033] Classify by data source and group data based on sensor type.

[0034] Step S202: data cleaning: data cleaning is performed based on the grouped environmental parameter data. Data cleaning includes filling in missing values ​​and removing erroneous values.

[0035] Step S203: outlier detection: using an autoencoder to perform outlier detection on the environmental parameter data after data cleaning.

[0036] Step S204, dimensionality reduction processing, using principal component analysis method, to perform dimensionality reduction processing on the environmental parameter data after removing abnormal data, to obtain pre-processed environmental parameter data.

[0037] In combination with the first aspect, in some specific embodiments, a deep learning algorithm is used to construct an oxygen demand prediction model, and the oxygen demand prediction model is trained and optimized based on the pre-processed environmental parameter data, specifically including:

[0038] Step S301, configuring an oxygen demand prediction model, using a standardized LSTM network architecture, the standardized LSTM network architecture includes an input layer, an LSTM hidden layer, a fully connected layer and an output layer.

[0039] Configuring the oxygen demand prediction model also includes configuring the LSTM hidden layer, setting the time step, setting the loss function and the activation function.

[0040] Configure LSTM hidden layers, the number is limited to 3 layers, and a single layer contains 64 neurons.

[0041] Set the time step to the past 24 hours.

[0042] Set the loss function and use the mean square error as the loss function, specifically:

[0043]

[0044] In the formula, y i is the actual oxygen demand value, is the model prediction value, n is the number of data points;

[0045] Step S302: train the oxygen demand prediction model, use the LSTM input features as input training data, use the historical oxygen demand values ​​as output training data, input the LSTM input features into the oxygen demand prediction model, obtain a preliminary prediction result, and then compare the prediction result of the model with the actual historical oxygen demand value to minimize the error.

[0046] The batch size of training data is set to 32.

[0047] Step S303, verification evaluation, verification includes selection of evaluation indicators and cross-validation;

[0048] The evaluation indicators are selected by using the mean absolute error, specifically:

[0049]

[0050] In the formula, n represents the number of samples, y i represents the true value of the i-th sample, represents the predicted value of the i-th sample, It represents the absolute error between the predicted value and the true value.

[0051] Step S304: Optimize the oxygen demand prediction model by using incremental learning. Based on the acquisition of new data, the LSTM model is continuously optimized by incremental learning. Each time new data arrives, a small-scale iterative update is performed based on the parameters of the original model, rather than completely retraining the oxygen demand prediction model.

[0052] In combination with the first aspect, in some specific implementations, based on the optimized oxygen demand prediction model obtained in step S300, the oxygen demand in the water body in the future period is predicted, specifically:

[0053] Step S400: Based on the optimized model, the pre-processed environmental parameter data is used as input features and input into the optimized LSTM model. The LSTM model will output the oxygen demand value in the water body in the future period of time, specifically:

[0054]

[0055] Among them, O d (t) represents the oxygen demand in the water body in the future period of time, f LSTM Represents the prediction function through the LSTM model, Represents the preprocessed environmental parameter data, represents the environmental parameter data vector at time t, n represents the time step, and h(t) represents the hidden state of the LSTM model at time t.

[0056] In combination with the first aspect, in some specific implementations, based on the prediction result obtained in step S400, the oxygen enrichment mode is automatically selected and switched to adjust the working intensity of the oxygen enrichment equipment, specifically including:

[0057] Step S501: based on the oxygen demand O in the water body within a period of time in the future d (t) and the dissolved oxygen concentration C(t) of the water body to select the oxygenation mode, specifically:

[0058]

[0059] Where M selectedis the selected oxygenation mode, O d (t) represents the oxygen demand in the water body in the future period of time, T H and T L They represent the oxygen thresholds of high demand and low demand, respectively, C(t) represents the dissolved oxygen concentration of the current water body, and C L and C H They represent the lower and upper thresholds of dissolved oxygen concentration, θ(t) represents the current water temperature, and θ optimal Indicates the preset water temperature; Δθ indicates the allowable range of water temperature deviation; G indicates gas oxygenation; E indicates electrolytic oxygenation; and B indicates microbubble oxygenation.

[0060] The oxygen enrichment mode selection is as follows:

[0061] A1. Gas oxygenation is suitable for scenarios where the oxygen demand in water is high and the dissolved oxygen concentration is low. It increases the oxygen content in water by quickly injecting oxygen or air.

[0062] A2. Electrolytic oxygenation is suitable for a stable environment and can maintain dissolved oxygen content for a long time. It produces oxygen by electrolyzing water with low energy consumption and high efficiency.

[0063] A3. Microbubble oxygenation is suitable for fine control and oxygen replenishment in local areas of water bodies. Microbubble generators produce tiny bubbles to improve oxygen transfer efficiency.

[0064] Step S502: activating the corresponding oxygenation equipment based on the oxygenation mode selected in step S501, and based on the oxygen demand in the water body within a period of time in the future. d (t) and the dissolved oxygen concentration C(t) of the water body automatically adjust the working intensity of the oxygenation equipment, specifically:

[0065] P(t)=α·(O d (t)-C(t))+β

[0066] Where P(t) represents the working intensity of the oxygenation equipment at time t, O d (t) represents the oxygen demand in the water body in the future, C(t) represents the dissolved oxygen concentration in the water body, O d (t)-C(t) represents the difference between the oxygen demand in the water body and the dissolved oxygen concentration in the water body in the future. If the dissolved oxygen concentration in the water body is lower than the demand, that is, O d (t)>C(t), the working intensity of the equipment is increased. If the dissolved oxygen concentration of the water body is close to the demand, the working intensity is reduced. α is the adjustment coefficient, which is used to control the response sensitivity of the working intensity to the demand and concentration difference. β is the basic working intensity, that is, the minimum working intensity of the equipment.

[0067] Step S503, emergency control, to ensure the continuity of oxygen supply, set the emergency mode, when equipment failure and oxygen demand forecast has a large error, start the backup equipment or select a more efficient oxygenation mode to ensure that the oxygen supply in the water is not interrupted.

[0068] When a failure of the oxygen enrichment equipment is detected or the prediction error exceeds the set threshold ΔT, the backup equipment is immediately started and automatically switches to the high-efficiency oxygen enrichment mode, specifically:

[0069] M emergency =G,ifΔO d (t)>ΔT

[0070] Where M emergency It is the oxygenation mode in emergency state. The oxygenation mode in emergency state is gas oxygenation. ΔO d (t) is the oxygen demand prediction error, and ΔT is the error tolerance threshold.

[0071] In a second aspect, the present application further provides an automatic adjustment system for aquaculture oxygenation equipment based on environmental parameters, including the following:

[0072] A parameter acquisition unit, used to acquire environmental parameter data in the aquaculture area;

[0073] A preprocessing classification unit, used for classifying and preprocessing the acquired environmental parameter data;

[0074] A model configuration unit, used to construct an oxygen demand prediction model using a deep learning algorithm, and to train and optimize the oxygen demand prediction model based on pre-processed environmental parameter data;

[0075] A prediction unit, used for predicting the oxygen demand in the water body within a future period of time based on the obtained optimized oxygen demand prediction model;

[0076] The mode switching unit is used to automatically select and switch the oxygen enrichment mode and adjust the working intensity of the oxygen enrichment equipment based on the prediction results obtained by the oxygen demand prediction model.

[0077] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0079] Figure 1 A flow chart of an automatic adjustment method for aquaculture oxygenation equipment based on environmental parameters provided in an embodiment of the present application.

[0080] Figure 2 A schematic diagram of an automatic adjustment system for aquaculture oxygenation equipment based on environmental parameters provided in an embodiment of the present application.

[0081] Description of the drawings: 11. Parameter acquisition unit; 12. Preprocessing classification unit; 13. Model configuration unit; 14. Prediction unit; 15. Mode switching unit. DETAILED DESCRIPTION

[0082] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0083] See also Figure 1 , Figure 1 A flow chart of an automatic adjustment method for aquaculture oxygenation equipment based on environmental parameters is provided for an embodiment of the present application.

[0084] In this embodiment, the automatic adjustment method of aquaculture oxygenation equipment based on environmental parameters may include step S100, step S200, step S300, step S400 and step S500.

[0085] Step S100: Acquire environmental parameter data in the aquaculture area.

[0086] Here, specifically include:

[0087] Step S101, obtaining environmental parameter data, deploying sensors to obtain environmental parameters of the aquaculture area, the environmental parameters including water temperature, dissolved oxygen concentration, pH value, ammonia nitrogen concentration, water flow velocity, atmospheric pressure and air temperature.

[0088] Step S102: Environmental parameter data transmission: The environmental parameter data acquired by the sensor is transmitted to a cloud database via a wireless communication network.

[0089] Step S200, performing data classification and preprocessing based on the environmental parameter data acquired in step S100;

[0090] Here, specifically include:

[0091] Preprocessing includes data cleaning, outlier detection and dimensionality reduction.

[0092] Step S201 , data classification, based on the environmental parameter data obtained in step S100 , classify the data into different categories, and perform group processing based on the data of different categories.

[0093] Grouping processing includes classification by time, by space, and by data source, specifically:

[0094] Classify by time: classify on the time axis based on the time point or time interval of data collection to separate real-time data from historical data. The time intervals include seconds, minutes and hours.

[0095] According to spatial classification, the data from multiple sensors in the breeding area are spatially classified based on the physical locations of the sensors, which include different depths and different water areas.

[0096] Classify by data source and group data based on sensor type.

[0097] Step S202: data cleaning: data cleaning is performed based on the grouped environmental parameter data. Data cleaning includes filling in missing values ​​and removing erroneous values.

[0098] Step S203: outlier detection: using an autoencoder to perform outlier detection on the environmental parameter data after data cleaning:

[0099] Step S204, dimensionality reduction processing, using principal component analysis method, to perform dimensionality reduction processing on the environmental parameter data after removing abnormal data, to obtain pre-processed environmental parameter data.

[0100] Step S300: construct an oxygen demand prediction model using a deep learning algorithm, and train and optimize the oxygen demand prediction model based on the pre-processed environmental parameter data.

[0101] Here, specifically include:

[0102] Step S301, configuring an oxygen demand prediction model, using a standardized LSTM network architecture, the standardized LSTM network architecture includes an input layer, an LSTM hidden layer, a fully connected layer and an output layer.

[0103] Configuring the oxygen demand prediction model also includes configuring the LSTM hidden layer, setting the time step, setting the loss function and the activation function.

[0104] Configure LSTM hidden layers, the number is limited to 3 layers, and a single layer contains 64 neurons.

[0105] Set the time step to the past 24 hours.

[0106] Set the loss function and use the mean square error as the loss function, specifically:

[0107]

[0108] In the formula, y i is the actual oxygen demand value, is the model prediction and n is the number of data points.

[0109] Step S302: train the oxygen demand prediction model, use the LSTM input features as input training data, use the historical oxygen demand values ​​as output training data, input the LSTM input features into the oxygen demand prediction model, obtain a preliminary prediction result, and then compare the prediction result of the model with the actual historical oxygen demand value to minimize the error.

[0110] The batch size of training data is set to 32.

[0111] Step S303, verification evaluation, verification includes selection of evaluation indicators and cross-validation;

[0112] The evaluation indicators are selected by using the mean absolute error, specifically:

[0113]

[0114] In the formula, n represents the number of samples, y i represents the true value of the i-th sample, represents the predicted value of the i-th sample, It represents the absolute error between the predicted value and the true value.

[0115] Step S304: Optimize the oxygen demand prediction model by using incremental learning. Based on the acquisition of new data, the LSTM model is continuously optimized by incremental learning. Each time new data arrives, a small-scale iterative update is performed based on the parameters of the original model, rather than completely retraining the oxygen demand prediction model.

[0116] Step S400: Based on the optimized oxygen demand prediction model obtained in step S300, the oxygen demand in the water body in the future period of time is predicted.

[0117] Here, specifically:

[0118] Based on the optimized model, the preprocessed environmental parameter data is used as input features and input into the optimized LSTM model. The LSTM model will output the oxygen demand value in the water body in the future, specifically:

[0119]

[0120] Among them, O d (t) represents the oxygen demand in the water body in the future period of time, f LSTM Represents the prediction function through the LSTM model, Represents the preprocessed environmental parameter data, represents the environmental parameter data vector at time t, n represents the time step, and h(t) represents the hidden state of the LSTM model at time t.

[0121] Step S500: Based on the prediction result obtained in step S400, the oxygen enrichment mode is automatically selected and switched to adjust the working intensity of the oxygen enrichment equipment.

[0122] Here, specifically include:

[0123] Step S501: based on the oxygen demand O in the water body within a period of time in the future d (t) and the dissolved oxygen concentration C(t) of the water body to select the oxygenation mode, specifically:

[0124]

[0125] Where M selected is the selected oxygenation mode, O d (t) represents the oxygen demand in the water body in the future period of time, T H and T L They represent the oxygen thresholds of high demand and low demand, respectively, C(t) represents the dissolved oxygen concentration of the water body, and C L and C H They represent the lower and upper thresholds of dissolved oxygen concentration, θ(t) represents the current water temperature, and θ optimal Indicates the preset water temperature; Δθ indicates the allowable range of water temperature deviation; G indicates gas oxygenation, E indicates electrolytic oxygenation, and B indicates microbubble oxygenation;

[0126] The oxygen enrichment mode selection is as follows:

[0127] A1. Gas oxygenation is suitable for scenarios where the oxygen demand in water is high and the dissolved oxygen concentration is low. It increases the oxygen content in water by quickly injecting oxygen or air.

[0128] A2. Electrolytic oxygenation is suitable for a stable environment and can maintain dissolved oxygen content for a long time. It produces oxygen by electrolyzing water with low energy consumption and high efficiency.

[0129] A3. Microbubble oxygenation is suitable for fine control and oxygen replenishment in local areas of water bodies. Microbubble generators produce tiny bubbles to improve oxygen transfer efficiency.

[0130] Step S502: activating the corresponding oxygenation equipment based on the oxygenation mode selected in step S501, and based on the oxygen demand in the water body within a period of time in the future. d (t) and the dissolved oxygen concentration C(t) of the water body automatically adjust the working intensity of the oxygenation equipment, specifically:

[0131] P(t)=α·(O d(t)-C(t))+β

[0132] Where P(t) represents the working intensity of the oxygenation equipment at time t, O d (t) represents the oxygen demand in the water body in the future, C(t) represents the dissolved oxygen concentration in the water body, O d (t)-C(t) represents the difference between the oxygen demand in the water body and the dissolved oxygen concentration in the water body in the future. If the dissolved oxygen concentration in the water body is lower than the demand, that is, O d (t)>C(t), the working intensity of the equipment is increased. If the dissolved oxygen concentration of the water body is close to the demand, the working intensity is reduced. α is the adjustment coefficient, which is used to control the response sensitivity of the working intensity to the demand and concentration difference. β is the basic working intensity, that is, the minimum working intensity of the equipment.

[0133] Step S503, emergency control, to ensure the continuity of oxygen supply, set the emergency mode, when equipment failure and oxygen demand forecast has a large error, start the backup equipment or select a more efficient oxygenation mode to ensure that the oxygen supply in the water is not interrupted.

[0134] When a failure of the oxygen enrichment equipment is detected or the prediction error exceeds the set threshold ΔT, the backup equipment is immediately started and automatically switches to the high-efficiency oxygen enrichment mode, specifically:

[0135] M emergency =G,ifΔO d (t)>ΔT

[0136] Where M emergency It is the oxygenation mode in emergency state. The oxygenation mode in emergency state is gas oxygenation. ΔO d (t) is the oxygen demand prediction error, and ΔT is the error tolerance threshold.

[0137] In the above content, in actual application, the environmental parameter data in the aquaculture area is first obtained, and the environmental parameter data obtained by the sensor is transmitted to the cloud database through the wireless communication network, and the sensor data is calibrated regularly.

[0138] After the transmission is completed, the data is classified into different categories, and grouped and processed based on the data of different categories, and classified according to time, space and data source. After the classification is completed, the data is cleaned, and the cleaned data is subjected to outlier detection and dimensionality reduction processing.

[0139] After the data is reduced in dimension, deep learning is used to build an oxygen demand prediction model based on the reduced data. Feature extraction is performed before modeling. The model is configured based on the extracted features. The configuration uses a standardized LSTM network architecture and the model with the completed architecture is trained. After the training is completed, the model is verified and evaluated. The model is optimized based on the verification and evaluation results, and the results are output based on the optimized model.

[0140] After the prediction results are obtained, the working intensity of the oxygen enrichment equipment is automatically adjusted based on the intensity of the water body's oxygen demand and the current water body environmental parameters.

[0141] Please refer to Figure 2 The present application further provides an automatic adjustment system for aquaculture oxygenation equipment based on environmental parameters, including the following:

[0142] The parameter acquisition unit 11 is used to acquire environmental parameter data in the aquaculture area.

[0143] The preprocessing classification unit 12 is used to classify and preprocess the acquired environmental parameter data.

[0144] The model configuration unit 13 is used to construct an oxygen demand prediction model using a deep learning algorithm, and to train and optimize the oxygen demand prediction model based on the pre-processed environmental parameter data.

[0145] The prediction unit 14 is used to predict the oxygen demand in the water body within a future period of time based on the obtained optimized oxygen demand prediction model.

[0146] The mode switching unit 15 is used to automatically select and switch the oxygen enrichment mode based on the prediction result obtained by the oxygen demand prediction model, and adjust the working intensity of the oxygen enrichment equipment.

[0147] It should be understood that the "one embodiment" or "an embodiment" or "a feasible implementation" or "some implementations" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the embodiments of the present invention. Therefore, "in one embodiment" or "in an embodiment" or "a feasible implementation" or "some implementations" appearing in various places throughout the specification do not necessarily refer to the same embodiment. In addition, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the steps and modules involved are not necessarily required for the embodiments of the present invention.

[0148] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by instructions. These instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0149] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0150] These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0151] The above is only a specific implementation of the present invention, but the protection scope of the embodiments of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed in the embodiments of the present invention should be included in the protection scope of the embodiments of the present invention. Therefore, the protection scope of the embodiments of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for automatically adjusting aquaculture oxygenation equipment based on environmental parameters, characterized in that: These include: S100, obtaining environmental parameter data in the aquaculture area; S200, performing data classification and preprocessing based on the environmental parameter data acquired in S100; S300, using a deep learning algorithm to build an oxygen demand prediction model, and training and optimizing the oxygen demand prediction model based on pre-processed environmental parameter data; S400, based on the optimized oxygen demand prediction model obtained in S300, predicting the oxygen demand in the water body in the future period of time; S500: Based on the prediction result obtained in S400, the oxygen enrichment mode is automatically selected and switched, and the working intensity of the oxygen enrichment equipment is adjusted.

2. The automatic adjustment method of aquaculture oxygenation equipment based on environmental parameters according to claim 1, characterized in that: Obtain environmental parameter data within the aquaculture area, including: S101, acquiring environmental parameter data, deploying environmental data sensors to acquire environmental parameter data of the aquaculture area, wherein the environmental parameter data include water temperature, dissolved oxygen concentration, pH value, ammonia nitrogen concentration, water flow velocity, atmospheric pressure and air temperature; S102, environmental parameter data transmission, transmitting the environmental parameter data acquired by the environmental data sensor to the cloud database via the wireless communication network.

3. The automatic adjustment method of aquaculture oxygenation equipment based on environmental parameters according to claim 1, characterized in that: Data classification and preprocessing are performed based on the environmental parameter data obtained by S100, including: The preprocessing includes data cleaning, outlier detection and dimensionality reduction; S201, data classification, classifying the environmental parameter data obtained in S100 into different categories, and performing group processing based on the data of different categories; S202, data cleaning, performing data cleaning based on the grouped environmental parameter data, wherein the data cleaning includes filling in missing values ​​and removing erroneous values; S203, outlier detection, using an autoencoder to perform outlier detection on the environmental parameter data after data cleaning; S204, dimensionality reduction processing, using principal component analysis method, to perform dimensionality reduction processing on the environmental parameter data after removing abnormal data, and obtain pre-processed environmental parameter data.

4. The automatic adjustment method of aquaculture oxygenation equipment based on environmental parameters according to claim 1, characterized in that: A deep learning algorithm is used to build an oxygen demand prediction model, and the oxygen demand prediction model is trained and optimized based on pre-processed environmental parameter data, including: S301, configuring an oxygen demand prediction model, using a standardized LSTM network architecture, wherein the standardized LSTM network architecture includes an input layer, an LSTM hidden layer, a fully connected layer, and an output layer; The configuring of the oxygen demand prediction model further includes configuring an LSTM hidden layer, setting a time step, setting a loss function and an activation function; Configure LSTM hidden layers, the number is limited to 3 layers, and a single layer contains 64 neurons; Set the time step to the past 24 hours; Set the loss function and use the mean square error as the loss function, specifically: In the formula, y i is the actual oxygen demand value, is the model prediction value, n is the number of data points; S302, training an oxygen demand prediction model, using LSTM input features as input training data, using historical oxygen demand values ​​as output training data, inputting LSTM input features into the oxygen demand prediction model, obtaining a preliminary prediction result, and then comparing the prediction result of the model with the actual historical oxygen demand value to minimize the error; S303, verification and evaluation, the verification and evaluation includes selecting evaluation indicators and cross-validation; S304. Optimize the oxygen demand prediction model and use incremental learning. Based on the acquisition of new data, the LSTM model is continuously optimized through incremental learning. Each time new data arrives, a small-scale iterative update is performed based on the parameters of the original model, rather than completely retraining the oxygen demand prediction model.

5. The automatic adjustment method of aquaculture oxygenation equipment based on environmental parameters according to claim 1, characterized in that: Based on the optimized oxygen demand prediction model obtained in S300, the oxygen demand in the water body in the future period is predicted, specifically: Based on the optimized model, the preprocessed environmental parameter data is used as input features and input into the optimized LSTM model. The LSTM model will output the oxygen demand value in the water body in the future, specifically: Among them, O d (t) represents the oxygen demand in the water body in the future period of time, f LSTM Represents the prediction function through the LSTM model, Represents the preprocessed environmental parameter data, represents the environmental parameter data vector at time t, n represents the time step, and h(t) represents the hidden state of the LSTM model at time t.

6. The automatic adjustment method of aquaculture oxygenation equipment based on environmental parameters according to claim 1, characterized in that: Based on the prediction results obtained by S400, the oxygen enrichment mode is automatically selected and switched to adjust the working intensity of the oxygen enrichment equipment, including: S501, based on the oxygen demand in the water body in the future d (t) and the dissolved oxygen concentration C(t) of the water body to select the oxygenation mode, specifically: Where M selected is the selected oxygenation mode, O d (t) represents the oxygen demand in the water body in the future period of time, T H and T L They represent the oxygen thresholds of high demand and low demand, respectively, C(t) represents the dissolved oxygen concentration of the water body, and C L and C H They represent the lower and upper thresholds of dissolved oxygen concentration in water, θ(t) represents the current water temperature, and θ optimal Indicates the preset water temperature, Δθ indicates the allowable range of water temperature deviation, G indicates gas oxygenation, E indicates electrolytic oxygenation, and B indicates microbubble oxygenation; S502: Activate the corresponding oxygenation equipment based on the oxygenation mode selected in S501, and based on the oxygen demand in the water body in the future period of time, d (t) and the dissolved oxygen concentration C(t) of the water body automatically adjust the working intensity of the oxygenation equipment, specifically: P(t)=α·(O d (t)-C(t))+β Where P(t) represents the working intensity of the oxygenation equipment at time t, O d (t) represents the oxygen demand in the water body in the future period of time, C(t) represents the dissolved oxygen concentration in the water body, O d (t)-C(t) represents the difference between the oxygen demand in the water body and the dissolved oxygen concentration in the water body in the future. If the dissolved oxygen concentration in the water body is lower than the demand, that is, O d (t)>C(t), then the working intensity of the equipment is increased. If the dissolved oxygen concentration of the water body is close to the demand, then the working intensity is reduced. α represents the adjustment coefficient, which is used to control the response sensitivity of the working intensity to the demand and concentration difference. β represents the basic working intensity, that is, the minimum working intensity of the equipment. S503, emergency control, to ensure the continuity of oxygen supply, set up emergency mode, when equipment failure and oxygen demand forecast has large errors, start backup equipment or select oxygen enrichment mode under emergency state to ensure uninterrupted oxygen supply in the water body.

7. The automatic adjustment method of aquaculture oxygenation equipment based on environmental parameters according to claim 6, characterized in that: Emergency control: To ensure the continuity of oxygen supply, an emergency mode is set. When equipment fails or there is a large error in the oxygen demand forecast, the backup equipment is started or the oxygen enrichment mode under the emergency state is selected to ensure that the oxygen supply in the water body is not interrupted. Specifically: When a failure of the oxygen enrichment equipment is detected or the prediction error exceeds the set threshold ΔT, the backup equipment is immediately started and automatically switched to the oxygen enrichment mode in the emergency state, specifically: M emergency =G,ifΔO d >ΔT Where M emergency It is the oxygenation mode in emergency state. The oxygenation mode in emergency state is gas oxygenation. ΔO d (t) is the oxygen demand prediction error, and ΔT is the error tolerance setting threshold.

8. Automatic adjustment system for aquaculture oxygenation equipment based on environmental parameters, characterized in that: These include: A parameter acquisition unit, used to acquire environmental parameter data in the aquaculture area; A preprocessing classification unit, used for classifying and preprocessing the acquired environmental parameter data; A model configuration unit, used to construct an oxygen demand prediction model using a deep learning algorithm, and to train and optimize the oxygen demand prediction model based on pre-processed environmental parameter data; A prediction unit, used for predicting the oxygen demand in the water body within a future period of time based on the obtained optimized oxygen demand prediction model; The mode switching unit is used to automatically select and switch the oxygen enrichment mode and adjust the working intensity of the oxygen enrichment equipment based on the prediction results obtained by the oxygen demand prediction model.

Images