Construction method of water bloom early warning model, and water bloom early warning method and device

Abstract

The invention discloses a construction method of a water bloom early warning model, a water bloom early warning method and a water bloom early warning device, and the construction method of the water bloom early warning model comprises the steps: obtaining a sample data set, the sample data set comprises mass spectrum characteristic data and first index data of volatile organic compounds released in the growth process of at least one kind of algae in at least one growth environment, and the first index data is algae density data, or the first index data can be used for calculating the algae density data; constructing a machine learning model by taking the mass spectrum characteristic data as input characteristic data and the first index data as an output target; and training and checking the machine learning model by using the sample data set to obtain a water bloom early warning model.

Description

Technical Field

[0001] This disclosure relates to, but is not limited to, the field of water environment protection technology, and in particular to a method for constructing an algal bloom early warning model, an algal bloom early warning method, and an apparatus. Background Technology

[0002] Currently, the combined effects of global climate change and continuously increasing nitrogen and phosphorus nutrient loads have significantly expanded the frequency, intensity, and geographical range of harmful algal blooms (HABs) in lakes, reservoirs, and drinking water sources. Phytoplankton such as cyanobacteria can proliferate rapidly in a short period, causing not only oxygen depletion in water bodies, filter clogging, and fish and shrimp mortality, but also releasing secondary metabolites such as microcystins and anabatin, severely damaging aquatic ecosystems and directly threatening human health.

[0003] Therefore, it is necessary to monitor algal blooms in water bodies such as rivers and lakes, and also to predict the development of algal blooms in order to achieve the effect of early warning. Summary of the Invention

[0004] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0005] This disclosure provides a method for constructing an algal bloom early warning model, comprising: acquiring a sample dataset, the sample dataset including mass spectrometry feature data of volatile organic compounds released by at least one algae during its growth in at least one growth environment and first index data, wherein the first index data is algal density data, or the first index data can be used to calculate the algal density data; constructing a machine learning model using the mass spectrometry feature data as input feature data and the first index data as output target; and training and testing the machine learning model using the sample dataset to obtain an algal bloom early warning model.

[0006] This disclosure also provides an apparatus for constructing an algal bloom early warning model, including a memory; and a processor connected to the memory, the memory being used to store instructions, the processor being configured to execute the steps of the algal bloom early warning model construction method as described in any embodiment of this disclosure based on the instructions stored in the memory.

[0007] This disclosure also provides a method for early warning of algal blooms, comprising: acquiring mass spectrometry characteristic data of volatile organic compounds released by algae in a target water body; inputting the mass spectrometry characteristic data of volatile organic compounds released by algae in the target water body into a pre-constructed algal bloom early warning model; and obtaining the algal density and / or algal bloom risk probability of the target water body based on the output of the algal bloom early warning model, wherein the algal bloom early warning model is constructed based on the construction method of the algal bloom early warning model as described in any embodiment of this disclosure.

[0008] This disclosure also provides an algal bloom early warning device, including a memory; and a processor connected to the memory, the memory being used to store instructions, and the processor being configured to execute the steps of the algal bloom early warning method according to any embodiment of this disclosure based on the instructions stored in the memory.

[0009] This disclosure also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method for constructing an algal bloom early warning model as described in any embodiment of this disclosure, or implements the algal bloom early warning method as described in any embodiment of this disclosure.

[0010] This disclosure also provides a program product including instructions, which, when executed by a computer, perform either the method for constructing an algal bloom warning model as described in any embodiment of this disclosure, or the method for issuing an algal bloom warning as described in any embodiment of this disclosure.

[0011] The method, apparatus, and device for constructing an algal bloom early warning model according to the present disclosure construct a machine learning model by using mass spectrometry feature data of volatile organic compounds released during algal growth as input feature data. The algal bloom early warning model can identify the algal density and / or algal bloom risk probability of a target water body. It has the advantages of high frequency, non-destructive, rapid, and sensitive characteristics, thus providing a fast, sensitive, and convenient method for monitoring and early warning of algal blooms. It can be applied to the field of algal bloom monitoring and control in freshwater lakes, reservoirs, and nearshore waters.

[0012] Other features and advantages of this disclosure will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the disclosure. Other advantages of this disclosure may be realized and obtained by means of the methods described in the description and the accompanying drawings. Attached Figure Description

[0013] The accompanying drawings are used to provide an understanding of the technical solutions of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the technical solutions of this disclosure and do not constitute a limitation on the technical solutions of this disclosure.

[0014] Figure 1A flowchart illustrating a method for constructing an algal bloom early warning model, provided as an exemplary embodiment of this disclosure.

[0015] Figure 2 A schematic diagram of a standard curve of Chlorella cell density and OD680 provided for an exemplary embodiment of this disclosure.

[0016] Figure 3 A schematic diagram of the growth curves of Chlorella over 27 days under different culture conditions provided for exemplary embodiments of this disclosure.

[0017] Figures 4A to 4G A schematic diagram illustrating the content of volatile organic compounds (AVOCs) in Chlorella during 27 days of growth under different growth conditions, as provided in an exemplary embodiment of this disclosure.

[0018] Figure 5 This is a schematic diagram illustrating the data set partitioning process for constructing an algal bloom early warning model, provided as an exemplary embodiment of this disclosure.

[0019] Figure 6 A schematic diagram illustrating the performance of an algal bloom early warning model for predicting algal density, provided as an exemplary embodiment of this disclosure.

[0020] Figure 7 A flowchart illustrating an algal bloom early warning method provided as an exemplary embodiment of this disclosure.

[0021] Figure 8 A schematic diagram of the predicted algae concentration probability distribution for a target water area, provided as an exemplary embodiment of this disclosure.

[0022] Figure 9 This is a schematic diagram of the entire process of early warning for algal blooms, provided as an exemplary embodiment of this disclosure.

[0023] Figure 10 A schematic diagram of a device for constructing an algal bloom early warning model provided as an exemplary embodiment of this disclosure.

[0024] Figure 11 This is a schematic diagram of the structure of an algal bloom early warning device provided as an exemplary embodiment of the present disclosure. Detailed Implementation

[0025] This disclosure describes several embodiments, but these descriptions are exemplary and not limiting, and it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically limited, any feature or element of any embodiment may be used in combination with, or may replace, any feature or element of any other embodiment.

[0026] This disclosure includes and contemplates combinations of features and elements known to those skilled in the art. The embodiments, features, and elements disclosed in this disclosure may also be combined with any conventional features or elements to form a unique inventive scheme as defined by the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive schemes to form another unique inventive scheme as defined by the claims. Therefore, it should be understood that any feature shown and / or discussed in this disclosure may be implemented individually or in any suitable combination. Therefore, the embodiments are not limited except by the limitations imposed by the appended claims and their equivalents. Furthermore, various modifications and changes may be made within the scope of the appended claims.

[0027] Furthermore, in describing representative embodiments, the specification may have presented methods and / or processes as a specific sequence of steps. However, the method or process should not be limited to the specific order of steps described herein, to the extent that the method or process does not depend on the specific order of steps described herein. As will be understood by those skilled in the art, other sequences of steps are also possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation of the claims. Moreover, the claims relating to the method and / or process should not be limited to the steps performed in the order written, and those skilled in the art will readily understand that these orders can be varied and still remain within the spirit and scope of the embodiments disclosed herein.

[0028] Algal blooms are a natural phenomenon caused by the proliferation and accumulation of algae in freshwater reservoirs. Algal blooms have become a global problem related to water pollution, and their severity is increasing with rapid economic development and human expansion.

[0029] To prevent and control algal blooms in advance, three main types of early warning technologies have been developed both domestically and internationally: First, the model-based method based on water quality and environmental factors. This method uses multi-parameter buoys or manual sampling to obtain data on water temperature, pH, dissolved oxygen, total nitrogen, total phosphorus, flow velocity, and chlorophyll a. Combined with machine learning, multivariate statistical, or ecodynamic models, it establishes a quantitative relationship between algal biomass and environmental factors to predict the probability of algal blooms. Second, the satellite / airborne remote sensing method uses spectral sensors to scan the water surface for color, retrieves chlorophyll a concentration, and estimates the area and location of algal blooms, enabling dynamic monitoring at a regional scale. Third, the molecular biology method uses fluorescence in situ hybridization, PCR / qPCR, enzyme-linked immunosorbent assay (ELISA), and Raman spectroscopy to detect key algal species, virulence genes, or intracellular toxins with high sensitivity and specificity, providing molecular evidence for early identification of algal blooms. These technologies have made progress in terms of accuracy, spatial coverage, or biological specificity, and have become important tools for current water environment management.

[0030] Despite significant progress in the accuracy, coverage, and sensitivity of related algal bloom early warning technologies, certain limitations remain. For example, while water quality and environmental factor-based early warning technologies can comprehensively analyze multiple physicochemical indicators, the direct correlation between these indicators and algal status is relatively weak, making it difficult to fully reflect algal physiological dynamics. Satellite remote sensing technology, although capable of covering large water bodies, is susceptible to weather conditions and the optical complexity of water bodies, and its monitoring effect is poor in water bodies with low chlorophyll concentrations. Molecular biology techniques, despite their high sensitivity and specificity, are expensive, complex to operate, and dependent on laboratory environments, limiting their feasibility for large-scale field applications. These limitations indicate that related technologies still fall short in terms of real-time performance, direct biological relevance, and ease of use in algal bloom early warning, urgently requiring the development of a rapid, sensitive, and convenient method for algal bloom early warning.

[0031] like Figure 1 As shown in the embodiments of this disclosure, a method for constructing an algal bloom early warning model is provided, including: Step 101: Obtain a sample dataset, which includes mass spectrometry feature data and first index data of volatile organic compounds (AVOCs) released by at least one algae during its growth in at least one growth environment. The first index data is algal density data, or the first index data can be used to calculate algal density data. Step 102: Using mass spectrometry feature data as input feature data and the first index data as output target, construct a machine learning model; Step 103: Use the sample dataset to train and test the machine learning model to obtain the algal bloom early warning model.

[0032] The method for constructing an algal bloom early warning model provided in this disclosure constructs a machine learning model by using mass spectrometry feature data of volatile organic compounds released during algal growth as input feature data. This algal bloom early warning model can identify the probability of algal bloom occurrence and predict the development trend of algal bloom. It has advantages such as high frequency, non-destructive, rapid and sensitive characteristics, thus providing a fast, sensitive and convenient method for monitoring and early warning of algal blooms. It can be applied to the field of algal bloom monitoring and control in freshwater lakes, reservoirs and nearshore waters.

[0033] This disclosure uses volatile organic compounds (AVOCs) released during algal growth as the core early warning indicator, replacing traditional indicators such as algal biomass and chlorophyll a. It directly correlates AVOCs with the probability of algal blooms, achieving early warning based on algal metabolic signals. As secondary metabolites constantly released during algal growth, AVOCs can reflect algal growth status signals more promptly and comprehensively than other algal growth status indicators. Online monitoring of AVOCs can be achieved using proton charge transfer time-of-flight mass spectrometry. This disclosure uses AVOCs as the core indicator for model training, identifying the probability of algal blooms and predicting future trends. It fully integrates intrinsic algal growth status indicators, achieving a faster and better true early warning of algal blooms compared to previous algal bloom early warning schemes.

[0034] In some exemplary embodiments, when the first indicator data can be used to calculate algal density data, the first indicator is any one of the following: chlorophyll a content, absorbance at a preset wavelength. However, this disclosure is not limited thereto. The first indicator data can also be set to other indicator data that can be used to calculate algal density data as needed.

[0035] In some exemplary embodiments, the preset wavelength can be 680 nm; however, this disclosure does not limit this. The following description uses absorbance (OD680) with a first index data of 680 nm as an example.

[0036] In some exemplary embodiments, the method further includes, prior to: The relationship curve between algal density and the first index is pre-constructed using the following method: multiple preset algal solutions of different concentrations are prepared; the number of algal cells in each preset algal solution and the value of the first index of each preset algal solution are measured; based on the number of algal cells in each preset algal solution and the value of the first index corresponding to each preset algal solution, the relationship curve between algal density and the first index of the preset algal solution is fitted.

[0037] In this embodiment of the disclosure, the preset algae can be set as needed. For example, the preset algae can be Chlorella vulgaris; however, this disclosure does not limit this. The following description uses Chlorella vulgaris as the preset algae.

[0038] For example, taking OD680 as the first indicator and Chlorella as the preset algae, the method for establishing the relationship curve between Chlorella density and OD680 is as follows: Using BG-11 medium as the background solution, a series of Chlorella solutions with concentration gradients were prepared (OD680 values ​​ranging from 0.04 to 0.27); flow cytometry was used to determine the accurate number of algal cells in each concentration solution, and the algal cell count was plotted on the ordinate and the OD680 value on the abscissa, as shown in the figure. Figure 2 The curve showing the relationship between Chlorella density and OD680 is shown.

[0039] like Figure 2 As shown, the relationship curve has a good fit, and the coefficient of determination R0 is [value missing]. 2 The value reaches 0.998, and the algal density can be directly calculated from the sample OD680 value.

[0040] In some exemplary implementations, the sample dataset can be constructed using the following methods: Pre-defined algae were cultured under multiple different combinations of initial experimental parameters, including TN:TP ratio, temperature, dissolved oxygen, and pH. Within a preset cultivation period, the first indicator data of the preset algal growth process are sampled, and the volatile organic compounds released during the preset algal growth process are collected; the collected volatile organic compounds are measured by mass spectrometry to obtain mass spectrometry data; the obtained mass spectrometry data are preprocessed to obtain mass spectrometry characteristic data.

[0041] For example, taking Chlorella as the first indicator and Chlorella as the preset algae, Chlorella is inoculated into BG-11 medium with an initial density of 8 × 10⁻⁶. 5 cell / mL; after inoculation, the culture medium was placed in a light incubator and cultured under the following conditions: 16 hours of light (light intensity 3000 Lux) / 8 hours of darkness, temperature 25°C, and the culture medium was shaken 3 times a day to prevent algal cells from settling.

[0042] To simulate the effects of different ecological conditions in natural water bodies on the growth of *Chlorella vulgaris*, multiple combinations of initial experimental parameters were set up, including: TN:TP ratio (35:7, 105:7, 247:7 mg / L), temperature (20 or 25°C), dissolved oxygen concentration (6, 8, or 17 mg / L), and pH value (7.5 or 8.2). The experimental conditions for each group are shown in Table 1, where *Algal species* refers to *Chlorella vulgaris*, *Initial experimental conditions* refers to the initial culture parameters, and *Group* is the group number. During the 27-day culture period, samples were taken 1 to 2 times daily, and the absorbance (OD680) of the algal solution at a wavelength of 680 nm was measured as follows: Figure 3 As shown, combined with Figure 2 The relationship curve between Chlorella density and OD680 can be used to calculate the algal density data of the sample using the OD680 value.

[0043]

[0044] Table 1 This embodiment of the disclosure simulates natural aquatic ecological conditions, setting multiple sets of parameter combinations such as TN:TP ratio (35:7, 105:7, 247:7 mg / L), temperature (20 / 25℃), dissolved oxygen (6 / 8 / 17 mg / L), and pH value (7.5 / 8.2), and collects AVOCs and algal density data during a 27-day cultivation period to ensure the comprehensiveness and universality of the dataset.

[0045] In some exemplary embodiments, the mass spectrometry feature data includes signal intensity at multiple mass-to-charge ratio positions within a preset mass-to-charge ratio range.

[0046] For example, the preset mass-to-charge ratio range can be between 15 and 249; however, this disclosure does not limit it. The preset mass-to-charge ratio range can be set as needed.

[0047] In this embodiment of the disclosure, the multiple mass-to-charge ratio positions can be any integer mass-to-charge ratio position within the range of 15 to 249; however, this disclosure does not limit this. For example, in other examples, the multiple mass-to-charge ratio positions can be multiple mass-to-charge ratio positions spaced apart by a preset step size within a preset mass-to-charge ratio range, wherein the preset step size can be any real number such as 1 or 2. When the initial mass-to-charge ratio position is 15, the ending mass-to-charge ratio position is 249, and the preset step size is 1, the multiple mass-to-charge ratio positions are all integer mass-to-charge ratio positions within the range of 15 to 249.

[0048] The AVOCs mass spectrometry feature data disclosed in this embodiment include feature numbers across the entire mass spectrometry range. These feature combinations fully reflect the growth status of algae, improving the stability and accuracy of algal bloom prediction methods.

[0049] For example, during the cultivation of Chlorella described above, while measuring the algal density each time, the headspace air of the culture flask can be collected simultaneously to detect the AVOCs of Chlorella. The specific steps are as follows: online measurement is performed using a mass spectrometer with a mass resolution of 2500 m / Δm (full width at half maximum), using H3O + As the reactant ion, the sample introduction system uses a PEEK capillary tube, heated to 70°C to prevent sample condensation. Headspace gas is directly introduced into the mass spectrometer through the injection port. The mass-to-charge ratio (M / Z) range is set from 15 to 249. The operating parameters include: ion source pressure 600V, drift tube voltage 800V, ion source current 3.00mA, RF voltage 150V, and signal accumulation time 20s.

[0050] The raw AVOCs mass spectrometry data of collected Chlorella are preprocessed to obtain mass spectrometry characteristic data. Specifically, preprocessing may include peak identification, integration, and calibration. In this embodiment, a Monte Carlo simulation algorithm can be used to integrate the peak area of ​​the material in regions at preset step sizes within a preset mass-to-charge ratio range to obtain the signal intensity at multiple mass-to-charge ratio positions.

[0051] For example, taking the seven groups of Chlorella (A1 to A7) mentioned above as examples, the changes in mass spectrometry characteristics of AVOCs of Chlorella under different growth conditions during the growth process are as follows: Figures 4A to 4G As shown, the sample dataset was obtained based on the M / Z signal intensity data and algal density data of samples under different ecological conditions.

[0052] In some exemplary implementations, training and testing machine learning models using sample datasets includes: The sample dataset is divided into a training set, a validation set, and a test set; The machine learning model is trained using the training set; the hyperparameters of the machine learning model are tuned using the validation set, where the tuned hyperparameters include: the number of base estimators, the learning rate, and the maximum depth, and the optimal parameter model is saved. Use the test set to evaluate the performance of the optimal parameter model.

[0053] In this embodiment of the disclosure, the machine learning model can be an XGBoost regression model, or other regression models, such as k-Nearest Neighbor (KNN), Support Vector Machine (SVM), etc., and this disclosure does not impose any limitations on it. The following description uses the XGBoost regression model as an example.

[0054] The machine learning model in this disclosure uses mass spectrometry feature data of AVOCs as input feature data and algal density data or other first indicator data that can represent algal density data (such as OD680 value or chlorophyll a content) as output target. For example, Figure 5 As shown, the dataset is randomly divided into a training set (80%) and a temporary set (20%) in an 8:2 ratio. The temporary set is further divided into a validation set (5%) and a test set (15%) in a 1:3 ratio. The specific division ratio of the dataset can be adjusted as needed, and this disclosure does not impose any restrictions on it.

[0055] In this embodiment, grid search and 5-fold cross-validation were used to systematically tune the hyperparameters of the XGBoost regressor. During the optimization of XGBoost, three hyperparameters were adjusted: the number of base estimators (n_estimators), the learning rate, and the maximum depth (max_depth). The range of n_estimators was set from 30 to 300 in increments of 10. For the learning rate hyperparameter, a specific set of discrete values ​​was investigated: 0.1, 0.15, 0.2, 0.25, and 0.3. The max_depth hyperparameter was also adjusted to values ​​of 3, 4, 5, 6, 7, and 8. This method ensures comprehensive coverage of parameter interactions crucial for algal bloom prediction while maintaining computational efficiency. The 5-fold validation strategy effectively reduces the risk of overfitting the dataset by maximizing the R-value of cross-validation. 2 The optimal parameters are selected based on the scores.

[0056] In one example, after adjusting the validation set, the optimal model parameters were determined to be: max_depth=5, learning_rate=0.1, n_estimators=300. The optimal parameter model was saved. The performance of this optimal parameter model was evaluated using the test set, and the results are as follows: Figure 6 As shown, the coefficient of determination R 2 The value is as high as 0.95, where N represents the number of samples.

[0057] In some exemplary embodiments, training a machine learning model using a training set includes: Multiple self-sampling subsets are created from the training set using sampling with replacement; A separate machine learning model is trained on each self-service resampling subset to obtain multiple algal bloom warning models. The outputs of these multiple algal bloom warning models are used to generate algal bloom risk probabilities.

[0058] In this embodiment, the uncertainty of bootstrap resampling is used to generate the probability of algal bloom risk. For example, when training the XGBoost regression model, a total of 100 bootstrap resampling subsets are generated from the training set (the number of bootstrap resampling subsets can be adjusted as needed, and this disclosure does not limit this), and a separate XGBoost model is trained on each bootstrap resampling subset. The prediction results on the test set are summarized to estimate the uncertainty of the model, and the coefficient of determination (R²) is used. 2 The model performance is evaluated using the mean squared error (MSE) and the final trained model for future use.

[0059] In this embodiment of the disclosure, the method for generating the algal bloom risk probability is as follows: A pre-set algal bloom risk threshold (e.g., it can be set to 1×10⁻⁶) is used. 6 (cell / mL or other arbitrary values), multiple algal bloom early warning models based on self-sampling are used to output multiple predicted algal densities of target water bodies. The ratio of the number of algal densities exceeding the algal bloom risk threshold to the total number of algal densities output is used as the algal bloom risk probability output.

[0060] The embodiments disclosed herein, combined with online algae monitoring equipment (such as a proton charge transfer time-of-flight mass spectrometer), can be flexibly applied to the field of algal bloom monitoring and have excellent practical prospects.

[0061] like Figure 7 As shown in the embodiments of this disclosure, an algal bloom early warning method is also provided, including: Step 701: Obtain mass spectrometry characteristic data of volatile organic compounds released by algae in the target water body; Step 702: Input the mass spectrometry characteristic data of volatile organic compounds released by algae in the target water body into the pre-constructed algal bloom early warning model, and obtain the algal density and / or algal bloom risk probability of the target water body based on the output of the algal bloom early warning model. The algal bloom early warning model is constructed using the construction method of algal bloom early warning model as described in any embodiment of this disclosure.

[0062] This embodiment utilizes the aforementioned constructed algal bloom early warning model to provide early warning of algal blooms in target water bodies. Based on the algal density and / or algal bloom risk probability of the target water body, an algal bloom early warning signal can be output. In this embodiment, specific monitoring output indicators may include: predicted algal density and / or algal bloom risk probability.

[0063] In some exemplary embodiments, the mass spectrometry characteristics of volatile organic compounds released by algae in the target water body are obtained by the following methods: Collect volatile organic compounds released by algae in the target water body; Mass spectrometry was used to measure the volatile organic compounds released by algae in the target water body to obtain mass spectrometry data of the volatile organic compounds released by algae in the target water body. The mass spectrometry data of volatile organic compounds released by algae in the target water body are preprocessed to obtain the mass spectrometry characteristic data of volatile organic compounds released by algae in the target water body.

[0064] For example, during field sampling, a dynamic enclosure method can be used, employing polytetrafluoroethylene (PTFE) bags to collect volatile organic compounds (AVOCs). The enclosure can be set up as follows: a cubic enclosure (e.g., 35 × 35 × 15 cm) is constructed using a stainless steel frame and covered with a PTFE membrane. Each enclosure is equipped with an inlet, an outlet, and a small fan to mix the air within. The enclosure is placed on a PTFE-lined ring mounted on a polyurethane foam board floating on the water surface. All photoactive substances are excluded from the enclosure area. During sampling, after sealing the sampling device for 20 minutes, 300 mL of gas is collected at the outlet and stored in a PTFE gas bag.

[0065] The gas collected from the target water body is analyzed using a mass spectrometer. After preprocessing, the mass spectrometry data of the target water body is used to obtain the mass spectrometry characteristic data of AVOCs. The mass spectrometry characteristic data of AVOCs in the target water body is then input into a pre-constructed algal bloom early warning model to obtain the algal density and / or algal bloom risk probability of the target water body.

[0066] We collected gas samples from four sampling points in a natural lake in Beijing, with three parallel samples collected from each sampling point. The probability of algal bloom risk in the water samples was obtained using a pre-built algal bloom early warning model, as shown below. Figure 8 As shown, Sample ij represents the prediction result of the j-th sample at the i-th sampling point, where i is between 1 and 4, and j is between 1 and 3. The red vertical line represents the pre-set algal bloom risk threshold. The horizontal axis represents the predicted algal concentration, and the vertical axis represents the number of times the predicted result occurs. Since this embodiment uses 100 self-sampling subsets as an example, the sum of all occurrences is 100. Based on this prediction result, a corresponding probability density distribution (the thin blue line in the figure) is generated. Figure 8 It can be determined that the actual algal density of the lake is between 30,928 and 266,941 cells per milliliter. -1Completely within the alert zone (10,000 to 1,000,000 cells / mL) -1 The model predicts an algal bloom risk probability of 67% to 81%. This indicates that the risk assessment in the actual samples is highly consistent with the actual situation, thus confirming the reliability of the algal bloom early warning based on AVOCs proposed in this disclosure.

[0067] In one example, such as Figure 9 As shown, the algal bloom early warning method of this disclosure is used to predict the probability of algal bloom risk, including the following steps: 1) Data collection: Collect AVOCs and algal biomass index data of algae growth process in typical freshwater lakes under different growth environments; 2) Construct a dataset based on the AVOCs and algal biomass index data under different growth environments; 3) Use the dataset to train and test a machine learning model (XGBoost) to obtain an algal bloom early warning model that is accurate, robust and universal; 4) Analyze the volatile organic compounds of algae in the water body to be tested, use the algal bloom early warning model to predict algal biomass, and obtain the probability of algal bloom risk.

[0068] The algal bloom early warning method of this disclosure uses mass spectrometry data of volatile organic compounds released by algae under different growth environments to train an algal bloom early warning model, and uses the mass spectrometry data of volatile organic compounds released by algae in the target water body to predict the algal biomass and algal growth status of the target water body, identify the probability of algal bloom occurrence and predict future trends, and has the advantages of high frequency, non-destructive, rapid and sensitive.

[0069] This disclosure also provides an apparatus for constructing an algal bloom early warning model, including a memory; and a processor connected to the memory, the memory being used to store instructions, the processor being configured to execute the steps of the algal bloom early warning model construction method as described in any embodiment of this disclosure based on the instructions stored in the memory.

[0070] like Figure 10As shown, in one example, the apparatus for constructing an algal bloom early warning model may include: a first processor 1010, a first memory 1020, a first bus system 1030, and a first transceiver 1040. The first processor 1010, first memory 1020, and first transceiver 1040 are connected via the first bus system 1030. The first memory 1020 stores instructions, and the first processor 1010 executes the instructions stored in the first memory 1020 to control the first transceiver 1040 to transmit and receive signals. Specifically, the first transceiver 1040, under the control of the first processor 1010, can acquire a sample dataset. The sample dataset includes mass spectrometry feature data and first index data of volatile organic compounds released by at least one algae during its growth under at least one growth environment. The first index data is algal density data or data that can be used to calculate the algal density. The first processor 1010 uses the mass spectrometry feature data as input feature data and the first index data as the output target to construct a machine learning model. The machine learning model is trained and tested using the sample dataset to obtain the algal bloom early warning model.

[0071] It should be understood that the first processor 1010 can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0072] The first memory 1020 may include read-only memory and random access memory, and provides instructions and data to the first processor 1010. A portion of the first memory 1020 may also include non-volatile random access memory. For example, the first memory 1020 may also store device type information.

[0073] The first bus system 1030 may include a power bus, control bus, and status signal bus, in addition to a data bus. However, for clarity, in... Figure 10 The general designated all buses as the first bus system 1030.

[0074] In implementation, the processing performed by the processing device can be accomplished through integrated logic circuits in the hardware of the first processor 1010 or through software instructions. That is, the method steps of this embodiment can be executed by the hardware processor, or by a combination of hardware and software modules within the processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other storage media. This storage medium is located in the first memory 1020. The first processor 1010 reads information from the first memory 1020 and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, further details are omitted here.

[0075] This disclosure also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for constructing an algal bloom warning model as described in any embodiment of this disclosure. The method for constructing an algal bloom warning model by executing executable instructions is essentially the same as the method for constructing an algal bloom warning model provided in the above embodiments of this disclosure, and will not be described in detail here.

[0076] In some possible implementations, various aspects of the method for constructing an algal bloom warning model provided in this disclosure can also be implemented in the form of a program product, which includes program code. When the program product is run on a computer device, the program code is used to cause the computer device to perform the steps in the method for constructing an algal bloom warning model according to various exemplary embodiments of this disclosure as described above. For example, the computer device can execute the method for constructing an algal bloom warning model as described in the embodiments of this disclosure.

[0077] This disclosure also provides an algal bloom early warning device, including a memory; and a processor connected to the memory, the memory being used to store instructions, the processor being configured to execute the steps of the algal bloom early warning method as described in any embodiment of this disclosure based on the instructions stored in the memory.

[0078] like Figure 11As shown, in one example, the algal bloom early warning device may include: a second processor 1110, a second memory 1120, a second bus system 1130, and a second transceiver 1140. The second processor 1110, the second memory 1120, and the second transceiver 1140 are connected via the second bus system 1130. The second memory 1120 stores instructions, and the second processor 1110 executes the instructions stored in the second memory 1120 to control the second transceiver 1140 to transmit and receive signals. Specifically, under the control of the second processor 1110, the second transceiver 1140 can acquire mass spectrometry data of volatile organic compounds released by algae in the target water body. The second processor 1110 inputs the mass spectrometry data of the volatile organic compounds released by algae in the target water body into a pre-constructed algal bloom early warning model. Based on the output of the algal bloom early warning model, the algal density and / or algal bloom risk probability of the target water body are obtained. The algal bloom early warning model is constructed using the construction method of the algal bloom early warning model described in any embodiment of this disclosure.

[0079] It should be understood that the second processor 1110 can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0080] The second memory 1120 may include read-only memory and random access memory, and provides instructions and data to the second processor 1110. A portion of the second memory 1120 may also include non-volatile random access memory. For example, the second memory 1120 may also store device type information.

[0081] In addition to the data bus, the second bus system 1130 may also include a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 11 The various buses were all labeled as the second bus system 1130.

[0082] In implementation, the processing performed by the processing device can be accomplished through integrated logic circuits in the hardware of the second processor 1110 or through software instructions. That is, the method steps of this embodiment can be executed by a hardware processor, or by a combination of hardware and software modules within the processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other storage media. This storage medium is located in the second memory 1120. The second processor 1110 reads information from the second memory 1120 and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, further details are omitted here.

[0083] This disclosure also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the algal bloom warning method as described in any embodiment of this disclosure. The algal bloom warning method driven by executing executable instructions is essentially the same as the algal bloom warning method provided in the above embodiments of this disclosure, and will not be described in detail here.

[0084] In some possible implementations, various aspects of the algal bloom warning method provided in this disclosure can also be implemented as a program product comprising program code that, when run on a computer device, causes the computer device to perform the steps in the algal bloom warning method according to various exemplary embodiments of this disclosure as described above. For example, the computer device can execute the algal bloom warning method described in the embodiments of this disclosure.

[0085] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0086] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0087] It should be noted that the above embodiments or implementation methods are merely exemplary and not restrictive. Therefore, this disclosure is not limited to the content specifically shown and described herein. Various modifications, substitutions, or omissions can be made to the form and details of the implementations without departing from the scope of this disclosure.

Claims

1. A method for constructing a water bloom early warning model, characterized in that, The method comprises the following steps: obtaining a sample data set, wherein the sample data set comprises mass spectrum feature data of volatile organic compounds released by at least one kind of algae in a growth process of the at least one kind of algae in at least one kind of growth environment and first index data, wherein the first index data is algae density data, or the first index data can be used to calculate the algae density data; constructing a machine learning model by taking the mass spectrum feature data as input feature data and taking the first index data as output target; training and testing the machine learning model by using the sample data set to obtain a water bloom early warning model.

2. The method of claim 1, wherein, The mass spectrum feature data comprises signal intensities of a plurality of mass-to-charge ratio positions within a preset mass-to-charge ratio range.

3. The method of claim 1, wherein, When the first index data can be used to calculate the algae density data, the first index is any one of chlorophyll a content and absorbance at a preset wavelength.

4. The method of claim 3, wherein, The method further comprises the following steps before the above steps: previously constructing a relationship curve between the algae density and the first index by the following method: configuring a plurality of preset algae solutions with different concentrations; determining the number of algae cells in each preset algae solution and the value of the first index of each preset algae solution; and fitting a relationship curve between the algae density and the first index of the preset algae solution according to the number of algae cells in each preset algae solution and the value of the first index corresponding to each preset algae solution.

5. The method of claim 1, wherein, The training and testing of the machine learning model by using the sample data set comprises the following steps: dividing the sample data set into a training set, a validation set and a test set; training the machine learning model by using the training set; adjusting hyperparameters of the machine learning model by using the validation set, wherein the adjusted hyperparameters of the machine learning model comprise the number of base estimators, the learning rate and the maximum depth, and the optimal parameter model is saved; evaluating the effect of the optimal parameter model by using the test set.

6. The method of claim 5, wherein, The training of the machine learning model by using the training set comprises the following steps: creating a plurality of bootstrap subsets from the training set by using a sampling method with replacement; training a single machine learning model on each bootstrap subset to obtain a plurality of water bloom early warning models, wherein the output results of the plurality of water bloom early warning models are used to generate a water bloom risk probability.

7. The method of claim 1, wherein, The sample data set is constructed by the following method: culturing preset algae under a plurality of different initial experimental parameter combinations, wherein the experimental parameters comprise the TN:TP ratio, temperature, dissolved oxygen and pH value; sampling the first index data of each group of preset algae in a growth process and collecting volatile organic compounds released by each group of preset algae in the growth process within a preset culture period; performing mass spectrum measurement on the volatile organic compounds to obtain mass spectrum data; and performing preprocessing on the mass spectrum data to obtain the mass spectrum feature data.

8. A method for early warning of water bloom, characterized in that, The method comprises the following steps: obtaining mass spectrum feature data of volatile organic compounds released by algae in a target water body; inputting the mass spectrum characteristic data of the volatile organic compounds released by algae in the target water body into a previously constructed water bloom early warning model, and obtaining the algae density and / or water bloom risk probability of the target water body based on the output of the water bloom early warning model, wherein the water bloom early warning model is constructed based on the construction method of the water bloom early warning model according to any one of claims 1 to 7.

9. The method of alarmin according to claim 8, characterized in that, The mass spectrum characteristic data of the volatile organic compounds released by algae in the target water body is obtained by the following method: collecting the volatile organic compounds released by algae in the target water body; performing mass spectrum measurement on the volatile organic compounds released by algae in the target water body to obtain mass spectrum data of the volatile organic compounds released by algae in the target water body; performing preprocessing on the mass spectrum data of the volatile organic compounds released by algae in the target water body to obtain mass spectrum characteristic data of the volatile organic compounds released by algae in the target water body. 10.A device for constructing a water bloom early warning model, characterized in that, The computer program product comprises instructions for implementing the steps of the construction method of the water bloom early warning model according to any one of claims 1 to 3 and 5 to 6, or the steps of the water bloom early warning method according to claim 8, when the computer program product is executed by a computer.

11. A device for early warning of water bloom, characterized in that, The computer program product comprises instructions for implementing the steps of the construction method of the water bloom early warning model according to any one of claims 1 to 3 and 5 to 6, or the steps of the water bloom early warning method according to claim 8, when the computer program product is executed by a computer.

12. A computer-readable storage medium, characterized in that, The computer program product comprises instructions for implementing the steps of the construction method of the water bloom early warning model according to any one of claims 1 to 3 and 5 to 6, or the steps of the water bloom early warning method according to claim 8, when the computer program product is executed by a computer.

13. A computer program product, characterised in that, The computer program product comprises instructions for implementing the steps of the construction method of the water bloom early warning model according to any one of claims 1 to 3 and 5 to 6, or the steps of the water bloom early warning method according to claim 8, when the computer program product is executed by a computer.

Images