Early warning device and method and deinsectization system

Abstract

The invention provides an early warning device and method and a deinsectization system. The early warning device comprises a data acquisition unit, a time sequence analysis unit, a data analysis unit and an early warning decision unit, the data acquisition unit is used for acquiring first detection data; the time sequence analysis unit is used for generating an insect quantity prediction result according to the first detection data; the data analysis unit is used for generating a factor association result and a risk classification result according to the first detection data; and the early warning decision unit is used for generating early warning information according to the insect quantity prediction result, the factor association result and the risk classification result.

Description

Technical Field

[0001] The present application relates to the technical field of insect control, and in particular to an early warning device, method and insect killing system. Background Art

[0002] At present, insect killing devices (such as solar insect killing lamps) are widely used in agricultural pest control. Solar insect killing lamps usually include solar panels, batteries, lamps, sensors and simple controllers. The controller is mainly used to control the opening or closing of the lamps to capture and kill pests, but it cannot provide timely and accurate warning information. Summary of the invention

[0003] A technical problem to be solved by the present application is to solve the problem that the existing insect killing devices cannot provide timely and accurate early warning information.

[0004] In order to solve the above technical problems, the embodiments of the present application provide an early warning device, method and insect extermination system.

[0005] An embodiment of the present application provides an early warning device, which includes: a data acquisition unit, a time series analysis unit, a data analysis unit and an early warning decision unit; the data acquisition unit is used to obtain first detection data; the time series analysis unit is used to generate an insect quantity prediction result based on the first detection data; the data analysis unit is used to generate a factor association result and a risk classification result based on the first detection data; the early warning decision unit is used to generate early warning information based on the insect quantity prediction result, the factor association result and the risk classification result.

[0006] In some embodiments, the time series analysis unit includes a first prediction module, which is used to obtain the corresponding relationship between different time periods and the change in insect quantity according to the first detection data, and obtain the insect quantity within a first preset time period according to the corresponding relationship.

[0007] In some embodiments, the time series analysis unit also includes a second prediction module, which is used to establish a long short-term memory neural network model based on the first detection data, and obtain the number of insects in a second preset time period based on the long short-term memory neural network model, and the span of the second preset time period is greater than the span of the first preset time period.

[0008] In some embodiments, the data analysis unit includes a first analysis module, which is used to construct a factor association model, and the factor association model is used to output a factor association result based on the first detection data. The factor association result includes an importance score for each feature in the first detection data, and multiple features sorted according to the importance score.

[0009] In some embodiments, the data analysis unit further includes a second analysis module, which is used to construct a risk classification model, and the risk classification model is used to calculate the risk score of each feature in the first detection data, thereby obtaining a risk classification result.

[0010] In some embodiments, the warning decision unit includes a first decision module, a second decision module and a warning generation module, the first decision module is used to calculate a first reference value, and confirm and generate a warning level based on the first reference value and the risk classification result, wherein the first reference value is the ratio of the insect quantity prediction result to the historical average value for the same period; the second decision module is used to determine the duration of the warning level based on the factor association result; the warning generation module is used to generate warning information based on the warning level and the duration of the warning level.

[0011] In some embodiments, the early warning device also includes a prevention and control suggestion generation unit, which stores a prevention and control information correspondence table; the prevention and control suggestion generation unit is used to query the corresponding prevention and control strategy from the prevention and control information correspondence table based on the insect quantity prediction results, factor association results and risk classification results, and generate prevention and control information based on the prevention and control strategy.

[0012] In some embodiments, the early warning device also includes a display unit, which is used to display early warning information and prevention and control information.

[0013] An embodiment of the present application provides an early warning method, which includes: acquiring first detection data; generating an insect quantity prediction result based on the first detection data; generating a factor association result and a risk classification result based on the first detection data; and generating early warning information based on the insect quantity prediction result, the factor association result and the risk classification result.

[0014] An embodiment of the present application provides an insect extermination system, which includes: an insect extermination device and an early warning device, and the insect extermination device is used to send first detection data to the early warning device.

[0015] An embodiment of the present application provides a computer device, including a processor and a memory storing a computer program, and when the processor executes the program, the method provided in the above embodiment is implemented.

[0016] An embodiment of the present application provides a computer-readable storage medium storing a computer program, which performs the steps of any one of the above-mentioned early warning methods when the computer program is executed by a processor.

[0017] Through the above technical scheme, the time series unit predicts the number of insects in the future period according to the first detection data collected by the data collection unit and generates an insect number prediction result. The data analysis unit analyzes the relationship between multiple factors and insect activities according to the first detection data and generates factor association results, and predicts the risk of insect risk outbreak and generates a risk classification result. The early warning decision unit is used to generate early warning information according to the insect number prediction results, factor association results and risk classification results. This can improve the utilization rate of data acquired by each sensor, and at the same time can provide more accurate early warning information. The false alarm rate can be reduced by at least 30%, which has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is a structural schematic diagram of the insect extermination system disclosed in the embodiment of the present application;

[0020] Figure 2 It is a schematic diagram of the structure of the insecticide device disclosed in the embodiment of the present application;

[0021] Figure 3 It is a structural schematic diagram of the early warning device disclosed in the embodiment of the present application;

[0022] Figure 4 is a structural schematic diagram of an early warning device disclosed in yet another embodiment of the present application;

[0023] Figure 5 is a structural schematic diagram of a data preprocessing unit disclosed in an embodiment of the present application;

[0024] Figure 6 is a schematic diagram of the structure of the time series analysis unit disclosed in the embodiment of the present application;

[0025] Figure 7 is a schematic diagram of the structure of the data analysis unit disclosed in the embodiment of the present application;

[0026] Figure 8 It is a structural diagram of the early warning decision unit disclosed in the embodiment of the present application;

[0027] Fig. 9 It is a structural diagram of a warning decision unit disclosed in another embodiment of the present application;

[0028] Fig.10is a structural schematic diagram of an early warning device disclosed in yet another embodiment of the present application;

[0029] Fig.11 is a schematic diagram of the structure of a display unit disclosed in an embodiment of the present application;

[0030] Fig.12 It is a flowchart of the early warning method disclosed in the embodiment of the present application;

[0031] Fig.13 A schematic diagram of the structure of a computer device disclosed in an embodiment of the present application.

[0032] Description of reference numerals:

[0033] 100, early warning device; 110, data acquisition unit; 120, time series analysis unit; 121, first prediction module; 122, second prediction module; 130, data analysis unit; 131, first analysis module; 132, second analysis module; 140, early warning decision unit; 141, first decision module; 142, second decision module; 143, early warning generation module; 144, early warning information storage module; 145, early warning information push module; 146, early warning effect tracking module; 150 , data preprocessing unit; 151, data cleaning module; 152, missing value filling module; 153, data standardization module; 154, feature extraction module; 200, insect killing device; 210, sensor; 201, insect counter; 202, temperature sensor; 203, humidity sensor; 204, light sensor; 205, wind speed sensor; 220, data collector; 300, server; 501, processor; 502, communication interface; 503, memory; 504, communication bus. DETAILED DESCRIPTION

[0034] The following is a further detailed description of the implementation methods of the present application in conjunction with the accompanying drawings and examples. The detailed descriptions of the following examples and the accompanying drawings are used to exemplarily illustrate the principles of the present application, but cannot be used to limit the scope of the present application. The present application can be implemented in many different forms and is not limited to the specific embodiments disclosed herein, but includes all technical solutions that fall within the scope of the claims.

[0035] The present application provides these embodiments to make the present application thorough and complete, and to fully express the scope of the present application to those skilled in the art. It should be noted that unless otherwise specifically stated, the relative arrangement of the parts and steps, the composition of the materials, the numerical expressions and the numerical values ​​set forth in these embodiments should be interpreted as being merely exemplary, and not as limiting.

[0036] In addition, words such as “include” or “comprises” and the like used in the present application mean that the elements before the word include the elements listed after the word, and do not exclude the possibility of also including other elements.

[0037] It should also be noted that in the description of this application, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to the specific circumstances. When a specific device is described as being located between a first device and a second device, there may or may not be an intermediate device between the specific device and the first device or the second device.

[0038] All terms used in this application have the same meaning as those understood by those of ordinary skill in the art to which this application belongs, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries, such as general dictionaries, should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined herein.

[0039] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0040] At present, insecticide devices have been widely used in agricultural production, which can effectively control crop pests, reduce the use of pesticides, and improve crop yields and quality. However, existing insecticide devices have the following problems and shortcomings: low data utilization: only recording the number of insects, not collecting and using environmental parameter data; limited analysis capabilities: lack of predictive analysis of insect activity trends; simple early warning mechanism: unable to provide timely and accurate early warning information; insufficient decision support: difficult to provide farmers with scientific prevention and control suggestions.

[0041] To this end, the present application embodiment provides an insect killing system, such as Figure 1 As shown, the insect extermination system includes an insect extermination device 200 and an early warning device 100 .

[0042] The network is used to provide a medium for a communication link between the pest control device 200 and the early warning device 100. The pest control device 200 and the early warning device 100 can communicate directly or indirectly through the network, and the embodiments of the present application are not limited here. For example, the pest control device 200 can directly send the first detection data to the early warning device 100 through the network. The network can include various connection types, such as wired, wireless communication links or optical fiber cables, etc. The embodiments of the present application are also not limited here.

[0043] like Figure 2 As shown, the insect killing device 200 includes multiple sensors 210 and a data collector 220 connected to the multiple sensors 210, and the data collector 220 is used to generate first detection data according to the data collected by the multiple sensors. For example, the multiple sensors 210 include: an insect counter 201, a temperature sensor 202, a humidity sensor 203, a light sensor 204 and a wind speed sensor 205.

[0044] The insect counter 201 adopts the infrared radiation principle and consists of a transmitter and a receiver. The transmitter emits infrared light with a wavelength of 940nm, and the receiver detects changes in light intensity. When an insect flies over, the light intensity decreases, triggering the count. The counter accumulates once every hour with an accuracy of ±1 insect / hour.

[0045] The temperature sensor 202 can be a PT100 platinum resistor. When the temperature changes, the resistance value of the PT100 platinum resistor also changes. Therefore, it can be connected to the analog input terminal of the data acquisition device 220 through a four-wire system and excited by a constant current source, so as to convert the measured resistance value change into temperature. The measurement range of the temperature sensor is -40℃~80℃, and the accuracy is ±0.5℃.

[0046] The humidity sensor 203 can be a capacitive humidity sensor. The humidity sensitive material in the capacitive humidity sensor changes its dielectric constant when the humidity changes. Therefore, the humidity sensor is connected to the analog input terminal of the data collector 220 to convert the measured capacitance value change into humidity. The measurement range of the humidity sensor is 0-100%RH, and the accuracy is ±3%RH.

[0047] The light sensor 204 can be a silicon photocell, which outputs a 4-20mA current signal, which is converted into a 0.4-2V voltage signal through a 100Ω precision resistor and then input into the data collector 220, thereby obtaining the light intensity of the environment where the insecticide 200 is located. The measurement range of the light sensor is 0-200000 lux, and the accuracy is ±5%.

[0048] The wind speed sensor 205 can be a three-cup structure, and the wind speed sensor outputs a frequency signal, which is connected to the digital input terminal of the data collector 220, and the wind speed is converted after the frequency is measured by the counter. The measurement range of the wind speed sensor is 0-30m / s, and the accuracy is ±0.5m / s.

[0049] The data collector 220 is provided with a real-time clock, and reads the data of each sensor at a preset sampling frequency (10 minutes / time under normal conditions, 5 minutes / time during peak hours) to generate the first detection data.

[0050] In some specific embodiments, the insect extermination device 200 also includes an insect attractant lamp, an insect killer, a power generation device and a power supply device.

[0051] Exemplarily, the insect-attracting lamp can emit light of a specific wavelength to attract insects, such as light with a wavelength in the ultraviolet range, and since many insects are very sensitive to such light, insects can be attracted to the vicinity of the insect-attracting lamp.

[0052] The insect killer is arranged around the insect trap light, and when the insect trap light attracts the insects into the insect killing range of the insect killer, the insect killer can capture the insects. For example, the insect killer may include at least one electric grid, and when the insects approach the electric shock range of the electric grid, the insects are given electric shocks, thereby killing them. Alternatively, the insecticide includes a sticky material, and the insects are attracted by the insect trap light and approach the sticky material and are captured by the sticky material, thereby achieving the effect of killing insects.

[0053] The power generation device may be a solar panel, through which the collected solar energy can be converted into electrical energy, and the electrical energy can be stored in the power supply device.

[0054] The power supply device can supply power to other electronic components such as insect traps, insect killers, multiple sensors, etc. The power supply device can be a battery that can store solar energy collected by solar panels during the day. The battery is usually a rechargeable battery to ensure normal operation at night or on cloudy days.

[0055] In one embodiment, the early warning device 100 in the above embodiment is further explained below.

[0056] like Figure 3 As shown, the early warning device 100 includes a data acquisition unit 110, a time series analysis unit 120, a data analysis unit 130 and an early warning decision unit 140; wherein the data acquisition unit 110 is used to obtain the first detection data from the data collector 220; the first detection data can be understood as the data obtained by each sensor in the insect killing device 200. The time series analysis unit 120 is used to generate an insect quantity prediction result according to the first detection data; the data analysis unit 130 is used to generate a factor association result and a risk classification result according to the first detection data; the early warning decision unit 140 is used to generate early warning information according to the insect quantity prediction result, the factor association result and the risk classification result.

[0057] In the implementation of the present application, the first detection data is collected by the data acquisition unit 110, and then the insect population in a future period is predicted according to the first detection data by the time series unit and an insect population prediction result is generated. The data analysis unit 130 analyzes the association between multiple factors and insect activities according to the first detection data and generates factor association results, as well as predicts the risk of insect outbreak and generates risk classification results. The early warning decision unit 140 is used to generate early warning information based on the insect population prediction results, factor association results and risk classification results. This can improve the utilization rate of the data acquired by each sensor, and can also provide more accurate early warning information. The false alarm rate can be reduced by at least 30%, which has broad application prospects.

[0058] The present application also provides an early warning device 100, see Figure 4 The early warning device 100 includes a data acquisition unit 110, a data preprocessing unit 150, a time series analysis unit 120, a data analysis unit 130 and an early warning decision unit 140.

[0059] It is understandable that the data acquisition unit 110, the data preprocessing unit 150, the time series analysis unit 120, the data analysis unit 130 and the early warning decision unit 140 in the above-mentioned embodiment can be integrated in the server 300. An independent microservice is created for each unit in the server 300. The embodiment of the present application does not limit the type of the server 300. The server 300 can be an independent physical server 300, or a server 300 cluster or distributed system composed of multiple physical servers 300, or a cloud server 300 that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0060] The data preprocessing unit 150 is used to clean, fill missing values, standardize data and extract features for the first detection data to generate second detection data.

[0061] In a specific example, if Figure 5 As shown, the data preprocessing unit 150 includes: a data cleaning module 151, a missing value filling module 152, a data standardization module 153 and a feature extraction module 154.

[0062] After receiving the first detection data, the data cleaning module 151 constructs an original data set X={x1, x2, . . . , xn} according to the first detection data, wherein x1-xn are individual data in the first detection data.

[0063] Calculate the sample mean μ and sample standard deviation σ. Among them,

[0064] μ=(1 / n)×Σxi;

[0065] σ=sqrt[(1 / (n-1))×Σ(xi-μ) 2 ].

[0066] Next, for each data xi in the original data set X, calculate the Z score: Zi = (xi-μ) / σ. And judge whether the data is abnormal based on the Z score of each data. For example, if |Zi|>3, mark the data Xi as an outlier. And judge whether the outlier is within a reasonable range, for example, by querying the expert knowledge base to judge whether the outlier is within a reasonable range. For unreasonable outliers, replace them with the average of the two time points before and after, so as to obtain the cleaned data set X'.

[0067] The missing value filling module 152 is used to generate missing data that may exist in the cleaned data set X' to form a complete data set.

[0068] Specifically, the missing value filling module 152 receives the cleaned data set X', which may contain missing data. Therefore, multiple possible filling values ​​can be generated for each missing data through multiple interpolation, and the generated filling values ​​are filled into the cleaned data set X', thereby forming multiple complete data sets Y. For example, 5 filling values ​​are generated for each missing data, and correspondingly, there are also 5 corresponding complete data sets, namely data set Y1, data set Y2, ... data set Y5. Then, each complete data set is analyzed, for example, regression analysis, mean calculation, hypothesis testing, etc. are performed on the complete data set, so as to obtain the parameter estimation value θi and variance Vi of each complete data set. The parameter estimation value θi can be the least squares solution of the parameters in the regression model.

[0069] Next, the comprehensive estimate θ and total variance VT of each complete data set are calculated, and finally the filled data set Y' and parameter estimate θ and its confidence interval are obtained.

[0070] Wherein, θ = (1 / 5) × Σθi.

[0071] VT=W+(1.2)B, where W is the within-group variance and B is the between-group variance.

[0072] The data standardization module 153 is used to perform standardization processing on the padded data set Y', so as to convert the features of different scales in the padded data set Y' into a unified scale.

[0073] Specifically, the data standardization module 153 calculates the mean μj and standard deviation σj of each feature j in the filled data set Y', and then calculates the standardized value z'ij according to the mean μj and standard deviation σj of each feature j, thereby obtaining the standardized data set Z'.

[0074] Among them, z'ij=(zij-μj) / σj, zij is each data in the filled data set Y'.

[0075] The feature extraction module 154 is used to generate an extended data set of derived features.

[0076] Specifically, the feature extraction module 154 receives the time series data set T = {t1, t2, ..., tn} and then calculates the 7-day moving average: MA7i = (1 / 7) × Σ (ti-6 + ti-5 + ... + ti), where i≥7.

[0077] Then the daily range DailyRangei and accumulated temperature AccumulatedTemp are calculated to obtain the extended data set T' containing the original data and derived features.

[0078] Among them, DailyRangei = DailyMaxi - DailyMini. DailyMaxi is the highest temperature during the day and night, and DailyMini is the lowest temperature during the day and night.

[0079] Calculating the accumulated temperature AccumulatedTemp includes the following steps:

[0080] Initialize AccumulatedTemp = 0

[0081] For the daily average temperature MeanTempi, if MeanTempi>10°C, AccumulatedTemp+=(MeanTempi-10).

[0082] In a specific embodiment, Figure 6 As shown, the time series analysis unit 120 includes a first prediction module 121, which is used to obtain the corresponding relationship between different time periods and the change in the number of insects according to the first detection data, and obtain the number of insects in the first preset time period according to the corresponding relationship.

[0083] Specifically, the first prediction module 121 can predict the short-term trend of insect population through an Auto Regressive Integrated Moving Average (ARIMA) model. The ARIMA model can predict the trend of insect population in the future by capturing the law of historical insect population changes.

[0084] The first prediction module 121 can receive a time series data set X={x1, x2...xn} from the preprocessing module, wherein the time series data set includes the number of insects in different time periods, such as the number of insects per hour, per day or per week. The time series data set is then divided into a training set and a test set, for example, 80% of the time series data set is used as a training set, and 20% of the time series data set is used as a test set.

[0085] The ARIMA model requires that the input time series data is stationary. Stationarity means that the mean, variance, and covariance of the data do not change over time. The unit root test (Augmented Dickey-Fuller, ADF) test can be used to test whether the data is stationary. If the test result is not stationary, the data can be made stationary through difference processing.

[0086] The ARIMA model is usually written as ARIMA(p,d,q), where: p is the order of the autoregressive term (AR part), d is the order of difference (the order that makes the series stationary), and q is the order of the moving average term (MA part).

[0087] Therefore, we can use the grid search method to traverse all combinations of p∈[0,5],d∈[0,2],q∈[0,5], and use each set of parameter combinations and the data in the training set to train the ARIMA model. At the same time, we calculate the Akaike information criterion (AIC) of each parameter combination and select the parameter combination (p, d, q) with the smallest AIC as the optimal parameter to build the ARIMA model.

[0088] Next, the number of insects in the future is predicted using the data in the test set and the ARIMA model, and the prediction results are evaluated. For example, the accuracy of the model can be evaluated by calculating the root mean square error (RMSE) and the mean absolute percentage error (MAPE). When the accuracy of the model meets the requirements, the corresponding relationship between different time periods and the change in the number of insects can be obtained. Then, the number of insects in the first preset time period can be predicted by the ARIMA model. For example, the number of insects within 1-7 days can be predicted.

[0089] In some specific embodiments, the time series analysis unit 120 also includes a second prediction module 122, which is used to establish a long short-term memory neural network model based on the first detection data, and obtain the number of insects in a second preset time period based on the long short-term memory neural network model, and the span of the second preset time period is greater than the span of the first preset time period.

[0090] Specifically, the second prediction module 122 can make a long-term trend prediction of the number of insects through a long short-term memory (LSTM) neural network model. Exemplarily, the second prediction module 122 can obtain a time series data set X={x1, x2,..., xn} from the preprocessing module, and the time series data set includes the number of insects in different time periods, such as the number of insects per hour, per day or per week. Since the input of the LSTM neural network model needs to be three-dimensional: [number of samples, time step, number of features], the time step can be adjusted according to the actual situation, for example, the window size is set to 30 days and the step size is 1 day, which means that each window will contain 30 days of historical data as feature input. The corresponding prediction is the number of insects in the next day.

[0091] The time series dataset is then divided into a training set, a validation set, and a test set. For example, 70% of the time series dataset is used as a training set, 15% of the time series dataset is used as a validation set, and 15% of the time series dataset is used as a test set.

[0092] The training set is input into the LSTM neural network model to train it. The LSTM neural network model can adopt the following structure: the input layer is set to shape = (30, num_features); the LSTM layer is set to 128 units, return_sequences = True; the random dropout (Dropout) layer is set to dropout_rate = 0.2; the LSTM layer is set to 64 units; the fully connected layer (Dense) layer is set to 32 neurons, and the activation function is ReLU; the output layer is set to 1 neuron, and the activation function is linear. The LSTM neural network model also uses the Adam optimizer and the mean square error (MSE) loss function to train the model until a preset number of times is reached, for example, the preset number of times can be 100 times.

[0093] After the training is completed, the prediction results of the LSTM neural network model can be verified using the validation set, for example, by calculating the mean absolute error (MAE) of the prediction results and other indicators to evaluate the prediction accuracy of the LSTM neural network model. When the accuracy of the LSTM neural network model meets the requirements, the corresponding relationship between different time periods and the change in the number of insects can be obtained. Then, the number of insects in the first preset time period can be predicted by the LSTM neural network model. For example, the number of insects within 7-30 days can be predicted.

[0094] In some specific embodiments, Figure 7 As shown, the data analysis unit 130 includes a first analysis module 131, which is used to construct a factor association model. The factor association model is used to output a factor association result based on the first detection data. The factor association result includes an importance score of each feature in the first detection data, and multiple features sorted according to the importance score.

[0095] Specifically, the first analysis module 131 obtains a feature data set and a target variable from the preprocessing module. Then the feature data set is divided into a training set and a test set, for example, 80% of the feature data set is divided into a training set, and 20% of the feature data set is divided into a test set. The feature data set includes multiple factors, such as temperature, humidity, precipitation, etc. The target variable is the result corresponding to the factor, for example, the number of insects corresponding to the factor.

[0096] Then the first analysis module 131 trains the random forest model by using the training set to obtain the factor association model. The parameters of the random forest model are set as follows: the number of trees is set to n_estimators=100; the maximum depth is set to max_depth=10; the minimum number of leaf node samples is set to min_samples_leaf=5; and the random seed is set to random_state=42.

[0097] Finally, after the first analysis module 131 inputs the test set into the factor association model, the factor association model can calculate the importance score of each feature, thereby clarifying which factors are most important for the prediction of the target variable. The features can also be sorted in descending order of importance score, and the top 10 most important features and their scores are output.

[0098] In some specific embodiments, the data analysis unit 130 further includes a second analysis module 132, and the second analysis module 132 is used to construct a risk classification model, and the risk classification model is used to calculate the risk score of each feature in the first detection data, so as to obtain a risk classification result.

[0099] Exemplarily, the second analysis module 132 obtains a feature data set and a target variable from the preprocessing module. The feature data set is then divided into a training set and a test set, for example, 70% of the feature data set is divided into a training set, and 30% of the feature data set is divided into a test set. The feature data set includes multiple factors, such as temperature, humidity, precipitation, etc. The target variable is a binary variable, representing the result corresponding to the factor, for example, the risk level of insect outbreak in the environment corresponding to the factor, and the risk level is divided into high risk and low risk.

[0100] The second analysis module 132 constructs a risk classification model through a support vector machine (SVM) algorithm. For example, a radial basis function (RBF) is used as the kernel function for constructing a risk classification model, which can effectively process nonlinear data. Then select appropriate penalty parameters C and gamma, because the penalty parameter C controls the penalty for classification errors. A smaller C value may cause the model to be underfitted, and a larger C value may cause overfitting. Gamma determines the width of the RBF kernel and affects the range of influence of high-dimensional mapping. Therefore, the second analysis module 132 can traverse all combination parameters of C∈[0.1,1,10,100] and gamma∈['scale','auto',0.1,1], select a suitable combination of penalty parameters C and gamma from them by using a grid search method and 5-fold cross validation, and use the selected penalty parameters C and gamma to train the risk classification model.

[0101] Finally, the second analysis module 132 inputs the data in the training set into the trained risk classification model, and then obtains the risk score of each feature, thereby obtaining a risk classification result.

[0102] In a specific embodiment, Figure 8 As shown, the warning decision unit 140 includes a first decision module 141 , a second decision module 142 and a warning generation module 143 .

[0103] The first decision module 141 is used to calculate the first reference value, and confirm the generation of the warning level according to the first reference value and the risk classification result, wherein the first reference value ratio is the ratio of the insect quantity prediction result to the historical average value of the same period. For example, if 1.5≤ratio<2, the warning level is yellow warning; if 2≤ratio<3, the warning level is orange warning; if ratio≥3 or the risk classification result predicts that the probability of insect outbreak is greater than 0.8, the warning level is red warning.

[0104] The second decision module 142 is used to determine the duration of the warning level according to the factor association result.

[0105] The warning generation module 143 is used to generate warning information according to the warning level and the duration of the warning level. Exemplarily, the warning message may include: warning level (yellow / orange / red), warning reason (such as: the number of pests is expected to exceed 3 times the historical period), impact range (affected area and crop type), expected duration, and one or more of the prevention and control measures.

[0106] In some embodiments, Fig. 9 As shown, the warning decision unit 140 also includes a warning information storage module 144 , a warning information push module 145 and a warning effect tracking module 146 .

[0107] The warning information storage module 144 is used to store warning information, such as the identification ID of the warning information, the generation time of the warning information, the warning level, the warning content, the impact range, the estimated duration, the warning status (active / relieved), etc.

[0108] The warning information push module 145 is used to push warning information to users according to preset push rules, such as through mobile phone APP push notifications, SMS notifications, emails, Web platform real-time updates and other channels.

[0109] The warning information tracking module is used to record the user's feedback on the warning information (such as: read, action taken, etc.). Alternatively, it can also be used to collect actual insect occurrences, compare them with the warning information, and calculate the warning accuracy and timeliness indicators.

[0110] In some specific embodiments, Fig.10 As shown, the early warning device 100 also includes a prevention and control suggestion generation unit 160, which stores a prevention and control information correspondence table; the prevention and control suggestion generation unit 160 is used to query the corresponding prevention and control strategy from the prevention and control information correspondence table according to the insect quantity prediction results, factor association results and risk classification results, and generate prevention and control information according to the prevention and control strategy.

[0111] Exemplarily, the prevention and control suggestion generating unit 160 establishes a SQL relational database to store a prevention and control information correspondence table of information such as prevention and control strategies, pesticide information, and natural enemy information. The prevention and control information correspondence table includes tables that record independent information such as strategy tables, pesticide tables, natural enemy tables, crop tables, and pest tables, and also includes an inter-table association correspondence table, which records the correspondence between tables such as pest-crop relationships and pest-pesticide applicability.

[0112] In this way, the prevention and control suggestion generation unit 160 can retrieve matching prevention and control strategies from the database according to crop type, growth stage, pest type, and weather conditions. And the search results are sorted according to the historical success rate of the strategy, and the impact of current environmental conditions (such as temperature and humidity) on the effectiveness of the strategy, and prevention and control information is generated. Exemplarily, the prevention and control information includes: recommended prevention and control measures (such as spraying pesticides, releasing natural enemies, etc.), specific operating steps, required materials (such as: pesticide type, dosage), precautions and safety tips, etc. In this way, accurate and personalized recommendations for prevention and control measures are achieved.

[0113] In some specific embodiments, the early warning device 100 also includes a knowledge graph unit, which is used to construct a knowledge graph and update the prevention and control information correspondence table in the prevention and control recommendation generation unit 160.

[0114] The knowledge graph unit extracts information from structured data sources (such as relational databases) and uses natural language processing technology to extract entities and relationships from unstructured texts (such as research papers). Among them, the types of entities can be divided into pests, crops, prevention and control measures, pesticides, natural enemies, etc., and the types of relationships can be divided into host relationships, natural enemy relationships, prevention and control effects, applicable conditions, etc.

[0115] The knowledge graph unit then infers the above entities and relationships based on preset rules, such as:

[0116] IF (pest A parasitizes crop B) AND (pesticide C controls pest A);

[0117] THEN (Pesticide C can be used on crop B)

[0118] The knowledge graph unit can also perform complex relationship analysis through graph algorithms, such as shortest path, centrality analysis, etc., thereby being able to automatically crawl the latest research results on a regular basis and generate a knowledge graph to update the prevention and control information correspondence table in the prevention and control recommendation generation unit 160.

[0119] In some specific embodiments, the warning device 100 further includes a display unit 170, such as Fig.11 As shown, the display unit 170 is used to display early warning information and prevention and control information, and can visualize data, such as using the ECharts library to draw insect population trend graphs, environmental factor correlation graphs, etc. At the same time, it can also implement multi-dimensional data screening based on user operations and interact with users.

[0120] The display unit 170 can display short-term and long-term prediction results, including confidence intervals, to provide users with prediction model performance indicators and explanatory analysis. It can also display the current warning status and detailed information in real time, and based on user operations, it can also filter historical warning records by time, region, severity, etc., while displaying targeted prevention and control measures and operation guides, and providing a knowledge graph interactive visualization interface.

[0121] The present application also provides an early warning method. Fig.12 As shown, the early warning method includes: S10: acquiring first detection data; S20: generating an insect quantity prediction result based on the first detection data; S30: generating a factor association result and a risk classification result based on the first detection data; S40: generating early warning information based on the insect quantity prediction result, the factor association result and the risk classification result.

[0122] In the implementation of this application, the number of insects in a future period of time is predicted based on the collected first detection data and an insect number prediction result is generated, and the correlation between multiple factors and insect activities is analyzed based on the collected first detection data and factor correlation results are generated, as well as the risk of insect risk outbreak is predicted and a risk classification result is generated. Thus, early warning information is generated based on the insect number prediction results, factor correlation results and risk classification results, thereby improving the utilization rate of data acquired by each sensor, and at the same time, it can also provide more accurate early warning information, and the false alarm rate can be reduced by at least 30%, which has broad application prospects.

[0123] Based on the above embodiments, the present application further provides a computer device, Fig.13 A schematic diagram of a computer device structure provided in an embodiment of the present application is shown in FIG. Fig.13 As shown, it includes: a processor 501 , a communication interface 502 , a memory 503 and a communication bus 504 , wherein the processor 501 , the communication interface 502 , and the memory 503 communicate with each other via the communication bus 504 .

[0124] The memory 503 stores a computer program. When the program is executed by the processor 501, the processor 501 performs the following steps:

[0125] S10: Acquire first detection data.

[0126] S20: Generate an insect quantity prediction result according to the first detection data.

[0127] S30: Generate a factor association result and a risk classification result according to the first detection data.

[0128] S40: Generate warning information based on the insect quantity prediction results, factor association results and risk classification results.

[0129] The communication bus mentioned in the above computer device can be PCI

[0130] (Peripheral Component Interconnect, peripheral component interconnect standard) bus or EISA (Extended Industry Standard Architecture, extended industry standard architecture) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0131] The communication interface 502 is used for communication between the above-mentioned computer device and other devices.

[0132] The memory may include RAM (Random Access Memory) or NVM (Non-Volatile Memory), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0133] The above-mentioned processor can be a general-purpose processor, which includes a central processing unit, a network processor (Network Processor, NP), etc.; it can also be a digital signal processing processor (Digital Signal Processing, DSP), an application-specific integrated circuit, a field programmable gate array or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.

[0134] Based on the above embodiments, an embodiment of the present application provides a computer-readable storage medium, in which a computer program executable by a computer device is stored. When the program is run on the computer device, the computer device implements the following steps when executing:

[0135] S10: Acquire first detection data.

[0136] S20: Generate an insect quantity prediction result according to the first detection data.

[0137] S30: Generate a factor association result and a risk classification result according to the first detection data.

[0138] S40: Generate warning information based on the insect quantity prediction results, factor association results and risk classification results.

[0139] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0140] So far, various embodiments of the present application have been described in detail. In order to avoid obscuring the concept of the present application, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solution disclosed herein.

[0141] Although some specific embodiments of the present application have been described in detail by way of examples, it should be understood by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present application. It should be understood by those skilled in the art that the above embodiments may be modified or some technical features may be replaced by equivalents without departing from the scope and spirit of the present application. In particular, the various technical features mentioned in the various embodiments may be combined in any manner as long as there is no structural conflict.

Claims

1. An early warning device, characterized in that: include: Data collection unit, time series analysis unit, data analysis unit and early warning decision unit; The data acquisition unit is used to obtain first detection data; The time series analysis unit is used to generate an insect quantity prediction result according to the first detection data; The data analysis unit is used to generate a factor association result and a risk classification result according to the first detection data; The early warning decision unit is used to generate early warning information according to the insect quantity prediction result, the factor association result and the risk classification result.

2. The device according to claim 1, characterized in that The time series analysis unit comprises a first prediction module, The first prediction module is used to obtain the corresponding relationship between different time periods and the change in the number of insects according to the first detection data, and to obtain the number of insects in a first preset time period according to the corresponding relationship.

3. The device according to claim 2, characterized in that The time series analysis unit further includes a second prediction module, The second prediction module is used to establish a long short-term memory neural network model based on the first detection data, and obtain the number of insects in a second preset time period based on the long short-term memory neural network model, and the span of the second preset time period is greater than the span of the first preset time period.

4. The device according to any one of claims 1 to 3, characterized in that The data analysis unit comprises a first analysis module, The first analysis module is used to construct a factor association model, and the factor association model is used to output a factor association result based on the first detection data. The factor association result includes an importance score of each feature in the first detection data, and multiple features sorted according to the importance score.

5. The device according to claim 4, characterized in that The data analysis unit further includes a second analysis module, The second analysis module is used to construct a risk classification model, and the risk classification model is used to calculate the risk score of each feature in the first detection data, so as to obtain a risk classification result.

6. The device according to claim 3, characterized in that The warning decision unit includes a first decision module, a second decision module and a warning generation module. The first decision module is used to calculate a first reference value, and confirm and generate an early warning level according to the first reference value and the risk classification result, wherein the first reference value is the ratio of the insect quantity prediction result to the historical average value for the same period; The second decision module is used to determine the duration of the warning level according to the factor association result; The warning generation module is used to generate warning information according to the warning level and the duration of the warning level.

7. The device according to claim 1, characterized in that The early warning device also includes a prevention and control suggestion generation unit, and the prevention and control suggestion generation unit stores a prevention and control information correspondence table; The prevention and control suggestion generating unit is used to query the corresponding prevention and control strategy from the prevention and control information corresponding table according to the insect quantity prediction result, the factor association result and the risk classification result, and generate prevention and control information according to the prevention and control strategy.

8. The device according to claim 7, characterized in that The early warning device also includes a display unit, and the display unit is used to display the early warning information and the prevention and control information.

9. An early warning method, characterized in that: include: Acquiring first detection data; generating an insect quantity prediction result according to the first detection data; generating a factor association result and a risk classification result according to the first detection data; Early warning information is generated according to the insect quantity prediction result, the factor association result and the risk classification result.

10. An insect extermination system, characterized in that: include: An insect extermination device and an early warning device according to any one of claims 1 to 9, wherein the insect extermination device is used to send the first detection data to the early warning device.

Images