An artificial wetland invasive snail monitoring method based on spatio-temporal autocorrelation analysis

Abstract

The present invention discloses a method for monitoring invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis. This method uses image acquisition devices to collect image data of constructed wetlands, and at the same time collects data on environmental variables, water quality microorganisms, and soil physical and chemical properties. By screening image samples and removing collinear environmental variables, effective data is obtained, and image recognition and principal component analysis algorithms are used to extract features. A monitoring model based on cellular automata is constructed, and the ANN algorithm is used to optimize parameters, and then the invasion risk is evaluated. Using spatio-temporal autocorrelation analysis combined with geographic information systems, the spatio-temporal distribution characteristics of snails are studied, and the diffusion paths and ranges are predicted. Compared with traditional monitoring methods, the present invention integrates multi-source data, uses advanced algorithms and models, can more accurately monitor invasive snails, provides a scientific basis for the prevention and control of invasive snails in constructed wetlands, and effectively protects the ecological system of constructed wetlands.

Description

Technical Field

[0001] The present invention relates to ecological protection and ecological distribution models, and particularly to a method for monitoring invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis. Background Art

[0002] In the field of ecological protection, the invasion of alien species is a severe challenge, posing a serious threat to the stability of ecosystems, biodiversity, and economic development. Constructed wetlands, as unique and important ecosystems, are not spared either. For example, Spartina alterniflora has invaded the wetlands in Tianjin, seriously disrupting the local ecological balance and affecting the habitats of native plants and wetland birds.

[0003] Currently, various technical means have been developed for monitoring invasive snails in constructed wetlands. Traditional monitoring methods such as field surveys require monitoring personnel to directly visit constructed wetlands and record information such as the species, quantity, and distribution range of snails by visual observation and simple tools. Although this method is intuitive, it is inefficient and vulnerable to the subjective factors of monitoring personnel, making it difficult to comprehensively monitor some hidden or widely distributed snails. Quadrat sampling is to set a certain number and size of quadrats in constructed wetlands and conduct detailed investigations and statistics on the snails in the quadrats to estimate the snail situation in the entire area. However, the rationality and representativeness of quadrat setting have a great impact on the accuracy of the results, and in large-scale constructed wetlands, the workload of quadrat sampling is huge.

[0004] With the development of technology, modern monitoring technologies have emerged continuously. Remote sensing monitoring technology analyzes satellite images of different time phases or high-resolution images obtained by unmanned aerial vehicles to grasp the distribution range and change trend of invasive snails. It uses the differences in spectral characteristics between snails and the surrounding environment to identify snails. For example, the reflectance of some invasive snails in specific bands is different from that of native organisms. However, this technology is greatly affected by factors such as weather and clouds, and the recognition accuracy for some tiny snails or snails in occluded areas needs to be improved. eDNA monitoring technology extracts DNA from environmental media such as water and soil in constructed wetlands, performs PCR amplification and high-throughput sequencing on specific DNA fragments of the genome, thereby monitoring invasive snails. This technology can detect trace amounts of snail DNA in the environment and can be monitored even when the number of snails is small or difficult to directly observe, but there are problems such as high detection costs and strict requirements for experimental environments and operations.

[0005] In recent years, artificial intelligence monitoring technology has used machine learning algorithms to analyze and identify a large amount of ecological environment data, and can also predict the diffusion trend of snails in combination with geographic information systems. However, this technology relies on high-quality training samples and appropriate algorithm models, and the accuracy and generalization ability of the models still need to be further improved.

[0006] These existing technologies all have certain limitations in the monitoring of invasive snails in constructed wetlands, and it is difficult to comprehensively, accurately, and efficiently monitor invasive snails. Therefore, it is of great practical significance and urgent need to develop a method that comprehensively considers spatio-temporal factors and can more accurately and comprehensively monitor invasive snails in constructed wetlands. Summary of the Invention

[0007] The main purpose of the present invention is to construct an efficient and accurate monitoring method for invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis to solve many technical problems existing in current monitoring means for invasive snails in constructed wetlands. For example, traditional monitoring methods such as field surveys are inefficient and easily interfered by subjective factors, remote sensing monitoring is greatly affected by weather and has insufficient recognition accuracy for tiny snails, eDNA monitoring has high costs and strict operation requirements, and the model accuracy and generalization ability of artificial intelligence monitoring need to be improved, etc. The present invention aims to comprehensively, accurately, and efficiently monitor the distribution, changes, and diffusion trends of invasive snails through multi-source data collection, scientific data processing and screening, advanced model construction and optimization, and in-depth spatio-temporal analysis, providing strong support for the protection and management of constructed wetland ecosystems, thereby effectively preventing and controlling the damage of invasive snails to the ecological balance of constructed wetlands.

[0008] The present invention provides a monitoring method for invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis, including:

[0009] S1, import a topographic map into a computer system and rasterize the monitoring area; use an image acquisition device to collect image data of the constructed wetland containing snails, and at the same time collect environmental variable data, water quality microorganism data, and soil physical and chemical property data;

[0010] S2, screen out representative image samples from images under the same environmental conditions in the same grid as effective image data, and use correlation analysis to remove environmental variables with collinearity exceeding a preset value in the environmental variable data to obtain effective environmental variable data;

[0011] S3, use image recognition to process the effective image data and extract the image features of the snails; use the principal component analysis algorithm to perform fusion feature extraction on the effective image data, effective environmental variable data, water quality microorganism data, and soil physical and chemical property data to obtain a comprehensive feature vector;

[0012] S4, construct an invasive snail monitoring model based on cellular automata, define the state, neighborhood relationship, and transition rules of the cells; use the ANN algorithm in machine learning to optimize the model parameters;

[0013] S5, input the real-time collected image data, environmental variable data, water quality microorganism data, and soil physical and chemical property data into the invasive snail monitoring model to evaluate the risk degree of invasive snails in the constructed wetland;

[0014] S6. Use spatio-temporal autocorrelation analysis to determine the spatial distribution characteristics and temporal variation trends of invasive snails in constructed wetlands, combine with Geographic Information System (GIS) to display the dynamic spatio-temporal distribution of snails, and predict the future invasion diffusion paths and ranges.

[0015] As a further preferred solution, in step S2, when screening representative image samples, calculate the clarity score through image gradient information; it includes:

[0016] Perform a square root operation on the sum of squares of the differences between the pixel values of each position and its adjacent positions in the image and then accumulate them.

[0017] Select the images with clarity scores in the top 30% and complete snail information as valid image data, and delete other images.

[0018] When removing the collinearity of environmental variables, use the Pearson correlation coefficient to calculate the correlation between the observed values of two environmental variables; this coefficient is obtained by dividing the sum of the products of the differences between the variable observed values and their means by the product of the square roots of the sums of the squares of their respective differences.

[0019] Remove the environmental variables with a correlation greater than 0.8 and less significance for snail monitoring.

[0020] As a further preferred solution, in step S3, use a convolutional neural network to extract snail image features; and: the convolutional layer calculates the feature map by multiplying the convolutional kernel with the corresponding position of the input image and adding a bias term; the pooling layer uses max pooling, that is, selects the maximum pixel value in the neighborhood.

[0021] As a further preferred solution, in step S3, when using the principal component analysis algorithm to fuse features, first standardize the data, subtract the variable mean from the original data and then divide by the standard deviation.

[0022] Then calculate the covariance matrix, which is the average of the sum of the products of the sample data and the differences between their means.

[0023] Perform eigenvalue decomposition on the covariance matrix, select the first several eigenvectors with larger eigenvalues to form a transformation matrix, and multiply the standardized data matrix by the transformation matrix to obtain the comprehensive eigenvector.

[0024] As a further preferred solution, in the cellular automaton model of step S4, the cell states are divided into three types: no snails, fewer snails, and more snails; and: the neighborhood adopts the Moore neighborhood, including the surrounding 8 cells.

[0025] The probability of cell state transition is jointly determined by a random term, a global transition probability, a constraint condition, and a neighborhood function; the random term is based on natural logarithm-related operations, and the global transition probability is the ratio of the number of cells with state transitions to the total number of cells;

[0026] The constraint condition is judged according to the environmental suitability index of the cell position, which is 1 when the index is greater than 0.5, and 0 otherwise; moreover, the neighborhood function is the sum of the cell state values in the neighborhood divided by the number of cells in the neighborhood;

[0027] When using the ANN algorithm to optimize the model parameters, the backpropagation algorithm is used to calculate the gradient, which is calculated by the sum of the products of the partial derivatives between the output of the output layer neurons, the input of the hidden layer neurons, and the weights, and the weights are continuously adjusted to minimize the loss function.

[0028] As a further preferred solution, in step S5, when evaluating the risk degree of invasive snails in the constructed wetland, a risk assessment index system is constructed, including:

[0029] The snail density index is the ratio of the number of snails per unit area to the area of the monitoring area;

[0030] The invasion and diffusion speed index is the ratio of the change in the number of snails over a certain period of time to the change in time;

[0031] The environmental suitability index is obtained by multiplying the ratio of the difference between the current values of environmental factors such as temperature, humidity, pH, and dissolved oxygen and the maximum and minimum values suitable for the survival of snails by their respective weight coefficients and then summing them up, and the sum of the weight coefficients is 1; and:

[0032] The weights of each index are determined by the analytic hierarchy process, and the final risk degree is the sum of the products of the scores of each index and their weights.

[0033] As a further preferred solution, in step S6, when using spatio-temporal autocorrelation analysis to study the spatio-temporal distribution characteristics and change trends of invasive snails in the constructed wetland, the following global Moran's I index and local Moran's I index are constructed;

[0034] The calculation formula of the global Moran's I index is:

[0035] ;

[0036] where n is the number of samples, is the sum of the elements of the spatial weight matrix, is the spatial weight matrix, 、 is the snail distribution data, is the mean value;

[0037] The calculation formula of the local Moran's I index is as follows:

[0038] ;

[0039] where , ;

[0040] When combining the geographic information system to display the dynamic changes in the spatio-temporal distribution of snails, the calculated Moran's I index results are visualized on the map, and different degrees of autocorrelation are represented by different colors and symbols;

[0041] When predicting the future invasion diffusion path and scope, a prediction method based on the cellular automaton model is adopted. According to the current state and transition rules of the cells, the state of the cells at future moments is iteratively calculated to predict the diffusion of snails.

[0042] As a further preferred solution, when using the principal component analysis algorithm for feature extraction, the method for determining the number of principal components is as follows: First, calculate the cumulative contribution rate, and the formula is:

[0043] ;

[0044] where is the th eigenvalue, is the number of original variables;

[0045] Select such that the cumulative contribution rate reaches more than 85%.

[0046] As a further preferred solution, in the invasion snail monitoring model based on cellular automata, the global transition probability is statistically obtained from historical data, and the formula is:

[0047] .

[0048] As a further preferred solution, when using the analytic hierarchy process to determine the weights of risk assessment indicators, a judgment matrix is constructed:

[0049] ;

[0050] where represents the importance degree of the ith indicator relative to the jth indicator, which is determined by expert scoring; then calculate the maximum eigenvalue and the corresponding eigenvector W of the judgment matrix, and the formula is:

[0051] ;

[0052] Finally, normalize the eigenvector to obtain the weights of each indicator 。

[0053] Compared with the prior art, the present invention focuses on the monitoring of invasive snails in constructed wetlands, integrates multiple technologies to build a comprehensive monitoring system, can comprehensively improve the monitoring level of invasive snails, and protect the ecosystem of constructed wetlands. Specifically as follows:

[0054] First, the present invention utilizes advanced image acquisition equipment and multi-source data acquisition technologies, combines image screening and correlation analysis methods to obtain high-quality images and environmental data. Through image recognition and principal component analysis algorithms, the image features and comprehensive feature vectors of snails are accurately extracted, and information such as the species, quantity, distribution range, and activity of snails can be more accurately identified. For example, in a complex constructed wetland environment, traditional methods are easily interfered with, resulting in large errors in snail recognition, while the present invention can effectively eliminate interference, improve the monitoring accuracy, and provide a reliable basis for subsequent prevention and control.

[0055] Second, the present invention constructs a monitoring model for invasive snails based on cellular automata and optimizes the model parameters by combining with the ANN algorithm. The cellular automata model takes into account the spatial heterogeneity and dynamic changes of snails in constructed wetlands. The ANN algorithm enables the model to automatically adjust parameters to adapt to different monitoring scenarios through learning a large amount of data. Compared with traditional fixed-parameter models, the model of the present invention can more accurately simulate the spatio-temporal distribution and change rules of snails, improve the accuracy of prediction, and effectively cope with the complexity of the constructed wetland environment and the uncertainty of snail invasion.

[0056] Third, the present invention uses spatio-temporal autocorrelation analysis combined with geographic information system to deeply study the spatio-temporal distribution characteristics and change trends of invasive snails. By calculating indicators such as Moran's I index, the aggregation and correlation of snail distribution can be accurately judged, and the dynamic changes of snail spatio-temporal distribution are intuitively displayed by combining with GIS technology to predict the future invasion diffusion path and range. Potential invasion risk areas can be discovered in advance, time can be gained for taking timely prevention and control measures, and the damage degree of invasive snails to the constructed wetland ecosystem can be reduced.

[0057] The present invention combines multi-source data and historical prevention and control cases, finds similar scenarios from historical cases and optimizes the prevention and control plan according to different snail invasion situations and environmental conditions, and makes decisions based on the set prevention and control rules. When the number of snails exceeds the threshold and the environmental suitability index is greater than 0.7, biological control combined with physical control methods are preferentially adopted to improve the scientificity and effectiveness of prevention and control decisions, and avoid resource waste and ecological damage caused by blind prevention and control.

[0058] The monitoring method of the present invention can timely detect the threat of invasive snails. Through precise monitoring, risk assessment, and scientific prevention and control decisions, it can effectively control the spread of invasive snails, protect the biodiversity and ecological balance of artificial wetlands. Reduce the damage of invasive snails to wetland plants and water ecosystems, maintain the normal functions of wetland ecosystems, ensure the ecological service value of artificial wetlands in aspects such as water purification, flood regulation, and biological habitats, promote the sustainable development of artificial wetlands, and provide strong support for ecological environment construction and protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 Shows the workflow diagram of the present invention;

[0060] Figure 2 Shows the plan view of an artificial wetland in an embodiment of the present invention;

[0061] Figure 3 Shows the schematic diagram of rasterizing the terrain in an embodiment of the present invention;

[0062] Figure 4 Shows the spatial distribution characteristics of invasive snails in an artificial wetland in an embodiment of the present invention;

[0063] Figure 5 Shows the schematic diagram of the dynamic change of the spatio-temporal distribution of invasive snails in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] The preferred embodiments of the present invention will be described in detail below to more clearly understand the purpose, features, and advantages of the present invention. It should be understood that the following embodiments are not intended to limit the scope of the present invention, but only to illustrate the essential spirit of the technical solution of the present invention.

[0065] In the following description, certain specific details are set forth in order to provide a thorough understanding of various disclosed embodiments. However, one of ordinary skill in the relevant art will recognize that the embodiments may be practiced without one or more of these specific details. In other instances, well-known technologies associated with the present application may not be shown or described in detail so as not to unnecessarily obscure the description of the embodiments.

[0066] References to "one embodiment" or "an embodiment" throughout the specification mean that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, the appearances of "in one embodiment" or "in an embodiment" throughout the specification are not necessarily all referring to the same embodiment. Additionally, the particular features, structures, or characteristics may be combined in any manner in one or more embodiments.

[0067] Such as Figure 1As shown in the figure, a monitoring method for invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis includes:

[0068] S1, Import a topographic map into the computer system and rasterize the monitoring area; Use an image acquisition device to collect image data of the constructed wetland containing snails, and at the same time collect environmental variable data, water quality microbial data, and soil physical and chemical property data;

[0069] S2, Screen out representative image samples from the images under the same environmental conditions in the same grid as effective image data, and use correlation analysis to remove environmental variables with collinearity exceeding a preset value in the environmental variable data to obtain effective environmental variable data;

[0070] S3, Use image recognition to process the effective image data and extract the image features of the snails; Use the principal component analysis algorithm to extract the fusion features of the effective image data, effective environmental variable data, water quality microbial data, and soil physical and chemical property data to obtain a comprehensive feature vector;

[0071] S4, Construct a monitoring model for invasive snails based on cellular automata, and define the state, neighborhood relationship, and transition rules of the cells; Use the ANN algorithm in machine learning to optimize the model parameters;

[0072] S5, Input the real-time collected image data, environmental variable data, water quality microbial data, and soil physical and chemical property data into the monitoring model for invasive snails to evaluate the risk degree of invasive snails in the constructed wetland;

[0073] S6, Use spatio-temporal autocorrelation analysis to determine the spatial distribution characteristics and temporal variation trends of invasive snails in the constructed wetland, combine with a geographic information system to display the dynamic changes of the spatio-temporal distribution of snails, and predict the future invasion diffusion path and scope.

[0074] As a further preferred solution, in step S2, when screening representative image samples, calculate the clarity score through image gradient information; including:

[0075] Perform a square root operation on the sum of the squares of the differences between the pixel values of each position in the image and its adjacent positions and then accumulate;

[0076] Select images with clarity scores in the top 30% and complete snail information as effective image data, and delete other images;

[0077] When removing the collinearity of environmental variables, use the Pearson correlation coefficient to calculate the correlation between the observed values of two environmental variables; This coefficient is obtained by dividing the sum of the products of the differences between the variable observed values and their means by the square root product of the sums of the squares of their respective differences;

[0078] Remove environmental variables with a correlation greater than 0.8 and less significance for snail monitoring.

[0079] The specific implementation method is as follows: The method for screening representative image samples from images under the same environmental conditions in the same grid is:

[0080] Calculate the clarity score of each image, and measure the clarity score through the gradient information of the image. The formula is:

[0081] ;

[0082] Where represents the pixel value of the image at the position, and M and N are the number of rows and columns of the image respectively;

[0083] Select the images with clarity scores in the top 30% and complete snail information as valid image data, and automatically delete other redundant images;

[0084] When using correlation analysis to remove environmental variables with collinearity exceeding the preset value in the environmental variable data, the Pearson correlation coefficient is used for calculation. The formula is:

[0085] ;

[0086] Where and are the observed values of two environmental variables respectively, and are the means of the corresponding variables, and n is the number of samples;

[0087] Remove the variables with little significance for snail monitoring among the environmental variables with a correlation greater than 0.8.

[0088] In step S3, use a convolutional neural network to extract snail image features; and: The convolutional layer calculates the feature map by multiplying the convolutional kernel with the corresponding position of the input image and adding the bias term; the pooling layer uses max pooling, that is, selects the maximum pixel value in the neighborhood. Among them, the convolutional calculation process of the convolutional layer is as follows:

[0089] ;

[0090] Where is the convolutional kernel, m, n are the sizes of the convolutional kernel, Image is the input image, is the bias term;

[0091] The pooling layer uses the max pooling operation. The formula is:

[0092] .

[0093] As a further preferred solution, when using the principal component analysis algorithm to fuse features, first standardize the data by subtracting the variable mean from the original data and then dividing by the standard deviation;

[0094] Next, calculate the covariance matrix, which is the average of the sum of the products of the sample data and the difference from its mean;

[0095] Perform eigenvalue decomposition on the covariance matrix, select the first several eigenvectors with larger eigenvalues to form a transformation matrix, and multiply the standardized data matrix by the transformation matrix to obtain the comprehensive eigenvector.

[0096] The specific implementation method is as follows: First, perform standardization processing on the data. The formula is:

[0097] ;

[0098] Among them, is the original data, is the mean of the i-th variable;

[0099] Then calculate the covariance matrix:

[0100] ;

[0101] Among them, is the i-th sample data, is the sample mean;

[0102] Perform eigenvalue decomposition on the covariance matrix to obtain eigenvectors and eigenvalues. Select the first k eigenvectors with eigenvalues greater than the set value to form a transformation matrix P. Finally, obtain the comprehensive eigenvector Y = XP, where X is the standardized data matrix.

[0103] As a further preferred solution, in the cellular automaton model of step S4, the cell states are divided into three types: no snails, few snails, and many snails; and: the neighborhood adopts the Moore neighborhood, including the surrounding 8 cells;

[0104] The cell state transition probability is jointly determined by a random term, a global transition probability, a constraint condition, and a neighborhood function; the random term is based on natural logarithm-related operations, and the global transition probability is the ratio of the number of cells with state transitions to the total number of cells;

[0105] The constraint condition is judged according to the environmental suitability index of the cell position. When the index is greater than 0.5, it is 1, otherwise it is 0; and the neighborhood function is the sum of the cell state values in the neighborhood divided by the number of cells in the neighborhood;

[0106] When optimizing the model parameters using the ANN algorithm, the backpropagation algorithm is adopted to calculate the gradient, which is calculated by the sum of the products of the partial derivatives between the outputs of the neurons in the output layer, the inputs of the neurons in the hidden layer, and the weights. The weights are continuously adjusted to minimize the loss function.

[0107] The specific implementation process includes the following steps:

[0108] Define the state of the cell as:

[0109] ;

[0110] The neighborhood relationship adopts the Moore neighborhood method, that is, the neighborhood of a cell includes its surrounding 8 cells;

[0111] The transition rule of the cell is determined by the following formula:

[0112] ;

[0113] Among them, is the probability of the state transition of cell at time t, represents the random term, is the global transition probability, represents the constraint condition of the cell unit. When the environmental suitability index of the location where the cell is located is greater than 0.5, , otherwise it is 0; represents the neighborhood function, which is determined by the following formula;

[0114] ;

[0115] Among them, is the neighborhood of the cell, is the number of cells in the neighborhood;

[0116] When using the ANN algorithm to optimize the model parameters, the backpropagation algorithm is adopted to calculate the gradient, and the formula is:

[0117] ;

[0118] Among them, E is the loss function, is the output of the k-th neuron in the output layer, is the input of the j-th neuron in the hidden layer, is the weight connecting the i-th neuron in the hidden layer and the j-th neuron in the output layer, and L is the number of neurons in the output layer;

[0119] Finally, the loss function is minimized by continuously adjusting the weights.

[0120] As a further preferred solution, in step S5, when evaluating the risk degree of invasive snails in the constructed wetland, a risk assessment index system is constructed, including:

[0121] The snail density index, which is the ratio of the number of snails per unit area to the area of the monitoring region;

[0122] The invasion and spread speed index, which is the ratio of the change in the number of snails over a certain period of time to the change in time;

[0123] The environmental suitability index, which is obtained by adding the products of the ratios of the differences between the current values of environmental factors such as temperature, humidity, pH, and dissolved oxygen and the minimum and maximum values suitable for the survival of snails, multiplied by their respective weight coefficients, and the sum of the weight coefficients is 1; and:

[0124] The weights of the various indicators are determined by the analytic hierarchy process, and the final risk degree is the sum of the products of the scores of the various indicators and their weights.

[0125] The specific calculation formulas for the risk assessment indicators are as follows:

[0126] Snail density index:

[0127] ;

[0128] where N is the number of snails per unit area, and A is the area of the monitoring region;

[0129] Invasion and spread speed index:

[0130] ;

[0131] where is the change in the number of snails over time;

[0132] Environmental suitability index:

[0133] ;

[0134] where 、 、 、 are the currently monitored values of temperature, humidity, pH, and dissolved oxygen respectively, 、 , 、 , 、 , 、 are the minimum and maximum values of the corresponding environmental factors suitable for the survival of snails respectively, 、 、 、 is a weight coefficient determined according to the importance of various environmental factors on snails, and + + + = 1;

[0135] The weights of each index are determined by the analytic hierarchy process, and the final risk level:

[0136] ;

[0137] are the scores of each index, are the weights of the corresponding indexes.

[0138] In some embodiments, in step S6, when using spatio-temporal autocorrelation analysis to study the spatio-temporal distribution characteristics and change trends of invasive snails in constructed wetlands, the following global Moran's I index and local Moran's I index are constructed;

[0139] The calculation formula of the global Moran's I index is:

[0140] ;

[0141] where n is the number of samples, is the sum of the elements of the spatial weight matrix, is the spatial weight matrix, , are the snail distribution data, is the mean value;

[0142] The calculation formula of the local Moran's I index is:

[0143] ;

[0144] where , ;

[0145] When combining the geographic information system to display the dynamic changes of the spatio-temporal distribution of snails, the calculated Moran's I index results are visualized on the map, and different autocorrelation degrees are represented by different colors and symbols;

[0146] When predicting the future invasion diffusion path and range, a prediction method based on the cellular automaton model is adopted. According to the current state and transition rules of the cells, the states of the cells at future moments are iteratively calculated, so as to predict the diffusion situation of snails.

[0147] As a further preferred solution, when calculating the image sharpness score, Gaussian weighting is performed on the gradient information, and the formula is:

[0148] ;

[0149] wherein, G(x, y) is a two-dimensional Gaussian function, , is the standard deviation.

[0150] As a further preferred solution, when using the principal component analysis algorithm for feature extraction, the method for determining the number of principal components is as follows: First, calculate the cumulative contribution rate, and the formula is:

[0151] ;

[0152] wherein, is the i-th eigenvalue, is the number of original variables;

[0153] Select such that the cumulative contribution rate reaches more than 85%.

[0154] As a further preferred solution, in the invasive snail monitoring model based on cellular automata, the global transition probability is statistically obtained from historical data, and the formula is:

[0155] .

[0156] As a further preferred solution, when using the analytic hierarchy process to determine the weights of risk assessment indicators, a judgment matrix is constructed:

[0157] ;

[0158] wherein, represents the importance degree of the i-th indicator relative to the j-th indicator, which is determined by expert scoring; then calculate the maximum eigenvalue and the corresponding eigenvector W of the judgment matrix, and the formula is:

[0159] ;

[0160] Finally, normalize the eigenvector to obtain the weights of each indicator .

[0161] As Figure 2 shown, a sewage treatment plant is located at the edge of a natural wetland (Guizhou Caohai Nature Reserve), and a large area of artificial wetland has been constructed. There is a vegetation-covered area between the artificial wetland and the natural wetland, and the main vegetation varieties are Iris siberica, Zizania latifolia, Oenanthe javanica, etc. In 2018, it was monitored that snail mollusks reproduced in the artificial wetland. After a long-term investigation, it was determined that they were invasive alien snail mollusks, including Tiphobia elegans Oxyloma elegans(Risso, 1826), Physa acuta Physella acuta raparnaud , 1805) and Euhadra luchuensis Acusta despecta (Gray, 1839), etc. Allowing these spiral mollusks to invade will cause significant ecological damage, such as breeding pathogens and endangering the drinking water safety of sewage treatment plants.

[0162] In response, the constructed wetland has carried out experimental prevention and control work according to the solution of the present invention and achieved good prevention and control effects. As Figure 3 shown, first, the terrain of the monitoring area is rasterized in grid units of 10 meters × 10 meters, and image data of the constructed wetland containing snails is collected using image acquisition equipment. In this embodiment, the image acquisition is carried out during the day when the weather is clear and the light is sufficient. During the breeding season of snails, it is continuously collected for 3 months with a 7-day cycle, and a combination of equipment such as a high-definition camera or a camera carried by a drone is used. Among them, most areas are continuously photographed by the drone according to the set flight trajectory, and some areas are photographed manually.

[0163] In addition, multi-source information such as environmental variable data, water quality microbial data, and soil physical and chemical property data is further combined. The distribution and characteristics of snails may vary under different environmental conditions. By comprehensively analyzing these data, it is possible to assist in judging the species and status of snails and improve the accuracy of identification. For example, some snails prefer specific water quality pH or soil types, and the identification range can be narrowed based on these environmental data.

[0164] In terms of image screening from the same grid under the same environmental conditions, after completing the image acquisition for 3 consecutive months with a 7-day cycle, for each 10-meter × 10-meter grid, for the images collected under the same environmental conditions (such as light intensity, water quality pH, water temperature, etc. at the same time period), calculate the clarity score of each image. Select the images with clarity scores in the top 30% and complete snail information as valid image data, and automatically delete other redundant images. In terms of environmental variable data processing, the collected environmental variable data covers light intensity, water temperature, water quality pH, soil humidity, soil nutrient content, etc. Use correlation analysis to remove environmental variables with collinearity exceeding the preset value (set to 0.8 in this embodiment). Calculate using the Pearson correlation coefficient. For environmental variables with a correlation greater than 0.8, compare and analyze their significance for snail monitoring, and remove the variables with less significance for snail monitoring, so as to obtain effective environmental variable data.

[0165] The valid image data of the same grid under the same environmental conditions screened in the previous step are input into the pre-built image recognition model based on the convolutional neural network (CNN). This model has been trained on a large number of snail image datasets and has the ability to recognize snail features. The image passes through the convolutional layer of the model in sequence. The convolutional kernels in the convolutional layer slide on the image to perform convolutional operations on local regions of the image, extracting primary features such as the edges, textures, and shapes of snails, such as recognizing the unique spiral texture of the snail shell and the general outline of the snail body. Then the image enters the pooling layer, and through the max-pooling operation, the feature map output by the convolutional layer is dimension-reduced, retaining key feature information while reducing the computational amount, enabling the model to process data more efficiently and improving the robustness of the model to a certain extent to prevent overfitting. After multiple alternating processes of the convolutional layer and the pooling layer, a feature vector containing detailed image features of snails is obtained.

[0166] Integrate the snail image feature vectors extracted from the valid image data with the valid environmental variable data, water quality microbial data, and soil physical and chemical property data obtained in the previous step. These data cover various aspects of information related to the survival of snails in the constructed wetland, such as light intensity, water temperature, types and quantities of microorganisms, soil pH, etc. Standardize the integrated multi-source data to ensure that different types and magnitudes of data are on the same comparable scale, eliminating the influence of data dimension differences on subsequent analysis.

[0167] Calculate the covariance matrix based on the standardized data to analyze the correlations between variables. By performing eigenvalue decomposition on the covariance matrix, eigenvectors and eigenvalues are obtained. Select the first several eigenvectors with larger eigenvalues according to the cumulative contribution rate to form a transformation matrix, and multiply the standardized multi-source data matrix by the transformation matrix to obtain the comprehensive feature vector. This comprehensive feature vector integrates key information in multiple aspects such as images, environment, water quality microorganisms, and soil physical and chemical properties, and more comprehensively reflects the characteristics related to invasive snails in the constructed wetland.

[0168] According to the 10 m × 10 m grid division of the constructed wetland, each grid is regarded as a cell. According to the actual presence of snails in each cell, and in accordance with the pre-set standards, the cell state is defined as: no snails present, snails present with a small quantity, and snails present with a large quantity. Through on-site investigations and image recognition statistics, if the number of snails in a cell is 0, its state is no snails present; if the number is between 1 and 100, it is determined that snails are present with a small quantity; if it exceeds this range, it is snails present with a large quantity.

[0169] Adopt the Moore neighborhood method, that is, the neighborhood of each cell includes 8 cells in the surrounding up, down, left, right and four diagonal directions. This neighborhood setting method can better simulate the spatial interaction relationship between cells, considering the possible diffusion directions and ranges of snails in the wetland environment. Combine the actual environmental conditions of the constructed wetland and the biological characteristics of snails to determine the cell transformation rules. The transformation rules comprehensively consider factors such as random factors, global transformation probability, environmental suitability index of the cell's location, and the states of cells in the neighborhood.

[0170] For example, the environmental suitability index is obtained by comprehensively evaluating environmental factors such as light, water temperature, water quality, and soil at the location of the cell. When the environmental suitability index is greater than 0.5, it is considered that the environment of the cell is more favorable for the survival and reproduction of snails, and corresponding weights are given in the transformation rules. The global transformation probability is obtained based on historical data statistics, reflecting the overall possibility of cell state transformation in the entire monitoring area. The neighborhood function measures the influence of the neighborhood on the current cell state transformation by calculating the ratio of the sum of the states of cells in the neighborhood to the number of cells in the neighborhood.

[0171] Take the comprehensive feature vector obtained in the previous step as the input data and input it into the artificial neural network (ANN) model. The ANN model has a hidden layer and an output layer composed of multiple neurons, and the neurons are connected by weights. Use the backpropagation algorithm to train the model. During the training process, the model calculates based on the input comprehensive feature vector and outputs the prediction results of the cell states. Compare the prediction results with the actual cell states and calculate the loss function value, which reflects the deviation degree between the model prediction results and the actual situation.

[0172] Through the backpropagation algorithm, starting from the output layer, reverse-propagate the gradient of the loss function along the connection path between neurons, calculate the influence degree of each weight on the loss function, that is, the gradient value. According to the calculated gradient values, adjust the weights between neurons to gradually reduce the loss function value. Continuously repeat this process, and after multiple iterative trainings, enable the ANN model to accurately predict the cell states based on the input comprehensive feature vector, thereby optimizing the parameters of the invasive snail monitoring model based on cellular automata and improving the simulation accuracy of the spatio-temporal distribution and changes of invasive snails in the constructed wetland.

[0173] Continuously use the same image acquisition device as in the previous period. During suitable weather periods, conduct image acquisition of the constructed wetland at fixed intervals to ensure capturing the latest snail distribution and activity conditions. At the same time, use various sensors to monitor environmental variable data in real-time, such as light intensity sensors, water temperature sensors, pH sensors, etc., and record the data regularly. Collect water quality microorganism data through professional water quality sampling tools and microbial detection kits, sample the soil regularly, and analyze the soil physical and chemical property data using laboratory instruments. Integrate these real-time collected data to ensure consistent timestamps for accurate input into the monitoring model.

[0174] Based on the previous research and monitoring experience of invasive snails, construct a risk assessment index system. This system includes a snail density index, which is measured by counting the number of snails per unit area; an invasion and spread speed index, which is determined according to the change in the number of snails and the change in the distribution range within different time periods; and an environmental suitability index, which comprehensively considers the suitability of environmental factors such as light, water temperature, water quality, and soil for the survival and reproduction of snails. For example, by analyzing historical data and experimental results, determine the suitable range of different environmental factors, and assign a higher suitability score when the actual environmental data is within the suitable range.

[0175] Input the integrated real-time data into an optimized invasive snail monitoring model based on cellular automata. The model simulates the state changes and distribution of snails in the current environment according to the input data and the established conversion rules. At the same time, calculate the scores of each index according to the risk assessment index system. Determine the weights of each index through the analytic hierarchy process, multiply the scores of each index by their weights and sum them up to obtain the risk degree assessment result of invasive snails in the constructed wetland. According to the pre-set risk level threshold, judge which level of low, medium, or high the current risk is.

[0176] On this basis, prepare data for spatio-temporal autocorrelation analysis. First, collect snail distribution data over a certain period of time (such as from the first discovery of invasive snails to the present), including the snail state information of each cell at different time points, as well as the corresponding environmental variable data, etc. Organize these data into a format suitable for spatio-temporal autocorrelation analysis to ensure the accuracy and integrity of the data.

[0177] Using the spatio-temporal autocorrelation analysis method, calculate the global Moran's I index and the local Moran's I index. During the calculation process, take the snail distribution data and the spatial weight matrix as inputs, and obtain the index values through the corresponding algorithms. These index values can reflect whether there is aggregation in the spatial distribution of invasive snails and which regions have significant spatial autocorrelation. A global Moran's I index greater than 0 indicates positive autocorrelation, that is, the snails show an aggregated distribution in space; the local Moran's I index can determine which specific cell regions have abnormal aggregation or dispersion situations.

[0178] Integrate the calculated spatio-temporal autocorrelation analysis results with the Geographic Information System. On platforms such as GIS, import the map data of the constructed wetland. Associate the spatio-temporal distribution data of the snails and the index values obtained from the spatio-temporal autocorrelation analysis with the map, as Figure 4 shown. By setting different colors, symbols, and transparencies, etc., visually display the dynamic changes in the spatio-temporal distribution of the snails on the map. For example, use red to represent the areas with a high aggregation degree of the snails, and the darker the color, the higher the aggregation degree; over time, demonstrate the expansion or contraction of the distribution range of the snails through dynamic display. As Figure 5 shown.

[0179] Based on the optimized invasive snail monitoring model and the spatio-temporal autocorrelation analysis results, use the spatial analysis function and model prediction function of the GIS platform to predict the future diffusion path and range of invasive snails. The model takes into account factors such as the reproductive characteristics of the snails, environmental suitability, and spatial autocorrelation, and simulates the possible diffusion direction and speed of the snails in the future for a period of time. By drawing the predicted diffusion range and path on the GIS map, provide forward-looking information for the management and prevention and control work of the constructed wetland, so as to formulate corresponding countermeasures in advance.

[0180] In Figure 4 and Figure 5 it can be seen that the main invasion paths of the snails are in two directions, namely the F8 grid direction and the H16 grid direction, especially the H16 grid direction is the most serious, and it has spread widely in the J16 direction and has invaded the hinterland of the constructed wetland. Combining with other environmental indicators, it is found that Iris sibirica provides a good medium for snail invasion. Therefore, the following measures are taken:

[0181] According to Figure 5Under the guidance of , the areas with darker colors in the figure are marked as key prevention and control areas. After all the Siberian irises in this area are removed, lime is evenly spread on this area to kill mollusks. Then, after being exposed to the sun for one day, water is introduced into this area, and lime is evenly spread again for snail control. The pH value of this area is adjusted, and after disinfection with potassium permanganate, four native plants, namely Scirpus validus, Zizania latifolia, Oenanthe javanica, and Epilobium hirsutum, are planted and rationally matched. The specific practices are as follows:

[0182] (1)Disinfection of native plants. After the four native plants, Scirpus validus, Zizania latifolia, Oenanthe javanica, and Epilobium hirsutum, are separately transplanted and retrieved, potassium permanganate is mixed with water at a ratio of 2‰, and they are soaked in a large basin for about one hour. After disinfection, they can be planted.

[0183] (2)Planting of native plants. The four native plants, Scirpus validus, Zizania latifolia, Oenanthe javanica, and Epilobium hirsutum, are planted in the test plot at a spacing of 50 cm * 50 cm, and each plant is cross-planted.

[0184] For Figure 5 the areas with lighter colors in are marked as secondary prevention and control areas. The plants are not replaced, and only killing and disinfection are carried out. Through the above treatments, good prevention and control effects have been achieved.

[0185] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of the present invention claimed is defined by the appended claims and their equivalents.

Claims

1. An artificial wetland invasive snail monitoring method based on spatio-temporal autocorrelation analysis, characterized in that, Including: S1. Import a topographic map into the computer system and rasterize the monitoring area; Use an image acquisition device to collect image data of the artificial wetland containing snails, and at the same time collect environmental variable data, water quality microbial data, and soil physical and chemical property data; S2. Screen representative image samples from the images under the same environmental conditions in the same grid as effective image data, and use correlation analysis to remove environmental variables with collinearity exceeding a preset value in the environmental variable data to obtain effective environmental variable data; S3. Use image recognition to process the effective image data and extract the image features of snails; use the principal component analysis algorithm to extract fusion features from the effective image data, effective environmental variable data, water quality microbial data, and soil physical and chemical property data to obtain a comprehensive feature vector; S4. Construct an invasive snail monitoring model based on cellular automata, regard each grid as a cell, and define the state, neighborhood relationship, and transition rules of the cell; the cell state is defined as: if the number of snails in a cell is 0, its state is no snails present; if the number is between 1 and 100, it is determined that there are snails present and the number is small; if it exceeds this range, there are snails present and the number is large; use the obtained comprehensive feature vector as input data and use the ANN algorithm in machine learning to optimize the model parameters; S5. Input the real-time collected image data, environmental variable data, water quality microbial data, and soil physical and chemical property data into the invasive snail monitoring model to evaluate the risk degree of invasive snails in the artificial wetland; S6. Collect snail distribution data within a certain period of time, where the snail distribution data includes the snail state information of each cell at different time points and the corresponding environmental variable data; use spatio-temporal autocorrelation analysis to determine the spatial distribution characteristics and temporal change trends of invasive snails in the artificial wetland, and combine with a geographic information system to display the spatio-temporal distribution dynamic changes of snails and predict the future invasion diffusion path and scope; In step S2, when screening representative image samples, calculate the clarity score by accumulating the square root of the sum of the squares of the differences between the pixel values of each position and adjacent positions in the image; Select images with clarity scores in the top 30% and complete snail information as effective image data, and delete other images; Remove environmental variables with collinearity exceeding a preset value in the environmental variable data, and the preset value is 0.8; use the Pearson correlation coefficient to calculate the correlation between the observed values of two environmental variables; This coefficient is obtained by dividing the sum of the products of the differences between the variable observed values and their means by the product of the square roots of the sums of the squares of their respective differences.

2. The artificial wetland invasive snail monitoring method based on spatio-temporal autocorrelation analysis according to claim 1, wherein, In step S3, use a convolutional neural network to extract snail image features; And: The convolutional layer calculates the feature map by multiplying the convolutional kernel with the corresponding position of the input image and adding a bias term; the pooling layer uses max pooling, that is, selects the maximum pixel value in the neighborhood.

3. The method for monitoring invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis according to claim 2, wherein, In step S3, when using the principal component analysis algorithm to fuse features, first standardize the data, subtract the variable mean from the original data and then divide by the standard deviation; Then calculate the covariance matrix, which is the average of the sum of the products of the sample data and the differences from their means; Perform eigenvalue decomposition on the covariance matrix, select the first several eigenvectors with larger eigenvalues to form a transformation matrix, and multiply the standardized data matrix by the transformation matrix to obtain the comprehensive eigenvector.

4. The method for monitoring invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis according to claim 1, characterized in that The neighborhood adopts the Moore neighborhood, and the neighborhood of each cell includes the 8 surrounding cells; The cell state transition probability is jointly determined by a random term, a global transition probability, a constraint condition, and a neighborhood function; the random term is based on natural logarithm-related operations, and the global transition probability is the ratio of the number of cells with state transitions to the total number of cells; The constraint condition is based on the environmental suitability index of the cell position Judge: ； Among them, , , , are the currently monitored temperature, humidity, pH value, and dissolved oxygen values respectively. , , , , , , , are the minimum and maximum values of the corresponding environmental factors suitable for the survival of snails respectively. , , , are the weight coefficients determined according to the importance of the influence of each environmental factor on snails, and + + + = 1; when the index is greater than 0.5, the constraint condition is 1, otherwise it is 0. Moreover, the neighborhood function is the sum of the cell state values in the neighborhood divided by the number of cells in the neighborhood; When using the ANN algorithm to optimize the model parameters, the backpropagation algorithm is used to calculate the gradient, which is calculated by the sum of the products of the partial derivatives between the output of the output layer neurons, the input of the hidden layer neurons, and the weights, and the weights are continuously adjusted to minimize the loss function.

5. The artificial wetland invasive snail monitoring method based on spatio-temporal autocorrelation analysis according to claim 4, characterized in that, In step S5, when evaluating the risk level of invasive snails in the constructed wetland, a risk assessment index system is constructed, including the snail density index, the invasion and diffusion speed index, and the environmental suitability index. Among them: The snail density index is the ratio of the number of snails per unit area to the area of the monitoring area; The invasion and diffusion speed index is the ratio of the change in the number of snails over a certain period of time to the change in time; And: The weights of each index are determined by the analytic hierarchy process, and the final risk level is the sum of the products of the scores of each index and their weights.

6. The method for monitoring invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis according to claim 1, wherein In step S6, when using spatio-temporal autocorrelation analysis to study the spatio-temporal distribution characteristics and change trends of invasive snails in the constructed wetland, the following global Moran's I index and local Moran's I index are constructed; The calculation formula of the global Moran's I index is: ； where n is the number of samples, is the sum of the elements of the spatial weight matrix, is the spatial weight matrix, , is the snail distribution data, is the mean; The calculation formula of the local Moran's I index is: ； Among them , ; where n is the number of samples, is the standardized value of the current cell i, is the standardized value of the current cell j; is the standard deviation, used to standardize the original data to obtain and ; through the original data is converted into a standard normal distribution, making the spatial autocorrelation of different variables comparable; When combining the geographic information system to display the dynamic changes in the spatio-temporal distribution of snails, the calculated Moran's I index results are visualized on the map, and different autocorrelation degrees are represented by different colors and symbols; When predicting the future invasion and diffusion path and scope, a prediction method based on the cellular automata model is adopted. According to the current state and transition rules of the cells, the state of the cells at future times is iteratively calculated to predict the diffusion of snails.

7. The method for monitoring invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis according to claim 3, characterized in that When using the principal component analysis algorithm for feature extraction, the method for determining the number of principal components is: first calculate the cumulative contribution rate, and the formula is: ； Among them, is the i-th eigenvalue, is the number of original variables; k represents the first k principal components selected from the principal component analysis results; Select So that the cumulative contribution rate reaches more than 85%.

8. The artificial wetland invasive snail monitoring method based on spatio-temporal autocorrelation analysis according to claim 4, characterized in that In the invasive snail monitoring model based on cellular automata, the global transition probability is obtained by statistical analysis of historical data, and the formula is: 。 9. The method for monitoring invasive snails in constructed wetlands based on spatio-temporal autocorrelation analysis according to claim 5, wherein When using the analytic hierarchy process to determine the weights of the risk assessment indicators, a judgment matrix is constructed: ； Among them, represents the importance degree of the \(i\)-th index relative to the \(j\)-th index, which is determined by expert scoring; then calculate the maximum eigenvalue of the judgment matrix and the corresponding eigenvector \(W\), and the formula is: ； Finally, the eigenvector is normalized to obtain the weights of each index .

Images