Ecological problem evaluation system and method, storage medium and electronic equipment

Abstract

The invention relates to the technical field of ecological environment assessment, in particular to an ecological problem assessment system, an ecological problem assessment method, a storage medium and electronic equipment. The ecological problem evaluation system comprises a data acquisition module; the data fusion module comprises a feature extraction model and a data fusion model, the feature extraction model is used for extracting a dynamic rule of spatial features changing along with time in the environment data to obtain space-time fusion feature data, and the data fusion model is used for compressing and reconstructing the space-time fusion feature data to obtain environment fusion feature data; the data analysis module comprises an ecological problem identification unit and a regional range identification unit, the ecological problem identification unit is used for identifying an ecological problem according to the environment fusion feature data, and the regional range identification unit is used for identifying a regional range corresponding to the ecological problem according to the environment data; and a report generation module. The ecological problem assessment system provided by the invention can efficiently, accurately and adaptively perform ecological problem assessment.

Description

Technical Field

[0001] The present disclosure relates to the technical field of ecological environment assessment, and in particular to an ecological problem assessment system, an ecological problem assessment method, a storage medium and an electronic device. Background Art

[0002] With the intensification of global climate change and the continuous expansion of human activities, ecological problems in geographical regions are becoming increasingly severe, and the degradation of the ecological environment has become an important factor restricting sustainable development.

[0003] At present, the identification and assessment of ecological and environmental problems rely on traditional manual monitoring and simplified statistical models, but these methods are often inadequate when faced with large-scale and complex ecological data, and lack efficient and intelligent analysis methods. In particular, cross-analysis of multiple dimensions of problems in ecological regions, such as the degradation of forests, grasslands, wetlands, soil erosion, water pollution, mining environments, and biodiversity, requires a huge workload and high labor costs, and traditional monitoring and assessment methods are difficult to complete efficiently and accurately.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] The purpose of the present disclosure is to provide an ecological problem assessment system, an ecological problem assessment method, a storage medium and an electronic device, aiming to solve the problem of low accuracy in ecological problem assessment.

[0006] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0007] According to one aspect of the present disclosure, an ecological problem assessment system is provided, comprising: a data acquisition module, for collecting environmental data from at least one data source; a data fusion module, comprising a feature extraction model and a data fusion model, the feature extraction model being used to extract the dynamic law of spatial features in the environmental data changing over time to obtain spatiotemporal fusion feature data, the data fusion model being used to compress and reconstruct the spatiotemporal fusion feature data to obtain environmental fusion feature data; a data analysis module, comprising an ecological problem identification unit and a regional scope identification unit, the ecological problem identification unit being used to identify ecological problems based on the environmental fusion feature data, the regional scope identification unit being used to identify the regional scope corresponding to the ecological problem based on the environmental data; and a report generation module being used to generate an ecological problem analysis report based on the data analysis process and data analysis results in the data analysis module.

[0008] Optionally, the feature extraction model is a deep space-time convolutional neural network, which includes: a convolution layer, which is used to extract the spatial features of the environmental data through convolution operations; a time convolution layer, which is used to process the convolved environmental data and realize long-term memory and forgetting of information through an LSTM gating mechanism to capture the temporal dependencies in the environmental data to obtain temporal features; and a fully connected layer, which is used to fuse the spatial features and the temporal features to generate the spatiotemporal fusion feature data.

[0009] Optionally, the data fusion model is an adaptive deep autoencoder network, and the data fusion model includes: an encoder, used to map the spatiotemporal fusion feature data to a low-dimensional space to extract high-order features of the spatiotemporal fusion feature data; a weight adjustment module, used to adaptively adjust the weights of each of the data sources using a self-attention mechanism; and a decoder, used to restore the high-order features in the low-dimensional space to the original dimension according to the adjusted weights to obtain the environmental fusion feature data.

[0010] Optionally, the ecological problem identification unit includes one or more of a first identification network, a second identification network, a third identification network and a fourth identification network; wherein the first identification network is used to perform image segmentation using a combined model of U-Net and CNN to identify ecological problems and obtain a first identification result; the second identification network is used to perform multivariate classification of ecological problems using a combined model of XGBoost and random forest to obtain a second identification result; the third identification network is used to identify the trend of water quality changes and the diffusion pattern of pollution sources using a combined model of TCN and CNN to obtain water pollution identification results in ecological problems; the fourth identification network is used to evaluate the correlation between the distribution area and habitat of organisms using a combined model of GNN and U-Net to obtain biodiversity identification results in ecological problems.

[0011] Optionally, the ecological problem identification unit further includes: an ecological problem integration unit, configured to integrate the first identification result, the second identification result, the water pollution identification result and the biodiversity identification result to obtain an ecological problem list.

[0012] Optionally, the area range identification unit is configured to: use a spatial clustering analysis method to perform spatial division and area positioning to identify the area range corresponding to the ecological problem.

[0013] Optionally, the environmental data includes one or more of remote sensing image data, drone aerial survey data, and ground monitoring data.

[0014] According to a second aspect of the present disclosure, there is provided an ecological problem assessment method, comprising: collecting environmental data from at least one data source; extracting the dynamic laws of spatial characteristics in the environmental data changing over time to obtain spatiotemporal fusion feature data, and compressing and reconstructing the spatiotemporal fusion feature data to obtain environmental fusion feature data; identifying ecological problems based on the environmental fusion feature data, and identifying the regional scope corresponding to the ecological problems based on the environmental data; and generating an ecological problem analysis report based on the regional scope corresponding to the ecological problems.

[0015] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the ecological problem assessment method in the above-mentioned embodiment is implemented.

[0016] According to the fourth aspect of an embodiment of the present disclosure, an electronic device is provided, characterized in that it includes: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the ecological problem assessment method as in the above-mentioned embodiment.

[0017] The exemplary embodiments of the present disclosure may have some or all of the following beneficial effects: In the technical solutions provided in some embodiments of the present disclosure, environmental data collected from multiple data sources are fused, fused environmental fusion feature data is obtained by spatiotemporal fusion feature data, and ecological problem assessment is performed based on the fused environmental fusion feature data. On the one hand, data features of multiple data sources are fused to perform environmental evaluation, thereby ensuring data diversity and improving the accuracy of environmental assessment; on the other hand, the introduction of feature extraction models and data fusion models based on deep learning models can improve the quality and effectiveness of data, meet the relevant needs of ecological problem assessment, and further improve the accuracy of environmental assessment.

[0018] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present disclosure, and together with the specification, are used to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings: Figure 1 A schematic diagram schematically shows the composition of an ecological problem assessment system in an exemplary embodiment of the present disclosure; Figure 2 A schematic diagram schematically illustrates a flow chart of a data preprocessing and fusion method in an exemplary embodiment of the present disclosure; Figure 3 A schematic diagram schematically illustrates a flow chart of an ecological problem assessment method in an exemplary embodiment of the present disclosure; Figure 4 The structure diagram of a computer system of an electronic device in an exemplary embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0020] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be more comprehensive and complete and will fully convey the concept of the example embodiments to those skilled in the art.

[0021] In addition, the described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the present disclosure.

[0022] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0023] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0024] With the intensification of global climate change and the continuous expansion of human activities, ecological problems in geographical regions are becoming increasingly severe, and the degradation of the ecological environment has become an important factor restricting sustainable development.

[0025] At present, the identification and assessment of ecological and environmental problems rely on traditional manual monitoring and simplified statistical models, but these methods are often inadequate when faced with large-scale and complex ecological data, and lack efficient and intelligent analysis methods. In particular, cross-analysis of multiple dimensions of problems in ecological regions, such as the degradation of forests, grasslands, wetlands, soil erosion, water pollution, mining environments, and biodiversity, requires a huge workload and high labor costs, and traditional monitoring and assessment methods are difficult to complete efficiently and accurately.

[0026] At present, although a variety of methods such as remote sensing technology, ground monitoring data and climate models have been widely used in the identification and assessment of ecological problems, their accuracy and timeliness still cannot meet the actual needs when faced with complex spatiotemporal characteristics and nonlinear, multi-factor intertwined ecological problems. Especially in the process of processing multi-dimensional ecological data, traditional methods cannot effectively deal with problems such as noise interference, data sparsity, and information loss, resulting in limited accuracy and reliability of identification results. In addition, existing technologies also have major deficiencies in processing the correlation, dynamic change characteristics and regional differences of spatiotemporal data, and cannot provide comprehensive and accurate ecological assessments. It is urgent to use a new generation of intelligent analysis technology to improve identification accuracy and efficiency.

[0027] Therefore, this disclosure proposes a smart geographical regional ecological problem identification and assessment method, which aims to solve the defects of existing methods in dealing with complex ecological problems. And by introducing a deep learning model, it realizes the intelligent processing and analysis of a large amount of ecological data, aiming to achieve efficient, accurate and adaptive ecological assessment and prediction. By integrating multi-source data, deep learning technology and complex mathematical calculations, it can analyze and predict ecological problems in geographical areas in real time, and help formulate scientific ecological restoration and protection strategies.

[0028] The implementation details of the technical solution of the embodiment of the present disclosure are elaborated in detail below.

[0029] Figure 1 The following is a schematic diagram showing the composition of an ecological problem assessment system in an exemplary embodiment of the present disclosure. Figure 1 As shown, the ecological problem assessment system 100 includes: A data collection module 101 is used to collect environmental data from at least one data source; The data fusion module 102 includes a feature extraction model and a data fusion model, wherein the feature extraction model is used to extract the dynamic law of the spatial features in the environmental data changing over time to obtain spatiotemporal fusion feature data, and the data fusion model is used to compress and reconstruct the spatiotemporal fusion feature data to obtain environmental fusion feature data; The data analysis module 103 includes an ecological problem identification unit and a regional range identification unit, wherein the ecological problem identification unit is used to identify ecological problems according to the environmental fusion feature data, and the regional range identification unit is used to identify the regional range corresponding to the ecological problem according to the environmental data; The report generation module 104 is used to generate an ecological problem analysis report according to the data analysis process and data analysis results in the data analysis module.

[0030] In the ecological problem assessment system provided by some embodiments of the present disclosure, environmental data collected from multiple data sources are fused through a data acquisition module and a data fusion module, and fused environmental fusion feature data are obtained through spatiotemporal fusion feature data. Ecological problem assessment is performed based on the fused environmental fusion feature data. On the one hand, data features of multiple data sources are fused to perform environmental evaluation, thereby ensuring data diversity and improving the accuracy of environmental assessment. On the other hand, the introduction of a feature extraction model and a data fusion model based on a deep learning model can improve the quality and effectiveness of the data, meet the relevant needs of ecological problem assessment, and further improve the accuracy of environmental assessment.

[0031] Below, each module of the ecological problem assessment system in this example implementation will be described in more detail with reference to the accompanying drawings and embodiments.

[0032] In one embodiment of the present disclosure, the data collection module 101 is used to collect environmental data from at least one data source.

[0033] Specifically, it is necessary to first collect environmental data from multiple data sources, such as remote sensing image data, drone aerial survey data, and ground monitoring data. Ground monitoring data includes terrain data, climate data, soil data, and water quality data. Compared with a single data source, multiple data sources can improve the accuracy of ecological problem assessment.

[0034] In one embodiment of the present disclosure, the data acquisition module 101 may also include a preprocessing unit for preprocessing the data set after the data is collected to improve the data quality and make it suitable for subsequent fusion processing. The preprocessing mainly includes: data cleaning: removing duplicate, missing or erroneous data points to ensure the integrity of the data; denoising: using relevant digital signal processing methods to remove data noise; data normalization: normalizing data of different units and scales to ensure that the data completes standardized comparison and analysis.

[0035] In one embodiment of the present disclosure, the data fusion module 102 includes a feature extraction model and a data fusion model. The feature extraction model is used to extract the dynamic law of the change of spatial features in the environmental data over time to obtain spatiotemporal fusion feature data, and the data fusion model is used to compress and reconstruct the spatiotemporal fusion feature data to obtain environmental fusion feature data.

[0036] Among them, the feature extraction model is a deep space-time convolutional neural network (DST-CNN), and the structure of DST-CNN includes: a convolution layer, which is used to extract the spatial features of the environmental data through convolution operations; a temporal convolution layer, which is used to process the environmental data after convolution, and realize the long-term memory and forgetting of information through the LSTM gating mechanism, so as to capture the temporal dependencies in the environmental data to obtain temporal features; a fully connected layer, which is used to fuse the spatial features and the temporal features to generate the spatiotemporal fusion feature data.

[0037] Specifically, first define the convolution kernel in the convolution layer K Extract local spatial features and transform environmental data X Input, the spatial feature results of the convolution layer output are:

[0038] In the time series convolution layer, the convolved environment data is processed by the set spatial convolution and LSTM gating mechanism to obtain the time series features, specifically:

[0039]

[0040] Where: is the hidden state at the current moment, i.e., the time series feature; is the current state; W and U are the weight matrices, b is the offset.

[0041] Finally, the fully connected layer performs weighted summation of the spatial and temporal features of the environmental data to obtain the spatiotemporal fusion feature data. X’ .

[0042] After the spatiotemporal fusion feature data extracted based on the DST-CNN model, the ADAE-Net model can be used to extract high-order features from the data through nonlinear mapping, perform efficient data fusion, and realize adaptive dynamic adjustment of weights between different data sources, thereby achieving more accurate feature extraction and data fusion.

[0043] The data fusion model is an adaptive deep autoencoder network (ADAE-Net), and the structure of ADAE-Net includes: an encoder, which is used to map the spatiotemporal fusion feature data to a low-dimensional space to extract high-order features of the spatiotemporal fusion feature data; a weight adjustment module, which is used to adaptively adjust the weights of each of the data sources using a self-attention mechanism; and a decoder, which is used to restore the high-order features in the low-dimensional space to the original dimension according to the adjusted weights to obtain the environmental fusion feature data.

[0044] Among them, it is assumed that the input data spatiotemporal fusion feature data is X’ , the output data reconstructed by the encoder and decoder is , the autoencoder objective function is:

[0045] The encoder is a variational autoencoder (VAE), and its variational lower bound (ELBO) formula is:

[0046] In the above formula: q ( z | x ) is the variational distribution; p ( x | z ) is the reconstruction distribution; is the prior distribution; for KL The divergence of .

[0047] For the self-attention mechanism, weighting is performed by calculating the self-attention score, specifically:

[0048] In the above formula: Q is the query matrix; K is the key matrix; V is the value matrix; The dimension of the key.

[0049] Figure 2 The following is a schematic diagram showing a flow chart of a data preprocessing and fusion method in an exemplary embodiment of the present disclosure. Figure 2 As shown in the figure, data preprocessing and fusion mainly include: Data preprocessing includes data cleaning, denoising and data normalization. Data cleaning: remove duplicate, missing or erroneous data points to ensure data integrity; denoising: use relevant digital signal processing methods to remove data noise; data normalization: normalize data of different units and scales to ensure standardized comparison and analysis of data.

[0050] Multi-source data fusion, including DST-CNN and ADAE-Net. DST-CNN is used to process spatial data (such as remote sensing images, terrain data) and time series data (such as climate change, water quality fluctuations, etc.), to extract the spatial characteristics of multi-source data and the dynamic laws of changes over time, so as to better capture the correlation in geographic spatial data. ADAE-Net is used to extract high-order features from data through nonlinear mapping, perform efficient data fusion, and realize adaptive dynamic adjustment of weights between different data sources, so as to achieve more accurate feature extraction and data fusion.

[0051] In a specific embodiment of the present disclosure, the data analysis module 103 includes an ecological problem identification unit and an area range identification unit. The ecological problem identification unit is used to identify ecological problems based on the environmental fusion feature data, and the area range identification unit is used to identify the area range corresponding to the ecological problem based on the environmental data.

[0052] Specifically, the ecological problem identification unit includes one or more of a first identification network, a second identification network, a third identification network, and a fourth identification network.

[0053] For the first recognition network, a combined model of U-Net and CNN is used to perform image segmentation to identify ecological problems and obtain the first recognition result.

[0054] Among them, CNN (Convolutional Neural Network) is a deep learning model that is particularly suitable for processing image data. The working principle of CNN is mainly based on convolution operation and feature extraction. In the convolution layer, the convolution kernel slides on the input image, and extracts local features in the image through weighted summation and bias operation of the local area. These features are nonlinearly transformed through activation functions to enhance the expressive power of the model. Then, the pooling layer downsamples the output of the convolution layer to reduce the dimension and calculation amount of the data while retaining important features. Finally, the fully connected layer further processes and classifies the output of the pooling layer.

[0055] The architecture of U-Net is in the shape of a "U", and is mainly composed of three parts: the encoder on the left, the decoder on the right, and the skip connections in the middle. It has unique advantages in processing biomedical image segmentation, such as random transformations such as elastic deformation, which effectively addresses the problem of scarce labeled data; avoids redundant calculations of sliding windows, improves positioning accuracy and segmentation efficiency; realizes the fusion of features of different scales, and improves the segmentation effect of small objects and edge areas.

[0056] Traditional image segmentation methods have problems such as insufficient local feature extraction and insufficient long-term dependency processing when processing remote sensing data. To this end, this paper combines the deep models of U-Net and CNN, and introduces the Depthwise Separable Convolution and Spatial Pyramid Pooling (SPP) modules on the basis of the original U-Net to improve the network's processing capabilities on high-resolution images while reducing computational complexity. In addition, combined with the self-attention mechanism, the model can automatically focus on important areas in the image, thereby improving the recognition and analysis capabilities of traditional ecological problems.

[0057] Among them, the convolution operation of the depth-separable convolution is specifically:

[0058] Where: is a depth-wise separable convolution operation, W is the convolution kernel, b is the offset.

[0059] The self-attention mechanism is as follows:

[0060] Where: , , are the weight matrices for query, key, and value, respectively.

[0061] For the second recognition network, the combined model of XGBoost and random forest is used to perform multivariate classification of ecological problems to obtain the second recognition results.

[0062] Among them, XGBoost is an optimized distributed gradient boosting library designed to implement efficient, flexible and portable machine learning algorithms. XGBoost is an improvement on the gradient boosting algorithm. It uses Newton's method to solve the extreme value of the loss function, expands the loss function Taylor to the second order, and adds a regularization term to the loss function. Its training objective function consists of two parts: the first part is the gradient boosting algorithm loss, and the second part is the regularization term. XGBoost builds a powerful ensemble model by combining multiple weak learners (usually decision trees). Its core principle involves the optimization of the loss function and the construction of the tree model. During the model training process, XGBoost uses a greedy algorithm to gradually build a tree model and optimizes the model performance by iteratively reducing the objective function.

[0063] Random Forest (RF) is an ensemble learning method that combines the classification results of multiple decision trees to improve the overall classification performance. RF is an extended variant of bagging. It uses decision trees as base learners and makes predictions by building multiple decision trees and combining their output results. When building each decision tree, random forest introduces two randomnesses: random selection of samples and random selection of features. The introduction of these two randomnesses makes random forests less likely to fall into overfitting and has good noise resistance.

[0064] Specifically, a method combining XGBoost and RF models is used to construct a multivariate classification of biological problems (XGBoost-RF). This model can effectively classify ecological problems, especially when the data features are complex or unevenly distributed. At the same time, the model has the ability to automatically optimize the classification tasks of multiple ecological problems, reduce conflicts in multi-task learning, and adaptively adjust the severity weights of different problems, thereby improving the model's ability to identify important ecological problems.

[0065] The loss function of the existing XGBoost model is optimized, and the optimized mathematical model is:

[0066] Where: is the loss function; is the regularization term; and are the regularization parameters respectively; Parameters for plants; function for each plant.

[0067] The weight of each plant in the weighted random forest can be expressed as:

[0068] Where: is the set of data points belonging to class k; For labels; is the number of data points of class k.

[0069] For the third identification network, the combined model of TCN and CNN is used to identify the trend of water quality changes and the diffusion pattern of pollution sources, so as to obtain the identification results of water pollution in ecological problems; Among them, TCN (Temporal Convolutional Network) is a convolutional neural network with causal convolution and dilated convolution characteristics, which is specially used to process time series data. Causal convolution ensures that the output at time t depends only on the input at time t and before, thus ensuring the causality of the model. Dilated convolution increases the receptive field of the convolution kernel by introducing holes, allowing the model to capture more distant temporal dependencies. CNN has been introduced before, so I will not go into details here.

[0070] In response to water pollution in ecological issues, this paper uses the causal convolution method in TCN to process time series data, and combines it with the CNN model to capture long-term dependencies while enhancing the extraction of local features in water quality data, identifying the trend of water quality changes and the diffusion pattern of pollution sources. In addition, reinforcement learning is combined with model training to enable the model to continuously optimize prediction accuracy through feedback.

[0071] First, the convolution process of the TCN model is optimized, and the optimized formula is:

[0072] Multi-scale convolution can be expressed as follows:

[0073]

[0074] Where: W 1. W 2 is the convolution kernel of different scales; b 1. b 2 is the bias term.

[0075] Then, the LSTM model is introduced to realize the recursive transfer of states through time steps. The formula is:

[0076] Where: is the current state; is the input value.

[0077] For the fourth identification network, the combined model of GNN and U-Net is used to evaluate the correlation between the distribution area and habitat of organisms to obtain biodiversity identification results in ecological problems.

[0078] GNN (Graph Neural Network) is a deep learning model specifically designed for processing graph-structured data. The core idea of ​​GNN is to update the representation of nodes through information propagation (or message passing) between nodes. In GNN, each node has a feature vector that represents the attribute or state of the node. The model updates the representation of the current node by iteratively aggregating the feature information of neighboring nodes. This information propagation mechanism enables GNN to capture the global structural information and local feature information of graph data. U-Net has been introduced before, so I will not go into details here.

[0079] The distribution of biodiversity and habitat assessment usually involve a large amount of graph structure data, especially the relationships between species and the spatial structure of habitats. Therefore, this disclosure introduces spectral convolution and spatial convolution in GCN to improve the learning ability of the model, especially the expression ability on high-dimensional and complex graph data. At the same time, the introduction of self-supervised learning enables the model to learn effectively without labeled data, further improving the generalization ability of the model, so as to more accurately evaluate the biodiversity of the ecosystem.

[0080] The propagation process in a neural network can be expressed as:

[0081] Where: For the k Characteristics of layer nodes; is the set of neighbor nodes; is the normalization coefficient; For the k The weight matrix of the layer.

[0082] The combination of upsampling and downsampling in the U-Net architecture can be expressed as:

[0083] Where: W is the convolution kernel; b is the bias term.

[0084] In one embodiment of the present disclosure, the ecological problem identification unit further includes: an ecological problem integration unit, which is used to integrate the first identification result, the second identification result, the water pollution identification result and the biodiversity identification result to obtain an ecological problem list.

[0085] Specifically, different ecological problem analysis methods can be used to analyze problems in various dimensions. After the analysis is completed, different analysis results will be obtained. The different analysis results can be deduplicated and supplemented, and then integrated into a list of ecological problems.

[0086] In one embodiment of the present disclosure, the area range identification unit is configured to: use a spatial clustering analysis method to perform spatial division and area positioning to identify the area range corresponding to the ecological problem.

[0087] Specifically, spatial clustering analysis methods can be used to spatially divide and regionally locate ecological problems, and accurately identify the distribution range of ecological problems. Spatial clustering methods include K-means and DBSCAN.

[0088] K-means is a classic unsupervised learning algorithm used to cluster data. The K-means algorithm treats the data set as an n-dimensional space with n features and attempts to divide the data points into K clusters by minimizing the sum of squared errors within the cluster. Each cluster contains a centroid, which is the mean of all data points in the cluster and is used to represent the center position of the cluster. The basic idea of ​​the algorithm is to iteratively assign each data point to the cluster with the nearest centroid and recalculate the centroid of each cluster until the centroid no longer changes or the maximum number of iterations is reached.

[0089] DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is a density-based clustering algorithm that can examine the connectivity between samples from the perspective of sample density and continuously expand clusters based on connectable samples.

[0090] The clustering process can be expressed as follows:

[0091] Where: is the i-th category; For Category The mean of is the sample point.

[0092] In one embodiment of the present disclosure, the report generation module 104 is used to generate an ecological problem analysis report according to the data analysis process and data analysis results in the data analysis module.

[0093] Specifically, natural language processing technology and generative models can be used to provide detailed analysis results. The report content will cover: data analysis process, images, distribution maps, trend forecasts, etc., and visual output for decision makers' reference.

[0094] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0095] Figure 3 The following is a schematic diagram showing a flow chart of an ecological problem assessment method in an exemplary embodiment of the present disclosure. Figure 3 As shown in the figure, the ecological problem assessment method mainly includes the following processes: Start by identifying and assessing ecological issues; Multi-source data preprocessing and fusion, that is, using deep learning models to realize the preprocessing of various types of data such as remote sensing images, drone data, terrain data, climate data, soil data and water quality data, as well as the fusion of data of different time-space dimensions.

[0096] Intelligent identification of ecological problems includes four parts: the first is intelligent identification of traditional ecological problems; the second is multi-dimensional classification of ecological problems; the third is time series identification of water pollution ecological problems; and the fourth is biodiversity distribution and habitat assessment.

[0097] Comprehensive assessment and spatial analysis are carried out, using spatial clustering and analysis methods (such as K-means and DBSCAN) to spatially divide and regionally locate ecological problems, and accurately identify the distribution range of ecological problems.

[0098] Finally, the report is automatically generated and visualized.

[0099] The present disclosure also provides an ecological problem assessment method, which includes the following steps: collecting environmental data from at least one data source; extracting the dynamic laws of spatial characteristics in the environmental data changing over time to obtain spatiotemporal fusion feature data, and compressing and reconstructing the spatiotemporal fusion feature data to obtain environmental fusion feature data; identifying ecological problems based on the environmental fusion feature data, and identifying the regional scope corresponding to the ecological problems based on the environmental data; and generating an ecological problem analysis report based on the regional scope corresponding to the ecological problems.

[0100] Each step in the above ecological problem assessment method has been described in detail in the corresponding system module, so it will not be repeated here. The ecological problem assessment method disclosed in this disclosure can be widely used in the fields of government, environmental protection agencies, scientific research institutions and corporate ecological management, providing comprehensive intelligent decision-making support for environmental protection, resource management, ecological restoration and land use planning, and promoting the efficient implementation of ecological protection work.

[0101] Based on the above method, the present invention has a versatile and intelligent ecological problem analysis and identification function. By inputting a variety of raw data such as remote sensing data, drone data, terrain data, etc., it can automatically and accurately identify ecological problems, predict trends, and optimize decisions. At the same time, it can automatically generate detailed ecological problem reports and provide intelligent decision-making suggestions, greatly improving the efficiency and accuracy of ecological management and reducing manual intervention and time costs.

[0102] In an exemplary embodiment of the present disclosure, a storage medium capable of implementing the above method is also provided. It can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a mobile phone. However, the program product of the present disclosure is not limited to this. In this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0103] In an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided. Figure 4 The structure diagram of a computer system of an electronic device in an exemplary embodiment of the present disclosure is schematically shown.

[0104] It should be noted that Figure 4 The computer system 400 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0105] like Figure 4 As shown, the computer system 400 includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 402 or the program loaded from the storage part 408 to the random access memory (RAM) 403. Various programs and data required for system operation are also stored in the RAM 403. The CPU 401, the ROM 402, and the RAM 403 are connected to each other through a bus 404. The input / output (I / O) interface 405 is also connected to the bus 404.

[0106] The following components are connected to the I / O interface 405: an input section 406 including a keyboard, a mouse, etc.; an output section 407 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the I / O interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 410 as needed so that a computer program read therefrom is installed into the storage section 408 as needed.

[0107] In particular, according to an embodiment of the present disclosure, the process described below with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 409, and / or installed from a removable medium 411. When the computer program is executed by the central processing unit (CPU) 401, various functions defined in the system of the present disclosure are executed.

[0108] It should be noted that the computer-readable medium shown in the embodiment of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0109] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0110] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, and the units described may also be arranged in a processor. The names of these units do not, in some cases, constitute limitations on the units themselves.

[0111] As another aspect, the present disclosure further provides a computer-readable medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the method described in the above embodiment.

[0112] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0113] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0114] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure.

[0115] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. An ecological problem assessment system, characterized in that: include: A data collection module, used to collect environmental data from at least one data source; A data fusion module, comprising a feature extraction model and a data fusion model, wherein the feature extraction model is used to extract the dynamic law of the spatial features in the environmental data changing over time to obtain spatiotemporal fusion feature data, and the data fusion model is used to compress and reconstruct the spatiotemporal fusion feature data to obtain environmental fusion feature data; A data analysis module, comprising an ecological problem identification unit and a regional range identification unit, wherein the ecological problem identification unit is used to identify ecological problems according to the environmental fusion feature data, and the regional range identification unit is used to identify the regional range corresponding to the ecological problem according to the environmental data; The report generation module is used to generate an ecological problem analysis report based on the data analysis process and data analysis results in the data analysis module.

2. The ecological problem assessment system according to claim 1, characterized in that: The feature extraction model is a deep spatial-temporal convolutional neural network, and the feature extraction model includes: A convolution layer, used for extracting spatial features of the environmental data through convolution operation; The temporal convolution layer is used to process the convolved environmental data and realize the long-term memory and forgetting of information through the LSTM gating mechanism, so as to capture the temporal dependency in the environmental data and obtain the temporal features; The fully connected layer is used to fuse the spatial features and the temporal features to generate the spatiotemporal fusion feature data.

3. The ecological problem assessment system according to claim 1, characterized in that: The data fusion model is an adaptive deep autoencoder network, and the data fusion model includes: An encoder, used for mapping the spatiotemporal fusion feature data to a low-dimensional space to extract high-order features of the spatiotemporal fusion feature data; A weight adjustment module, used to adaptively adjust the weight of each of the data sources using a self-attention mechanism; A decoder is used to restore the high-order features in the low-dimensional space to the original dimension according to the adjusted weights to obtain the environment fusion feature data.

4. The ecological problem assessment system according to claim 1, characterized in that: The ecological problem identification unit includes one or more of a first identification network, a second identification network, a third identification network and a fourth identification network; wherein, The first recognition network is used to perform image segmentation using a combined model of U-Net and CNN to identify ecological problems and obtain a first recognition result; The second recognition network is used to perform multivariate classification of ecological problems using a combination model of XGBoost and random forest to obtain a second recognition result; The third identification network is used to identify the trend of water quality changes and the diffusion pattern of pollution sources by using a combined model of TCN and CNN, so as to obtain water pollution identification results in ecological problems; The fourth identification network is used to evaluate the correlation between the distribution area and habitat of organisms by using a combined model of GNN and U-Net, so as to obtain biodiversity identification results in ecological problems.

5. The ecological problem assessment system according to claim 4, characterized in that: The ecological problem identification unit also includes: The ecological problem integration unit is used to integrate the first identification result, the second identification result, the water pollution identification result and the biodiversity identification result to obtain an ecological problem list.

6. The ecological problem assessment system according to claim 1, characterized in that: The area range identification unit is configured to: Spatial cluster analysis methods were used to perform spatial division and regional positioning in order to identify the regional scope corresponding to the ecological problems.

7. The ecological problem assessment system according to claim 1, characterized in that: The environmental data includes one or more of remote sensing image data, drone aerial survey data, and ground monitoring data.

8. A method for assessing ecological problems, characterized in that: include: Collecting environmental data from at least one data source; Extracting the dynamic law of the spatial features in the environmental data changing over time to obtain spatiotemporal fusion feature data, and compressing and reconstructing the spatiotemporal fusion feature data to obtain environmental fusion feature data; Identifying ecological problems based on the environmental fusion feature data, and identifying the area scope corresponding to the ecological problems based on the environmental data; Generate an ecological problem analysis report based on the regional scope corresponding to the ecological problem.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the ecological problem assessment method according to claim 8 is implemented.

10. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, when the one or more computer programs are executed by the one or more processors, enables the one or more processors to implement the ecological problem assessment method as described in claim 8.

Images