Ecological environment parameter monitoring method, device and system

Abstract

The invention relates to the technical field of data processing, in particular to an ecological environment parameter monitoring method, device and system, and the method comprises the steps: obtaining an ecological environment parameter sequence of a plurality of initial dimensions, and obtaining the dynamic change characteristics of each position according to the change trend of each position in the ecological environment parameter sequence; on the basis of the association of the dynamic change characteristics between the initial dimensions, the dominant effect of each initial dimension on the ecological environment is obtained by combining the number of the associated initial dimensions associated with each initial dimension, so that the ecological contribution degree of each position in each initial dimension is obtained by combining the dynamic change characteristics of each position in each initial dimension; according to the method, the ecological contribution degree serves as a weight, an initial covariance matrix in principal component analysis is weighted, correction of the covariance matrix and dimension reduction of an initial dimension are achieved, data redundancy is reduced, the complexity of ecological environment parameter data is remarkably reduced while main characteristics of data are kept, and the data processing efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a method, device and system for monitoring ecological environment parameters. Background Art

[0002] In the process of monitoring ecological environment parameters, existing monitoring methods mostly have problems such as insufficient real-time performance, limited coverage, and low data processing efficiency, and it is difficult to meet the multi-dimensional monitoring requirements of complex ecological environments. Currently, it mainly relies on sensor networks, remote sensing technologies, and Internet of Things architectures. By collecting key parameters such as temperature, humidity, and atmospheric components in the environment, combined with big data processing, the accuracy and efficiency of ecological environment monitoring are improved, providing a scientific basis for environmental protection decision-making.

[0003] In the prior art, ecological environment monitoring usually involves real-time monitoring of multiple ecological environment parameters. Ecological environment parameters involve a large number of multi-dimensional variables, and the amount of data is extremely large, resulting in high data processing difficulty. Summary of the Invention

[0004] In order to solve the technical problem that the large amount of data of ecological environment parameters leads to high data processing difficulty, the purpose of the present invention is to provide a method, device and system for monitoring ecological environment parameters, and the specific technical solutions adopted are as follows: In the first aspect of the present invention, a method for monitoring ecological environment parameters is provided, including: Obtain ecological environment parameter sequences of multiple initial dimensions, and obtain dynamic change characteristics of each position according to the change trends of each position in the ecological environment parameter sequences; Based on the difference between the correlation degree of the overall data of each initial dimension and other initial dimensions and the correlation degree of the local data, combined with the number of associated initial dimensions associated with each initial dimension, obtain the leading role of each initial dimension in the ecological environment; the correlation degree is obtained from the dynamic change characteristics; According to the leading role of each initial dimension in the ecological environment and the dynamic change characteristics of each position in each initial dimension, obtain the ecological contribution degree of each position in each initial dimension; Use the ecological contribution degree of each position in each initial dimension as a weight to weight the initial covariance matrix in principal component analysis and reduce the dimension of the initial dimension.

[0005] In an exemplary embodiment, obtaining the dynamic change characteristics of each position according to the change trends of each position in the ecological environment parameter sequences includes: Segment the first ecological environment parameter sequence to obtain multiple data segments; the first ecological environment parameter sequence is the ecological environment parameter sequence of the first initial dimension, and the first initial dimension is any one of the initial dimensions; Based on the change situation of the data in the first data segment, obtain the long-term change situation of the first position; the first data segment is any data segment in the first ecological environment parameter sequence; the first position is the position of any parameter in the first data segment; Based on the difference between the actual value and the fitted value of the first position, obtain the short-term volatility of the first position; Fuse the long-term change situation and the short-term volatility of the first position to obtain the dynamic change characteristics of the first position.

[0006] In an exemplary embodiment, based on the change situation of the data in the first data segment, obtaining the long-term change situation of the first position includes: Perform linear fitting on the first data segment to obtain the first slope of the fitted line of the first data segment; Taking the first position as the demarcation point, perform linear fitting on the data before the first position and the data after the first position in the first data segment respectively to obtain the second slope of the fitted line of the previous data and the third slope of the fitted line of the subsequent data; According to the first slope and the slope difference, obtain the long-term change situation of the first position; the long-term change situation is directly proportional to the slope difference and directly proportional to the first slope; the slope difference is the difference between the second slope and the third slope.

[0007] In an exemplary embodiment, based on the difference between the overall data correlation degree and the local data correlation degree of each initial dimension and other initial dimensions, and combining the number of associated initial dimensions associated with each initial dimension, obtaining the leading role of each initial dimension in the ecological environment includes: Obtain the first correlation degree between the first initial dimension and the second initial dimension, where the first correlation degree is the correlation degree between the first initial dimension and the overall sequence of dynamic change characteristics of the second initial dimension; the overall sequence of dynamic change characteristics includes the dynamic change characteristics of all positions corresponding to the initial dimension; the first initial dimension is any initial dimension, and the second initial dimension is any other initial dimension different from the first initial dimension; Obtain the difference in the trend deviation degree between two adjacent positions within the preset neighborhood range of the first position, and then obtain the overall situation of the difference in the trend deviation degree within the preset neighborhood range of the first position to obtain the short-term volatility of the first position.

[0008] In an exemplary embodiment, based on the difference between the overall data correlation degree and the local data correlation degree of each initial dimension and other initial dimensions, and combining the number of associated initial dimensions associated with each initial dimension, obtaining the leading role of each initial dimension in the ecological environment includes: Obtain the first correlation degree between the first initial dimension and the second initial dimension, where the first correlation degree is the correlation degree between the first initial dimension and the overall sequence of dynamic change characteristics of the second initial dimension; the overall sequence of dynamic change characteristics includes the dynamic change characteristics of all positions corresponding to the initial dimension; the first initial dimension is any initial dimension, and the second initial dimension is any other initial dimension different from the first initial dimension; Obtain the first local sequence of dynamic change features corresponding to the preset neighborhood range of the second position for the first initial dimension, and the second local sequence of dynamic change features corresponding to the second initial dimension; the second position is the position of any parameter in the first initial dimension; Obtain the second correlation degree between the first initial dimension and the second initial dimension at the second position, where the second correlation degree is the correlation degree between the first local sequence of dynamic change features and the second local sequence of dynamic change features; Obtain the true variability of the second position of the first initial dimension based on the difference between the first correlation degree and the second correlation degree; Obtain the leading role of the first initial dimension in the ecological environment based on the true variability and the number of associated initial dimensions associated with the first initial dimension.

[0009] In an exemplary embodiment, obtaining the true variability of the second position of the first initial dimension based on the difference between the first correlation degree and the second correlation degree includes: Obtain the first correlation degree between the first initial dimension and each other initial dimension, and the difference between the first correlation degree and the second correlation degree between the first initial dimension and each other initial dimension at the second position, and calculate the mean value to obtain the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions; Obtain the true variability of the second position of the first initial dimension based on the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions; the true variability is inversely proportional to the overall correlation difference.

[0010] In an exemplary embodiment, the process of obtaining the associated initial dimensions associated with the first initial dimension includes: taking the initial dimensions corresponding to the first correlation degrees greater than a preset threshold among the first correlation degrees between the first initial dimension and each other initial dimension as the associated initial dimensions associated with the first initial dimension; Obtaining the leading role of the first initial dimension in the ecological environment based on the true variability and the number of associated initial dimensions associated with the first initial dimension includes: Calculate the mean value of the true variabilities of each position of the first initial dimension to obtain the overall true variability of the first initial dimension; Obtain the leading role of the first initial dimension in the ecological environment based on the overall true variability of the first initial dimension and the number of associated initial dimensions associated with the first initial dimension; the leading role is directly proportional to the overall true variability and directly proportional to the number of associated initial dimensions.

[0011] In an exemplary embodiment, weighting the initial covariance matrix in the principal component analysis includes: Construct an ecological contribution degree matrix according to the ecological contribution degrees of each position in each initial dimension, and obtain the transposed matrix of the ecological contribution degree matrix; Calculate the product of the ecological contribution degree matrix, the transposed matrix and the initial covariance matrix to obtain the weighted covariance matrix.

[0012] In the second aspect of the present invention, an ecological environment parameter monitoring device is provided. The ecological environment parameter monitoring device includes a unit for executing the above-mentioned ecological environment parameter monitoring method.

[0013] In the third aspect of the present invention, an ecological environment parameter monitoring system is provided, including: a memory and a processor; the memory is connected to the processor; the memory is used to store program instructions; the processor is used to implement the above-mentioned ecological environment parameter monitoring method when the program instructions are executed.

[0014] The present invention has the following beneficial effects: By analyzing the change trends of each position in the ecological environment parameter sequences of each initial dimension, the dynamic change characteristics of each position are obtained, which helps to capture the time trends and fluctuation laws of the ecological environment parameters, and can more comprehensively understand the dynamic evolution of the ecosystem; then analyze the differences between the overall data correlation degrees and the local data correlation degrees of each initial dimension and other initial dimensions, and combine the number of associated initial dimensions associated with each initial dimension to obtain the dominant role of each initial dimension in the ecological environment, and further analyze the interaction relationships between the ecological environment parameters to provide support for dimensionality reduction; According to the dominant role of each initial dimension in the ecological environment and the dynamic change characteristics of each position in each initial dimension, the ecological contribution degrees of each position in each initial dimension are obtained. The ecological contribution degrees highlight the core parameter indicators. Finally, the principal initial covariance matrix is weighted according to the ecological contribution degrees of each position in each initial dimension to realize the correction of the covariance matrix, reduce data redundancy through dimensionality reduction, significantly reduce the complexity of the ecological environment parameter data while maintaining the main features of the data, and improve the data processing efficiency. Description of the Drawings

[0015] Figure 1 is a flowchart of an ecological environment parameter monitoring method provided by an embodiment of the present invention; Figure 2 is a flowchart for obtaining the dynamic change characteristics provided by an embodiment of the present invention; Figure 3 is a flowchart for obtaining the long-term change situation provided by an embodiment of the present invention; Figure 4 is a flowchart for obtaining the short-term volatility provided by an embodiment of the present invention; Figure 5 is the overall acquisition flowchart of the leading role provided by an embodiment of the present invention; Figure 6 is the acquisition flowchart of the true variability provided by an embodiment of the present invention; Figure 7 is the calculation flowchart of the leading role provided by an embodiment of the present invention; Figure 8 is the flowchart for weighting the initial covariance matrix in the principal component analysis provided by an embodiment of the present invention. Detailed implementation manners

[0016] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following combines the accompanying drawings and preferred embodiments to detail the specific implementation manners, structures, features and effects of the present invention. In the following description, different "an embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The data information collected in this application has been obtained with full consent and authorization, and the collection, use and processing of relevant information need to comply with the relevant laws, regulations and standards of the relevant regions.

[0018] The application scenario of an ecological environment parameter monitoring method provided in this embodiment is as follows: A variety of types of sensors are set in the monitoring area to detect various types of ecological environment parameters. The specific types and quantities of sensors are set according to actual needs. For example: air quality sensors (such as PM2.5 sensors, PM10 sensors), temperature sensors, humidity sensors, wind speed sensors, harmful gas sensors (such as CO sensors, CO 2 sensors, SO 2 sensors, NO 2 sensors, O 3 sensors), water body dissolved oxygen sensors, soil moisture content sensors, etc. Representative sites are determined as monitoring points in the monitoring area, and the above various types of sensors are set at the monitoring points. Each sensor is signal-connected to a data processing device through a data transmission module, which can be a wireless signal connection or a wired signal connection. In an exemplary embodiment, the data transmission module is a wireless communication module and is powered by solar energy to ensure reliable signal transmission.

[0019] Various sensors collect ecological environment parameters at the same sampling frequency, and various sensors collect data synchronously. Then, at a certain sampling moment, ecological environment parameters of each type can be obtained. The specific value of the sampling frequency is set according to actual needs, for example: once an hour. The data collection period is set according to actual needs, for example, one month. Therefore, the data collection period includes multiple sampling moments. Then, for any sensor, multiple ecological environment parameters can be obtained during the data collection period.

[0020] As Figure 1 shown, this embodiment provides a method for monitoring ecological environment parameters, including: Step 1: Obtain ecological environment parameter sequences of multiple initial dimensions, and obtain dynamic change characteristics of each position according to the change trends of each position in the ecological environment parameter sequences. Step 2: Based on the differences between the overall data correlation degree and the local data correlation degree of each initial dimension and other initial dimensions, and combined with the number of associated initial dimensions associated with each initial dimension, obtain the leading role of each initial dimension in the ecological environment. Step 3: According to the leading role of each initial dimension in the ecological environment and the dynamic change characteristics of each position in each initial dimension, obtain the ecological contribution degree of each position in each initial dimension. Step 4: Use the ecological contribution degree of each position in each initial dimension as a weight to weight the initial covariance matrix in the principal component analysis and reduce the dimension of the initial dimension.

[0021] The following specifically explains each step.

[0022] Step 1: Obtain ecological environment parameter sequences of multiple initial dimensions, and obtain dynamic change characteristics of each position according to the change trends of each position in the ecological environment parameter sequences.

[0023] Set various types of sensors as each initial dimension, and the data information of as many initial dimensions as there are types of sensors will be obtained.

[0024] Obtain the ecological environment parameter sequences of each initial dimension, that is, the ecological environment parameter sequences detected by various sensors. For the ecological environment parameter sequence of any initial dimension, the ecological environment parameter sequence is composed of ecological environment parameters at each sampling moment in time series. Moreover, the number of ecological environment parameters included in the ecological environment parameter sequences of each initial dimension is the same. Since the ecological environment parameter sequence is a data sequence composed of multiple data in time series, each position in the ecological environment parameter sequence represents each data point, and the ecological environment parameters at the same position in the ecological environment parameter sequences of each initial dimension are data at the same sampling moment.

[0025] There are numerous ecological environment parameters that need to be monitored in the ecological environment. During the process of dimensionality reduction, when compressing high-dimensional data into a few key feature dimensions, it is necessary to maintain the core description ability of the ecological environment changes. By analyzing the dynamic change characteristics of ecological environment parameters, identifying the time series trends and fluctuation rules of the parameters, we can more comprehensively understand the dynamic evolution of the ecosystem in the monitoring area and the performance of the parameters in the ecological environment, capture the environmental change patterns, and timely identify potential ecological problems. Therefore, according to the change trends at each position in the ecological environment parameter sequence, the dynamic change characteristics at each position are obtained.

[0026] In an exemplary embodiment, as Figure 2 shown, a specific acquisition process of the dynamic change characteristics is given, including: Step 1-1: Segment the first ecological environment parameter sequence to obtain multiple data segments.

[0027] Since the processing process for the ecological environment parameter sequences of each initial dimension is the same, for the convenience of description, the first initial dimension is set as any one of the initial dimensions as follows. The ecological environment parameter sequence of the first initial dimension is set as the first ecological environment parameter sequence.

[0028] Segment the first ecological environment parameter sequence to obtain multiple data segments. In an exemplary embodiment, in order to improve the accuracy and reliability of data processing, a specific segmentation method is given as follows: The changes of ecological environment parameters have natural periodic characteristics. Affected by the natural laws of the earth, through the periodic analysis of ecological environment parameters, the natural fluctuation patterns on a long time scale can be identified, providing a basis for the long-term monitoring and prediction of the ecological environment.

[0029] Therefore, the fast Fourier transform is used to obtain the periodic characteristics of the first ecological environment parameter sequence. Specifically: use the fast Fourier transform to process the first ecological environment parameter sequence, then calculate the amplitude of each frequency, identify the frequency with the largest amplitude, and the corresponding period is the main period of the first ecological environment parameter sequence.

[0030] According to the obtained main period, segment the first ecological environment parameter sequence to obtain multiple data segments, and the length of each data segment is the data length corresponding to the main period.

[0031] Step 1-2: Based on the change situation of the data in the first data segment, obtain the long-term change situation at the first position.

[0032] The long-term change trend of ecological environment parameters, that is, the long-term change situation, can reflect the evolution of the ecosystem, predict future environmental changes, and reveal the regular changes of ecological environment parameters (such as temperature, precipitation, pollutant concentration, etc.) over time, so as to identify potential environmental problems. For example, the long-term downward trend of water resources may indicate the coming of a regional water crisis. Therefore, it is necessary to analyze the long-term change trend of ecological environment parameters.

[0033] For the sake of convenience of description, it is assumed that the first data segment is any data segment in the first ecological environment parameter sequence, and the first position is the position of any parameter in the first data segment, that is, the first position is any data point in the first data segment.

[0034] In an exemplary embodiment, as Figure 3 shown, the following gives a specific acquisition process of the long-term change situation: Step 1-2-1: Perform linear fitting on the first data segment to obtain the first slope of the fitting line of the first data segment.

[0035] Perform linear fitting on the first data segment, use the most common least squares method to draw the fitting line of the first data segment, and then obtain the linear slope of the fitting line of the first data segment, which is defined as the first slope. The first slope characterizes the overall change trend of the first data segment.

[0036] Step 1-2-2: Taking the first position as the demarcation point, perform linear fitting on the data before and after the first position in the first data segment respectively to obtain the second slope of the fitting line of the data before the first position and the third slope of the fitting line of the data after the first position.

[0037] In the first data segment, taking the first position as the demarcation point, the data in the first data segment can be divided into two parts of data. Among them, the data before the first position in the first data segment is used as the first segmented data, and the data after the first position in the first data segment is used as the second segmented data.

[0038] Use the least squares method to perform linear fitting on the first segmented data to obtain the linear slope of the fitting line, which is defined as the second slope; use the least squares method to perform linear fitting on the second segmented data to obtain the linear slope of the fitting line, which is defined as the third slope. The second slope characterizes the change trend of the data before the first position in the first data segment, and the third slope characterizes the change trend of the data after the first position in the first data segment.

[0039] It should be understood that if the first position is the first position in the first data segment, since there is no data before it, then the second slope is set to 0 at this time. If the first position is the second position in the first data segment, since there is only one data before it, then the second slope is also set to 0 at this time. Similarly, if the first position is the last position or the penultimate position in the first data segment, then the third slope is set to 0 at this time.

[0040] Step 1-2-3: Obtain the long-term change situation of the first position according to the first slope and the slope difference.

[0041] Obtain the difference between the second slope and the third slope, which is defined as the slope difference. In an exemplary embodiment, the difference between the second slope and the third slope is specifically the absolute value of the difference between the second slope and the third slope.

[0042] Since the larger the absolute value of the difference between the second slope and the third slope, the greater the difference in the change trends of the data before and after the first position in the first data segment, and the greater the change in the long-term change trend of the first position, that is, the greater the long-term change situation. Since the first slope represents the overall change trend of the first data segment, the greater the overall change trend, the greater the long-term change situation of the first position. Therefore, the long-term change situation is proportional to the slope difference and proportional to the first slope.

[0043] In an exemplary embodiment, obtain the absolute value of the first slope, and then calculate the product of the absolute value of the first slope and the slope difference, and use this product as the long-term change situation of the first position.

[0044] By using the above method, obtain the long-term change situation of each position in the first data segment, and then obtain the long-term change situation of each position in each data segment.

[0045] Step 1-3: Obtain the short-term volatility of the first position based on the difference between the actual value and the fitted value of the first position.

[0046] Analyzing the short-term fluctuation law of ecological environment parameters can reflect the changes in the ecological environment, predict and mitigate sudden environmental events, and usually reflects the direct impact of weather changes, climate phenomena or local natural events. For example, the short-term fluctuations of parameters such as temperature, humidity or air pollutant concentration can reveal instantaneous weather changes, temporary fluctuations in air quality, or the impacts brought by short-term climate phenomena (such as heatwaves, precipitation, etc.).

[0047] To make the short-term change characteristics of ecological environment parameters more obvious, obtain the short-term volatility of each position based on the difference between the actual value and the fitted value of each position.

[0048] In an exemplary embodiment, such as Figure 4As shown below is a specific process for obtaining short-term volatility: Step 1-3-1: Obtain the trend deviation degree of the first position according to the difference between the actual value and the fitted value of the first position.

[0049] Since the first position is the position of any parameter in the first data segment, and the first data segment is any data segment in the first ecological environment parameter sequence, it can be understood that the first position is the position of any parameter in the first ecological environment parameter sequence.

[0050] Obtain the actual value of the first position. The actual value of the first position is the numerical value of the ecological environment parameter at the first position in the first ecological environment parameter sequence, which is the actual numerical value detected by the sensor.

[0051] Obtain the fitted value of the first position. The fitted value is the numerical value on the fitted straight line corresponding to the first position after linearly fitting the first data segment.

[0052] Then, there may be a certain difference between the actual value and the fitted value of the first position. The greater the difference, the greater the deviation between the fitted value and the actual value, that is, the greater the trend deviation degree of the first position. Therefore, obtain the trend deviation degree of the first position according to the difference between the actual value and the fitted value of the first position.

[0053] In an exemplary embodiment, calculate the absolute value of the difference between the actual value and the fitted value of the first position, and this absolute value of the difference is defined as the trend deviation degree of the first position. Thus, obtain the trend deviation degrees of each position in the first ecological environment parameter sequence.

[0054] Step 1-3-2: Obtain the difference in the trend deviation degrees of two adjacent positions within the preset neighborhood range of the first position, and further obtain the overall situation of the trend deviation degree difference within the preset neighborhood range of the first position, so as to obtain the short-term volatility of the first position.

[0055] Preset a neighborhood range, and the length of this preset neighborhood range is set according to the actual situation, such as 21 data points. Obtain the preset neighborhood ranges of each position. In an exemplary embodiment, each position is at the center of the corresponding preset neighborhood range. Characterize the short-term volatility of each position through the overall situation of the trend deviation degrees within the preset neighborhood range corresponding to each position.

[0056] It should be understood that for a number of starting positions in the first ecological environment parameter sequence, since the number of positions before them is very small and it is impossible to ensure that all data within the preset neighborhood range can be obtained centered on them, then all positions from them to the first position will be used as the first half of their preset neighborhood range, and the second half will still be obtained in the original way. For example: if the preset neighborhood range is 21, then for the 5th position, normally, data of 10 positions before and after it need to be obtained. Since there are only 4 positions before the 5th position, then obtain the first 4 positions before the 5th position, the 5th position, and the 10 positions after the 5th position as the preset neighborhood range of the 5th position. Another example: for the 1st position, normally, data of 10 positions before and after it need to be obtained. Since there are no other positions before the 1st position, then obtain the 1st position and the 10 positions after the 1st position as the preset neighborhood range of the 1st position.

[0057] Similarly, for a number of end positions in the first ecological environment parameter sequence, since the number of positions after them is very small and it is impossible to ensure that all data within the preset neighborhood range can be obtained centered on them, then all positions from them to the last position will be used as the second half of their preset neighborhood range, and the first half will still be obtained in the original way. For example: if the preset neighborhood range is 21, then for the penultimate 5th position, normally, data of 10 positions before and after it need to be obtained. Since there are only 4 positions after the penultimate 5th position, then obtain the last 4 positions after the penultimate 5th position, the penultimate 5th position, and the first 10 positions before the penultimate 5th position as the preset neighborhood range of the penultimate 5th position. Another example: for the last 1st position, normally, data of 10 positions before and after it need to be obtained. Since there are no other positions after the last 1st position, then obtain the last 1st position and the first 10 positions before the last 1st position as the preset neighborhood range of the last 1st position.

[0058] Obtain the difference in trend deviation degrees between two adjacent positions within the preset neighborhood range of the first position. In an exemplary embodiment, the difference in trend deviation degrees is the absolute value of the difference in trend deviation degrees between two adjacent positions within the preset neighborhood range of the first position. It should be understood that the trend deviation degree of the first position is also involved in the calculation of the difference in trend deviation degrees between two adjacent positions within its preset neighborhood range.

[0059] Then, obtain the overall situation of the difference in trend deviation degrees within the preset neighborhood range of the first position. In an exemplary embodiment, the overall situation is the average value of the difference in trend deviation degrees, and the short-term volatility of the first position is obtained.

[0060] In an exemplary embodiment, the calculation formula for short-term volatility is as follows: ; wherein, represents the short-term volatility at the th position; represents the number of positions within the preset neighborhood range of the th position; represents the trend deviation degree of the th position within the preset neighborhood range of the th position; represents the trend deviation degree of the th position within the preset neighborhood range of the th position.

[0061] represents the absolute value of the difference between the trend deviation degree of the th position and the trend deviation degree of the th position within the preset neighborhood range of the th position.

[0062] represents the average value of the absolute values of the differences in trend deviation degrees within the preset neighborhood range of the th position, representing the short-term volatility at the th position. The larger this value, the greater the short-term volatility.

[0063] Steps 1 - 4: Integrate the long-term change situation and short-term volatility of the first position to obtain the dynamic change characteristics of the first position.

[0064] Comprehensively consider the long-term change situation and short-term volatility of the first position to obtain the dynamic change characteristics of the first position. In an exemplary embodiment, normalize the long-term change situation of the first position and normalize the short-term volatility of the first position, and then calculate the product of the normalized long-term change situation of the first position and the normalized short-term volatility of the first position, and use this product as the dynamic change characteristics of the first position.

[0065] The normalization in this embodiment, as well as the norm normalization function, can be specifically set according to the actual situation. For example: The maximum-minimum normalization method can be used, or the following common method can be used: , represents the processing object, and exp represents the exponential function with the natural constant e as the base.

[0066] Step 2: Based on the difference between the degree of association of each initial dimension with the overall data of other initial dimensions and the degree of association with the local data, and combined with the number of associated initial dimensions associated with each initial dimension, obtain the dominant role of each initial dimension in the ecological environment.

[0067] Based on the results of the dynamic characteristic analysis, further explore the internal correlations among the ecological environment parameters, such as the impact of meteorological parameters on pollutant diffusion or the interactions among water quality indicators. Incorporate the dynamic change characteristics into the overall ecological environment system and construct an interaction model among the parameters to identify the key parameters that have the most significant impact on ecological environment monitoring.

[0068] In an exemplary embodiment, as Figure 5 shown, the process of obtaining the dominant role of each initial dimension in the ecological environment includes: Step 2-1: Obtain the first degree of correlation between the first initial dimension and the second initial dimension.

[0069] For ease of explanation, it is assumed that the second initial dimension is any other initial dimension different from the first initial dimension.

[0070] Obtain the overall sequence of the dynamic change characteristics of the first initial dimension, defined as the first overall sequence of dynamic change characteristics. The first overall sequence of dynamic change characteristics includes the dynamic change characteristics of all positions of the first initial dimension and is obtained by arranging the dynamic change characteristics of all positions of the first initial dimension according to the sorting of each position.

[0071] Obtain the overall sequence of the dynamic change characteristics of the second initial dimension, defined as the second overall sequence of dynamic change characteristics. The second overall sequence of dynamic change characteristics includes the dynamic change characteristics of all positions of the second initial dimension and is obtained by arranging the dynamic change characteristics of all positions of the second initial dimension according to the sorting of each position.

[0072] Obtain the first degree of correlation between the first initial dimension and the second initial dimension, that is, the degree of correlation between the first overall sequence of dynamic change characteristics and the second overall sequence of dynamic change characteristics.

[0073] The degree of correlation is the correlation situation between the two overall sequences of dynamic change characteristics and can be calculated through calculation methods such as correlation coefficients and similarities. In an exemplary embodiment, a specific calculation process is provided as follows: First, obtain the consistency of the dynamic change laws of the first overall sequence of dynamic change characteristics and the second overall sequence of dynamic change characteristics: In the monitoring and analysis of ecological environment parameters, the dynamic change laws between parameters and their consistency help to reveal the internal correlations in the ecological environment system. The changes in the dynamic change characteristics of parameters reflect their fluctuations over time and the connections among them, thereby extracting the correlations among ecological environment parameters: ; wherein, represents the dynamic change consistency between the th initial dimension and the th initial dimension; represents the The overall sequence of dynamic change characteristics of the initial dimension Denote the Overall sequence of dynamic change characteristics of the initial dimension Denote the DTW distance (Dynamic Time Warping distance) between the overall sequence of dynamic change characteristics of the th initial dimension and the overall sequence of dynamic change characteristics of the

[0074] Then obtain the parameter correlation, that is, ecological interaction, between the first overall sequence of dynamic change characteristics and the second overall sequence of dynamic change characteristics, which is also the degree of correlation: measure the correlation strength between parameters by analyzing the consistency of the dynamic change laws of parameters: ; Among them, Denote the Ecological interaction, that is, the degree of correlation, between the th initial dimension and the Denote the Pearson correlation coefficient between the overall sequence of dynamic change characteristics of the th initial dimension and the overall sequence of dynamic change characteristics of the Denote the normalization function

[0075] In other embodiments, the Pearson correlation coefficient or cosine similarity of the two overall sequences of dynamic change characteristics can also be directly normalized as the degree of correlation between the two overall sequences of dynamic change characteristics

[0076] Through the above process, obtain the first degree of correlation between the first initial dimension and each of the other initial dimensions

[0077] Step 2-2: Obtain the first local sequence of dynamic change characteristics corresponding to the first initial dimension and the second local sequence of dynamic change characteristics corresponding to the second initial dimension within the preset neighborhood range at the second position

[0078] The preset second position is the position of any parameter in the first initial dimension, and its essential meaning is the same as that of the first position in the above text, both representing the position of any parameter in the first initial dimension

[0079] Since there are multiple positions within the preset neighborhood range of the second position, and each position has dynamic change characteristics, therefore, obtaining multiple dynamic change characteristics of the preset neighborhood range of the second position in the first initial dimension, that is, the preset neighborhood range of the second position corresponds to multiple dynamic change characteristics in the first initial dimension. These dynamic change characteristics form a first local sequence of dynamic change characteristics according to the sorting of each position. Since the positions in each initial dimension correspond to each other, therefore, obtaining multiple dynamic change characteristics of the preset neighborhood range of the second position corresponding to the second initial dimension, that is, multiple dynamic change characteristics of the corresponding positions of the preset neighborhood range of the second position in the second initial dimension. These dynamic change characteristics form a second local sequence of dynamic change characteristics according to the sorting of each position.

[0080] Step 2-3: Obtain the second correlation degree between the first initial dimension and the second initial dimension at the second position.

[0081] Obtaining the second correlation degree between the first initial dimension and the second initial dimension at the second position, that is, obtaining the correlation degree between the first local sequence of dynamic change characteristics and the second local sequence of dynamic change characteristics. The obtaining process of the second correlation degree also adopts the obtaining method of the first correlation degree in Step 2-1, and will not be repeated here.

[0082] Through the above process, the second correlation degree between the first initial dimension and each of the other initial dimensions at the second position can be obtained, and then the second correlation degree between the first initial dimension and each of the other initial dimensions at each position can be obtained.

[0083] Step 2-4: Obtain the true variability of the second position of the first initial dimension according to the difference between the first correlation degree and the second correlation degree.

[0084] In ecological environment monitoring, systematically analyzing various ecological environment parameters to clarify which ecological environment parameters play a leading role in environmental changes and which ecological environment parameter changes can significantly affect the state of the ecosystem not only helps to optimize the monitoring system but also improves the accuracy of decision-making. On this basis, dimensionality reduction processing is carried out to reduce data redundancy and improve data processing efficiency.

[0085] In the process of dimensionality reduction of ecological environment parameters, it is first necessary to ensure the accuracy of ecological environment parameters. When there are abnormal fluctuations in the data values of ecological environment parameters in a certain initial dimension, the reasons may be equipment failures, interference from external factors leading to changes in the real environment, etc. In the case of equipment failures or data collection errors, it is unlikely that monitoring devices across the entire range of the ecological environment will malfunction simultaneously, which is accidental. When the environment actually changes, due to the interrelationships among ecological environment parameters, such as an increase in temperature leading to a decrease in air flow and less diffusion of pollutants, resulting in an increase in the concentration of a certain gas. Therefore, by observing the correlation relationships among ecological environment parameters, it can be determined whether the ecological environment parameters represent real ecological environment changes. Then, based on the differences between the first correlation degree and the second correlation degree, the true variability of the second position in the first initial dimension is obtained. In an exemplary embodiment, as Figure 6 shown, a specific process for obtaining the true variability is given: Step 2-4-1: Obtain the first correlation degree between the first initial dimension and each other initial dimension, as well as the difference in the second correlation degree between the first initial dimension and each other initial dimension at the second position, and calculate the average value to obtain the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions.

[0086] For any other initial dimension, obtain the first correlation degree between the first initial dimension and that other initial dimension, as well as the second correlation degree between the first initial dimension and that other initial dimension at the second position. Then, obtain the difference between the first correlation degree between the first initial dimension and that other initial dimension and the second correlation degree between the first initial dimension and that other initial dimension at the second position, to obtain the correlation difference of the second position of the first initial dimension with respect to that other initial dimension. Then, calculate the average value of the correlation differences of the second position of the first initial dimension with respect to each other initial dimension to obtain the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions.

[0087] Step 2-4-2: Based on the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions, obtain the true variability of the second position of the first initial dimension.

[0088] The smaller the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions, the less the relationship of the first initial dimension has changed, indicating that the first initial dimension can better represent real ecological environment changes and the true variability is greater. Therefore, based on the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions, the true variability of the second position of the first initial dimension is obtained, and the true variability is inversely proportional to the overall correlation difference.

[0089] In an exemplary embodiment, the calculation formula for the true variability is given as follows: ; where represents the true variability of the th position of the th initial dimension; represents the number of other initial dimensions except the th initial dimension; represents the degree of association between the th initial dimension and the th initial dimension at the th position; here is the general level used as a reference in this formula; represents the overall association difference of the th position of the th initial dimension with respect to all other initial dimensions. The smaller it is, the more the th initial dimension represents the real ecological environment change, and the greater the true variability of the th initial dimension.

[0090] Step 2-5: Obtain the dominant role of the first initial dimension in the ecological environment based on the true variability and the number of associated initial dimensions associated with the first initial dimension.

[0091] By analyzing the impact of ecological environment parameters on the ecological environment, key ecological environment parameters important for ecological environment monitoring can be obtained. Each ecological environment parameter in ecological environment monitoring represents the change of the ecological environment. If there is a strong association between a certain ecological environment and multiple other ecological environment parameters, it indicates that this ecological environment parameter is an important evaluation index in the monitoring of the ecological environment and is representative. At the same time, the overall level of the true variability of the ecological environment parameter represents the real environmental change and is of great significance for evaluating the ecological environment.

[0092] First, it is necessary to obtain the associated initial dimensions associated with the first initial dimension. In an exemplary embodiment, a preset threshold is set. This preset threshold is used to compare with the first degree of association between the first initial dimension and each other initial dimension to obtain the associated initial dimensions having an association relationship with the first initial dimension. The numerical range of this preset threshold is 0-1, and the specific value is set according to actual needs, such as 0.7.

[0093] Compare the first-degree of association between the first initial dimension and each of the other initial dimensions with a preset threshold, obtain the first-degree of association greater than the preset threshold from them, then obtain the initial dimension corresponding to the first-degree of association greater than the preset threshold. There is a strong correlation between the obtained initial dimensions and the first dimension. Take the obtained initial dimensions as the associated initial dimensions associated with the first initial dimension. Then, obtain the number of associated initial dimensions. Therefore, the more the number of associated initial dimensions, the more representative the corresponding initial dimension is, and it is of great significance for evaluating the ecological environment.

[0094] According to the true variability and the number of associated initial dimensions associated with the first initial dimension, obtain the dominant role of the first initial dimension in the ecological environment. In an exemplary embodiment, as Figure 7 shown, the following gives a specific obtaining process of the dominant role: Step 2-5-1: Calculate the mean of the true variability of each position of the first initial dimension to obtain the overall true variability of the first initial dimension; Step 2-5-2: According to the overall true variability of the first initial dimension and the number of associated initial dimensions associated with the first initial dimension, obtain the dominant role of the first initial dimension in the ecological environment.

[0095] The overall true variability of the first initial dimension reflects the overall situation of the true variability of each position of the first initial dimension. The greater the overall true variability, the more dominant the first initial dimension is in the ecological environment. Therefore, the dominant role is proportional to the overall true variability and proportional to the number of associated initial dimensions.

[0096] In an exemplary embodiment, the following gives a calculation formula for the dominant role: ; Among them, represents the dominant role of the th initial dimension in the ecological environment; represents the number of associated initial dimensions of the th initial dimension. The larger this value is, the more strongly associated initial dimensions there are with the th initial dimension in the ecological environment. Then, through the change of the ecological environment parameters of the th initial dimension, the possible changes of more ecological environment parameters in the ecological environment can be inferred, and the dominant role in the ecological environment is greater; represents the number of ecological environment parameters in the ecological environment parameter sequence of the th initial dimension.

[0097] represents the The overall true variability of the initial dimensions. The larger the value, the more reference significance the th initial dimension has for monitoring the ecological environment.

[0098] Step 3: According to the dominant role of each initial dimension in the ecological environment and the dynamic change characteristics of each position in each initial dimension, obtain the ecological contribution degree of each position in each initial dimension.

[0099] Analyzing the contribution degree of each initial dimension to the ecological environment can highlight the core indicators in the ecological environment and reduce data redundancy. If the short-term dynamic change characteristics of the initial dimension are strong, it indicates that the ecological environment may change and needs to be focused on for analysis. Therefore, according to the dominant role of each initial dimension in the ecological environment and the dynamic change characteristics of each position in each initial dimension, obtain the ecological contribution degree of each position in each initial dimension. In an exemplary embodiment, the calculation formula of the ecological contribution degree is given as follows: ; Wherein, represents the ecological contribution degree of the th position in the th initial dimension, represents the dynamic change characteristic of the th position in the th initial dimension.

[0100] By adopting the above process, the ecological contribution degree of each position in each initial dimension is obtained.

[0101] Step 4: Use the ecological contribution degree of each position in each initial dimension as a weight to weight the initial covariance matrix in the principal component analysis and reduce the dimension of the initial dimension.

[0102] To reduce the complexity and redundancy of ecological environment parameters while retaining the information most representative of changes in the ecological environment, data reduction is performed on the ecological environment parameters. Ecological environment monitoring usually involves a large number of ecological environment parameters, and there are often strong correlations and redundant information among these ecological environment parameters. Directly processing these high-dimensional data will increase the computational difficulty and storage pressure. Dimensionality reduction can extract the most important features in the ecological environment parameters. Principal component analysis can effectively discover the principal components in the data and reduce the dimension of the feature space by projecting the high-dimensional data into a new low-dimensional space, thereby removing redundancy and improving data processing efficiency. Therefore, using the ecological contribution degree of each position in each initial dimension as a weight and introducing it into the principal component analysis to weight the initial covariance matrix in the principal component analysis and reduce the dimension of the initial dimension can preferentially retain the information of ecological environment parameters that have a greater impact on the ecological environment during the dimensionality reduction process.

[0103] In an exemplary embodiment, such asFigure 8 As shown in Figure 8 , a specific implementation process for weighting the initial covariance matrix in principal component analysis is given: Step 4-1: Construct an ecological contribution degree matrix according to the ecological contribution degrees of each position in each initial dimension, and obtain the transpose matrix of the ecological contribution degree matrix.

[0104] Construct an ecological environment parameter matrix according to each position in each initial dimension. For example, each row of the matrix represents each initial dimension, and each column represents each position. Then, construct an ecological contribution degree matrix according to the ecological contribution degrees of each position in each initial dimension in the same format. The ecological contribution degree matrix characterizes that different positions in different initial dimensions have different influences on the dimensionality reduction result.

[0105] Step 4-2: Calculate the product of the ecological contribution degree matrix, the transpose matrix, and the initial covariance matrix to obtain the weighted covariance matrix.

[0106] After obtaining the weighted covariance matrix, perform dimensionality reduction according to the implementation process of the principal component analysis algorithm. Specifically: calculate the eigenvalues and eigenvectors of the weighted covariance matrix, select the eigenvectors corresponding to the first preset number of largest eigenvalues as the principal components, and project the original data onto these principal components, thereby realizing dimensionality reduction. This not only retains the key information of the original data but also takes into account the contribution degrees of each initial dimension through weighting, thus realizing efficient and high-quality dimensionality reduction for ecological environment monitoring.

[0107] Therefore, the dimensionality reduction process compresses the data in the ecological environment parameters from a high-dimensional space to a lower dimension, retains the most representative feature information in the parameters, and reflects the main change trends of the ecological environment, thereby realizing high-quality monitoring of the ecological environment parameters. The dimensionality-reduced data has a lower dimension and can more intuitively display the change rules of the ecological environment parameters. The principal components after dimensionality reduction can be used as new monitoring indicators for visualization and subsequent analysis. Through forms such as charts, heatmaps, and trend charts, the changes in ecological environment parameters can be clearly displayed, helping decision-makers quickly identify the key driving factors of environmental changes. The dimensionality-reduced data can not only assist in analyzing the current ecological environment status but also be used for long-term dynamic monitoring and trend analysis. Through time series analysis of the principal components, the periodicity, long-term trends, and sudden changes in the changes of ecological environment parameters can be identified, thereby providing early warnings for the ecological environment.

[0108] This embodiment also provides an ecological environment parameter monitoring device, which includes units for executing the above-mentioned ecological environment parameter monitoring method. Each unit can be a software unit corresponding to each method step or a hardware circuit for executing each method step, which is not limited in this embodiment.

[0109] This embodiment also provides an ecological environment parameter monitoring system, including: a memory and a processor; the memory is connected to the processor, and the memory is used to store program instructions; the processor is used to implement the steps in the above-mentioned embodiment of the ecological environment parameter monitoring method when the program instructions are executed.

[0110] In an exemplary embodiment, the present invention provides a computer-readable storage medium storing a computer program, which when executed by a processor implements the steps in the above-mentioned embodiment of the ecological environment parameter monitoring method.

[0111] It should be noted that: the above sequence of embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0112] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. The key point of each embodiment is to illustrate the differences from other embodiments.

Claims

1. A method for monitoring ecological environment parameters, characterized in that: include: Acquire ecological environment parameter sequences of multiple initial dimensions, and obtain dynamic change characteristics of each position according to the change trend of each position in the ecological environment parameter sequence; Based on the difference between the degree of association of each initial dimension with the overall data and the degree of association of the local data of each other initial dimension, combined with the number of associated initial dimensions associated with each initial dimension, the dominant role of each initial dimension on the ecological environment is obtained; the degree of association is obtained by the dynamic change characteristics; According to the leading role of each initial dimension on the ecological environment and the dynamic change characteristics of each position in each initial dimension, the ecological contribution of each position in each initial dimension is obtained; The ecological contribution of each position in each initial dimension is used as a weight to weight the initial covariance matrix in the principal component analysis and reduce the dimensionality of the initial dimension; According to the change trend of each position in the ecological environment parameter sequence, the dynamic change characteristics of each position are obtained, including: Segmenting a first ecological environment parameter sequence to obtain a plurality of data segments; the first ecological environment parameter sequence is an ecological environment parameter sequence of a first initial dimension, and the first initial dimension is any initial dimension; Based on the change of data in the first data segment, the long-term change of the first position is obtained; the first data segment is any data segment in the first ecological environment parameter sequence; the first position is the position of any parameter in the first data segment; Based on the difference between the actual value and the fitted value of the first position, obtaining the short-term volatility of the first position; The long-term changes and short-term fluctuations of the first position are integrated to obtain the dynamic change characteristics of the first position; The calculation formula of ecological contribution is: ； in, Indicates The initial dimension The ecological contribution of each location, Indicates The initial dimension The dynamic change characteristics of the position, Indicates The dominant role of the initial dimension on the ecological environment, norm represents the normalization function.

2. The method for monitoring ecological environment parameters according to claim 1, characterized in that: Based on the change of data in the first data segment, the long-term change of the first position is obtained, including: Performing straight line fitting on the first data segment to obtain a first slope of the fitting straight line of the first data segment; Taking the first position as the dividing point, respectively performing straight line fitting on the data before and after the first position in the first data segment to obtain a second slope of the fitting straight line of the data before and a third slope of the fitting straight line of the data after; According to the first slope and the slope difference, the long-term change of the first position is obtained; the long-term change is proportional to the slope difference and proportional to the first slope; the slope difference is the difference between the second slope and the third slope.

3. A method for monitoring ecological environment parameters as claimed in claim 2, characterized in that: Based on the difference between the actual value and the fitted value of the first position, the short-term volatility of the first position is obtained, including: Obtaining a trend deviation of the first position according to a difference between an actual value and a fitted value of the first position; The trend deviation difference between two adjacent positions within a preset neighborhood range of the first position is obtained, and then the overall situation of the trend deviation difference within the preset neighborhood range of the first position is obtained to obtain the short-term volatility of the first position.

4. The method for monitoring ecological environment parameters according to claim 1, characterized in that: Based on the difference between the degree of association between each initial dimension and the overall data of other initial dimensions and the degree of association between the local data, combined with the number of associated initial dimensions associated with each initial dimension, the dominant role of each initial dimension on the ecological environment is obtained, including: Obtaining a first correlation degree between a first initial dimension and a second initial dimension, wherein the first correlation degree is a correlation degree between an overall sequence of dynamic change features of the first initial dimension and the second initial dimension; the overall sequence of dynamic change features includes dynamic change features of all positions of the corresponding initial dimension; the first initial dimension is any initial dimension, and the second initial dimension is any other initial dimension different from the first initial dimension; Acquire a preset neighborhood range of a second position corresponding to a first dynamic change feature local sequence of the first initial dimension, and a second dynamic change feature local sequence corresponding to the second initial dimension; the second position is the position of any parameter in the first initial dimension; Acquire a second correlation degree between the first initial dimension and the second initial dimension at the second position, wherein the second correlation degree is a correlation degree between the first dynamic change feature local sequence and the second dynamic change feature local sequence; According to the difference between the first correlation degree and the second correlation degree, the real variability of the second position of the first initial dimension is obtained; According to the real variability and the number of associated initial dimensions associated with the first initial dimension, the dominant effect of the first initial dimension on the ecological environment is obtained.

5. The method for monitoring ecological environment parameters as claimed in claim 4, characterized in that: According to the difference between the first correlation degree and the second correlation degree, the real variability of the second position of the first initial dimension is obtained, including: Obtain the first correlation degree between the first initial dimension and each of the other initial dimensions, and the difference between the second correlation degrees of the first initial dimension and each of the other initial dimensions at the second position, and calculate the average to obtain the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions; According to the overall correlation difference of the second position of the first initial dimension with respect to all other initial dimensions, the real variability of the second position of the first initial dimension is obtained; the real variability is inversely proportional to the overall correlation difference.

6. The method for monitoring ecological environment parameters according to claim 4, characterized in that: The process of acquiring the associated initial dimension associated with the first initial dimension includes: taking the initial dimension corresponding to the first correlation degree greater than a preset threshold among the first correlation degrees between the first initial dimension and each other initial dimension as the associated initial dimension associated with the first initial dimension; According to the real variability and the number of associated initial dimensions associated with the first initial dimension, the dominant effect of the first initial dimension on the ecological environment is obtained, including: Calculate the mean of the true variability of each position of the first initial dimension to obtain the overall true variability of the first initial dimension; According to the overall real variability of the first initial dimension and the number of associated initial dimensions associated with the first initial dimension, the dominant effect of the first initial dimension on the ecological environment is obtained; the dominant effect is proportional to the overall real variability and proportional to the number of associated initial dimensions.

7. The method for monitoring ecological environment parameters according to claim 1, characterized in that: Weighting the initial covariance matrix in principal component analysis includes: According to the ecological contribution of each position in each initial dimension, an ecological contribution matrix is ​​constructed, and a transposed matrix of the ecological contribution matrix is ​​obtained; The product of the ecological contribution matrix, the transposed matrix and the initial covariance matrix is ​​calculated to obtain a weighted covariance matrix.

8. An ecological environment parameter monitoring device, characterized in that: The ecological environment parameter monitoring device includes a unit for executing the ecological environment parameter monitoring method described in any one of claims 1-7.

9. An ecological environment parameter monitoring system, characterized in that it includes: Memory and processor; The memory is connected to the processor; The memory is used to store program instructions; The processor is used to implement the ecological environment parameter monitoring method described in any one of claims 1 to 7 when the program instructions are executed.

Images