Lake ecology intelligent monitoring method and system based on Internet of Things

Abstract

The invention relates to a lake ecology intelligent monitoring method and system based on the Internet of Things. The method comprises the following steps: collecting multi-dimensional parameter data of a target lake area; preprocessing the multi-dimensional parameter data to obtain standardized multi-dimensional parameter data; according to the standardized multi-dimensional parameter data, calculating correlation strength among different parameters through a Pearson's correlation coefficient, and obtaining a parameter correlation graph; key factors influencing the lake ecosystem are extracted according to a parameter set of target relation strength in the parameter association relation graph; and constructing a lake ecological health assessment model through a fuzzy comprehensive evaluation algorithm, calculating membership degree values of the key factors to different health levels, and obtaining a comprehensive health index of the target lake ecological system. According to the invention, comprehensive monitoring, health assessment and risk early warning of the lake ecosystem are realized, and a scientific basis and decision support are provided for lake ecological protection and management.

Description

Technical Field

[0001] The present invention relates to the technical field of water ecology monitoring, and in particular to an Internet of Things-based lake ecology intelligent monitoring method and system. Background Art

[0002] As a vital component of Earth's water cycle and biodiversity, the health of lake ecosystems directly impacts regional environmental quality and human well-being. With the acceleration of industrialization and intensification of climate change, lake ecosystems are facing unprecedented pressure. Establishing an effective ecological monitoring system has become an urgent need for environmental protection and sustainable development.

[0003] Traditional lake ecological monitoring relies primarily on manual sampling and periodic testing, an approach with significant drawbacks. The limited frequency of manual monitoring makes it difficult to capture dynamic changes in ecosystems, and the restricted scope of monitoring prevents comprehensive coverage of the entire lake area. Furthermore, traditional methods capture a single type of data and lack the ability to comprehensively analyze multi-dimensional information such as water quality, meteorology, hydrology, and biology.

[0004] The complexity of lake ecosystems means that their monitoring faces unique technical challenges. Water quality parameters, meteorological conditions, hydrological conditions, and biological activities in lake environments influence each other, forming a complex ecological network. Monitoring of a single parameter cannot accurately reflect the overall ecological health status. The complex correlation of this multi-dimensional data further leads to difficulties in data processing and analysis. Traditional statistical analysis methods are unable to handle large amounts of heterogeneous data, let alone identify potential correlation patterns between data. When there is noise interference and outliers in the monitoring data, it becomes extremely difficult to accurately extract effective information from massive amounts of complex data and conduct reliable ecological health assessments. Data quality issues not only affect the accuracy of analysis results, but may also lead to incorrect judgments about ecological conditions, thereby affecting the scientific nature of environmental protection decisions.

[0005] How to build a comprehensive monitoring system that can collect multi-dimensional lake ecological data in real time and accurately assess the health status of lake ecosystems through intelligent data processing and analysis technologies has become a key issue that needs to be urgently addressed in the current field of lake ecological protection. Summary of the Invention

[0006] The purpose of this invention is to provide a lake ecological intelligent monitoring method and system based on the Internet of Things, which can realize comprehensive monitoring, health assessment and risk warning of lake ecosystems, and provide a scientific basis and decision-making support for lake ecological protection and management.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A lake ecological intelligent monitoring method based on the Internet of Things, comprising:

[0009] Collect multi-dimensional parameter data of the target lake area;

[0010] Preprocessing the multidimensional parameter data to obtain standardized multidimensional parameter data;

[0011] According to the standardized multidimensional parameter data, the correlation strength between different parameters is calculated by using the Pearson correlation coefficient to obtain a parameter correlation relationship map;

[0012] Extracting key factors affecting the lake ecosystem based on a parameter set of target relationship strength in the parameter association relationship map;

[0013] A lake ecological health assessment model was constructed through a fuzzy comprehensive evaluation algorithm, and the membership values ​​of key factors to different health levels were calculated to obtain the comprehensive health index of the target lake ecosystem.

[0014] Optionally, the multi-dimensional parameter data collected for the target lake area include:

[0015] A distributed sensor network is used to deploy water quality detection sensors, meteorological monitoring equipment, hydrological measurement devices, and image acquisition devices in different areas of the target lake. Multi-dimensional parameter data including water temperature, dissolved oxygen, pH value, total nitrogen, total phosphorus, turbidity, wind speed, rainfall, water level changes, flow rate, and algae density are collected in real time through wireless communication protocols.

[0016] Optionally, preprocessing the multidimensional parameter data to obtain standardized multidimensional parameter data includes:

[0017] A wavelet denoising algorithm is used to denoise the multidimensional parameter data. If a high-frequency noise component is detected in the multidimensional parameter data, the signal is decomposed into different frequency components using wavelet decomposition, and the signal is reconstructed after filtering out the noise frequency band to obtain the denoised multidimensional parameter data.

[0018] The denoised multidimensional parameter data is subjected to data standardization processing to obtain standardized multidimensional parameter data.

[0019] Optionally, the Pearson correlation coefficient is used to calculate the correlation strength between different parameters, and the parameter correlation relationship map is obtained, including:

[0020] Calculating the correlation coefficient between parameters in the standardized multidimensional parameter matrix using the Pearson correlation coefficient, and obtaining a significant correlation relationship set based on the correlation coefficient;

[0021] According to the significant correlation set, the parameter correlation graph is constructed, wherein nodes represent parameters, edges represent significant correlations, and the weights of edges are the absolute values ​​of correlation coefficients.

[0022] Optionally, based on the parameter set of target relationship strength in the parameter association relationship map, extracting key factors affecting the lake ecosystem includes:

[0023] According to the correlation coefficient, a set of strongly correlated parameters is obtained;

[0024] The principal component analysis algorithm is used to calculate the eigenvectors and eigenvalues ​​of the parameters in the strongly correlated parameter set to obtain the key factors.

[0025] Optionally, the membership values ​​of key factors to different health levels are calculated to obtain the comprehensive health index of the target lake ecosystem, including:

[0026] Calculate the membership value of each key factor to different health levels and obtain the membership matrix;

[0027] If the membership value of the key factor in the membership matrix is ​​the largest at the excellent health level, then the key factor is determined to contribute positively to the ecological health of the lake; if the membership value is the largest at the deteriorated health level, then the key factor is determined to contribute negatively to the ecological health of the lake, and the health contribution determination results of each key factor are obtained;

[0028] The comprehensive health index of the target lake ecosystem is calculated by combining the health contribution judgment results of each key factor and the membership matrix through the weighted average method.

[0029] Optionally, after obtaining the comprehensive health index of the target lake ecosystem, the following items may be included:

[0030] Inputting the standardized multidimensional parameter data into a health index change trend model to obtain a health index change trend, wherein the health index change trend model obtains a comprehensive health index sequence from historical data through a time series analysis method, adopts an ARIMA model to fit the change trend, and establishes the health index change trend model;

[0031] If the health index shows a downward trend during the preset monitoring period and the decline exceeds the preset threshold, the ecological early warning mechanism will be triggered and a lake ecological risk warning signal will be obtained.

[0032] The present invention also provides a lake ecological intelligent monitoring system based on the Internet of Things, comprising:

[0033] Data acquisition module, used to collect multi-dimensional parameter data of the target lake area;

[0034] A preprocessing module, used to preprocess the multidimensional parameter data to obtain standardized multidimensional parameter data;

[0035] A map building module is used to calculate the correlation strength between different parameters using the Pearson correlation coefficient based on the standardized multi-dimensional parameter data to obtain a parameter correlation relationship map;

[0036] An analysis module, configured to extract key factors affecting the lake ecosystem based on a parameter set of target relationship strength in the parameter association relationship map;

[0037] The ecological monitoring module is used to construct a lake ecological health assessment model through a fuzzy comprehensive evaluation algorithm, calculate the membership values ​​of each key factor to different health levels, and obtain the comprehensive health index of the target lake ecosystem.

[0038] The beneficial effects of the present invention include: It uses a distributed sensor network to collect multidimensional water quality, meteorological, and hydrological parameter data in real time, employs wavelet denoising and data standardization for preprocessing, uses correlation analysis to identify patterns in the associations between parameters, extracts key ecological influencing factors through principal component analysis, constructs a fuzzy comprehensive evaluation model to calculate the ecological health index, and establishes a health index trend model based on time series analysis. When the index continuously declines and exceeds a threshold, an early warning mechanism is triggered. This invention enables comprehensive monitoring, health assessment, and risk warning of lake ecosystems, providing a scientific basis and decision-making support for lake ecological protection and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 This is a flow chart of an Internet of Things-based lake ecological intelligent monitoring method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0043] Example 1:

[0044] like Figure 1 This embodiment provides a lake ecological intelligent monitoring method based on the Internet of Things, including:

[0045] Collect multi-dimensional parameter data of the target lake area;

[0046] Preprocessing the multidimensional parameter data to obtain standardized multidimensional parameter data;

[0047] Based on the standardized multidimensional parameter data, the correlation strength between different parameters is calculated using the Pearson correlation coefficient to obtain the parameter correlation relationship map;

[0048] Extract key factors affecting lake ecosystems based on the parameter set of target relationship strength in the parameter association relationship map;

[0049] A lake ecological health assessment model was constructed through a fuzzy comprehensive evaluation algorithm, and the membership values ​​of key factors to different health levels were calculated to obtain the comprehensive health index of the target lake ecosystem.

[0050] Furthermore, the multi-dimensional parameter data collected for the target lake area include:

[0051] A distributed sensor network is used to deploy water quality detection sensors, meteorological monitoring equipment, hydrological measurement devices, and image acquisition devices in different areas of the target lake. Multi-dimensional parameter data including water temperature, dissolved oxygen, pH value, total nitrogen, total phosphorus, turbidity, wind speed, rainfall, water level changes, flow rate, and algae density are collected in real time through wireless communication protocols.

[0052] Specifically, the distributed network architecture achieves full coverage in lake water quality monitoring through the rational deployment of sensor nodes. For example, a lake with an area of ​​10 square kilometers can be equipped with 50 water quality sensor nodes, 20 meteorological devices, 15 hydrological devices, and image acquisition devices, with node spacing of approximately 400 meters to ensure uniform coverage. Each node is equipped with a GPS module to obtain geographic coordinates. For example, the coordinates of node A are (30.5, 114.3). The LoRa protocol is used to check communication status and ensure stable data transmission. Machine vision technology is used to identify algae density in images captured by the image acquisition device.

[0053] Preferably, LoRa's low power consumption supports long-distance communication, reduces energy consumption, and improves network reliability. Data collection is achieved through wireless communication protocols.

[0054] Specifically, water quality sensors collect water temperature, dissolved oxygen, pH, and turbidity every hour. For example, at a certain node, the water temperature is 25.3°C, the dissolved oxygen is 6.8 mg / L, the pH is 7.2, and the turbidity is 15 NTU. Meteorological equipment records wind speeds of, for example, 5 m / s and rainfall of, for example, 2 mm / h. Hydrological devices monitor water levels at 1.2 m and flow rates at 0.3 m / s. Collected data is transmitted to the cloud via 5G or LoRa, ensuring real-time performance and meeting dynamic monitoring requirements. Data cleaning algorithms remove noise and outliers.

[0055] Furthermore, preprocessing the multidimensional parameter data to obtain standardized multidimensional parameter data includes:

[0056] The wavelet denoising algorithm is used to denoise the multidimensional parameter data. If high-frequency noise components are detected in the multidimensional parameter data, the wavelet decomposition is used to decompose the signal into different frequency components, and the signal is reconstructed after filtering out the noise frequency band to obtain the denoised multidimensional parameter data.

[0057] The denoised multidimensional parameter data is subjected to data standardization processing to obtain standardized multidimensional parameter data.

[0058] Specifically, wavelet denoising decomposes data into frequency components, retaining low-frequency signals that reflect the true data trend while filtering out high-frequency noise. The frequency decomposition results can be analyzed by examining the characteristics of the high-frequency noise components. For example, if the frequency of the noise component in the turbidity data during a particular monitoring session exceeds a threshold of 10Hz, noise band filtering can be used to remove these high-frequency components, retaining the low-frequency components that reflect the true turbidity changes. These filtered frequency components can then be used to reconstruct the signal, generating denoised data.

[0059] For the denoised multidimensional monitoring data, data standardization is used to convert parameters of different dimensions into a unified scale. The standardized value of each parameter Z = (X-μ) / σ is calculated using the Z-score standardization method, where X represents the original value, μ represents the mean, and σ represents the standard deviation.

[0060] In one embodiment, median filtering is used to process turbidity data. If the turbidity at a node suddenly increases to 100 NTU, significantly deviating from the normal range of 10-20 NTU, this value is discarded. If missing data is detected, such as if a pH value at a node has not been uploaded for two consecutive hours, linear interpolation can be used to fill in the missing value, estimating it to 7.2 based on the previous and next data values ​​of 7.1 and 7.3. This method improves data integrity and ensures accurate subsequent analysis.

[0061] Furthermore, the correlation strength between different parameters is calculated using the Pearson correlation coefficient, and the parameter correlation relationship map is obtained, including:

[0062] The correlation coefficient between the parameters in the standardized multidimensional parameter matrix is ​​calculated by the Pearson correlation coefficient, and a significant correlation relationship set is obtained according to the correlation coefficient;

[0063] According to the set of significant correlation relationships, a parameter correlation relationship map is constructed, in which nodes represent parameters, edges represent significant correlation relationships, and the weight of the edge is the absolute value of the correlation coefficient.

[0064] Specifically, the Pearson correlation coefficient measures the strength of linear associations between parameters. For example, the calculated correlation coefficient between water temperature and dissolved oxygen is -0.75, with an absolute value exceeding the preset threshold of 0.6. This indicates that as water temperature increases, dissolved oxygen levels decrease significantly, marking a significant association. Similarly, rainfall and flow rate may exhibit a positive correlation, with a correlation coefficient of 0.82, also identified as a significant association. The correlation coefficient matrix summarizes the strength of associations across all parameter pairs, clearly demonstrating potential connections between variables. This collection of significant associations lays the foundation for subsequent exploration. In constructing a parameter association map, nodes can be parameters such as pH, dissolved oxygen, and rainfall, while edges represent significant associations, with edge weights equal to the absolute value of the correlation coefficient. In one possible implementation, when optimizing the parameter association map, key parameter nodes and their associated edges are retained. For example, only nodes such as water temperature, rainfall, and flow rate, along with their strongly associated edges with parameters such as dissolved oxygen, are retained in the final map. This optimized map is more concise, highlighting the relationships between core parameters and facilitating rapid identification of key influencing factors. This approach helps to gain a deeper understanding of the interactions among water quality, meteorological, and hydrological parameters, providing a clear direction for subsequent monitoring and analysis.

[0065] Furthermore, based on the parameter set of the target relationship strength in the parameter association relationship map, the key factors affecting the lake ecosystem are extracted, including:

[0066] According to the correlation coefficient, a set of strongly correlated parameters is obtained;

[0067] The principal component analysis algorithm is used to calculate the eigenvectors and eigenvalues ​​of the parameters in the strongly correlated parameter set to obtain the key factors.

[0068] Specifically, strongly correlated parameter combinations are obtained from the parameter correlation relationship map. Correlation coefficient calculation methods are used to determine the strength of the correlation between each parameter, resulting in a set of strongly correlated parameters. For this set of strongly correlated parameters, a principal component analysis algorithm is then used to calculate the eigenvectors and eigenvalues ​​of each parameter to obtain a set of principal components. Based on this set of principal components, the contribution rate of each principal component is calculated. By summing the contribution rates, it is determined whether the cumulative contribution rate reaches a preset threshold and the number of principal components is determined. The corresponding principal components are then extracted from the set of principal components to obtain a reduced set of ecological impact factors.

[0069] Suppose the analysis finds a correlation coefficient of 0.75 between dissolved oxygen and temperature, exceeding the preset threshold of 0.6, and marks this parameter combination as strongly correlated. This method quantifies the linear relationship between parameters and identifies parameter pairs with significant impacts on lake ecosystems, providing a data foundation for subsequent analysis.

[0070] Furthermore, the membership values ​​of each key factor to different health levels are calculated to obtain the comprehensive health index of the target lake ecosystem, including:

[0071] Calculate the membership value of each key factor to different health levels and obtain the membership matrix;

[0072] If the membership value of the key factor in the membership matrix is ​​the largest at the excellent health level, then the key factor is judged to contribute positively to the ecological health of the lake. If the membership value is the largest at the deteriorating health level, then the key factor is judged to contribute negatively to the ecological health of the lake. The health contribution judgment results of each key factor are obtained.

[0073] The comprehensive health index of the target lake ecosystem is calculated by combining the health contribution judgment results of each key factor and the membership matrix through the weighted average method.

[0074] Specifically, a fuzzy comprehensive evaluation model is constructed based on key factors, with membership functions assigned to four health levels: excellent, fair, poor, and deteriorating. The membership function parameters for each factor are then derived. Using a fuzzy comprehensive evaluation algorithm, the membership values ​​of each factor at the four health levels of excellent, fair, poor, and deteriorating are calculated for the reduced set of ecological impact factors, resulting in a membership matrix. For example, the membership function for dissolved oxygen can be set based on its concentration range: 8 mg / L and above is excellent, 6-8 mg / L is fair, 4-6 mg / L is poor, and below 4 mg / L is deteriorating.

[0075] Then, based on the comprehensive health index value, a comparison is made using the preset threshold interval. If the comprehensive health index value falls into the excellent interval, the lake ecosystem is judged to be in an excellent state; if it falls into the deteriorating interval, it is judged to be in a deteriorating state, and the health status classification of the lake ecosystem is obtained.

[0076] Furthermore, the comprehensive health index of the target lake ecosystem is obtained, including:

[0077] Input the standardized multi-dimensional parameter data into the health index change trend model to obtain the health index change trend. The health index change trend model obtains the comprehensive health index sequence from historical data through the time series analysis method, adopts the ARIMA model to fit the change trend, and establishes the health index change trend model;

[0078] If the health index shows a downward trend during the preset monitoring period and the decline exceeds the preset threshold, the ecological early warning mechanism will be triggered and a lake ecological risk warning signal will be obtained.

[0079] Specifically, the ARIMA model is used to fit trends in the health index. The ARIMA model identifies patterns in time series by analyzing the autocorrelation and seasonality of data. For example, for a lake's health index series, the model might identify a periodic decline in the health index due to increased rainfall in the summer, resulting in excessive nutrient input. After model fitting, a trend model can be generated to reflect the long-term direction of the health index, such as a gradual decline or cyclical fluctuations, providing a basis for subsequent forecasts. Assuming a lake is monitored monthly, the ARIMA model is used to predict the health index for the next three months. For example, if the rate of change from the first cycle to the second cycle is -4.4%, and the rate of change from the second cycle to the third cycle is -4.6%. If the preset threshold is -3%, a continuous decline and the rate of change exceeds the threshold triggers a logical decision. This logical decision is used to generate ecological warning signals. For example, if the health index declines for three consecutive cycles with a rate of change exceeding -3%, the system will automatically generate a warning signal. Such a signal can alert managers to potential risks to the lake ecosystem.

[0080] Example 2:

[0081] This embodiment implements an IoT-based lake ecological intelligent monitoring system according to the method of the first embodiment, including:

[0082] Data acquisition module, used to collect multi-dimensional parameter data of the target lake area;

[0083] A preprocessing module is used to preprocess the multidimensional parameter data to obtain standardized multidimensional parameter data;

[0084] A map building module is used to calculate the correlation strength between different parameters using the Pearson correlation coefficient based on standardized multi-dimensional parameter data to obtain a parameter correlation relationship map;

[0085] An analysis module is used to extract key factors affecting the lake ecosystem based on the parameter set of target relationship strength in the parameter association relationship map;

[0086] The ecological monitoring module is used to construct a lake ecological health assessment model through a fuzzy comprehensive evaluation algorithm, calculate the membership values ​​of each key factor to different health levels, and obtain the comprehensive health index of the target lake ecosystem.

[0087] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A lake ecological intelligent monitoring method based on the Internet of Things, characterized in that: include: Collect multi-dimensional parameter data of the target lake area; Preprocessing the multidimensional parameter data to obtain standardized multidimensional parameter data; According to the standardized multidimensional parameter data, the correlation strength between different parameters is calculated by using the Pearson correlation coefficient to obtain a parameter correlation relationship map; Extracting key factors affecting the lake ecosystem based on a parameter set of target relationship strength in the parameter association relationship map; A lake ecological health assessment model was constructed through a fuzzy comprehensive evaluation algorithm, and the membership values ​​of key factors to different health levels were calculated to obtain the comprehensive health index of the target lake ecosystem.

2. The lake ecological intelligent monitoring method based on the Internet of Things according to claim 1 is characterized in that: The multi-dimensional parameter data collected for the target lake area include: A distributed sensor network is used to deploy water quality detection sensors, meteorological monitoring equipment, hydrological measurement devices, and image acquisition devices in different areas of the target lake. Multi-dimensional parameter data including water temperature, dissolved oxygen, pH value, total nitrogen, total phosphorus, turbidity, wind speed, rainfall, water level changes, flow rate, and algae density are collected in real time through wireless communication protocols.

3. The lake ecological intelligent monitoring method based on the Internet of Things according to claim 1 is characterized in that: Preprocessing the multidimensional parameter data to obtain standardized multidimensional parameter data includes: A wavelet denoising algorithm is used to denoise the multidimensional parameter data. If a high-frequency noise component is detected in the multidimensional parameter data, the signal is decomposed into different frequency components using wavelet decomposition, and the signal is reconstructed after filtering out the noise frequency band to obtain the denoised multidimensional parameter data. The denoised multidimensional parameter data is subjected to data standardization processing to obtain standardized multidimensional parameter data.

4. The lake ecological intelligent monitoring method based on the Internet of Things according to claim 1 is characterized in that: The correlation strength between different parameters is calculated by the Pearson correlation coefficient, and the parameter correlation relationship map is obtained, including: Calculating the correlation coefficient between parameters in the standardized multidimensional parameter matrix using the Pearson correlation coefficient, and obtaining a significant correlation relationship set based on the correlation coefficient; According to the significant correlation set, the parameter correlation graph is constructed, wherein nodes represent parameters, edges represent significant correlations, and the weights of edges are the absolute values ​​of correlation coefficients.

5. The lake ecological intelligent monitoring method based on the Internet of Things according to claim 4 is characterized in that: According to the parameter set of the target relationship strength in the parameter association relationship map, the key factors affecting the lake ecosystem are extracted, including: According to the correlation coefficient, a set of strongly correlated parameters is obtained; The principal component analysis algorithm is used to calculate the eigenvectors and eigenvalues ​​of the parameters in the strongly correlated parameter set to obtain the key factors.

6. The lake ecological intelligent monitoring method based on the Internet of Things according to claim 1 is characterized in that: Calculate the membership values ​​of each key factor to different health levels and obtain the comprehensive health index of the target lake ecosystem, including: Calculate the membership value of each key factor to different health levels and obtain the membership matrix; If the membership value of the key factor in the membership matrix is ​​the largest at the excellent health level, then the key factor is determined to contribute positively to the ecological health of the lake; if the membership value is the largest at the deteriorated health level, then the key factor is determined to contribute negatively to the ecological health of the lake, and the health contribution determination results of each key factor are obtained; The comprehensive health index of the target lake ecosystem is calculated by combining the health contribution judgment results of each key factor and the membership matrix through the weighted average method.

7. The lake ecological intelligent monitoring method based on the Internet of Things according to claim 1 is characterized in that: Obtaining the comprehensive health index of the target lake ecosystem includes: Inputting the standardized multidimensional parameter data into a health index change trend model to obtain a health index change trend, wherein the health index change trend model obtains a comprehensive health index sequence from historical data through a time series analysis method, adopts an ARIMA model to fit the change trend, and establishes the health index change trend model; If the health index shows a downward trend during the preset monitoring period and the decline exceeds the preset threshold, the ecological early warning mechanism will be triggered and a lake ecological risk warning signal will be obtained.

8. An Internet of Things-based lake ecological intelligent monitoring system, the system being used to implement the method according to any one of claims 1 to 7, characterized in that: include: Data acquisition module, used to collect multi-dimensional parameter data of the target lake area; A preprocessing module, used to preprocess the multidimensional parameter data to obtain standardized multidimensional parameter data; A map building module is used to calculate the correlation strength between different parameters using the Pearson correlation coefficient based on the standardized multi-dimensional parameter data to obtain a parameter correlation relationship map; An analysis module, configured to extract key factors affecting the lake ecosystem based on a parameter set of target relationship strength in the parameter association relationship map; The ecological monitoring module is used to construct a lake ecological health assessment model through a fuzzy comprehensive evaluation algorithm, calculate the membership values ​​of each key factor to different health levels, and obtain the comprehensive health index of the target lake ecosystem.

Images