Landscaping maintenance monitoring and early warning system

Abstract

The invention relates to the field of monitoring and early warning, and discloses a landscaping maintenance monitoring and early warning system which comprises a garden information acquisition module, a garden image processing module, a normal area analysis module, an abnormal area analysis module, a monitoring and early warning analysis module and a monitoring and early warning output module. The garden information acquisition module is used for acquiring environment data and plant growth data of a garden greening area in real time; the garden image processing module obtains an image of a landscaping area through a camera or an unmanned aerial vehicle, and carries out image processing and analysis; the normal area analysis module analyzes and defines the standard of a healthy area according to the collected data and historical data; the abnormal area analysis module compares the normal area with other areas, and analyzes an abnormal area; and the monitoring and early warning analysis module synthesizes data from normal area analysis and abnormal area analysis to perform intelligent early warning analysis. The method has the advantage of providing accurate early warning.

Description

Technical Field

[0001] The present invention relates to the field of monitoring and early warning, and in particular to a garden greening maintenance monitoring and early warning system. Background Art

[0002] As an important part of urban ecological construction, gardening and greening not only provides beautiful landscapes for cities, but also plays an important role in improving air quality, regulating climate, reducing noise pollution, etc. With the continuous acceleration of urbanization, the area of ​​gardening and greening has gradually expanded, making the task of gardening and greening maintenance and management more arduous. The traditional gardening and greening management model mostly relies on manual inspections, regular fertilization, watering and other methods to ensure the greening effect, but this management method has many limitations.

[0003] The traditional manual inspection method relies on a large amount of manpower input, with low management efficiency. In addition, due to the vast area of ​​garden greening, manual inspection often cannot achieve 24-hour comprehensive monitoring without blind spots. This method cannot effectively discover potential problems, such as pests and diseases, environmental abnormalities, etc., resulting in the spread of pests and diseases or plant water shortage, improper fertilization and other problems that cannot be solved in time. In addition, manual inspections also have the problems of inaccurate and unrealistic data recording, and lack of effective feedback and analysis of the greening maintenance status. With the expansion of the scale of garden greening management, traditional means such as regular irrigation and fertilization face the problem of inefficiency. Especially in extreme weather conditions, manual management cannot respond quickly and adjust maintenance measures in time. For example, under long-term high temperature, drought or rainy weather, traditional irrigation and fertilization methods often cannot meet the actual needs of plants, thus affecting plant growth and greening effects.

[0004] With the development of modern information technologies such as the Internet of Things, sensor technology, and remote sensing technology, more and more intelligent landscaping monitoring systems have emerged. These systems monitor environmental parameters such as soil moisture, temperature, light intensity, and air quality in real time through various sensors installed in landscaping areas, collect a large amount of data, and upload the data to the cloud platform through wireless networks for centralized processing and analysis. The introduction of such technologies has improved the real-time nature of landscaping management and the accuracy of data. However, the existing landscaping monitoring systems still have certain shortcomings. Most existing systems can provide basic alarm functions, but usually only warn of abnormal situations in one aspect. Intelligent early warning and decision support after comprehensive analysis of multiple factors have not yet been fully realized. For example, factors such as climate change and outbreaks of pests and diseases often require multiple data cross-analysis to detect potential problems in a timely manner, but the existing systems lack such comprehensive analysis and intelligent early warning capabilities. Therefore, it is necessary to design a landscaping maintenance monitoring and early warning system that provides accurate early warnings. Summary of the invention

[0005] 1. Technical issues to be solved

[0006] In view of the deficiencies in the prior art, the present invention provides a garden greening maintenance monitoring and early warning system, which has the advantage of providing accurate early warnings and solves the problems in the above-mentioned background technology.

[0007] (II) Technical solution

[0008] In order to achieve the above-mentioned purpose of providing accurate early warning, the present invention provides the following technical solutions: a garden greening maintenance monitoring and early warning system, including a garden information collection module, a garden image processing module, a normal area analysis module, an abnormal area analysis module, a monitoring and early warning analysis module and a monitoring and early warning output module;

[0009] The garden information collection module is responsible for collecting environmental data and plant growth data of the garden greening area in real time;

[0010] The garden information collection module is further responsible for real-time collection of environmental data and plant growth data of the garden greening area, including temperature, humidity, light intensity, soil pH, soil moisture, precipitation, wind speed, and plant leaf color, leaf area, stem height, and root growth, and data collection is performed through various sensors installed in the park.

[0011] The garden image processing module obtains images of the garden greening area through a camera or a drone, performs image processing and analysis, identifies vegetation health conditions, pests and diseases, and leaf color changes, and helps determine the overall health status of the garden area;

[0012] The garden image processing module further includes using a threshold segmentation method, an edge detection method, a region growing method or a semantic segmentation method based on deep learning to separate the plant area in the image from the background. For images with complex backgrounds, a deep learning algorithm is used to accurately segment plant areas of different categories, and image processing is performed through color space to extract the green component of the leaves and analyze the color changes of the leaves.

[0013] The normal area analysis module analyzes and defines the standards of the healthy area based on the collected data and historical data, and determines the standards of the normal area by combining the environmental data and the plant status data through the data analysis method;

[0014] The normal area analysis module further includes analyzing the normal range of different areas based on historical data using statistical analysis methods, estimating the standards of normal areas using statistical models, determining the standard values ​​of historical normal areas through data distribution and trend analysis methods, establishing the standards of normal areas using single environmental factors or plant growth data, comparing real-time collected environmental data with historical data, using threshold detection, data deviation analysis and other methods to determine whether the real-time data meets the normal area standards, calculating the deviation value between the real-time data and the historical health standards, and determining whether the area is a normal area based on the deviation value, and dividing the entire landscaping area into normal areas and abnormal areas based on the established standard model.

[0015] The abnormal area analysis module compares the normal area with other areas, analyzes the area with abnormalities, and identifies the abnormal area based on the abnormality detection algorithm and the output of the garden image processing module;

[0016] The abnormal area analysis module further includes a specific method of analyzing the abnormal area:

[0017] Each data item is processed using Z-Score standardization, and the calculation formula is:

[0018]

[0019] Where, X i is the collected data value of a certain area, μ is the mean value of the area in the historical data, and σ is the standard deviation of the area in the historical data. Through Z-Score conversion, all data are unified to the same scale.

[0020] Use the Mahalanobis distance to detect whether there are abnormal areas in the data. The Mahalanobis distance calculation formula is:

[0021] D 2 =(X-μ) T S -1 (X-μ)

[0022] In the formula, X is the vector of the data to be tested, μ is the mean vector of each variable, S is the covariance matrix of the data, and D 2 is the Mahalanobis distance;

[0023] Each abnormal area is scored and its abnormal severity is calculated. The severity is evaluated by comprehensively considering the values ​​of Z-Score and Mahalanobis distance. The calculation formula is:

[0024] Abnormal severity = α*|Z|+β*D 2

[0025] In the formula, α and β are weighting coefficients, indicating the weights of Z-Score and Mahalanobis distance in anomaly scoring.

[0026] The monitoring and early warning analysis module integrates the data from normal area analysis and abnormal area analysis to perform intelligent early warning analysis, evaluate potential problems and generate early warning information, and use multivariate analysis algorithms to conduct comprehensive analysis on the collected multidimensional data to generate early warning prediction results;

[0027] The monitoring and early warning analysis module further includes using a multivariate analysis algorithm to conduct a comprehensive analysis of the collected multidimensional data to generate early warning prediction results. Based on the results of the multivariate analysis, combined with historical data and real-time data, it is evaluated whether the current environment and plant status are close to the boundary of a potential abnormal area or a healthy area, and the risk factor of each area is calculated. The severity and probability of occurrence of potential problems are evaluated, and an early warning prediction model is constructed using machine learning or statistical modeling technology. Based on historical data, real-time collected data and comprehensive analysis results, future anomalies are predicted, and early warning thresholds are set for each potential problem.

[0028] The monitoring and early warning output module feeds back early warning information to management personnel in real time according to the output of the monitoring and early warning analysis module.

[0029] (III) Beneficial effects

[0030] Compared with the prior art, the present invention provides a garden greening maintenance monitoring and early warning system, which has the following beneficial effects:

[0031] 1. Analyze and define the standards of healthy areas based on the collected data and historical data. Determine the standards of normal areas through data analysis methods, combined with environmental data and plant status data. Compare normal areas with other areas to analyze abnormal areas. Based on the anomaly detection algorithm, combined with the output of the garden image processing module, identify abnormal areas. Use the anomaly detection algorithm to process the collected data, which can effectively detect abnormal areas that are significantly different from healthy areas. The algorithm can automatically identify abnormal conditions caused by environmental changes or pests and diseases.

[0032] 2. Integrate the data from normal area analysis and abnormal area analysis to conduct intelligent early warning analysis, evaluate potential problems and generate early warning information, use multivariate analysis algorithms to conduct comprehensive analysis on the collected multidimensional data, generate early warning prediction results, combine historical data and real-time data to evaluate whether the current environment and plant status are close to the boundary of potential abnormal areas or healthy areas, calculate the risk factor of each area, and evaluate the severity and probability of potential problems.

[0033] 3. Through risk assessment, the severity of potential problems can be quantified to provide a basis for decision-making. According to the risk factor, the system can prioritize high-risk areas and take preventive measures in advance.

[0034] 4. Through training models, it is possible to predict and generate high-precision early warning signals to help garden managers take timely action. When the data exceeds the preset threshold, the system can automatically generate early warning information, reduce human intervention, and improve response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the structure of the present invention. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0037] The present invention provides a technical solution: a garden greening maintenance monitoring and early warning system, including a garden information acquisition module, a garden image processing module, a normal area analysis module, an abnormal area analysis module, a monitoring and early warning analysis module and a monitoring and early warning output module;

[0038] The garden information collection module is responsible for collecting the environmental data and plant growth data of the garden greening area in real time;

[0039] Responsible for real-time collection of environmental data and plant growth data in the garden greening area, including temperature, humidity, light intensity, soil pH, soil moisture, precipitation, wind speed, and plant leaf color, leaf area, stem height, and root growth. Data is collected through various sensors installed in the park to ensure coverage of all key environmental variables and plant growth factors, so as to obtain comprehensive data. Sensors are deployed in key locations to ensure comprehensive monitoring of different areas and different types of plants. Sensors continuously collect environmental and plant growth data, and the frequency of data collection is set according to demand to ensure that the collected environmental and plant data can reflect changes in the park in real time, support timely response to emergencies, ensure uninterrupted data collection, avoid missed or delayed collection, and ensure the timeliness and continuity of data.

[0040] The garden image processing module obtains images of the garden greening area through cameras or drones, performs image processing and analysis, identifies vegetation health conditions, pests and diseases, and leaf color changes, and helps determine the overall health status of the garden area;

[0041] The plant area in the image is separated from the background by using threshold segmentation, edge detection, region growing or semantic segmentation based on deep learning. For images with complex backgrounds, the deep learning algorithm is used to accurately segment different types of plant areas. The plant area is accurately extracted so that subsequent analysis of health status, pests and diseases, etc. can be focused on the plant. The segmentation method based on deep learning can automatically identify complex plant morphology and environment, greatly improving processing efficiency and accuracy. Image processing is performed through color space to extract the green component of the leaves and analyze the color changes of the leaves. Color changes usually reflect the health status of the plant. For example, yellowish green may mean lack of water or fertilizer. The area and morphology of the plant leaves in the image are combined with machine learning algorithms to judge the growth of the plant. Through the analysis of features such as color changes and leaf morphology, it helps to judge whether the plant lacks water, fertilizer or has abnormal growth. Color analysis can identify problems such as yellowing and withering of leaves in advance, so that corresponding measures can be taken as soon as possible. Common pests and diseases can be identified by analyzing features such as lesions and traces of pests on the leaves. For example, powdery mildew, red spider mites, etc., use image recognition technology to detect whether there are foreign objects, spots or abnormal shapes on the leaf surface, use deep learning to classify pests and diseases, and train models to identify different types of pests and diseases. It can accurately identify the presence and types of pests and diseases and provide targeted solutions. Deep learning methods can identify subtle symptoms of pests and diseases, detect potential problems early, and improve garden management efficiency.

[0042] The normal area analysis module analyzes and defines the standards of healthy areas based on the collected data and historical data, and determines the standards of normal areas through data analysis methods, combined with environmental data and plant status data;

[0043] Based on historical data, statistical analysis methods are used to analyze the normal range of different regions, and statistical models are used to estimate the standard of normal areas. The standard values ​​of normal areas in history are determined through data distribution and trend analysis methods, such as soil moisture between 30%-60% is normal, and plant height within a certain range is normal. According to the distribution of historical data, the normal range is defined for each environmental factor and plant growth indicator, providing a standard basis for subsequent judgments. Historical data provides a reference model that helps to identify whether the current collected data is within the normal range.

[0044] Use a single environmental factor or plant growth data to establish the standard for normal areas. For example, for temperature, set a normal range, such as 25℃±3℃. Combine multiple data dimensions and use multivariate statistical analysis to integrate multiple environmental data and plant data into a comprehensive scoring model to determine the standard for comprehensive healthy areas. Train the healthy area classification model through machine learning algorithms to further optimize the standard for healthy areas. Through the comprehensive analysis of single and multiple factors, establish an accurate normal area standard model, combine multiple environmental factors and plant status data, and be able to more comprehensively evaluate the health status of landscaping areas and improve accuracy.

[0045] Compare the real-time collected environmental data with historical data, use threshold detection, data deviation analysis and other methods to determine whether the real-time data meets the normal area standard, calculate the deviation value between the real-time data and the historical health standard, and determine whether the area is a normal area based on the deviation value. By comparing real-time data with historical data, it is possible to identify in real time whether the current area is in a normal state. By judging the deviation value, it is possible to accurately identify abnormalities in the environment or plant status and make timely adjustments.

[0046] According to the established standard model, the entire garden greening area is divided into normal areas and abnormal areas. Cluster analysis is used to divide the area into several categories, and each area is scored for health. Areas with scores below the set threshold are classified as abnormal areas, and vice versa. Based on the health area standard, the garden area is accurately divided into normal areas and abnormal areas, which is convenient for subsequent maintenance decisions. While classifying the areas, the system automatically generates a health score report to help garden managers give priority to abnormal areas.

[0047] The abnormal area analysis module compares the normal area with other areas, analyzes the abnormal area, and identifies the abnormal area based on the anomaly detection algorithm and the output of the garden image processing module;

[0048] The specific method by which the abnormal area analysis module analyzes the abnormal area is as follows:

[0049] Each data item is processed using Z-Score standardization, and the calculation formula is:

[0050]

[0051] Where, X i is the collected data value of a certain area, μ is the mean value of the area in the historical data, and σ is the standard deviation of the area in the historical data. Through Z-Score conversion, all data are unified to the same scale, which is convenient for subsequent comprehensive analysis and judgment.

[0052] Multidimensional anomaly detection: Use the Mahalanobis distance to detect whether there are abnormal areas in multidimensional data. The Mahalanobis distance calculation formula is:

[0053] D 2 =(X-μ) T S -1 (X-μ)

[0054] In the formula, X is the vector of the data to be tested, μ is the mean vector of each variable, S is the covariance matrix of the data, and D 2 is the Mahalanobis distance;

[0055] Each abnormal area is scored and its abnormal severity is calculated. The severity is evaluated by comprehensively considering the values ​​of Z-Score and Mahalanobis distance. The calculation formula is:

[0056] Abnormal severity = α*|Z|+β*D 2

[0057] In the formula, α and β are weighting coefficients, indicating the weights of Z-Score and Mahalanobis distance in anomaly scoring.

[0058] According to the calculated severity of the abnormality, the abnormal area is graded:

[0059] Mild abnormality: severity score in [0,2);

[0060] Moderate abnormality: severity score in [2,5);

[0061] Highly abnormal: severity score in [5,10);

[0062] Extremely abnormal: Severity score greater than 10.

[0063] The scoring results of abnormal areas are intuitively presented to garden management personnel through charts, heat maps or GIS maps, so that they can be quickly located and processed. According to the scores of abnormal areas, the early warning mechanism is automatically triggered. For high-severity abnormal areas, the system should issue emergency processing warnings and recommend appropriate maintenance measures.

[0064] Assume that the Z-Score of soil moisture data at different locations in a certain park are:

[0065] Region A: Z = 2.5

[0066] Region B: Z = 3.2

[0067] Region C: Z = 0.5

[0068] After analyzing using Mahalanobis distance, we get:

[0069] Region A:D 2 =4.8

[0070] Region B: D 2 =7.1

[0071] Region C:D 2 =1.3

[0072] Assume that the thresholds are set as follows: Z-Score threshold is 3, Mahalanobis distance threshold is 5, weight coefficients α = 0.7, β = 0.3;

[0073] Calculate severity:

[0074] Area A: Severity = 0.7 × 2.5 + 0.3 × 4.8 = 2.88 (mild abnormality)

[0075] Region B: Severity = 0.7 × 3.2 + 0.3 × 7.1 = 4.61 (moderately abnormal)

[0076] Region C: Severity = 0.7 × 0.5 + 0.3 × 1.3 = 0.74 (normal)

[0077] Based on the scores, the system draws the following conclusions:

[0078] Area A: Mild abnormality, need attention.

[0079] Area B: Moderately abnormal, further examination is required.

[0080] Region C: Normal.

[0081] The monitoring and early warning analysis module integrates the data from normal area analysis and abnormal area analysis to conduct intelligent early warning analysis, evaluate potential problems and generate early warning information. It uses multivariate analysis algorithms to conduct comprehensive analysis on the collected multidimensional data to generate early warning prediction results.

[0082] The multivariate analysis algorithm is used to conduct a comprehensive analysis of the collected multidimensional data to generate early warning prediction results. Based on the results of the multivariate analysis, combined with historical data and real-time data, the current environment and plant status are evaluated to see if they are close to the boundaries of potential abnormal areas or healthy areas, and the risk factor of each area is calculated to assess the severity and probability of potential problems. Through risk assessment, the severity of potential problems is quantified to provide a basis for decision-making. According to the risk factor, the system can prioritize high-risk areas and take preventive measures in advance.

[0083] The formula for calculating the risk factor for each area is:

[0084]

[0085] Where W i is the weight of the ith factor, S i is the standardized score of the ith factor;

[0086] Use machine learning or statistical modeling technology to build an early warning prediction model. Based on historical data, real-time data collection and comprehensive analysis results, predict future anomalies and set early warning thresholds for each potential problem, such as when the temperature exceeds a certain threshold or when the humidity fluctuates greatly, and generate an alarm signal. By training the model, it is possible to predict and generate high-precision early warning signals to help garden managers take timely action. When the data exceeds the preset threshold, the system can automatically generate early warning information, reduce human intervention and improve response speed.

[0087] The monitoring and early warning output module provides real-time feedback of early warning information to management personnel based on the output of the monitoring and early warning analysis module.

[0088] The monitoring and early warning output module feeds back early warning information to garden management personnel in real time based on the output of the monitoring and early warning analysis module, ensuring that they can promptly understand the abnormal conditions and potential risks in the garden greening area. The early warning results are conveyed to relevant personnel through multiple information transmission channels, providing detailed abnormal descriptions, affected areas, possible causes, recommended treatment measures, and urgency. The module can also generate visual reports to help managers quickly understand the early warning situation and automatically adjust subsequent management strategies according to the early warning level, so as to take timely and effective maintenance measures to ensure the health and sustainable development of the garden greening area. At the same time, the module supports the historical recording and query functions of early warning information, which is convenient for later analysis and decision support.

[0089] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0090] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A landscaping maintenance monitoring and early warning system, characterized in that: It includes a garden information collection module, a garden image processing module, a normal area analysis module, an abnormal area analysis module, a monitoring and early warning analysis module and a monitoring and early warning output module; The garden information collection module is responsible for collecting environmental data and plant growth data of the garden greening area in real time; The garden image processing module obtains images of the garden greening area through a camera or a drone, performs image processing and analysis, identifies vegetation health conditions, pests and diseases, and leaf color changes, and helps determine the overall health status of the garden area; The normal area analysis module analyzes and defines the standards of the healthy area based on the collected data and historical data, and determines the standards of the normal area by combining the environmental data and the plant status data through the data analysis method; The abnormal area analysis module compares the normal area with other areas, analyzes the area with abnormalities, and identifies the abnormal area based on the abnormality detection algorithm and the output of the garden image processing module; The monitoring and early warning analysis module integrates the data from normal area analysis and abnormal area analysis to perform intelligent early warning analysis, evaluate potential problems and generate early warning information, and use multivariate analysis algorithms to conduct comprehensive analysis on the collected multidimensional data to generate early warning prediction results; The monitoring and early warning output module feeds back early warning information to management personnel in real time according to the output of the monitoring and early warning analysis module.

2. A garden greening maintenance monitoring and early warning system according to claim 1, characterized in that: The garden information collection module is further responsible for real-time collection of environmental data and plant growth data of the garden greening area, including temperature, humidity, light intensity, soil pH, soil moisture, precipitation, wind speed, and plant leaf color, leaf area, stem height, and root growth, and data collection is performed through various sensors installed in the park.

3. A garden greening maintenance monitoring and early warning system according to claim 1, characterized in that: The garden image processing module further includes using a threshold segmentation method, an edge detection method, a region growing method or a semantic segmentation method based on deep learning to separate the plant area in the image from the background. For images with complex backgrounds, a deep learning algorithm is used to accurately segment plant areas of different categories, and image processing is performed through color space to extract the green component of the leaves and analyze the color changes of the leaves.

4. A garden greening maintenance monitoring and early warning system according to claim 1, characterized in that: The normal area analysis module further includes analyzing the normal range of different areas based on historical data using statistical analysis methods, estimating the standards of normal areas using statistical models, determining the standard values ​​of historical normal areas through data distribution and trend analysis methods, establishing the standards of normal areas using single environmental factors or plant growth data, comparing real-time collected environmental data with historical data, using threshold detection, data deviation analysis and other methods to determine whether the real-time data meets the normal area standards, calculating the deviation value between the real-time data and the historical health standards, and determining whether the area is a normal area based on the deviation value, and dividing the entire landscaping area into normal areas and abnormal areas based on the established standard model.

5. A garden greening maintenance monitoring and early warning system according to claim 1, characterized in that: The abnormal area analysis module further includes a specific method of analyzing the abnormal area: Each data item is processed using Z-Score standardization, and the calculation formula is: Where, X i is the collected data value of a certain area, μ is the mean value of the area in the historical data, and σ is the standard deviation of the area in the historical data. Through Z-Score conversion, all data are unified to the same scale. Use the Mahalanobis distance to detect whether there are abnormal areas in the data. The Mahalanobis distance calculation formula is: D 2 =(X-μ) T S -1 (X-μ) In the formula, X is the vector of the data to be tested, μ is the mean vector of each variable, S is the covariance matrix of the data, and D 2 is the Mahalanobis distance; Each abnormal area is scored and its abnormal severity is calculated. The severity is evaluated by comprehensively considering the values ​​of Z-Score and Mahalanobis distance. The calculation formula is: Abnormal severity = α*|Z|+β*D 2 In the formula, α and β are weighting coefficients, indicating the weights of Z-Score and Mahalanobis distance in anomaly scoring.

6. A garden greening maintenance monitoring and early warning system according to claim 1, characterized in that: The monitoring and early warning analysis module further includes using a multivariate analysis algorithm to conduct a comprehensive analysis of the collected multidimensional data to generate early warning prediction results. Based on the results of the multivariate analysis, combined with historical data and real-time data, it is evaluated whether the current environment and plant status are close to the boundary of a potential abnormal area or a healthy area, and the risk factor of each area is calculated. The severity and probability of occurrence of potential problems are evaluated, and an early warning prediction model is constructed using machine learning or statistical modeling technology. Based on historical data, real-time collected data and comprehensive analysis results, future anomalies are predicted, and early warning thresholds are set for each potential problem.

Images