Data monitoring method, system and device suitable for garden environment

Abstract

This invention relates to the field of data monitoring technology, specifically to a data monitoring method, system, and device suitable for garden environments. The method includes: deploying multiple sensor categories in the garden environment, each sensor category including multiple monitoring sensors; collecting garden environment data from the monitoring sensors to construct a historical garden environment sequence; analyzing the historical garden environment sequence, filtering distorted environmental data segments, and identifying and defining the sensor category corresponding to the distorted sensors as the disturbed distorted sensor category; obtaining the distorted environmental data segments corresponding to the distorted sensors in the disturbed distorted sensor category using smoothing coefficient prediction to obtain a joint predicted environmental data segment; performing time analysis on the distorted environmental data segment and the joint predicted environmental data segment to determine the environmental dimension distortion of the joint predicted environmental data segment; and re-obtaining the smoothing coefficient prediction joint predicted environmental data segment based on the environmental dimension distortion to obtain the true predicted data segment, thus completing the data monitoring of the garden environment.

Description

Technical Field

[0001] This invention relates to the field of data monitoring technology, and specifically to a data monitoring method, system, and equipment suitable for garden environments. Background Technology

[0002] Gardens are experiential scenic areas built to simulate natural environments. Gardeners plant a large number of trees according to a pre-designed plan, and some gardens also have many rare tree species, rocks, or artificial hills permanently placed within the garden for visitors to experience. Because gardens are artificial ecological parks, their ability to maintain ecological balance is relatively weaker compared to purely natural environments. Without regular monitoring and maintenance of the garden environment, environmental problems such as ecological imbalance and eutrophication of soil and water may occur, damaging the original experiential value and economic benefits of the garden. Therefore, it is necessary to construct a data monitoring system suitable for garden environments to achieve timely monitoring of garden conditions.

[0003] Existing technologies typically involve installing various monitoring sensors, such as soil moisture detectors and temperature sensors, throughout the garden to create a garden environmental data monitoring platform. These sensors collect environmental data and wirelessly upload it to the platform, which then remotely monitors the transmitted data. However, in real-world scenarios, limitations such as construction area and maintenance costs mean that gardens are often located in remote urban areas. Facilities within garden areas often lack stable mains power and cellular network signals, leading to distortion in the environmental monitoring data uploaded by the sensors and affecting the effectiveness of the monitoring. Summary of the Invention

[0004] To address the technical problem that existing data monitoring devices for garden environments suffer from data distortion, affecting the effectiveness of monitoring, the present invention aims to provide a data monitoring method suitable for garden environments. The specific technical solution adopted is as follows: Multiple sensor categories are deployed in the garden environment, with each category including multiple monitoring sensors. Garden environment data is collected from the monitoring sensors to construct a historical garden environment sequence. Analyze the historical garden environment sequence, filter out distorted environment data segments, and determine and define the sensor category corresponding to the distorted sensor as the disturbed distorted sensor category. Obtain the distortion environment data segment corresponding to the distortion sensor in the category of disturbed distortion sensor predicted by the smoothing coefficient, and obtain the joint predicted environment data segment; A time analysis is performed on the combined distorted environmental data segment and the joint prediction environmental data segment to determine the environmental dimension distortion of the joint prediction environmental data segment. Based on the distortion of the environmental dimension, the smoothing coefficient is re-acquired to predict the joint prediction of the environmental data segment, and the real prediction data segment is obtained to complete the data monitoring of the garden environment.

[0005] Preferably, multiple sensor types are deployed in the garden environment, with each sensor type including multiple monitoring sensors. Garden environment data is collected from these monitoring sensors to construct a historical garden environment sequence, including: Multiple monitoring sensors are deployed in the garden environment to monitor data, and a data monitoring platform is built. Each monitoring sensor stores its corresponding geographic coordinates in the data monitoring platform. Based on the data monitoring platform, the garden environment data collected by each monitoring sensor is exported, and a corresponding historical garden environment sequence is constructed.

[0006] Preferably, the historical garden environment sequence is analyzed, distorted environment data segments are filtered out, and the sensor category corresponding to the distorted sensor is determined and defined as the disturbed distorted sensor category, including: The historical garden environment sequence was divided to obtain multiple local environmental data segments, and distorted environmental data segments were filtered out. The monitoring sensors corresponding to the distorted environmental data segments are identified as suspected distorted sensors, and the distance measurement is determined by the geographical coordinates of each pair of suspected distorted sensors. Multiple clusters are obtained based on distance metrics. Distortion sensors are then screened through these clusters, and the sensor category corresponding to each distortion sensor is defined as the disturbed distortion sensor category.

[0007] Preferably, the historical garden environment sequence is divided into multiple local environmental data segments, and distorted environmental data segments are filtered out, including: The fluctuation standard value of the historical garden environment sequence was evaluated by analysis, and the historical garden environment sequence was divided into multiple local environmental data segments; Analyze each local environmental data segment to assess the local fluctuation value, and combine the fluctuation standard value to obtain the degree of instability. Distorted environmental data segments are filtered from local environmental data segments based on the degree of instability.

[0008] Preferably, multiple clusters are obtained based on distance metrics, and distortion sensors are filtered out through these clusters. Specifically: All suspected distorted sensors are clustered based on distance metrics to obtain multiple clusters. The mean distance metric of each cluster is determined, and the suspected distorted sensors in the cluster corresponding to the minimum mean distance metric are selected and defined as distorted sensors.

[0009] Preferably, the distortion environment data segment corresponding to the distortion sensor in the category of disturbed distortion sensor predicted by the smoothing coefficient is obtained to obtain the joint prediction environment data segment, including: Determine the perturbation correlation of each perturbation distortion sensor category, and define the currently analyzed perturbation distortion sensor category as the target category. Based on the difference in perturbation correlation between the target category and the remaining non-target categories of perturbation distortion sensor categories, determine the associated distortion sensor category corresponding to the target category. The correlation weight is obtained by combining the differences in the perturbation correlation between the target category and the associated distortion sensor category. The Holt-Winters seasonality method is used to obtain the smoothing coefficient. The correlation weight is added to the smoothing coefficient to predict the distortion environmental data segment corresponding to any distortion sensor in the target category, so as to obtain the jointly predicted environmental data segment.

[0010] Preferably, a time analysis is performed on the combined distorted environmental data segment and the jointly predicted environmental data segment to determine the environmental dimension distortion of the jointly predicted environmental data segment, including: Define the distorted environment data segment corresponding to the associated distorted sensor category as the prediction reference data segment, and filter the time strongly correlated reference data segment based on any joint prediction environment data segment combined with the prediction reference data segment; Determine the overlap between time-correlated reference data segments and corresponding joint prediction environment data segments to determine the temporal similarity. By analyzing the joint prediction environmental data segment and the time-correlated reference data segment separately, the degree of instability is obtained, and the environmental dimension distortion of the joint prediction environmental data segment is determined by combining the time similarity.

[0011] Preferably, based on the distortion of the environmental dimension, the smoothing coefficient prediction and joint prediction environmental data segment are re-acquired to obtain the true prediction data segment, thus completing the data monitoring of the garden environment, including: By integrating and analyzing joint forecast environmental data segments, the potential period of forecast disturbance at the current moment can be determined. Based on the predicted period of disturbance, the Holt-Winters seasonality method is used to obtain a new smoothing coefficient by combining the smoothing coefficient before reacquisition and the environmental dimension distortion. The new smoothing coefficient is used to predict the joint prediction environment data segment to obtain the true prediction data segment; The distorted environmental data segments are replaced with real predicted data segments, and then spliced ​​with normal environmental data segments from historical garden environmental sequences to form a new data sequence. The garden environment is then monitored using this new data sequence.

[0012] To address the above problems, the present invention further provides: a data monitoring system suitable for garden environments, the system comprising: The data acquisition module is used to: deploy multiple sensor types in the garden environment, with each sensor type including multiple monitoring sensors, collect garden environment data from the monitoring sensors, and construct a historical garden environment sequence; The data classification module is used to: analyze historical garden environment sequences, filter distorted environment data segments, and determine and define the sensor category corresponding to the distorted sensor as the disturbed distorted sensor category. The data processing module is used to: obtain the distortion environment data segment corresponding to the distortion sensor in the category of distortion sensor that is predicted by the smoothing coefficient, and obtain the joint prediction environment data segment; The data analysis module is used to: perform time analysis on the combined distorted environmental data segment and the joint prediction environmental data segment, and determine the environmental dimension distortion of the joint prediction environmental data segment; The data monitoring module is used to: re-acquire smoothing coefficient predictions based on environmental dimension distortion, jointly predict environmental data segments, obtain real prediction data segments, and complete data monitoring of the garden environment.

[0013] To address the aforementioned problems, the present invention also provides a data monitoring device suitable for garden environments, the device comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus, and the processor calls logical instructions in the memory to execute the data monitoring method suitable for garden environments described in any of the preceding claims.

[0014] The present invention has the following beneficial effects: 1. By analyzing the historical garden environment sequences constructed by various monitoring sensors within the garden environment, and combining the correlation between the stability of their electrical signals and the characteristics of their respective monitoring scenarios, prediction and correction of environmental monitoring data can be achieved. Specifically, firstly, distorted environmental data segments are screened, distorting sensors are identified, and the categories of disturbed distorting sensors are determined. Then, a joint predicted environmental data segment is obtained from the distorted environmental data segments. A time analysis is performed on both types of environmental data segments to determine the distortion in the environmental dimension. Finally, a smoothing coefficient is re-obtained based on the environmental dimension distortion to predict the joint predicted environmental data segment, resulting in a true predicted data segment that replaces the original distorted environmental data segment. While ensuring data integrity, this method can effectively reduce the distortion of garden environment data uploaded by monitoring sensors, making the data monitoring process for the garden environment more complete and accurate, and can be widely applied in scenarios such as intelligent sensors, intelligent sensing systems, and intelligent sensing elements.

[0015] 2. The data monitoring system and equipment for garden environments provided by this invention have the same beneficial effects as the data monitoring method for garden environments provided by this invention, and will not be described in detail here. Attached Figure Description

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating the steps of a data monitoring method suitable for garden environments, provided in one embodiment of the present invention; Figure 2 This is a block diagram illustrating the implementation structure of a data monitoring system suitable for garden environments, provided as an embodiment of the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a data monitoring method, system, and device suitable for garden environments proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] The following description, in conjunction with the accompanying drawings, details a specific solution for a data monitoring method, system, and equipment suitable for garden environments provided by the present invention.

[0021] If the monitoring sensors placed throughout the garden are affected by power supply interference due to the garden's offset location, it will impact the monitoring sensors in various environmental dimensions in terms of time and space, leading to data acquisition deviations and reducing the accuracy of the monitoring data. Therefore, by analyzing the historical garden environment sequences constructed by each monitoring sensor in the garden environment, and combining the analysis of the stability of their electrical signals and the correlation between their respective monitoring scene characteristics, we can achieve predictive correction of environmental monitoring data. That is, we can make accurate predictions on the distorted monitoring data in a second stage, ensuring the effectiveness of the monitoring sensors in monitoring the garden environment, and ensuring that the actual operating status of the monitoring sensors can be clearly displayed. This directly corresponds to the technical classification of intelligent sensing systems, reflecting the technological innovation and industrial application value in the field of intelligent sensor industry.

[0022] Please see Figure 1The diagram illustrates a flowchart of a data monitoring method for garden environments according to an embodiment of the present invention, the method comprising: Step S1: Deploy multiple sensor types in the garden environment, with each sensor type including multiple monitoring sensors, collect garden environment data from the monitoring sensors, and construct a historical garden environment sequence; Step S2: Analyze the historical garden environment sequence, filter out distorted environment data segments, and determine and define the sensor category corresponding to the distorted sensor as the disturbed distorted sensor category; Step S3: Obtain the distortion environment data segment corresponding to the distortion sensor in the category of distortion sensor predicted by the smoothing coefficient, and obtain the joint prediction environment data segment; Step S4: Perform time analysis on the combined distorted environmental data segment and the joint prediction environmental data segment to determine the distortion of the environmental dimension of the joint prediction environmental data segment; Step S5: Based on the distortion of the environmental dimension, re-acquire the smoothing coefficient prediction and joint prediction environmental data segment to obtain the real prediction data segment, and complete the data monitoring of the garden environment.

[0023] The garden environment refers to an outdoor space system within a specific area, planned, designed, constructed, and managed artificially. It is primarily composed of elements such as plants, water bodies, topography, buildings, roads, and various landscape facilities, possessing certain ecological functions, aesthetic value, and cultural connotations. It belongs to the category of man-made ecological gardens. Compared to purely natural environments without human intervention, its ability to maintain ecological balance and stability is limited. Therefore, real-time monitoring of the garden environment is usually necessary to prevent environmental problems such as frequent plant diseases and pests, soil degradation, water eutrophication, and invasive alien species. However, due to the often remote location of garden environments and the lack of stable municipal power supply and reliable cellular network signal coverage, various monitoring sensors frequently experience signal interruptions, data loss, or transmission delays during data collection, resulting in a certain degree of distortion and bias in the garden environment data. Therefore, a data monitoring method suitable for garden environments is proposed, aiming to ensure the effectiveness and practical application value of the final monitoring results by predicting and correcting the collected garden environment data.

[0024] Further, step S1 includes: Step S11: Deploy multiple monitoring sensors in the garden environment to implement data monitoring, build a data monitoring platform, and store the corresponding geographic coordinates of each monitoring sensor in the data monitoring platform.

[0025] As an optional implementation method, the sensor categories include various types such as soil moisture monitoring equipment, temperature sensors, humidity sensors, light intensity sensors, carbon dioxide concentration sensors, wind speed and direction sensors, and rainfall sensors. Among them, soil moisture monitoring equipment is used to monitor soil moisture content in real time; temperature sensors are used to monitor the temperature of multiple environmental parameters; humidity sensors are used to accurately measure air humidity; light intensity sensors are used to sense light intensity; carbon dioxide concentration sensors are used to monitor the carbon dioxide content in the air; wind speed and direction sensors are used to monitor the garden environment and measure wind speed and direction; and rainfall sensors are used to record precipitation to comprehensively and accurately collect environmental data.

[0026] It can be explained that each sensor category corresponds to multiple monitoring sensors, which together build a comprehensive data monitoring platform and store the geographical coordinates of each monitoring sensor, that is, the coordinates of the monitoring sensor's installation location, providing a data foundation for subsequent analysis.

[0027] Step S12: Export the garden environment data collected by each monitoring sensor based on the data monitoring platform, and construct the corresponding historical garden environment sequence.

[0028] Specifically, the garden environment data of each monitoring sensor in various sensor categories at the current moment is exported from the data monitoring platform and arranged into a sequence according to time order, which serves as the historical garden environment sequence of each monitoring sensor. Each historical garden environment sequence contains multiple historical garden environment data and one garden environment data, and the garden environment data is located at the last position in the historical garden environment sequence. That is, the garden environment data is the data corresponding to the current moment, and the historical garden environment data is the data corresponding to the historical moment.

[0029] To better illustrate this, since the monitoring sensors are manufactured by different companies and the environmental dimensions being monitored are different, the frequency at which they perform data monitoring is usually not completely consistent. Therefore, in this embodiment, it is assumed that the sampling frequency of each monitoring sensor is based on its own lowest sampling frequency.

[0030] Understandably, within a garden environment, environmental characteristics such as temperature, humidity, and pH value at the same dimension will be monitored using the same type of sensor. In actual monitoring, the garden environment is usually divided into multiple zones, with various types of monitoring sensors deployed in each zone. When the circuit connected to the monitoring sensor experiences unstable changes, it usually affects the monitoring sensors deployed in the corresponding zone within the garden environment, resulting in significant unstable fluctuations in the collected garden environment data. Therefore, by analyzing the spatial range of unstable fluctuations formed by different sensor types based on historical garden environment sequences, we can identify monitoring sensors that may cause distortion and construct a category of disturbed and distorted sensors.

[0031] Further, step S2 includes: Step S21: Divide the historical garden environment sequence to obtain multiple local environmental data segments, and filter out distorted environmental data segments.

[0032] It is explained that distortion refers to the presence of outliers, noise, or missing information in the garden environment data that do not conform to the actual situation. It may manifest as data mutations, abnormal patterns, or incomplete information, affecting the accuracy and reliability of the data.

[0033] Further, step S21 includes: Step S211: Analyze the historical garden environment sequence to assess the standard value of fluctuation, and divide the historical garden environment sequence to obtain multiple local environmental data segments.

[0034] Specifically, based on any monitoring sensor, denoted as monitoring sensor A, the analysis is conducted by calculating the standard deviation of the corresponding historical garden environment sequence, which is then used as the standard value of fluctuation, denoted as . This is used to reflect the overall degree of change in garden environment data within the historical garden environment sequence. Then, the historical garden environment sequence is divided equally according to the original sequence length to obtain multiple local environment data segments. Preferably, in this embodiment, the historical garden environment sequence is divided into 10 local environment data segments according to 10% of the original sequence length, which can be adjusted according to the actual situation. In particular, if there is insufficient data in a local environment data segment during the division process, resulting in inconsistencies in the local environment data segments, the local environment data segments with insufficient data can be directly omitted.

[0035] Step S212: Analyze each local environmental data segment to evaluate the local fluctuation value, and combine the fluctuation standard value to obtain the degree of instability.

[0036] Specifically, based on the local environmental data segments obtained in step S211, analysis is performed, and the standard deviation of each local environmental data segment is calculated as the local fluctuation value of the corresponding local environmental data segment, denoted as... This is used to reflect the degree of drastic change in local environmental data segments and determine the degree of instability in each local environmental data segment. The corresponding calculation formula is: in, This indicates the degree of instability in a local environmental data segment; Represents the normalization function; This represents the local fluctuation value of a local environmental data segment; Standard values ​​representing the fluctuations in the historical garden environment sequence; This represents absolute value operations.

[0037] It can be noted that both the local fluctuation value and the standard fluctuation value are known; therefore, the normalization function... A minimum-maximum normalization function is used. Here, the normalization range is the set of numerical values ​​calculated from all local environmental data segments divided under the historical garden environment sequence, linearly transforming the data to... Within this scope, the performance of each local environmental data segment can be analyzed more intuitively. The degree of instability of the local environmental data segment reflects the possibility that there may be significant distortion in the corresponding local environmental data segment. The greater the degree of instability of the local environmental data segment, the more obvious and unstable the fluctuation of the garden environment data is compared with the overall historical garden environment sequence, reflecting that the corresponding local environmental data segment is more likely to have distortion.

[0038] Step S213: Filter distorted environment data segments from local environment data segments based on the degree of instability.

[0039] It is explained that a preset screening threshold is set to 1. In this embodiment, the screening threshold is 0.4, which can be adjusted according to the actual situation. Similarly, based on step S212, the degree of instability of all local environmental data segments is obtained and compared with the screening threshold. Local environmental data segments with a degree of instability greater than the screening threshold are regarded as distorted environmental data segments; conversely, local environmental data segments with a degree of instability less than or equal to the screening threshold are regarded as normal environmental data segments.

[0040] Step S22: Define the monitoring sensor corresponding to the distorted environmental data segment as a suspected distorted sensor, and determine the distance measurement by the geographical coordinates of each pair of suspected distorted sensors.

[0041] Specifically, the monitoring sensor to which the distorted environmental data segment belongs is regarded as a suspected distorted sensor, and the absolute value of the difference between the geographic coordinates of two suspected distorted sensors is used as the distance metric. The absolute value of the difference is the Euclidean distance between the two geographic coordinates to avoid distance metric calculation errors caused by the different positive and negative directions of the geographic coordinates.

[0042] Step S23: Based on the distance metric, obtain multiple clusters, filter out distorted sensors through clusters, and define the sensor category corresponding to the distorted sensor as the disturbed distorted sensor category.

[0043] To clarify, a suspected distorted sensor refers to a monitoring sensor that may have signal distortion after analysis in step S21; a distorted sensor indicates a monitoring sensor that has signal distortion.

[0044] Further, in step S23, multiple clusters are obtained based on distance metrics, and distortion sensors are filtered out through these clusters, specifically as follows: All suspected distorted sensors are clustered based on distance metrics to obtain multiple clusters. The mean distance metric of each cluster is determined, and the suspected distorted sensors in the cluster corresponding to the minimum mean distance metric are selected and defined as distorted sensors.

[0045] Specifically, based on the distance metric, all suspected distorted sensors are clustered using DBSCAN (Density-Based Spatial Clustering of Applications with Noise) to obtain several clusters. Then, the mean of the overall distance metric for each cluster is calculated based on the distance metric in each cluster. After comparison, the suspected distorted sensors in the cluster corresponding to the smallest mean of the overall distance metric are identified as distorted sensors, and the sensor category to which the distorted sensor belongs is identified as the disturbed distorted sensor category. That is, at the current moment, there may be multiple disturbed distorted sensor categories, each disturbed distorted sensor category contains multiple distorted sensors, and each distorted sensor contains multiple distorted environmental data segments.

[0046] Understandably, based on the categorized perturbed distortion sensors, since various environmental characteristics such as temperature, humidity, light, and soil moisture within the garden environment are interconnected and influence each other, it is difficult to fully reflect the true trend of environmental changes using a single monitoring sensor. Therefore, it is necessary to combine the historical garden environment sequences of different monitoring sensors to jointly predict the data. This ensures that the prediction data acquired by the corresponding monitoring sensors does not consider too many factors, i.e., it does not only consider the historical garden environment sequence of its own monitoring sensor for prediction, thus avoiding prediction bias caused by a single data source and improving the prediction accuracy of each monitoring sensor.

[0047] Furthermore, step S3 includes: Step S31: Determine the perturbation correlation of each perturbation distortion sensor category, and define the currently analyzed perturbation distortion sensor category as the target category. Based on the difference in perturbation correlation between the target category and the remaining non-target categories of perturbation distortion sensor categories, determine the associated distortion sensor category corresponding to the target category.

[0048] Specifically, taking the target category as an example, which is the category of disturbed distortion sensors currently being analyzed, the mean of the Pearson correlation coefficients among all distorted environment data segments in the target category is calculated. First, the Pearson correlation coefficient between each pair of distorted environment data segments is calculated. This involves standardizing each pair of distorted environment data segments by subtracting their respective means and dividing by their standard deviation to eliminate the influence of dimensions. The covariance of each pair of standardized data segments is then calculated and divided by the product of the standard deviations of the two distorted environment data segments to obtain the Pearson correlation coefficient. Next, the Pearson correlation coefficients are summed and divided by the total number of pairs to obtain the mean. Finally, the mean of the Pearson correlation coefficients is used as the disturbed correlation of the target category. Similarly, the disturbed correlation of all disturbed distortion sensor categories is obtained. The Pearson correlation coefficient is a statistical indicator used to quantify the degree of linear correlation between two variables, and its numerical range is... A result of 1 indicates a perfect positive linear correlation; a result of -1 indicates a perfect negative linear correlation; and a result of 0 indicates no linear correlation.

[0049] Next, the perturbation correlation difference value is calculated between the target category and the remaining non-target categories of perturbed distortion sensor categories. For example, assuming the perturbation correlation of the target category is... The perturbation correlation of any non-target category of perturbation distortion sensor category is: The perturbation association difference value between the two categories is , This represents absolute value operation, used to ensure that the disturbance correlation difference value is always greater than or equal to 0. A preset screening threshold of 2 is used. In this embodiment, the screening threshold of 2 is 0.6, which can be adjusted according to the actual situation. Disturbed distortion sensor categories that are not in the target category and whose disturbance correlation difference value is less than the screening threshold of 2 are regarded as the target category of the associated distortion sensor category. Conversely, disturbed distortion sensor categories that are not in the target category and whose disturbance correlation difference value is greater than or equal to the screening threshold of 2 are not further analyzed.

[0050] Step S32: Obtain the correlation weight by combining the perturbation correlation difference between the target category and the associated distortion sensor category, obtain the smoothing coefficient by using the Holt-Winters seasonality method, and add the correlation weight to the smoothing coefficient to predict the distortion environment data segment corresponding to any distortion sensor in the target category, so as to obtain the jointly predicted environment data segment.

[0051] To better illustrate, the Holt-Winters seasonality method is a statistical technique widely used in time series forecasting. It is specifically designed to handle data with significant seasonal fluctuations and trend changes. It introduces a seasonal component adjustment mechanism on the basis of traditional exponential smoothing, which can more accurately capture and predict the periodic changes in data. That is, it decomposes the time series into three parts: horizontal, trend and seasonal, and dynamically updates and fits these components to achieve data forecasting.

[0052] Specifically, the mean of the perturbation correlation difference between the target category and all associated distortion sensor categories is calculated. That is, the perturbation correlation difference between the target category and the associated distortion sensor categories is obtained according to step S31, and the sum is divided by the total number of pairs to obtain the mean of the perturbation correlation difference, which is used as the correlation weight of the target category. Taking any distortion sensor in the target category as an example, the correlation weight is added to the horizontal smoothing coefficient, trend smoothing coefficient, and seasonal smoothing coefficient to predict the distortion environment data segment of the distortion sensor. That is, for any distortion environment data segment of the distortion sensor currently being analyzed, the horizontal, trend, and seasonal components are identified in the adjacent normal local environment data segments. Here, the horizontal refers to the current garden environment data of the data segment, which represents the value of the last result; the trend refers to the direction and magnitude of the change of garden environment data over time; and the seasonality refers to the periodic fluctuation of garden environment data. The three components are smoothed by the Holt-Winters seasonality method to obtain the corresponding smoothing coefficients, and then the correlation weight is applied to obtain the predicted data segment of the corresponding distortion environment data segment, which is used as the joint predicted environment data segment. Similarly, the joint predicted environment data segment of all distortion sensors is obtained.

[0053] It should be noted that in practical applications, the method of applying correlation weights can be adjusted according to the actual situation. When there is an interaction or nonlinear relationship between the three smoothing coefficients, the correlation weights are applied to each smoothing coefficient by multiplication. Conversely, when the three smoothing coefficients are independent and linear, the correlation weights are applied by addition. For example, when the garden environment data in the local environmental data segment shows obvious seasonal changes, its fluctuation may increase with the increase of the level component. In this case, the correlation weights are applied by multiplication.

[0054] Specifically, when performing data prediction for each distorted environmental data segment, the input garden environmental data is the normal data preceding the corresponding distorted environmental data segment. Therefore, the length of the resulting joint predicted environmental data segment is consistent with the length of the corresponding distorted environmental data segment, that is, one distorted environmental data segment corresponds to one joint predicted environmental data segment.

[0055] Understandably, based on the aforementioned steps, monitoring sensors that may exhibit distortion were identified, and preliminary prediction results for the corresponding monitoring sensors were obtained, i.e., the joint prediction environmental data segment. Since the monitoring sensors differ in model and data sampling frequency, their response times on the corresponding monitoring data also differ. These response times, to a certain extent, represent the degree of instability in the corresponding garden environment dimension when the power grid is unstable. In other words, the response times reflect the degree of instability of the monitoring sensors in different garden environment dimensions when the power grid is unstable. Therefore, based on the joint prediction environmental data segment, the distribution of fluctuation breakpoints in the garden environment data collected by the corresponding monitoring sensors when the power grid is unstable is analyzed, thereby assessing the environmental dimension distortion in the joint prediction environmental data segment of the corresponding monitoring sensors.

[0056] Further, step S4 includes: Step S41: Define the distorted environment data segment corresponding to the associated distorted sensor category as the prediction reference data segment, and filter the time strongly correlated reference data segment based on any joint prediction environment data segment combined with the prediction reference data segment.

[0057] Specifically, taking the target category determined in step S3 above as an example, each distortion environment data segment covered in each associated distortion sensor category corresponding to the target category is used as the prediction reference data segment of the target category; then, taking any joint prediction environment data segment as an example, based on each associated distortion sensor category, the prediction reference data segment with the closest time distance to the joint prediction environment data segment is selected as the time strongly associated reference data segment, and the time strongly associated reference data segment for each associated distortion sensor category is determined accordingly.

[0058] Step S42: Determine the overlap between the time-correlated reference data segment and the corresponding joint prediction environment data segment to determine the temporal similarity.

[0059] Specifically, taking any strongly correlated time reference data segment determined in step S41 as an example, the time similarity is determined by comparing it with the joint prediction environment data segment: when the strongly correlated time reference data segment and the joint prediction environment data segment do not overlap, it means that the two data segments are not in the same time period, but are independently distributed one after the other. The inverse proportional value of the duration of the remaining time between the two data segments is obtained, that is, the inverse proportional value of the duration of a single time. Here, the duration of the remaining time between the two data segments refers to the time interval between the two data segments. For example, assuming that the time period corresponding to the strongly correlated time reference data segment is 10:03-10:06 and the time period corresponding to the joint prediction environment data segment is 10:08-10:11, the duration is 10:06-10:08, with an interval of 2 minutes. Taking the reciprocal of this time interval, the inverse proportional value is obtained. That is, by using the inverse proportional value to reflect the inverse relationship between time length and proportion, this value is used as the temporal similarity between the joint prediction environmental data segment and the corresponding strongly correlated reference data segment, denoted as . .

[0060] When there is partial overlap between the strongly correlated time reference data segment and the joint prediction environmental data segment, it indicates that some of the landscape environment data in the two data segments fall within the same time period, but are still distributed sequentially. In this case, the inverse proportional value of the duration between the starting points of the two time periods is obtained as the temporal similarity between the joint prediction environmental data segment and the corresponding strongly correlated time reference data segment. Here, the duration refers to the time interval between the starting points of the two time periods. For example, assuming the time period corresponding to the strongly correlated time reference data segment is 10:03-10:06 and the time period corresponding to the joint prediction environmental data segment is 10:05-10:07, then the duration is 10:03-10:05, with an interval of 2 minutes, resulting in the inverse proportional value. .

[0061] It should be added that the inverse proportional value, i.e. the acquisition of time similarity, is only an example. When collecting garden environment data in actual situations, the time interval is usually accurate to the second to capture subtle changes. Therefore, it is almost impossible to determine the ideal case of an inverse proportional value of 1 for the time interval.

[0062] When the time-strongly correlated reference data segment and the joint prediction environment data segment completely overlap, it means that the two data segments are in the same time period. The time similarity between the joint prediction environment data segment and the corresponding time-strongly correlated reference data segment is directly set to 1.

[0063] It can be noted that due to the instability of the power grid, the response time of the monitoring sensors collecting garden environment data varies. Therefore, the starting time of each constructed historical garden environment sequence may be different. Consequently, for the segmented local environment data, the jointly predicted environment data segment determined according to the aforementioned steps may have three possibilities: non-overlap, partial overlap, and complete overlap with the corresponding strongly correlated reference data segment. Therefore, corresponding analysis is required according to different situations to obtain the temporal similarity of each situation, the range of which is... The greater the temporal similarity, the more likely there is a significant correlation between the joint prediction environmental data segment and the corresponding strongly temporally correlated reference data segment in terms of temporal changes. This reflects that the joint prediction environmental data segment and the corresponding strongly temporally correlated reference data segment are more likely to be caused by the same unstable situation. In other words, the temporal similarity indicates the degree to which the prediction environmental data segment and the strongly temporally correlated reference data segment may be caused by the same unstable situation.

[0064] Step S43: Analyze the joint prediction environmental data segment and the time-correlated reference data segment respectively to obtain the degree of instability, and determine the environmental dimension distortion of the joint prediction environmental data segment by combining the time similarity.

[0065] Specifically, following step S21, the degree of instability of the joint prediction environment data segment and the time-strongly correlated reference data segment is obtained similarly, and the degree of instability of the joint prediction environment data segment is denoted as... The degree of instability of a strongly correlated time-series reference data segment is denoted as... By combining the time similarity, the environmental dimension distortion of the currently analyzed joint prediction environmental data segment is determined, in order to reflect the content of distortion information left by environmental influences in the joint prediction environmental data segment obtained after the initial data prediction. The corresponding calculation formula is as follows: in, This indicates the distortion of the environmental dimension in the joint prediction environmental data segment currently being analyzed; Represents the normalization function; This indicates the number of time-correlated reference data segments corresponding to the joint prediction environment data segment currently being analyzed; This indicates the first segment of the joint forecast environmental data segment currently being analyzed. A time-strongly correlated reference data segment; Indicates the first The degree of instability of a strongly correlated time-series reference data segment; This indicates the degree of instability in the current analysis of the joint forecast environmental data segment; This indicates the first segment of the joint forecast environmental data segment currently being analyzed. The temporal similarity of strongly correlated reference data segments.

[0066] It can be explained that the normalization function The data is linearly transformed using a maximum-minimum normalization function. Within the range; if the environmental dimension distortion is greater, it indicates that the power grid instability causes interference and impact on the garden environment, and the interference on the joint prediction environmental data segment after the initial data prediction is greater, reflecting the more distortion information left by the environmental influence on the joint prediction environmental data segment after the initial data prediction; similarly, the environmental dimension distortion of all joint prediction environmental data segments can be obtained.

[0067] Understandably, environmental distortion can be used to clearly represent the degree of distortion caused by the instability of mains power and cellular network signals in the collected garden environment data by the corresponding monitoring sensors. Garden environments, being scenic areas, experience peak and off-peak seasons throughout the year. Therefore, when making data predictions, the Holt-Winters seasonality method is usually chosen to quantify the seasonal components of the garden environment and incorporate them into the predictive analysis. However, in actual data prediction, the Holt-Winters seasonality method itself includes horizontal and trend components in addition to seasonal components, which need to be considered together for data prediction. The distortion effect of environmental distortion in different components varies depending on the component category, resulting in different prediction interferences. Therefore, it is necessary to calculate the smoothing coefficients of the horizontal, trend, and seasonal components in the Holt-Winters seasonality method based on environmental distortion.

[0068] Furthermore, step S5 includes: Step S51: Integrate and analyze the joint prediction environmental data segments to determine the possible period of the prediction that may be disturbed at the current moment.

[0069] Specifically, the union of the time periods occupied by all joint prediction environmental data segments is taken as the disturbed prediction time period at the current moment. Therefore, based on the comprehensive coverage of multiple joint prediction environmental data segments on the time axis, there may be multiple disturbed prediction time periods at the current moment, so as to reflect the time intervals that may have an impact.

[0070] Step S52: Based on the disturbed prediction time period, the Holt-Winters seasonality method is used to obtain a new smoothing coefficient by combining the smoothing coefficient before reacquisition and the environmental dimension distortion.

[0071] Specifically, the current predicted time period is used as the data object for analyzing the seasonal components using the Holt-Winters seasonality method. The seasonal cycle length is then modified to more accurately capture the periodic changes in the garden environment data, resulting in a new seasonal smoothing coefficient. This is then combined with the smoothing coefficient from step S32 to obtain new smoothing coefficients for the horizontal or trend-related components. The corresponding calculation formula is as follows: in, This indicates the new horizontal smoothing coefficient or trend smoothing coefficient; This represents the horizontal smoothing coefficient or trend smoothing coefficient before re-acquisition (the smoothing coefficient in step S32). This indicates the distortion of the environmental dimension in the joint prediction environmental data segment currently being analyzed; Represents the normalization function; This represents the newly acquired seasonal smoothing coefficient; This represents the seasonal smoothing coefficient before re-acquisition.

[0072] It can be explained that the normalization function The data is linearly transformed using a maximum-minimum normalization function. Within the range; similarly, new horizontal smoothing coefficients, trend smoothing coefficients, and seasonal smoothing coefficients are obtained for each joint forecast environmental data segment.

[0073] Step S53: Use the new smoothing coefficient to predict the joint prediction environment data segment to obtain the true prediction data segment.

[0074] Specifically, by using the newly acquired smoothing coefficients, each joint prediction environment data segment is predicted again to obtain the corresponding prediction data sequence, which is then used as the actual prediction data segment.

[0075] Step S54: Replace the distorted environmental data segment with the real predicted data segment, and splice it with the normal environmental data segment of the historical garden environment sequence to form a new data sequence. Use the new data sequence to monitor the garden environment.

[0076] Specifically, the original distorted environmental data segments are replaced with real predicted data segments, and then spliced ​​with other normal environmental data segments from the historical garden environment sequence to form a new garden environment sequence; similarly, new garden environment sequences corresponding to all monitoring sensors are obtained; preferably, the new garden environment sequence is evaluated based on the data monitoring platform, a pre-set warning threshold is set, and the new garden environment sequence is compared with the warning threshold. If the garden environment data in the new garden environment sequence exceeds the warning threshold, the data monitoring platform issues a warning display to prompt relevant personnel to take timely countermeasures.

[0077] Understandably, by analyzing the historical garden environment sequences constructed by various monitoring sensors within the garden environment, and combining the analysis of the stability of their electrical signals and the correlation between their respective monitoring scene characteristics, prediction and correction of environmental monitoring data can be achieved. Specifically, firstly, distorted environmental data segments are screened, distorting sensors are identified, and the category of disturbed distorting sensors is determined. Then, a joint predicted environmental data segment is obtained from the distorted environmental data segments. A time analysis is performed on both types of environmental data segments to determine the distortion in the environmental dimension. Finally, a smoothing coefficient is re-obtained based on the environmental dimension distortion to predict the joint predicted environmental data segment, resulting in a true predicted data segment that replaces the original distorted environmental data segment. While ensuring data integrity, this approach can effectively reduce the distortion of garden environment data uploaded by monitoring sensors, making the data monitoring process for the garden environment more complete and accurate, and thus enabling widespread application in scenarios such as intelligent sensors, intelligent sensing systems, and intelligent sensing elements.

[0078] Please see Figure 2 The second embodiment of the present invention provides a data monitoring system suitable for garden environments, the system comprising: The data acquisition module is used to: deploy multiple sensor types in the garden environment, with each sensor type including multiple monitoring sensors, collect garden environment data from the monitoring sensors, and construct a historical garden environment sequence; The data classification module is used to: analyze historical garden environment sequences, filter distorted environment data segments, and determine and define the sensor category corresponding to the distorted sensor as the disturbed distorted sensor category. The data processing module is used to: obtain the distortion environment data segment corresponding to the distortion sensor in the category of distortion sensor that is predicted by the smoothing coefficient, and obtain the joint prediction environment data segment; The data analysis module is used to: perform time analysis on the combined distorted environmental data segment and the joint prediction environmental data segment, and determine the environmental dimension distortion of the joint prediction environmental data segment; The data monitoring module is used to: re-acquire smoothing coefficient predictions based on environmental dimension distortion, jointly predict environmental data segments, obtain real prediction data segments, and complete data monitoring of the garden environment.

[0079] The data acquisition module is used to acquire raw data and establish a data foundation; the data classification module filters distorted environmental data segments based on the raw data to classify sensor categories; the data processing module, based on the obtained categories of disturbed distorted sensors, processes the distorted environmental data segments to obtain joint predicted environmental data segments; the data analysis module comprehensively analyzes the distorted environmental data segments obtained by the data classification module and the joint predicted environmental data segments obtained by the data processing module to assess the distortion of environmental dimensions; the data monitoring module performs secondary prediction based on the distortion of environmental dimensions, replacing the original distorted environmental data segments and completing the data monitoring work; through the synergistic effect of the data acquisition module, data classification module, data processing module, data analysis module, and data monitoring module, timely monitoring of garden environmental data is achieved.

[0080] The third embodiment of the present invention provides a data monitoring device suitable for garden environments. The device includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory communicate with each other through the communication bus. The processor calls logical instructions in the memory to execute the data monitoring method suitable for garden environments provided in any of the foregoing embodiments.

[0081] It can be noted that both the system and the equipment have the same beneficial effects as the aforementioned data monitoring method for garden environments, and will not be elaborated upon here. The data monitoring system or equipment for garden environments requires the use of the data monitoring method for garden environments provided in any of the aforementioned embodiments during operation. Therefore, integrating the system or equipment with program data or configuring different hardware to produce functions similar to those achieved by this invention falls within the scope of protection of this invention.

[0082] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0083] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method of data monitoring suitable for a garden environment, characterized in that, The method includes: Multiple sensor categories are deployed in the garden environment, with each category including multiple monitoring sensors. Garden environment data is collected from the monitoring sensors to construct a historical garden environment sequence. Analyze the historical garden environment sequence, filter out distorted environment data segments, and determine and define the sensor category corresponding to the distorted sensor as the disturbed distorted sensor category. Obtain the distortion environment data segment corresponding to the distortion sensor in the category of disturbed distortion sensor predicted by the smoothing coefficient, and obtain the joint predicted environment data segment; A time analysis is performed on the combined distorted environmental data segment and the joint prediction environmental data segment to determine the environmental dimension distortion of the joint prediction environmental data segment. Based on the distortion of the environmental dimension, the smoothing coefficient is re-acquired to predict the joint prediction of the environmental data segment, and the real prediction data segment is obtained to complete the data monitoring of the garden environment. The step of obtaining the distortion environment data segment corresponding to the distortion sensor in the category of disturbed distortion sensor for predicting the smoothing coefficient, and obtaining the joint prediction environment data segment, includes: Determine the perturbation correlation of each perturbation distortion sensor category, and define the currently analyzed perturbation distortion sensor category as the target category. Based on the difference in perturbation correlation between the target category and the remaining non-target categories of perturbation distortion sensor categories, determine the associated distortion sensor category corresponding to the target category. The correlation weight is obtained by combining the differences in the perturbation correlation between the target category and the associated distortion sensor category. The Holt-Winters seasonality method is used to obtain the smoothing coefficient. The correlation weight is added to the smoothing coefficient to predict the distortion environmental data segment corresponding to any distortion sensor in the target category, so as to obtain the jointly predicted environmental data segment.

2. The data monitoring method for garden environments according to claim 1, characterized in that, Multiple sensor categories are deployed in the garden environment, with each category including multiple monitoring sensors. Garden environmental data is collected from these sensors to construct a historical garden environmental sequence, including: Multiple monitoring sensors are deployed in the garden environment to monitor data, and a data monitoring platform is built. Each monitoring sensor stores its corresponding geographic coordinates in the data monitoring platform. Based on the data monitoring platform, the garden environment data collected by each monitoring sensor is exported, and a corresponding historical garden environment sequence is constructed.

3. The data monitoring method for garden environments according to claim 2, characterized in that, Analyze historical garden environment sequences, filter distorted environment data segments, and identify and define the sensor categories corresponding to the distorted sensors as disturbed distorted sensor categories, including: The historical garden environment sequence was divided to obtain multiple local environmental data segments, and distorted environmental data segments were filtered out. The monitoring sensors corresponding to the distorted environmental data segments are identified as suspected distorted sensors, and the distance measurement is determined by the geographical coordinates of each pair of suspected distorted sensors. Multiple clusters are obtained based on distance metrics. Distortion sensors are then screened through these clusters, and the sensor category corresponding to each distortion sensor is defined as the disturbed distortion sensor category.

4. The data monitoring method for garden environments according to claim 3, characterized in that, The historical garden environment sequence was divided into multiple local environmental data segments. Distorted environmental data segments were then filtered out, including: The fluctuation standard value of the historical garden environment sequence was evaluated by analysis, and the historical garden environment sequence was divided into multiple local environmental data segments; Analyze each local environmental data segment to assess the local fluctuation value, and combine the fluctuation standard value to obtain the degree of instability. Distorted environmental data segments are filtered from local environmental data segments based on the degree of instability.

5. The data monitoring method for garden environments according to claim 3, characterized in that, Multiple clusters are obtained based on distance metrics, and distortion sensors are filtered out through these clusters. Specifically: All suspected distorted sensors are clustered based on distance metrics to obtain multiple clusters. The mean distance metric of each cluster is determined, and the suspected distorted sensors in the cluster corresponding to the minimum mean distance metric are selected and defined as distorted sensors.

6. The data monitoring method for garden environments according to claim 1, characterized in that, A time analysis is performed on the combined distorted environmental data segment and the jointly predicted environmental data segment to determine the environmental dimension distortion of the jointly predicted environmental data segment, including: Define the distorted environment data segment corresponding to the associated distorted sensor category as the prediction reference data segment, and filter the time strongly correlated reference data segment based on any joint prediction environment data segment combined with the prediction reference data segment; Determine the overlap between time-correlated reference data segments and corresponding joint prediction environment data segments to determine the temporal similarity. By analyzing the joint prediction environmental data segment and the time-correlated reference data segment separately, the degree of instability is obtained, and the environmental dimension distortion of the joint prediction environmental data segment is determined by combining the time similarity.

7. The data monitoring method for garden environments according to claim 1, characterized in that, Based on the distortion of the environmental dimension, the smoothing coefficient is re-acquired to predict the joint prediction of the environmental data segment, thus obtaining the true prediction data segment and completing the data monitoring of the garden environment, including: By integrating and analyzing joint forecast environmental data segments, the potential period of forecast disturbance at the current moment can be determined. Based on the predicted period of disturbance, the Holt-Winters seasonality method is used to obtain a new smoothing coefficient by combining the smoothing coefficient before reacquisition and the environmental dimension distortion. The new smoothing coefficient is used to predict the joint prediction environment data segment to obtain the true prediction data segment; The distorted environmental data segments are replaced with real predicted data segments, and then spliced ​​with normal environmental data segments from historical garden environmental sequences to form a new data sequence. The garden environment is then monitored using this new data sequence.

8. A data monitoring system suitable for garden environments, characterized in that, The system includes: The data acquisition module is used to: deploy multiple sensor types in the garden environment, with each sensor type including multiple monitoring sensors, collect garden environment data from the monitoring sensors, and construct a historical garden environment sequence; The data classification module is used to: analyze historical garden environment sequences, filter distorted environment data segments, and determine and define the sensor category corresponding to the distorted sensor as the disturbed distorted sensor category. The data processing module is used to: acquire the distortion environment data segment corresponding to the distortion sensor in the category of distortion sensor for smoothing coefficient prediction, and obtain the joint prediction environment data segment; the acquisition of the distortion environment data segment corresponding to the distortion sensor in the category of distortion sensor for smoothing coefficient prediction and obtaining the joint prediction environment data segment includes: Determine the perturbation correlation of each perturbation distortion sensor category, and define the currently analyzed perturbation distortion sensor category as the target category. Based on the difference in perturbation correlation between the target category and the remaining non-target categories of perturbation distortion sensor categories, determine the associated distortion sensor category corresponding to the target category. The correlation weight is obtained by combining the perturbation correlation difference between the target category and the associated distortion sensor category. The Holt-Winters seasonality method is used to obtain the smoothing coefficient. The correlation weight is added to the smoothing coefficient to predict the distortion environment data segment corresponding to any distortion sensor in the target category, and the joint predicted environment data segment is obtained. The data analysis module is used to: perform time analysis on the combined distorted environmental data segment and the joint prediction environmental data segment, and determine the environmental dimension distortion of the joint prediction environmental data segment; The data monitoring module is used to: re-acquire smoothing coefficient predictions based on environmental dimension distortion, jointly predict environmental data segments, obtain real prediction data segments, and complete data monitoring of the garden environment.

9. A data monitoring device suitable for garden environments, characterized in that, The device includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory communicate with each other through the communication bus. The processor calls logical instructions in the memory to execute the data monitoring method for garden environments as described in any one of claims 1 to 7.

Images