Granary side wall environment abnormity monitoring method and system based on cloud computing

Abstract

The invention discloses a granary side wall environment abnormity monitoring method and system based on cloud computing, and belongs to the technical field of grain storage management, and the method comprises the steps: S10, collecting environment data of a granary side wall through different sensor nodes, and generating a data set related to the environment data; the environment data comprises temperature, humidity, air pressure and gas, and uploading the collected environment data to a cloud platform; s20, receiving, storing, analyzing and early warning the environmental data by the cloud platform; wherein the receiving of the environmental data is to receive the environmental data and perform verification and format conversion. According to the granary side wall environment monitoring system and method, environmental data monitoring, intelligent analysis, an accurate early warning mechanism and risk management are carried out on the granary side wall, the scientificity and the intelligent level of granary side wall environment monitoring are improved, damage to grains caused by environment abnormity is effectively prevented, and a scientific basis and decision support are provided for granary management.

Description

Technical Field

[0001] The present invention relates to the technical field of grain storage management, and in particular to a method and system for monitoring abnormalities in a grain silo side wall environment based on cloud computing. Background Art

[0002] Food security, a crucial foundation for national security, is crucial to the well-being of the nation and the stable development of society. Granaries, as crucial storage locations for grain, ensure that their internal environmental stability is crucial to grain quality and safety. Changes in silo wall conditions, such as temperature, humidity, and air pressure, can lead to food spoilage and pest infestation. Therefore, silo wall environmental monitoring is essential.

[0003] Regarding this research, application document CN202411253604.7 provides a method and system for monitoring grain storage environment anomalies. This technical solution includes building a multimodal sensor network and collecting grain storage environment indicator data and grain images from the grain warehouse based on the multimodal sensor network; establishing a dynamic integration model and then using the grain storage environment indicator data and grain images to construct a multi-source environmental dataset; identifying the grain storage environment and determining the type of grain storage environment anomaly based on the multi-source environmental dataset, and adjusting the grain storage environment indicators when an anomaly occurs in the grain storage environment. This technical solution can monitor and identify grain storage environment anomalies and anomaly types, improve grain storage management efficiency, and provide strong support for optimizing grain storage management strategies.

[0004] Another application, filed with application number CN202211620062.3, provides a method, system, terminal, and medium for monitoring the environment in a grain silo. This technical solution involves acquiring detection data from various detection points within a grain pile, including carbon dioxide and temperature values. In response to a carbon dioxide value exceeding a preset carbon dioxide threshold or / and a temperature value exceeding a preset temperature threshold, the detection data is determined to be abnormal and an abnormality signal is generated. Based on the abnormal signal, the detection point where the abnormal signal occurred is determined to determine the location of the abnormality. This technical solution can rapidly locate the location of abnormal changes in the grain pile, improving the accuracy of detecting abnormal changes.

[0005] However, the anomaly detection models used in these technical solutions are often based on simple threshold judgments, making it difficult to accurately identify subtle abnormal changes in complex environments. For example, the coordinated changes in temperature and humidity on the side walls of a granary can cause grain spoilage, but existing models struggle to accurately capture this correlation and lack the ability to deeply mine and intelligently analyze data, resulting in reduced monitoring accuracy. Summary of the Invention

[0006] In view of the above-mentioned problems existing in the technical field of existing grain storage management, the present invention is proposed.

[0007] Therefore, one of the purposes of the present invention is to provide a cloud computing-based granary side wall environmental anomaly monitoring method and system, which improves the scientificity and intelligence level of granary side wall environmental monitoring by monitoring environmental data of the granary side wall, intelligent analysis, and precise early warning mechanism and risk management, effectively prevents damage to grain caused by environmental anomalies, and provides scientific basis and decision-making support for granary management.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In one aspect, the present invention provides a method for monitoring abnormalities in a granary sidewall environment based on cloud computing, comprising the following steps:

[0010] Step S10: Collect environmental data of the granary sidewall through different sensor nodes and generate a data set of environmental data; the environmental data includes temperature, humidity, air pressure and gas, and upload the collected environmental data to the cloud platform;

[0011] Step S20: The cloud platform includes the reception, storage, analysis and early warning of environmental data; wherein,

[0012] Receiving environmental data, including receiving environmental data and performing verification and format conversion;

[0013] Storage of environmental data, which is to store and manage environmental data after verification and format conversion;

[0014] Analysis of environmental data, including trend analysis, anomaly detection, and correlation analysis;

[0015] For early warning of environmental data, a safety threshold is preset based on the results of environmental data analysis. When the monitored environmental data exceeds the safety threshold, the environmental data on the granary side wall is judged to be abnormal and an early warning is issued; otherwise, no judgment is made. At the same time, the sensor nodes corresponding to the environmental data exceeding the safety threshold are obtained;

[0016] Step S30: When the monitored environmental data exceeds the safety threshold, the difference between the environmental data of the remaining sensor nodes and the safety threshold is calculated, and the environmental data of at least four sensor nodes closest to the safety threshold are obtained based on the calculated difference, and the environmental data of the four sensor nodes are marked as key monitored sensor nodes;

[0017] Step S40: Count the environmental data with the most historical abnormalities on the granary side wall, obtain the sensor node corresponding to the environmental data, mark the sensor node as a reference sensor node, analyze the changes in the environmental data of the sensor nodes adjacent to the reference sensor node, and make corresponding judgments on the reference sensor node based on the analysis results.

[0018] As a preferred solution of the present invention, in step S40, the reference sensor node is judged accordingly based on the analysis results, including judging the local environmental abnormality and sensor failure based on the reference sensor node. If it is judged that the local environmental abnormality is present, corresponding processing is performed according to the type and / or degree of the local environmental abnormality, and the corresponding processing includes ventilation, cooling and dehumidification.

[0019] As a preferred embodiment of the present invention, in step S20, trend analysis is performed on the environmental data, including by linear regression analysis, and the steps include data preprocessing, feature selection, model building, model evaluation and trend prediction;

[0020] Data preprocessing includes data cleaning, data smoothing and data standardization;

[0021] Data cleaning, to remove missing values, outliers and / or duplicate values ​​from environmental data;

[0022] Data smoothing is to use the moving average method to smooth the environmental data;

[0023] Data standardization is to standardize environmental data to the same dimension;

[0024] Feature selection is to select time as the independent variable δ and environmental data as the dependent variable λ;

[0025] The model was established using the linear regression model λ = aδ + b, where a represents the slope, which is the rate of change of the trend, and b represents the intercept;

[0026] Fit the linear regression model using the least squares method and calculate the fitted straight line;

[0027] Model evaluation is to calculate the goodness of fit of the linear regression model and evaluate the degree of fit of the linear regression model to the environmental data; based on the goodness of fit, check whether the residuals meet the normal distribution and independence assumptions;

[0028] Trend forecasting is to use the fitted linear regression model to predict environmental data for future periods.

[0029] As a preferred solution of the present invention, the characteristic changes of the environmental data of the reference sensor node in different time periods are obtained, and the correlation changes of the environmental data of other sensor nodes are calculated based on the characteristic changes, and the correlation coefficient is calculated as follows:

[0030] Among them, r xy represents the correlation coefficient between variables x and y;

[0031] Where xi and y i Represent the values ​​of variables x and y at the i-th time point respectively;

[0032] and represent the means of variables x and y respectively;

[0033] n represents the number of environmental data;

[0034] The range of the correlation coefficient is [-1, 1]. xy When it is close to 1, it means there is a positive correlation between x and y;

[0035] When r xy When it is close to -1, it means there is a negative correlation between x and y;

[0036] When r xy When it is close to 0, it means there is no correlation between x and y.

[0037] As a preferred solution of the present invention, the calculation of the correlation change of the environmental data of other sensor nodes based on the feature change also includes calculating it according to the cross-correlation function of the time series, as shown below:

[0038]

[0039] Where r xy (τ) represents the cross-correlation function between the changes x and y, where (τ) represents the time delay;

[0040] x i and y i Represent the values ​​of variables x and y at the i-th time point respectively;

[0041] and represent the means of variables x and y respectively;

[0042] n represents the number of environmental data;

[0043] When r xy When the value of (τ) is the largest, it means that the correlation between x and y is the strongest when the time delay (τ) is reached.

[0044] As a preferred solution of the present invention, the environmental data with the largest change in other sensor nodes is obtained based on the calculated associated changes. When the environmental data corresponding to the reference sensor node exceeds the safety threshold, the environmental data with the largest change is also determined to be in an abnormal change state; otherwise, no determination is made.

[0045] As a preferred solution of the present invention, the environmental data with the largest change is marked as risky environmental data, and the regular changes of the risky environmental data when the environmental data corresponding to the reference sensor node exceeds the safety threshold are calculated in different time periods, and the results are calculated according to the seasonal decomposition method, as shown below:

[0046] Q t =W t +S t +G t Among them, Q t represents the monitoring value at time t;

[0047] Where W t represents the trend component at time t, S t represents the seasonal component at time t, G t represents the random component at time t.

[0048] As a preferred solution of the present invention, it further includes calculating according to a periodic analysis method as shown below:

[0049] Q t =Aft+φ)+∈ t Among them, Q t represents the monitoring value at time t;

[0050] Where A represents the amplitude of the periodic component, f represents the frequency of the periodic component, φ represents the phase of the periodic component, and ∈ t represents the random error at time t;

[0051] According to the calculation results, the 6 to 8 environmental data that appear the most times in different time periods for the sensor nodes corresponding to the risky environmental data are obtained, and the minimum environmental data is collected among the 6 to 8 environmental data. If the minimum environmental data appears in the corresponding time period in the future, it is determined that the environmental data corresponding to the reference sensor node will exceed the safety threshold; otherwise, no determination is made.

[0052] In another aspect, the present invention provides a system for the above-mentioned cloud computing-based granary sidewall environmental anomaly monitoring method, comprising:

[0053] The data acquisition module collects environmental data from the granary sidewall through different sensor nodes and generates a data set of environmental data, including temperature, humidity, air pressure, and gas, and uploads the collected environmental data to the cloud platform;

[0054] A data fusion processing module, which includes a data receiving unit, a data storage unit, a data analysis unit and a determination unit;

[0055] The data receiving unit is used to receive environmental data in the cloud platform and perform verification and format conversion on the environmental data;

[0056] The data storage unit is used to store and manage the environmental data after verification and format conversion;

[0057] The data analysis unit is used to perform trend analysis, anomaly detection and correlation analysis on the stored and managed environmental data;

[0058] The judgment unit is used to preset a safety threshold based on the results of the environmental data analysis. When the monitored environmental data exceeds the safety threshold, it determines that the environmental data of the granary side wall is abnormal and issues an early warning; otherwise, no judgment is made. At the same time, the sensor node corresponding to the environmental data exceeding the safety threshold is obtained;

[0059] A calculation module is used to calculate the difference between the environmental data of the remaining sensor nodes and the safety threshold when the monitored environmental data exceeds the safety threshold, obtain the environmental data of at least four sensor nodes closest to the safety threshold based on the calculated difference, and mark the environmental data of the four sensor nodes as key monitored sensor nodes;

[0060] The statistical module is used to count the environmental data with the most historical abnormalities on the granary side wall, obtain the sensor nodes corresponding to the environmental data, mark the sensor nodes as reference sensor nodes, analyze the changes in the environmental data of the sensor nodes adjacent to the reference sensor nodes, and make corresponding judgments on the reference sensor nodes based on the analysis results.

[0061] Beneficial effects:

[0062] 1. This invention uses a cloud platform to analyze environmental data in real time. Combined with preset safety thresholds, it can promptly and accurately detect abnormal changes in environmental data and issue early warning signals. This precise early warning mechanism helps to identify potential environmental problems in advance and prevent food damage caused by environmental changes.

[0063] 2. By analyzing the correlation changes between different sensor nodes using methods such as correlation coefficients and cross-correlation functions, we can quickly locate the source of abnormal data and identify other environmental factors associated with the abnormal data. This helps us gain a deeper understanding of the changing patterns of the granary environment and provides a basis for scientific decision-making.

[0064] 3. By calculating the difference between the environmental data of the sensor nodes and the safety threshold, the sensor nodes closest to the safety threshold are selected as key monitoring targets. This strategy can concentrate resources, improve monitoring efficiency, and ensure the environmental safety of key areas.

[0065] 4. Based on seasonal decomposition and periodic analysis, the regular changes of risk environment data in different time periods are analyzed, which helps to predict possible environmental anomalies in the future in advance and provide a scientific basis for risk management and preventive measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0067] in:

[0068] Figure 1 This is a schematic diagram of the modular structure of a granary side wall environmental anomaly monitoring system based on cloud computing according to an embodiment of the present invention;

[0069] Figure 2 Schematic diagram of a method flow in an embodiment of the present invention;

[0070] Numbers in the figure: 110 - data acquisition module; 120 - data fusion processing module; 1201 - data receiving unit; 1202 - data storage unit; 1203 - data analysis unit; 1204 - determination unit; 130 - calculation module; 140 - statistics module. DETAILED DESCRIPTION

[0071] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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 described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of the present invention.

[0072] Because existing anomaly detection models often rely on simple thresholds, they struggle to accurately identify subtle anomalies in complex environments. For example, the synergistic changes in silo wall temperature and humidity can cause grain spoilage, but existing models struggle to accurately capture this correlation and lack the ability to deeply mine and intelligently analyze data, resulting in reduced monitoring accuracy.

[0073] Based on this, the present invention proposes a cloud computing-based granary side wall environmental anomaly monitoring method and system, which improves the scientificity and intelligence level of granary side wall environmental monitoring by monitoring environmental data of the granary side wall, intelligent analysis, and precise early warning mechanism and risk management, effectively prevents damage to grain caused by environmental anomalies, and provides scientific basis and decision-making support for granary management.

[0074] The present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0075] Reference Figures 1 to 2 , is an embodiment of the present invention, which provides a method for monitoring abnormalities in the side wall environment of a granary based on cloud computing, comprising the following steps:

[0076] Step S10: Collect environmental data of the granary sidewall through different sensor nodes and generate a data set of environmental data; the environmental data includes temperature, humidity, air pressure and gas, and upload the collected environmental data to the cloud platform;

[0077] In this embodiment, in a feasible implementation scheme, 50 sensor nodes are evenly arranged on the side wall of a medium-sized granary at intervals of 5 meters. Each sensor node includes a temperature sensor, a humidity sensor, an air pressure sensor, and a gas composition sensor.

[0078] On the side wall of a large granary, 100 sensor nodes are arranged. The type and arrangement of sensor nodes are the same as above.

[0079] Step S20: The cloud platform includes the reception, storage, analysis and early warning of environmental data; wherein,

[0080] Receiving environmental data, including receiving environmental data and performing verification and format conversion;

[0081] Storage of environmental data, which is to store and manage environmental data after verification and format conversion;

[0082] In this embodiment, it supports fast query and backup of environmental data;

[0083] Analysis of environmental data, including trend analysis, anomaly detection, and correlation analysis;

[0084] In this embodiment, for example, by establishing a normal range model for environmental data, when data is detected to be outside the normal range, it is determined to be an abnormal situation; at the same time, the correlation between different environmental factors is analyzed, such as the impact of the coordinated changes in temperature and humidity on food.

[0085] For early warning of environmental data, a safety threshold is preset based on the results of environmental data analysis. When the monitored environmental data exceeds the safety threshold, the environmental data on the granary side wall is judged to be abnormal and an early warning is issued; otherwise, no judgment is made. At the same time, the sensor nodes corresponding to the environmental data exceeding the safety threshold are obtained;

[0086] Step S30: When the monitored environmental data exceeds the safety threshold, the difference between the environmental data of the remaining sensor nodes and the safety threshold is calculated, and the environmental data of at least four sensor nodes closest to the safety threshold are obtained based on the calculated difference, and the environmental data of the four sensor nodes are marked as key monitored sensor nodes;

[0087] Step S40: Counting the environmental data with the most historical abnormalities on the granary sidewall, obtaining the sensor node corresponding to the environmental data, marking the sensor node as a reference sensor node, analyzing changes in the environmental data of the sensor nodes adjacent to the reference sensor node, and making corresponding judgments on the reference sensor node based on the analysis results;

[0088] This embodiment further illustrates that, by collecting multi-dimensional environmental data such as temperature, humidity, air pressure, and gas through a variety of sensor nodes, comprehensive monitoring is ensured, capable of covering all aspects of the granary side wall environment. The cloud platform integrates data reception, storage, analysis, and early warning functions to form a complete monitoring system, thereby improving the efficiency and reliability of monitoring. In step S40, a corresponding determination is made on the reference sensor node based on the analysis results, including determining whether a local environmental anomaly and a sensor failure are determined based on the reference sensor node. If a local environmental anomaly is determined, corresponding processing is performed based on the type and / or degree of the local environmental anomaly, and the corresponding processing includes ventilation, cooling, and dehumidification.

[0089] In this embodiment, if the sensor fails, the maintenance personnel are promptly notified to repair or replace it;

[0090] It can distinguish between local environmental anomalies and sensor failures, avoid misjudgment, and improve the accuracy and reliability of the monitoring system;

[0091] In step S20 , trend analysis is performed on the environmental data, including by linear regression analysis, and the steps include data preprocessing, feature selection, model building, model evaluation, and trend prediction;

[0092] Data preprocessing includes data cleaning, data smoothing and data standardization;

[0093] Data cleaning, to remove missing values, outliers and / or duplicate values ​​from environmental data;

[0094] Data smoothing is to use the moving average method to smooth the environmental data;

[0095] In this embodiment, the effect of short-term fluctuations on the trend is reduced;

[0096] Data standardization is to standardize environmental data to the same dimension;

[0097] In this embodiment, to facilitate subsequent modeling;

[0098] Feature selection is to select time as the independent variable δ and environmental data as the dependent variable λ;

[0099] The model was established using the linear regression model λ = aδ + b, where a represents the slope, which is the rate of change of the trend, and b represents the intercept;

[0100] Fit the linear regression model using the least squares method and calculate the fitted straight line;

[0101] Model evaluation is to calculate the goodness of fit of the linear regression model and evaluate the degree of fit of the linear regression model to the environmental data; based on the goodness of fit, check whether the residuals meet the normal distribution and independence assumptions;

[0102] Trend forecasting is to use the fitted linear regression model to predict environmental data for future periods;

[0103] In this embodiment, data quality is improved through data cleaning, smoothing and standardization, providing a reliable data basis for subsequent analysis;

[0104] The linear regression model is used for trend analysis, the model is fitted by the least squares method, and the goodness of fit of the model is evaluated to ensure the scientificity and accuracy of trend prediction;

[0105] Furthermore, the characteristic changes of the environmental data of the reference sensor node in different time periods are obtained, and the correlation changes of the environmental data of other sensor nodes are calculated based on the characteristic changes, which are calculated in the form of correlation coefficients as shown below: Among them, r xy represents the correlation coefficient between variables x and y;

[0106] Where x i and y i Represent the values ​​of variables x and y at the i-th time point respectively;

[0107] and represent the means of variables x and y respectively;

[0108] n represents the number of environmental data;

[0109] The range of the correlation coefficient is [-1, 1]. xy When it is close to 1, it means there is a positive correlation between x and y;

[0110] When r xy When it is close to -1, it means there is a negative correlation between x and y;

[0111] When r xyWhen it is close to 0, it means there is no correlation between x and y;

[0112] In this embodiment, the correlation coefficient (PearsonCorrelationCoefficient) is used to measure the correlation (linear correlation) between two variables x and y. Its value is between -1 and 1, and is also called the Pearson correlation coefficient.

[0113] Different time periods include the four seasons of spring, summer, autumn and winter;

[0114] This embodiment further illustrates that by calculating the correlation coefficient, the correlation between different sensor nodes can be quantified to help identify other environmental factors associated with abnormal data;

[0115] On the basis of the above, the correlation change of the environmental data of other sensor nodes is calculated based on the feature change, and is also calculated according to the cross-correlation function (CCF) of the time series, as shown below:

[0116]

[0117] Where r xy (τ) represents the cross-correlation function between the changes x and y, where (τ) represents the time delay;

[0118] x i and y i Represent the values ​​of variables x and y at the i-th time point respectively;

[0119] and represent the means of variables x and y respectively;

[0120] n represents the number of environmental data;

[0121] When r xy When the value of (τ) is the largest, it means that the correlation between x and y is the strongest at the time delay (τ);

[0122] In this embodiment, the cross-correlation function can be used to analyze the delay correlation between different sensor nodes, helping to identify potential time correlation. By analyzing the time delay correlation, the trend of environmental changes can be more accurately predicted, further improving the prediction ability of the system.

[0123] Furthermore, based on the calculated correlation changes, the environmental data with the largest change among other sensor nodes is obtained. When the environmental data corresponding to the reference sensor node exceeds the safety threshold, the environmental data with the largest change is also determined to be in an abnormal change state; otherwise, no determination is made.

[0124] In this embodiment, by calculating the correlation changes, the environmental data with the largest changes are identified and marked as risky environmental data, thereby improving the accuracy of risk identification. When the environmental data of the reference sensor node exceeds the safety threshold, it can also be determined whether the risky environmental data is in an abnormal change state, thereby enhancing the overall monitoring capability of the system and improving the comprehensiveness and effectiveness of early warning.

[0125] Based on the above, the environmental data with the largest changes are marked as risky environmental data. The regular changes of risky environmental data when the environmental data corresponding to the reference sensor node exceeds the safety threshold are calculated in different time periods. The results are calculated based on the seasonal decomposition method as follows:

[0126] Q t =W t +S t +G t Among them, Q t represents the monitoring value at time t;

[0127] Where W t represents the trend component at time t, S t represents the seasonal component at time t, G t represents the random component at time t;

[0128] In this embodiment, seasonal decomposition can decompose the time series into trend, seasonal and random components, so as to analyze the regular changes in different time periods. In the four seasons of spring, summer, autumn and winter, we can calculate the average value of each season to understand the seasonal variation pattern of environmental data.

[0129] In this embodiment, seasonal decomposition can be used to analyze the regular changes in risk environment data in different time periods, helping to identify the impact of seasonal factors on environmental data. At the same time, trend components, seasonal components, and random components can be separated, providing support for in-depth analysis of the changing patterns of environmental data.

[0130] Specifically, this embodiment also includes calculating based on periodic analysis, as shown below:

[0131] Q t =A(ft+φ)+∈ t Among them, Q t represents the monitoring value at time t;

[0132] Where A represents the amplitude of the periodic component, f represents the frequency of the periodic component, φ represents the phase of the periodic component, and ∈ t represents the random error at time t;

[0133] Based on the calculation results, the 6 to 8 environmental data that appear the most frequently in different time periods for the sensor node corresponding to the risky environmental data are obtained. The minimum environmental data is collected from the 6 to 8 environmental data. If the minimum environmental data appears in the corresponding time period in the future, it is determined that the environmental data corresponding to the reference sensor node exceeds the safety threshold; otherwise, no determination is made.

[0134] Periodicity analysis can be used to analyze periodic changes in time series. In the four seasons of spring, summer, autumn and winter, we can calculate the periodic components of each season separately to understand the periodic change pattern of environmental data.

[0135] In this embodiment, through periodic analysis, the periodic variation pattern of environmental data can be identified, helping to predict environmental anomalies that may occur in the future.

[0136] Based on the above, the present invention improves the scientificity and intelligence level of environmental monitoring of the granary side walls by monitoring environmental data of the granary side walls, intelligent analysis, and precise early warning mechanism and risk management, effectively preventing damage to grain caused by environmental abnormalities, and providing scientific basis and decision-making support for granary management.

[0137] This embodiment combines the above-mentioned cloud computing-based granary side wall environmental anomaly monitoring method and also proposes a system applied to the method, as follows:

[0138] The data acquisition module 110 collects environmental data of the granary side wall through different sensor nodes and generates a data set of environmental data; the environmental data includes temperature, humidity, air pressure and gas, and uploads the collected environmental data to the cloud platform; the data fusion processing module 120 includes a data receiving unit 1201, a data storage unit 1202, a data analysis unit 1203 and a determination unit 1204;

[0139] The data receiving unit 1201 is used to receive environmental data in the cloud platform and perform verification and format conversion on the environmental data;

[0140] The data storage unit 1202 is used to store and manage the environment data after verification and format conversion;

[0141] The data analysis unit 1203 is used to perform trend analysis, anomaly detection, and correlation analysis on the stored and managed environmental data. The determination unit 1204 is used to preset a safety threshold based on the results of the environmental data analysis. When the monitored environmental data exceeds the safety threshold, it determines that the environmental data of the granary side wall is abnormal and issues an early warning. Otherwise, no determination is made. At the same time, the sensor nodes corresponding to the environmental data exceeding the safety threshold are obtained.

[0142] The calculation module 130 is configured to calculate the difference between the environmental data of the remaining sensor nodes and the safety threshold when the monitored environmental data exceeds the safety threshold, obtain the environmental data of at least four sensor nodes closest to the safety threshold based on the calculated difference, and mark the environmental data of the four sensor nodes as key monitored sensor nodes;

[0143] The statistical module 140 is used to count the environmental data with the most historical abnormalities on the granary side wall, obtain the sensor nodes corresponding to the environmental data, mark the sensor nodes as reference sensor nodes, analyze the changes in the environmental data of the sensor nodes adjacent to the reference sensor nodes, and make corresponding judgments on the reference sensor nodes based on the analysis results.

[0144] In summary, the present invention improves the scientificity and intelligence level of environmental monitoring of granary side walls by monitoring environmental data of granary side walls, intelligent analysis, and precise early warning mechanism and risk management, effectively prevents damage to grain caused by environmental anomalies, and provides scientific basis and decision-making support for granary management.

[0145] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A cloud computing-based method for monitoring abnormalities in the granary side wall environment, characterized in that: The following steps are involved: step S10: Collect environmental data of the granary side wall through different sensor nodes and generate a data set of the environmental data; the environmental data includes temperature, humidity, air pressure and gas, and upload the collected environmental data to the cloud platform; Step S20: The cloud platform includes receiving, storing, analyzing and warning environmental data; wherein, Receiving the environmental data includes receiving the environmental data and performing verification and format conversion; The storage of the environmental data is to store and manage the environmental data after verification and format conversion; Analysis of the environmental data, including trend analysis, anomaly detection, and correlation analysis of the environmental data; The early warning of the environmental data is to preset a safety threshold based on the results of the environmental data analysis. When the monitored environmental data exceeds the safety threshold, the environmental data of the granary side wall is determined to be abnormal and an early warning is issued; otherwise, no determination is made. At the same time, the sensor node corresponding to the environmental data exceeding the safety threshold is obtained; Step S30: When the monitored environmental data exceeds the safety threshold, the difference between the environmental data of the remaining sensor nodes and the safety threshold is calculated, and the environmental data of at least four sensor nodes closest to the safety threshold are obtained according to the calculated difference, and the environmental data of the four sensor nodes are marked as key monitored sensor nodes; Step S40: Count the environmental data with the most historical abnormalities on the granary side wall, obtain the sensor node corresponding to the environmental data, mark the sensor node as a reference sensor node, analyze the changes in the environmental data of the sensor nodes adjacent to the reference sensor node, and make corresponding judgments on the reference sensor node based on the analysis results.

2. The method for monitoring abnormalities of granary sidewall environment based on cloud computing according to claim 1, characterized in that: In step S40, the reference sensor node is judged accordingly based on the analysis results, including judging local environmental anomalies and sensor failures based on the reference sensor node. If it is judged that the local environmental anomaly is present, corresponding processing is performed based on the type and / or degree of the local environmental anomaly, and the corresponding processing includes ventilation, cooling, and dehumidification.

3. The method for monitoring abnormalities of granary sidewall environment based on cloud computing according to claim 1, characterized in that: In step S20, trend analysis is performed on the environmental data, including by linear regression analysis, and the steps include data preprocessing, feature selection, model building, model evaluation and trend prediction; The data preprocessing includes data cleaning, data smoothing and data standardization; The data cleaning is to remove missing values, abnormal values ​​and / or duplicate values ​​from the environmental data; The data smoothing is to use the moving average method to smooth the environmental data; The data standardization is to standardize the environmental data to the same dimension; The feature selection is to select time as the independent variable δ and environmental data as the dependent variable λ; The model is established by using a linear regression model λ=aδ+b, wherein a represents the slope, the slope is the rate of change of the trend, and b represents the intercept; Fit the linear regression model using the least squares method and calculate the fitted straight line; The model evaluation is to calculate the goodness of fit of the linear regression model and evaluate the degree of fit of the linear regression model to the environmental data; Based on the stated goodness of fit, check whether the residuals meet the assumptions of normal distribution and independence; The trend prediction is to use the fitted linear regression model to predict the environmental data of the future period.

4. The method for monitoring abnormalities of granary side wall environment based on cloud computing according to claim 1, characterized in that: Obtain characteristic changes of the environmental data of the reference sensor node in different time periods, and calculate the associated changes of the environmental data of other sensor nodes based on the characteristic changes, and calculate the correlation coefficient as follows: Among them, r xy represents the correlation coefficient between variables x and y; Where x i and y i Represent the values ​​of variables x and y at the i-th time point respectively; and represent the means of variables x and y respectively; n represents the number of environmental data; The range of the correlation coefficient is [-1, 1]. xy When it is close to 1, it means there is a positive correlation between x and y; When r xy When it is close to -1, it means there is a negative correlation between x and y; When r xy When it is close to 0, it means there is no correlation between x and y.

5. The method for monitoring abnormalities of granary sidewall environment based on cloud computing according to claim 4, characterized in that: Calculating the correlation changes of the environmental data of other sensor nodes based on the feature changes also includes calculating the correlation changes according to the cross-correlation function of the time series, as shown below: Where r xy (τ) represents the cross-correlation function between the changes x and y, where (τ) represents the time delay; x i and y i Represent the values ​​of variables x and y at the i-th time point respectively; and represent the means of variables x and y respectively; n represents the number of environmental data; When r xy When the value of (τ) is the largest, it means that the correlation between x and y is the strongest when the time delay (τ) is reached.

6. The method for monitoring abnormalities of granary side wall environment based on cloud computing according to any one of claims 4 to 5, characterized in that: Acquire the environmental data with the largest change among other sensor nodes based on the calculated correlation change, and when the environmental data corresponding to the reference sensor node exceeds the safety threshold, determine that the environmental data with the largest change is in an abnormal change state; Otherwise, no judgment is made.

7. The method for monitoring abnormalities of granary sidewall environment based on cloud computing according to claim 6, characterized in that: The environmental data with the largest change is marked as risky environmental data. The regular changes of the risky environmental data when the environmental data corresponding to the reference sensor node exceeds the safety threshold are calculated in different time periods. The seasonal decomposition method is used to calculate the regular changes, as shown below: Q t =W t +S t +G t Among them, Q t represents the monitoring value at time t; Where W t represents the trend component at time t, S t represents the seasonal component at time t, G t represents the random component at time t.

8. The method for monitoring abnormalities of granary sidewall environment based on cloud computing according to claim 7, characterized in that: It also includes calculations based on periodic analysis as follows: Q t =Aft+φ)+∈ t Among them, Q t represents the monitoring value at time t; Where A represents the amplitude of the periodic component, f represents the frequency of the periodic component, φ represents the phase of the periodic component, and ∈ t represents the random error at time t; According to the calculation results, the 6 to 8 environmental data that appear the most times in different time periods for the sensor node corresponding to the risk environment data are obtained, and the minimum environmental data is collected from the 6 to 8 environmental data. If the minimum environmental data appears in the corresponding time period in the future, it is determined that the environmental data corresponding to the reference sensor node will exceed the safety threshold; otherwise, no determination is made.

9. The system used in the method for monitoring abnormalities of granary side wall environment based on cloud computing as claimed in claim 1 is characterized in that: include: A data acquisition module collects environmental data from the granary sidewall through different sensor nodes and generates a data set about the environmental data; the environmental data includes temperature, humidity, air pressure and gas, and uploads the collected environmental data to the cloud platform; A data fusion processing module, the data fusion processing module includes a data receiving unit, a data storage unit, a data analysis unit and a determination unit; The data receiving unit is used to receive the environmental data in the cloud platform and perform verification and format conversion on the environmental data; The data storage unit is used to store and manage the environmental data after verification and format conversion; The data analysis unit is used to perform trend analysis, anomaly detection and correlation analysis on the stored and managed environmental data; The determination unit is used to preset a safety threshold based on the results of the environmental data analysis. When the monitored environmental data exceeds the safety threshold, it is determined that the environmental data of the granary side wall is abnormal and an early warning is issued; Otherwise, no judgment is made; at the same time, the sensor nodes corresponding to the environmental data exceeding the safety threshold are obtained; a calculation module configured to, when the monitored environmental data exceeds the safety threshold, calculate the difference between the environmental data of the remaining sensor nodes and the safety threshold, obtain the environmental data of at least four sensor nodes closest to the safety threshold based on the calculated difference, and mark the environmental data of the four sensor nodes as key monitored sensor nodes; The statistical module is used to count the environmental data with the most historical abnormalities on the granary side wall, obtain the sensor node corresponding to the environmental data, mark the sensor node as a reference sensor node, analyze the changes in the environmental data of the sensor nodes adjacent to the reference sensor node, and make corresponding judgments on the reference sensor node based on the analysis results.

Images