Edge computing type cultural relic warehouse environment multi-parameter online monitoring method based on POE power supply

Abstract

The application provides a POE power supply-based edge computing type cultural relic warehouse environment multi-parameter online monitoring method, which comprises the following steps: performing protocol unification processing, preprocessing and cleaning on multi-source sensing data to obtain clean data; performing multi-source data fusion correlation analysis and constructing a cultural relic warehouse multi-parameter linkage analysis decision table; obtaining a quantitative cultural relic damage risk assessment and early warning level to obtain a cultural relic damage risk index and an early warning level, so as to output a data uploading scheme and a local backup according to a network state; and constructing a multi-gas cross interference model by using standard data flow to output high-precision gas concentration data compensated by cross interference. The application can simplify wiring, improve data processing efficiency, system reliability and maintenance convenience, and provides an efficient, reliable and easy-to-maintain intelligent solution for cultural relic warehouse environment monitoring.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring technology for cultural relic protection, and in particular to a method for online monitoring of multiple parameters of the environment in a cultural relic warehouse based on PoE-powered edge computing. Background Technology

[0002] In cultural relic preservation, the long-term stability of environmental parameters in the storage facilities directly affects the preservation status of the artifacts. Inappropriate temperature, humidity, light exposure, and pollutant concentrations can lead to irreversible damage such as corrosion, discoloration, and embrittlement. Traditional manual inspections and discrete monitoring methods are insufficient for continuous tracking and comprehensive analysis of environmental parameters, making it impossible to effectively assess the cumulative impact of environmental factors on cultural relics. Furthermore, existing monitoring systems suffer from technical bottlenecks such as complex wiring, data transmission delays, lack of local decision-making capabilities, maintenance difficulties, and interference from multiple parameters. Summary of the Invention

[0003] To overcome the shortcomings of existing technologies, the purpose of this invention is to provide a PoE-powered edge computing-based online monitoring method for multiple parameters of the cultural relic storage environment. This method simplifies wiring, improves data processing efficiency, system reliability, and ease of maintenance, providing an efficient, reliable, and easy-to-maintain intelligent solution for monitoring the environment of cultural relic storage facilities.

[0004] To achieve the above objectives, the present invention provides the following solution: a multi-parameter online monitoring method for the environment of a cultural relic storage facility based on PoE power supply, comprising:

[0005] Acquire multi-source sensor data from the cultural relics warehouse, and perform protocol unification processing, preprocessing, and cleaning on the multi-source sensor data to obtain clean data;

[0006] Based on the cleaning data, multi-source data fusion and correlation analysis are performed, and a multi-parameter linkage analysis decision table for cultural relics warehouses is constructed to obtain cultural relics damage risk assessment results and linkage control instructions.

[0007] Based on the clean data and multi-parameter linkage analysis decision table, a quantitative assessment of cultural relic damage risk and early warning level are obtained, resulting in a cultural relic damage risk index and early warning level. According to the network status, a data upload plan and local backup are output.

[0008] Using the standard data stream, a radial basis function neural network is constructed and trained to obtain a multi-gas cross-interference model, so as to output high-precision gas concentration data after cross-interference compensation.

[0009] Optionally, multi-source sensor data from the artifact storage facility is acquired, and the multi-source sensor data undergoes protocol unification processing, preprocessing, and cleaning to obtain clean data, including:

[0010] Choose a PoE switch with an AI acceleration chip built into it for edge computing, select CAT6A Ethernet cable for power and data transmission, and connect all sensor nodes through a single network cable to complete the infrastructure construction.

[0011] Based on the aforementioned infrastructure construction, multi-source data is collected using temperature sensors, humidity sensors, carbon dioxide sensors, and wireless sensors. A protocol parser is embedded in the edge gateway to parse multi-source data from different protocols and unify them into a standard JSON format. Then, a timestamp and data source tag are added to each data item to obtain a standard data stream.

[0012] Based on the standard data stream, outliers are automatically discarded by setting a physical range threshold, thus eliminating invalid values. Data is then aligned to a unified timestamp sequence using linear interpolation or nearest neighbor methods. Continuously changing data is processed using a Kalman filter algorithm, and gas sensor data is processed using a sliding window averaging method to achieve noise reduction and obtain clean data.

[0013] Optionally, based on the cleaning data, multi-source data fusion and correlation analysis is performed, and a multi-parameter linkage analysis decision table for the cultural relic storage is constructed to obtain the cultural relic damage risk assessment results and linkage control instructions, including:

[0014] Based on the cleaning data, a correlation analysis was performed on temperature, relative humidity, ultraviolet light intensity, fine particulate matter concentration, carbon dioxide concentration, ozone concentration, nitrogen dioxide concentration, sulfur dioxide concentration, formaldehyde concentration, and TVOC concentration to obtain correlation information.

[0015] Based on the aforementioned correlation information, analyses are conducted on heat source interference patterns, humid and hot air intrusion patterns, compound pollutant intrusion patterns, organic pollutant release patterns, personnel activity characteristics patterns, light radiation hazard patterns, and particulate matter pollution patterns to construct a multi-parameter linkage analysis and decision table for cultural relic storage, thereby obtaining cultural relic damage risk assessment results and linkage control instructions.

[0016] Optional,

[0017] The heat source interference mode is used to combine ultraviolet light intensity and illuminance data to analyze the uneven heating caused by the heat source, obtain the first linkage analysis result between temperature rise and humidity drop, and trigger the first linkage control based on the first linkage analysis result.

[0018] The humid and hot air intrusion mode is used to analyze whether hot air containing moisture has been introduced by checking the flow of people in the warehouse, and to obtain the second linkage analysis result between the increase in humidity and the increase in temperature. The second linkage control is triggered based on the second linkage analysis result.

[0019] The composite pollutant intrusion mode is used to compare the monitoring station data inside and outside the warehouse, as well as whether there is external pollution intrusion, to obtain the third linkage analysis result between the increase of sulfur dioxide, the increase of nitrogen dioxide and the increase of O3, and to trigger the third linkage control based on the third linkage analysis result.

[0020] The organic pollutant release mode is used to investigate whether there is internal organic material release, and to obtain the fourth linkage analysis result between the increase in formaldehyde and the increase in TOVC concentration. The fourth linkage control is triggered based on the fourth linkage analysis result.

[0021] The personnel activity characteristic pattern is used to analyze personnel activities in the warehouse and obtain the fifth linkage analysis results between the increase in carbon dioxide and the increase in TOVC concentration. The fifth linkage control is triggered based on the fifth linkage analysis results.

[0022] The light irradiation hazard mode is used to analyze the lighting system and obtain the sixth linkage analysis result between the increase in ultraviolet light intensity and the increase in temperature. The sixth linkage control is triggered based on the sixth linkage analysis result.

[0023] The particulate matter pollution mode is used to analyze particulate matter that is intruded from the outside and generated internally, and to obtain the seventh linkage analysis result of PM2.5 concentration rising alone. The seventh linkage control is triggered based on the seventh linkage analysis result.

[0024] Optionally, based on the cleaning data and the multi-parameter linkage analysis decision table, a quantitative assessment and early warning level of cultural relic damage risk is obtained, resulting in a cultural relic damage risk index and early warning level. This allows for the output of a data upload plan and local backup based on network conditions, including:

[0025] Based on the aforementioned cleaning data and multi-parameter linkage analysis decision table, the damage risk index for metallic cultural relics is calculated and warning levels are assigned; the damage risk index for organic cultural relics is also calculated and warning levels are assigned. Wherein:

[0026] The formula for calculating the damage risk index of metal cultural relics is as follows:

[0027] ;

[0028] in, The current relative humidity. The critical corrosion humidity. Humidity activation coefficient, The sulfur dioxide corrosion coefficient is... This refers to the concentration of sulfur dioxide. The coefficient of nitrogen dioxide corrosion is given. This refers to the concentration of nitrogen dioxide.

[0029] The formula for calculating the damage risk index of organic cultural relics is as follows:

[0030] ;

[0031] in, The current Kelvin temperature, For reference Kelvin temperature, The current relative humidity. For reference relative humidity, The activation energy for hydrolysis, Let be the ideal gas constant. Humidity index Ozone damage coefficient, This refers to the ozone concentration.

[0032] Determine the network status. When the network status is determined to be in normal mode, 90% of the data analysis is completed on the edge computing terminal. When the network status is determined to be in offline mode, the local storage mechanism is activated and the key monitoring functions are run at a reduced frequency.

[0033] Data is lightweighted using Delta encoding and a general compression algorithm. The compressed data and event logs are then stored in a local embedded multimedia card in the form of an SQLite database, completing local disaster recovery backup of the data. Once the network is restored, the data will be automatically uploaded.

[0034] Optionally, using the standard data stream, a radial basis function neural network is constructed and trained to obtain a multi-gas cross-interference model, which outputs high-precision gas concentration data after cross-interference compensation, including:

[0035] By selecting a radial basis function neural network, the number of nodes for the original sensor readings is set as the input layer, the number of nodes for the radial basis function neurons is set as the hidden layer, and the number of nodes for the compensated concentration values ​​is set as the output layer, the network structure is obtained.

[0036] Based on the network structure, the standard data stream is set as training data. The training data is processed using the K-means clustering algorithm to find the optimal center point of the radial basis function neurons. The optimal weights from the hidden layer to the output layer are calculated using the least squares method to complete the parameter optimization operation and obtain the multi-gas cross-interference model. The multi-gas cross-interference model is then verified and its performance is evaluated to output high-precision gas concentration data after cross-interference compensation.

[0037] This invention discloses the following technical effects by providing a PoE-powered edge computing-based online monitoring method for multiple parameters of an artifact storage environment:

[0038] 1) Advantages of simplified cabling: Change from dual-line (220V + low voltage) to single-line (CAT6A Ethernet); reduce material costs by 20-30%; speed up construction by 50%; reduce visual interference during warehouse renovation.

[0039] 2) Data processing efficiency advantages: Edge computing processes 90% of the data, reducing system latency from seconds to milliseconds; real-time response to critical events, without relying on cloud connections; supports synchronous correlation analysis of up to 10 parameters.

[0040] 3) System reliability advantages: The offline mode ensures uninterrupted monitoring; local storage ensures no data loss; and it supports fault self-diagnosis and isolation.

[0041] 4) Advantages in ease of maintenance: The single-line topology is clear and fault diagnosis is easy; the sensor is plug-and-play and supports hot-swapping; remote firmware upgrades are possible.

[0042] 5) The approach shifted from passive emergency response to proactive prevention, quantified risk assessment, and avoidance of reliance on experience-based judgments, thereby enhancing the effectiveness of cultural relic protection and improving management efficiency.

[0043] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments 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.

[0045] Figure 1 This is a schematic diagram of the method flow provided in an embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] like Figure 1 As shown, this invention provides a PoE-powered edge computing-based online monitoring method for multiple parameters of an artifact storage environment, including:

[0049] Step 1: Acquire multi-source sensor data from the cultural relic storage facility, and perform protocol unification processing, preprocessing, and cleaning on the multi-source sensor data to obtain clean data; Step 1 includes:

[0050] Choose a PoE switch with an AI acceleration chip built into it for edge computing, select CAT6A Ethernet cables for power and data transmission, and connect all sensor nodes with a single network cable to complete the infrastructure construction.

[0051] Based on the aforementioned infrastructure, multi-source data is collected using temperature sensors, humidity sensors, carbon dioxide sensors, and wireless sensors. A protocol parser is embedded in the edge gateway to parse the multi-source data from different protocols and unify it into a standard JSON format. Then, each data item is tagged with a timestamp and a data source label to obtain a standard data stream.

[0052] Based on the standard data stream, outliers are automatically discarded by setting a physical range threshold, thus eliminating invalid values. Data is then aligned to a unified timestamp sequence using linear interpolation or nearest neighbor methods. Continuously changing data is processed using a Kalman filter algorithm, and gas sensor data is processed using a sliding window averaging method to achieve noise reduction and obtain clean data.

[0053] Step 2: Based on the cleaning data, perform multi-source data fusion and correlation analysis, and construct a multi-parameter linkage analysis decision table for the cultural relic warehouse to obtain the cultural relic damage risk assessment results and linkage control instructions; Step 2 includes:

[0054] Based on the cleaning data, correlation analysis was performed on temperature, relative humidity, ultraviolet light intensity, fine particulate matter concentration, carbon dioxide concentration, ozone concentration, nitrogen dioxide concentration, sulfur dioxide concentration, formaldehyde concentration, and TVOC concentration to obtain correlation information.

[0055] Based on the aforementioned correlation information, analyses are conducted on heat source interference patterns, humid and hot air intrusion patterns, compound pollutant intrusion patterns, organic pollutant release patterns, personnel activity characteristics patterns, light radiation hazard patterns, and particulate matter pollution patterns to construct a multi-parameter linkage analysis and decision table for cultural relic storage, thereby obtaining cultural relic damage risk assessment results and linkage control instructions.

[0056] The heat source interference mode is used to combine ultraviolet light intensity and illuminance data to analyze the uneven heating caused by the heat source, obtain the first linkage analysis result between temperature rise and humidity drop, and trigger the first linkage control based on the first linkage analysis result.

[0057] The humid and hot air intrusion mode is used to analyze whether hot air containing moisture has been introduced by checking the flow of people in the warehouse, and to obtain the second linkage analysis result between the increase in humidity and the increase in temperature. The second linkage control is triggered based on the second linkage analysis result.

[0058] The composite pollutant intrusion mode is used to compare the monitoring station data inside and outside the warehouse, as well as whether there is external pollution intrusion, to obtain the third linkage analysis result between the increase of sulfur dioxide, the increase of nitrogen dioxide and the increase of O3, and to trigger the third linkage control based on the third linkage analysis result.

[0059] The organic pollutant release mode is used to investigate whether there is internal organic material release, and to obtain the fourth linkage analysis result between the increase in formaldehyde and the increase in TOVC concentration. The fourth linkage control is triggered based on the fourth linkage analysis result.

[0060] The personnel activity characteristic pattern is used to analyze personnel activities in the warehouse and obtain the fifth linkage analysis results between the increase in carbon dioxide and the increase in TOVC concentration. The fifth linkage control is triggered based on the fifth linkage analysis results.

[0061] The light irradiation hazard mode is used to analyze the lighting system and obtain the sixth linkage analysis result between the increase in ultraviolet light intensity and the increase in temperature. The sixth linkage control is triggered based on the sixth linkage analysis result.

[0062] The particulate matter pollution mode is used to analyze particulate matter that is intruded from the outside and generated internally, and to obtain the seventh linkage analysis result of PM2.5 concentration rising alone. The seventh linkage control is triggered based on the seventh linkage analysis result.

[0063] The decision-making table for multi-parameter linkage analysis of cultural relics storage is shown in Table 1 below:

[0064] Table 1. Decision-making table for multi-parameter linkage analysis of cultural relic storage facilities

[0065]

[0066]

[0067]

[0068] Step 3: Based on the clean data and multi-parameter linkage analysis decision table, obtain a quantitative assessment of cultural relic damage risk and early warning level, and obtain the cultural relic damage risk index (the relative damage rate of the cultural relic relative to its most ideal preservation conditions under the current environmental conditions, dimensionless) and early warning level, so as to output a data upload plan and local backup according to the network status; Step 3 includes:

[0069] Based on the aforementioned cleaning data and multi-parameter linkage analysis decision table, the damage risk index for metallic cultural relics is calculated and warning levels are assigned; the damage risk index for organic cultural relics is also calculated and warning levels are assigned. Wherein:

[0070] The formula for calculating the damage risk index of metal cultural relics is as follows:

[0071] ;

[0072] in, The current relative humidity (%) The critical corrosion humidity (%). Humidity activation coefficient, The sulfur dioxide corrosion coefficient is... This refers to the concentration of sulfur dioxide (ppb). The coefficient of nitrogen dioxide corrosion is given. This refers to the concentration of nitrogen dioxide (ppb).

[0073] The risk levels of damage to metal cultural relics are classified as shown in Table 2 below:

[0074] Table 2 Classification of Damage Risk Levels for Metal Cultural Relics

[0075]

[0076] The formula for calculating the damage risk index of organic cultural relics is as follows:

[0077] ;

[0078] in, The current Kelvin temperature (K) is given. For reference Kelvin temperature (K). The current relative humidity (%) For reference relative humidity (%) The activation energy for hydrolysis is (KJ / mol). Let be the ideal gas constant. Humidity index Ozone damage coefficient, This represents the ozone concentration (ppb).

[0079] The risk levels of damage to organic cultural relics are classified as shown in Table 3 below:

[0080] Table 3 Classification of Damage Risk Levels for Organic Cultural Relics

[0081]

[0082] The system determines the network status. When the network status is determined to be in normal mode, 90% of the data analysis is completed on the edge computing terminal. When the network status is determined to be in offline mode, the local storage mechanism is activated and the critical monitoring functions are run at reduced frequency.

[0083] Data is lightweighted using Delta encoding and a general compression algorithm. The compressed data and event logs are then stored in a local embedded multimedia card in the form of an SQLite database, completing local disaster recovery backup of the data. Once the network is restored, the data will be automatically uploaded.

[0084] Step 4: Using the standard data stream, construct and train a radial basis function neural network to obtain a multi-gas cross-interference model, and output high-precision gas concentration data after cross-interference compensation. Step 4 includes:

[0085] By selecting a radial basis function neural network, the number of nodes representing the original sensor readings is set as the input layer, the number of nodes representing the radial basis function neurons is set as the hidden layer, and the number of nodes representing the compensated concentration values ​​is set as the output layer, the network structure is obtained.

[0086] Based on the network structure, the standard data stream is set as training data. The training data is processed using the K-means clustering algorithm to find the optimal center point of the radial basis function neurons. The optimal weights from the hidden layer to the output layer are calculated using the least squares method to complete the parameter optimization operation and obtain the multi-gas cross-interference model. The multi-gas cross-interference model is then verified and its performance is evaluated to output high-precision gas concentration data after cross-interference compensation.

[0087] Therefore, this invention provides a PoE-powered edge computing-based online monitoring method for multiple parameters of the cultural relic storage environment, which simplifies wiring, improves data processing efficiency, system reliability, and ease of maintenance, providing an efficient, reliable, and easy-to-maintain intelligent solution for monitoring the environment of cultural relic storage facilities.

[0088] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0089] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A multi-parameter online monitoring method for the environment of a cultural relic storage facility based on PoE-powered edge computing, characterized in that, include: Acquire multi-source sensor data from the cultural relics storage facility, perform protocol unification processing, preprocessing and cleaning on the multi-source sensor data to obtain clean data, including: Choose a PoE switch with an AI acceleration chip built into it for edge computing, select CAT6A Ethernet cable for power and data transmission, and connect all sensor nodes through a single network cable to complete the infrastructure construction. Based on the aforementioned infrastructure construction, multi-source data is collected using temperature sensors, humidity sensors, carbon dioxide sensors, and wireless sensors. A protocol parser is embedded in the edge gateway to parse multi-source data from different protocols and unify them into a standard JSON format. Then, a timestamp and data source tag are added to each data item to obtain a standard data stream. Based on the standard data stream, outliers are automatically discarded by setting a physical range threshold, thus eliminating invalid values. Data is then aligned to a unified timestamp sequence using linear interpolation or the nearest neighbor method. Continuously changing data is processed using a Kalman filter algorithm, and gas sensor data is processed using a sliding window averaging method to achieve noise reduction and obtain clean data. Based on the cleaning data, multi-source data fusion and correlation analysis are performed, and a multi-parameter linkage analysis decision table for the cultural relics warehouse is constructed to obtain the cultural relics damage risk assessment results and linkage control instructions, including: Based on the cleaning data, a correlation analysis was performed on temperature, relative humidity, ultraviolet light intensity, fine particulate matter concentration, carbon dioxide concentration, ozone concentration, nitrogen dioxide concentration, sulfur dioxide concentration, formaldehyde concentration, and TVOC concentration to obtain correlation information. Based on the aforementioned correlation information, analyses are performed on heat source interference patterns, humid and hot air intrusion patterns, compound pollutant intrusion patterns, organic pollutant release patterns, personnel activity characteristics patterns, light radiation hazard patterns, and particulate matter pollution patterns to construct a multi-parameter linkage analysis and decision table for cultural relic storage, and to obtain cultural relic damage risk assessment results and linkage control instructions. The heat source interference mode is used to combine ultraviolet light intensity and illuminance data to analyze the uneven heating caused by the heat source, obtain the first linkage analysis result between temperature rise and humidity drop, and trigger the first linkage control based on the first linkage analysis result. The humid and hot air intrusion mode is used to analyze whether hot air containing moisture has been introduced by checking the flow of people in the warehouse, and to obtain the second linkage analysis result between the increase in humidity and the increase in temperature. The second linkage control is triggered based on the second linkage analysis result. The composite pollutant intrusion mode is used to compare the monitoring station data inside and outside the warehouse, as well as whether there is external pollution intrusion, to obtain the third linkage analysis result between the increase of sulfur dioxide, the increase of nitrogen dioxide and the increase of O3, and to trigger the third linkage control based on the third linkage analysis result. The organic pollutant release mode is used to investigate whether there is internal organic material release, and to obtain the fourth linkage analysis result between the increase in formaldehyde and the increase in TOVC concentration. The fourth linkage control is triggered based on the fourth linkage analysis result. The personnel activity characteristic pattern is used to analyze personnel activities in the warehouse and obtain the fifth linkage analysis results between the increase in carbon dioxide and the increase in TOVC concentration. The fifth linkage control is triggered based on the fifth linkage analysis results. The light irradiation hazard mode is used to analyze the lighting system and obtain the sixth linkage analysis result between the increase in ultraviolet light intensity and the increase in temperature. The sixth linkage control is triggered based on the sixth linkage analysis result. The particulate matter pollution mode is used to analyze particulate matter that is intruded from the outside and generated internally, and to obtain the seventh linkage analysis result of the PM2.5 concentration rising alone. The seventh linkage control is triggered based on the seventh linkage analysis result. Based on the cleanliness data and multi-parameter linkage analysis decision table, a quantitative assessment and early warning level of cultural relic damage risk are obtained, resulting in a cultural relic damage risk index and early warning level. According to network conditions, a data upload plan and local backup are output, including: Based on the aforementioned cleaning data and multi-parameter linkage analysis decision table, the damage risk index for metallic cultural relics is calculated and warning levels are assigned; the damage risk index for organic cultural relics is also calculated and warning levels are assigned. Wherein: The formula for calculating the damage risk index of metal cultural relics is as follows: ； in, The current relative humidity. The critical corrosion humidity. Humidity activation coefficient, The sulfur dioxide corrosion coefficient is... This refers to the concentration of sulfur dioxide. The coefficient of nitrogen dioxide corrosion is given. This refers to the concentration of nitrogen dioxide. The formula for calculating the damage risk index of organic cultural relics is as follows: ； in, The current Kelvin temperature, For reference Kelvin temperature, The current relative humidity. For reference relative humidity, The activation energy for hydrolysis, Let be the ideal gas constant. Humidity index Ozone damage coefficient, This refers to the ozone concentration. Determine the network status. When the network status is determined to be in normal mode, 90% of the data analysis is completed on the edge computing terminal. When the network status is determined to be in offline mode, the local storage mechanism is activated and the key monitoring functions are run at a reduced frequency. Data is lightweighted using Delta encoding and a general compression algorithm. The compressed data and event logs are then stored in a local embedded multimedia card in the form of an SQLite database to complete local disaster recovery backup of the data. Once the network is restored, the data will be automatically uploaded. Using the standard data stream, a radial basis function neural network is constructed and trained to obtain a multi-gas cross-interference model, so as to output high-precision gas concentration data after cross-interference compensation.

2. The method for online monitoring of multiple parameters of an artifact storage environment based on PoE power supply in edge computing, as described in claim 1, is characterized in that... Using the standard data stream, a radial basis function neural network is constructed and trained to obtain a multi-gas cross-interference model, which outputs high-precision gas concentration data after cross-interference compensation, including: By selecting a radial basis function neural network, the number of nodes for the original sensor readings is set as the input layer, the number of nodes for the radial basis function neurons is set as the hidden layer, and the number of nodes for the compensated concentration values ​​is set as the output layer, the network structure is obtained. Based on the network structure, the standard data stream is set as training data. The training data is processed using the K-means clustering algorithm to find the optimal center point of the radial basis function neurons. The optimal weights from the hidden layer to the output layer are calculated using the least squares method to complete the parameter optimization operation and obtain the multi-gas cross-interference model. The multi-gas cross-interference model is then verified and its performance is evaluated to output high-precision gas concentration data after cross-interference compensation.

Images