Equipment management method based on data acquisition

Abstract

The invention discloses an equipment management method based on data acquisition, and relates to the technical field of equipment management, and the method comprises the following steps: S1, selecting and configuring a sensor; s2, establishing a data acquisition protocol; s3, deploying data acquisition hardware; s4, designing and applying a data storage system; s5, performing data cleaning and preprocessing; s6, deploying an analysis model; s7, establishing a real-time monitoring system; s8, optimizing a maintenance strategy; and S9, performing regular evaluation and optimization. According to the equipment management method based on data acquisition, the running state of the equipment is monitored in real time through real-time data acquisition and analysis, and potential faults and performance problems are found in time. In combination with an advanced data processing technology, such as a machine learning analysis model, future behaviors of the equipment can be learned and predicted from historical data, so that preventive maintenance is realized, and risks of equipment failure and production interruption are reduced. In addition, a condition-based maintenance policy replaces traditional periodic maintenance.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment management, and in particular to an equipment management method based on data collection. Background Art

[0002] In modern industrial production, equipment management and maintenance are key to ensuring production efficiency and long-term equipment operation. Traditional equipment management methods rely on scheduled maintenance schedules and manual inspections. Although this method is simple, it has obvious shortcomings. First, periodic maintenance may not detect equipment failures or performance degradation in a timely manner, resulting in unexpected production downtime and high maintenance costs. Secondly, this method usually cannot make full use of equipment operation data, resulting in a lack of data support for maintenance decisions and difficulty in optimizing equipment management. In addition, traditional methods are usually backward in data collection and processing, and cannot monitor equipment status in real time, which is particularly insufficient in a rapidly changing industrial environment.

[0003] The main disadvantage of existing technologies is their reactive maintenance strategy, which relies on intervention only after equipment failure or significant performance degradation, and cannot prevent failures from occurring. In addition, the lack of efficient data utilization strategies means that a large amount of potentially valuable operational data is not fully mined, resulting in a lack of scientific basis for maintenance decisions and the inability to dynamically adjust maintenance plans to adapt to the actual operating conditions of the equipment. Therefore, a more intelligent and data-driven equipment management method is needed to improve the ability of preventive maintenance, reduce equipment failure rates, and improve overall operational efficiency. Summary of the invention

[0004] In view of the deficiencies of the prior art, the present invention provides a device management method based on data collection to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a device management method based on data collection, comprising the following steps:

[0006] S1. Select and configure sensors

[0007] According to the working principle and key performance indicators of industrial automation equipment, select sensors to measure parameters such as temperature, pressure, and vibration, and configure the sensors. The installation location and installation method of the sensors ensure maximum data accuracy and representativeness;

[0008] S2. Establish data collection protocol

[0009] Develop data collection protocols, including collection frequency, data format, and communication protocol. The data collection frequency is adjusted according to the operating speed and failure rate of the equipment to capture sufficient operating data. The communication protocol used is determined to ensure timely and secure data transmission. At the same time, a communication protocol efficiency evaluation model is constructed to optimize the selection of collection frequency, data format, and communication protocol in a quantitative form.

[0010] S3. Deploy data collection hardware

[0011] Install data acquisition hardware, such as PLC and data acquisition cards. The data acquisition hardware is compatible with the control systems of sensors and equipment to automatically collect and forward data. At the same time, the data transmission efficiency of different data acquisition hardware is regularly analyzed, and hardware maintenance and replacement are regularly performed based on data transmission efficiency;

[0012] S4. Design and application of data storage systems

[0013] Design a stable and scalable data storage system using a time series database that supports fast read and write operations and data recovery capabilities;

[0014] S5. Data cleaning and preprocessing

[0015] Before using the collected data, the data is subjected to operations including filtering out outliers, filling in missing data and normalizing the data;

[0016] S6. Deployment of analytical models

[0017] Deploy analytical models applied to equipment data to identify equipment operating status and failure modes, and then learn from historical data to predict future performance and potential problems of equipment;

[0018] S7. Establish a real-time monitoring system

[0019] Based on the collected data and analysis models, a real-time monitoring system is established to display equipment status, alarms and performance indicators in real time;

[0020] S8. Optimize maintenance strategy

[0021] Utilize the collected data and analysis results to optimize maintenance strategies, and adopt condition-based maintenance strategies to replace traditional periodic maintenance;

[0022] S9. Regular evaluation and optimization

[0023] Regularly evaluate the efficiency and effectiveness of data collection, storage and analysis, and continuously optimize the configuration of equipment management based on equipment upgrades, the emergence of new technologies and the accumulation of operating experience.

[0024] To further optimize the technical solution, in step S2, the communication protocol efficiency evaluation model (CPEM) is as follows:

[0025]

[0026] Among them, (N) represents the network stability coefficient, (R) represents the data collection frequency, (P) represents the data packet loss rate, (L) represents the fixed network delay, (D) represents the size of the data unit, (S) represents the distance between the device and the server, and (C) represents the complexity index of the communication protocol;

[0027] (λ) is a function used to adjust the impact of data size and distance on delay, and the function is expressed as:

[0028] λ(D, S) = D / S 2 ;

[0029] The communication protocol efficiency evaluation model is used to evaluate the efficiency of different communication protocols. According to the actual operating environment and requirements of industrial automation equipment, the model matches the most suitable protocol:

[0030] Determine the acquisition frequency: By adjusting the data acquisition frequency (R), observe the (CPEM) values ​​at different frequencies, and find the optimal frequency that does not overload the storage system while ensuring data quality;

[0031] Select communication protocol: Compare the (CPEM) values ​​of different communication protocols, such as Modbus and OPC UA, under the same network conditions, and select the protocol with the highest (CPEM) value to ensure timely and secure data transmission;

[0032] Network optimization: By adjusting network stability (N) and solving the data packet loss problem (P), the (CPEM) value is improved, thereby optimizing the entire data transmission process through the communication protocol efficiency evaluation model.

[0033] To further optimize the technical solution, in step S3, when regularly analyzing the data transmission efficiency of different data acquisition hardware, a data transmission efficiency model is used to analyze the data transmission efficiency. The data transmission efficiency model is shown in the following formula:

[0034]

[0035] Among them, (R d ) represents the data acquisition rate, (E f ) represents the efficiency factor of the hardware, (L t ) represents the transmission delay, (P n ) represents the network congestion parameter, (γ) is the coefficient for adjusting the impact of network congestion;

[0036] The data transmission efficiency model is used to analyze the data transmission efficiency of different data acquisition hardware, and the data acquisition hardware with the lowest DTEM value or DTEM value lower than the preset value is maintained or replaced.

[0037] To further optimize the technical solution, in step S4, the data storage system includes:

[0038] The data receiving module is used to receive input data from various sensors and data acquisition hardware, use message queues to buffer the received data stream, and eliminate obvious erroneous data through preliminary data verification;

[0039] Data processing and cleaning module, which is used to clean and preprocess the received data, including outlier detection, missing value processing and data format conversion, combining statistical methods and machine learning algorithms to identify and process outliers, and using interpolation or regression models to fill in missing data;

[0040] The data storage module is used for persistent storage of data. It uses a time series database to optimize read and write performance, optimizes the storage and query of time series data, and supports high-speed data insertion and real-time query.

[0041] The data indexing and query module uses the built-in indexing function of the database and customizes the indexing strategy to optimize specific types of queries and support complex query requirements, such as time range queries and aggregate queries;

[0042] The data security and permission management module implements TLS / SSL encrypted transmission, uses OAuth or JWT for user authentication and authorization, and database-level access control.

[0043] The data backup and recovery module is used to back up data regularly and provide a recovery mechanism when data is lost.

[0044] The data analysis support module is used to provide API and data export functions, support real-time access and batch processing of data, allow seamless integration with external analysis tools, and provide necessary data support and interfaces for upper-level data analysis and decision support.

[0045] To further optimize the technical solution, in step S5:

[0046] Before filtering outliers, it is necessary to detect outliers. The detection model is as follows:

[0047]

[0048] Among them, (x i ) is a single data point, (μ) is the calculated data mean, (σ) is the standard deviation, (z i) is the calculated Z-score, which is used to determine the degree of abnormality of a data point;

[0049] The detection model is based on the assumption of a standard normal distribution and a data point is considered anomalous if:

[0050] [|z i |>θ]

[0051] Where (θ) is the threshold value, which is adjusted according to the specific business scenario, for example:

[0052] (θ = 3.0), indicating that points beyond three standard deviations from the mean are outliers. In the temperature or pressure readings of the equipment, the detection model is used to identify and delete abnormal readings, thereby avoiding the use of abnormal data caused by sensor errors or temporary noise in the analysis;

[0053] When filling in missing data, the following formula model is used:

[0054] [x′ i =x i-1 +r·(x i+1 -x i-1 )]

[0055] Among them, (x′ i ) is the value after padding, (x i-1 ) and (x i+1 ) are the data points before and after the missing value, respectively, (r) is a random factor ranging from 0 to 1, which is selected according to the stability of the equipment operation to provide randomness to simulate the changes in real operating conditions;

[0056] This model assumes that the data has a certain continuity and linear relationship in the time series. Missing values ​​are estimated by interpolating between previous and subsequent data points, and a random factor (r) is added to increase the authenticity and naturalness of the estimate.

[0057] Further optimizing the technical solution, in step S6, the deployed analysis model includes an equipment failure prediction model and a predictive maintenance model;

[0058] The equipment failure prediction model is directly integrated into the real-time monitoring system in step S7, calculates the failure probability of the equipment in real time, and triggers an alarm when a predetermined threshold is reached;

[0059] The output of the predictive maintenance model directly affects the alarm logic of the real-time monitoring system and optimizes the operation and maintenance of the equipment based on the feedback of the model;

[0060] The equipment failure prediction model is as follows:

[0061] [P fail (t) = σ(β 0 +β1 x temp (t)+β 2 x press (t)+β 3 x vib (t) 2 )]

[0062] in,

[0063] (P fail (t)) is the probability of equipment failure at time (t),

[0064] (x temp (t), x press (t), x vib (t)) are the measured values ​​of temperature, pressure and vibration, respectively,

[0065] (β 0 , β 1 , β 2 , β 3 ) is the model parameter, which is learned through historical fault data, (σ) is the Sigmoid function, which is used to convert the linear output into a probability value;

[0066] Among them, the model parameters (β 3 ) is designed to make the impact of vibration data nonlinear, which can better capture the failure risk under enhanced vibration;

[0067] The predictive maintenance model is as follows:

[0068]

[0069] in,

[0070] (RUL) is the predicted remaining useful life,

[0071] (x temp (t), x prss (t)) are the current temperature and pressure measurements,

[0072] (γ 0 , γ 1 , γ 2 ) are model parameters, optimized by data-driven methods,

[0073] (∈) is a small constant that prevents the denominator from being zero and increases the stability of the model;

[0074] The predictive maintenance model uses inverse temperature and logarithmic pressure as input variables, and nonlinear transformation helps the model better capture the impact of complex physical processes on equipment life.

[0075] To further optimize the technical solution, in step S7, the real-time monitoring system includes:

[0076] Data visualization module, which is used to provide real-time graphical representation of equipment operation data, such as real-time charts of temperature, pressure, and vibration parameters, to intuitively monitor equipment status and identify any abnormalities or signs of deviation from normal operating modes;

[0077] The alarm generation and management module automatically triggers alarms based on preset thresholds and the output of the analysis model;

[0078] Performance indicator calculation module, used to calculate key performance indicators such as equipment efficiency, operating time and downtime;

[0079] The equipment health assessment module uses machine learning algorithms to evaluate the health status of equipment in real time and predict the remaining service life and potential failure points of the equipment;

[0080] The automatic response module controls the equipment to take preset emergency measures, such as slowing down or completely shutting down, when the detection system issues a fault warning.

[0081] To further optimize the technical solution, in step S8, the condition-based maintenance strategy determines the time and method of maintenance according to the actual working condition and performance data of the equipment, rather than following a fixed maintenance schedule, and specifically includes the following process:

[0082] Using the real-time monitoring system established in step S7, the key performance indicators of the equipment are continuously tracked, and the operating status of the equipment is understood in real time through the data;

[0083] Set alert thresholds for each key performance indicator. The thresholds are set based on historical data analysis and equipment manufacturer specifications. When the monitored data exceeds the threshold, a maintenance signal will be triggered.

[0084] Identify unusual patterns in data through analytical models;

[0085] Perform necessary maintenance activities according to maintenance strategy;

[0086] After maintenance, re-evaluate equipment performance to ensure the effectiveness of maintenance activities, and use performance data comparison before and after maintenance to evaluate whether the maintenance activities have achieved the expected results.

[0087] To further optimize the technical solution, in step S9, an evaluation model is used to perform regular evaluation, and the evaluation model is as follows:

[0088] P(t)=α·log(D(t))+β·S(t)+γ·L(t)

[0089] Among them, (P(t)) is the system performance score at time (t), (D(t)) is the data capture rate at time (t), that is, the ratio of the amount of data successfully captured and stored to the amount of data that should be captured, (S(t)) is the data storage availability at time (t), which measures the stability and scalability of data storage, (L(t)) is the data latency rate at time (t), that is, the average time delay from data acquisition to data availability for analysis, and (α, β, γ) are weight parameters that reflect the importance of different indicators in the overall performance score.

[0090] To further optimize the technical solution, the evaluation model periodically calculates (P(t)) to evaluate the overall performance, compares (P(t)) with the set performance benchmark to determine whether a system upgrade or optimization is needed, and adjusts the values ​​of (α, β, γ) based on operating experience and the application of new technologies to reflect the impact of changes on the overall performance, such as faster data transmission protocols or more efficient storage solutions, and updates model parameters to adapt to these changes. Based on the model results, specific operations are implemented, such as increasing the data buffer area, optimizing the data collection frequency, or improving the data preprocessing process.

[0091] Compared with the prior art, the present invention provides a device management method based on data collection, which has the following beneficial effects:

[0092] This data acquisition-based equipment management method monitors the operating status of equipment in real time through real-time data collection and analysis, and promptly detects potential failures and performance problems. Combined with advanced data processing technologies, such as machine learning analysis models, it can learn and predict the future behavior of equipment from historical data, thereby achieving preventive maintenance and reducing the risk of equipment failure and production interruption. In addition, condition-based maintenance strategies replace traditional periodic maintenance, making maintenance work more precise and economical, and arranging maintenance activities according to the actual condition of the equipment, thereby effectively improving maintenance efficiency and the overall reliability of the equipment. This method not only improves production efficiency, but also reduces maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 A schematic diagram of a flow chart of a device management method based on data collection proposed by the present invention;

[0094] Figure 2 A structural schematic diagram of a data storage system in a device management method based on data acquisition proposed by the present invention;

[0095] Figure 3 A schematic diagram of the structure of a real-time monitoring system in a device management method based on data acquisition proposed by the present invention;

[0096] Figure 4This is a flow chart of a condition-based maintenance strategy in step S8 of a device management method based on data collection proposed by the present invention. DETAILED DESCRIPTION

[0097] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0098] Embodiment 1:

[0099] See also Figure 1 , a device management method based on data collection, comprising the following steps:

[0100] S1. Select and configure sensors

[0101] According to the working principle and key performance indicators of industrial automation equipment, select sensors to measure parameters such as temperature, pressure, and vibration, and configure the sensors. The installation location and installation method of the sensors ensure maximum data accuracy and representativeness.

[0102] When carrying out practical applications, it is necessary to first evaluate the characteristics and environment of industrial automation equipment:

[0103] Before selecting an installation location, first evaluate the equipment's operating environment and operating characteristics. For equipment that operates in high-temperature environments, it is necessary to select sensors that can withstand high temperatures and consider heat dissipation. Understand the physical layout of the equipment and identify key monitoring points, such as bearings, gearboxes, and pressure vessels.

[0104] And determine the most important measurement points for the key performance indicators of various industrial automation equipment. For example:

[0105] Temperature: Install near the heat exchanger outlet, motor windings or bearings.

[0106] Pressure: Placed at the inlet / outlet of a pressure vessel or pipeline.

[0107] Vibration: Installed on the bearing seat or casing, close to the main shaft of the rotating machinery.

[0108] At the same time, choose the appropriate installation method according to the type of equipment:

[0109] Non-intrusive, such as using a clamping device to fix to the pipe or equipment housing, suitable for temperature and vibration sensors.

[0110] Intrusive, such as threading or welding the sensor directly onto the equipment component, is common in pressure and flow sensors.

[0111] Choose appropriate sensor materials and shielding measures to resist electromagnetic interference, chemical corrosion or mechanical wear. For example, for electromagnetically sensitive areas, choose sensors with shielded cables and metal housings.

[0112] Finally, the installation position of the sensor should be precisely located. For example, when measuring the vibration of rotating equipment, the sensor should be close to the bearing and parallel to the rotating axis. For example, for pressure and temperature, the static area of ​​the fluid should be selected to avoid interference from eddy currents and turbulence.

[0113] S2. Establish data collection protocol

[0114] Develop a data collection protocol, including collection frequency, data format, and communication protocol. The data collection frequency is adjusted according to the operating speed and failure rate of the equipment to capture sufficient operating data, and the communication protocol used is determined to ensure timely and secure data transmission. At the same time, a communication protocol efficiency evaluation model is constructed to optimize the selection of collection frequency, data format, and communication protocol in a quantitative form.

[0115] In this embodiment, the communication protocol efficiency evaluation model (CPEM) is as follows:

[0116]

[0117] Among them, (N) represents the network stability coefficient. High stability means more reliable data transmission and less network fluctuation.

[0118] (R) represents the data collection frequency. A higher frequency may result in more data transmission requirements.

[0119] (P) represents the packet loss rate. If the packet loss rate is high, data integrity will be affected and a more reliable protocol is needed.

[0120] (L) represents fixed network delay;

[0121] (D) represents the size of the data unit;

[0122] (S) represents the distance between the device and the server;

[0123] (C) represents the complexity index of the communication protocol. Simple protocols are usually easier to implement and maintain, but may not support advanced functions.

[0124] (λ) is a function used to adjust the impact of data size and distance on delay, and the function is expressed as:

[0125] λ(D, S) = D / S 2 ;

[0126] The communication protocol efficiency evaluation model is used to evaluate the efficiency of different communication protocols. According to the actual operating environment and requirements of industrial automation equipment, the model matches the most suitable protocol:

[0127] Determine the acquisition frequency: By adjusting the data acquisition frequency (R), observe the (CPEM) values ​​at different frequencies, and find the optimal frequency that does not overload the storage system while ensuring data quality;

[0128] Select communication protocol: Compare the (CPEM) values ​​of different communication protocols, such as Modbus and OPC UA, under the same network conditions, and select the protocol with the highest (CPEM) value to ensure timely and secure data transmission;

[0129] Network optimization: By adjusting network stability (N) and solving the data packet loss problem (P), the (CPEM) value is improved, thereby optimizing the entire data transmission process through the communication protocol efficiency evaluation model.

[0130] It not only provides a quantitative method to evaluate and select communication protocols, but also enables flexible adjustment and optimization based on specific data characteristics and business scenarios to achieve the optimal design of the data acquisition system.

[0131] In this embodiment, the data acquisition protocol may also be established according to the configuration parameters in Table 1.

[0132]

[0133]

[0134] Table 1

[0135] S3. Deploy data collection hardware

[0136] Install data acquisition hardware, such as PLC and data acquisition cards. The data acquisition hardware should be compatible with the control systems of sensors and equipment to automatically collect and forward data. At the same time, the data transmission efficiency of different data acquisition hardware should be regularly analyzed, and hardware maintenance and replacement should be performed regularly based on the data transmission efficiency.

[0137] In this embodiment, when the data transmission efficiency of different data acquisition hardware is analyzed periodically, a data transmission efficiency model is used to analyze the data transmission efficiency. The data transmission efficiency model is shown in the following formula:

[0138]

[0139] Among them, (R d ) represents the data acquisition rate, (E f ) represents the efficiency factor of the hardware, which ranges from 0 to 1 according to the hardware performance score. (L t) represents the transmission delay, (P n ) represents the network congestion parameter, which affects the network congestion of data transmission, and its value range is 0-100. (γ) is the coefficient for adjusting the impact of network congestion, which is an assumed value.

[0140] The data transmission efficiency model is used to analyze the data transmission efficiency of different data acquisition hardware, and the data acquisition hardware with the lowest DTEM value or DTEM value lower than the preset value is maintained or replaced.

[0141] S4. Design and application of data storage systems

[0142] Design a stable and scalable data storage system using a time series database that supports fast read and write operations and data recovery capabilities.

[0143] S5. Data cleaning and preprocessing

[0144] Before using the collected data, operations including filtering outliers, filling missing data and normalizing the data are performed.

[0145] In this embodiment, before filtering out the outliers, it is necessary to detect the outliers. The detection model is as follows:

[0146]

[0147] Among them, (x i ) is a single data point, (μ) is the calculated data mean, (σ) is the standard deviation, (z i ) is the calculated Z-score, which is used to determine the degree of abnormality of a data point;

[0148] The detection model is based on the assumption of a standard normal distribution and a data point is considered anomalous if:

[0149] [|z i |>θ]

[0150] Where (θ) is the threshold value, which is adjusted according to the specific business scenario, for example:

[0151] (θ = 3.0), indicating that points beyond three standard deviations from the mean are outliers. In the temperature or pressure readings of the equipment, the detection model is used to identify and delete abnormal readings, thereby avoiding the use of abnormal data caused by sensor errors or temporary noise in the analysis;

[0152] When filling in missing data, the following formula model is used:

[0153] [x′ i =x i-1 +r·(x i+1 -x i-1 )]

[0154] Among them, (x′ i ) is the value after padding, (x i-1 ) and (x i+1 ) are the data points before and after the missing value, respectively, (r) is a random factor ranging from 0 to 1, which is selected according to the stability of the equipment operation to provide randomness to simulate the changes in real operating conditions;

[0155] This model assumes that the data has a certain continuity and linear relationship in the time series. Missing values ​​are estimated by interpolating between previous and subsequent data points, and a random factor (r) is added to increase the authenticity and naturalness of the estimate.

[0156] During the implementation process, the original data should first be tested for outliers and cleaned, then the missing value filling model should be applied to the result data, and finally the data should be normalized to meet the needs of the subsequent analysis model. This process ensures the comprehensiveness and effectiveness of data preprocessing and supports the efficient implementation of equipment management and fault prediction.

[0157] S6. Deployment of analytical models

[0158] Deploy analytical models applied to equipment data to identify equipment operating status and failure modes, thereby learning from historical data and predicting future equipment performance and potential problems.

[0159] In this embodiment, the deployed analytical models include equipment failure prediction models and predictive maintenance models;

[0160] The equipment failure prediction model is directly integrated into the real-time monitoring system in step S7, calculates the failure probability of the equipment in real time, and triggers an alarm when a predetermined threshold is reached;

[0161] The output of the predictive maintenance model directly affects the alarm logic of the real-time monitoring system and optimizes the operation and maintenance of the equipment based on the feedback of the model;

[0162] The equipment failure prediction model is as follows:

[0163] [P fail (t) = σ(β 0 +β 1 x temp (t)+β 2 x press (t)+β 3 x vib (t) 2 )]

[0164] in,

[0165] (P fail(t)) is the probability of equipment failure at time (t),

[0166] (x temp (t), x press (t), x vib (t)) are the measured values ​​of temperature, pressure and vibration, respectively,

[0167] (β 0 , β 1 , β 2 , β 3 ) is the model parameter, which is learned through historical fault data, (σ) is the Sigmoid function, which is used to convert the linear output into a probability value;

[0168] Among them, the model parameters (β 3 ) is designed to make the impact of vibration data nonlinear, which can better capture the failure risk under enhanced vibration;

[0169] The predictive maintenance model is as follows:

[0170]

[0171] in,

[0172] (RUL) is the predicted remaining useful life,

[0173] (x temp (t), x press (t)) are the current temperature and pressure measurements,

[0174] (γ 0 , γ 1 , γ 2 ) are model parameters, optimized by data-driven methods,

[0175] (∈) is a small constant that prevents the denominator from being zero and increases the stability of the model;

[0176] The predictive maintenance model uses inverse temperature and logarithmic pressure as input variables, and nonlinear transformation helps the model better capture the impact of complex physical processes on equipment life.

[0177] S7. Establish a real-time monitoring system

[0178] Based on the collected data and analysis models, a real-time monitoring system is established to display equipment status, alarms and performance indicators in real time.

[0179] S8. Optimize maintenance strategy

[0180] Utilize the collected data and analysis results to optimize maintenance strategies, and adopt condition-based maintenance strategies to replace traditional periodic maintenance.

[0181] In this embodiment, the condition-based maintenance strategy determines the time and method of maintenance based on the actual working condition and performance data of the equipment, rather than following a fixed maintenance schedule, and specifically includes the following processes:

[0182] like Figure 4 As shown, the real-time monitoring system established in step S7 is used to continuously track the key performance indicators of the equipment and understand the operating status of the equipment in real time through the data;

[0183] Set alert thresholds for each key performance indicator. The thresholds are set based on historical data analysis and equipment manufacturer specifications. When the monitored data exceeds the threshold, a maintenance signal will be triggered.

[0184] Identify unusual patterns in data through analytical models;

[0185] Perform necessary maintenance activities according to maintenance strategy;

[0186] After maintenance, re-evaluate equipment performance to ensure the effectiveness of maintenance activities, and use performance data comparison before and after maintenance to evaluate whether the maintenance activities have achieved the expected results.

[0187] At the same time, based on the results of equipment maintenance and new data analysis, threshold settings and maintenance strategies are continuously optimized. As time goes by and technology develops, strategies are adjusted and improved to adapt to new operating environments and equipment conditions.

[0188] S9. Regular evaluation and optimization

[0189] Regularly evaluate the efficiency and effectiveness of data collection, storage and analysis, and continuously optimize the configuration of equipment management based on equipment upgrades, the emergence of new technologies and the accumulation of operating experience.

[0190] In this embodiment, an evaluation model is used for periodic evaluation, and the evaluation model is as follows:

[0191] P(t)=α·log(D(t))+β·S(t)+γ·L(t)

[0192] Among them, (P(t)) is the system performance score at time (t), (D(t)) is the data capture rate at time (t), that is, the ratio of the amount of data successfully captured and stored to the amount of data that should be captured, (S(t)) is the data storage availability at time (t), which measures the stability and scalability of data storage, (L(t)) is the data latency rate at time (t), that is, the average time delay from data acquisition to data availability for analysis, and (α, β, γ) are weight parameters that reflect the importance of different indicators in the overall performance score.

[0193] The evaluation model periodically calculates (P(t)) to evaluate the overall performance, compares (P(t)) with the set performance benchmark to determine whether a system upgrade or optimization is needed, adjusts the values ​​of (α, β, γ) based on operating experience and the application of new technologies to reflect the impact of changes on the overall performance, such as faster data transmission protocols or more efficient storage solutions, updates model parameters to adapt to these changes, and implements specific operations based on the model results, such as increasing the data buffer area, optimizing the data collection frequency, or improving the data preprocessing process.

[0194] Embodiment 2:

[0195] See also Figure 2 Based on the device management method based on data collection described in Example 1, the data storage system designed in step S4 includes:

[0196] The data receiving module is used to receive input data from various sensors and data acquisition hardware. It needs to process high-concurrency data streams and ensure preliminary verification and caching of data. It uses message queues to buffer the received data streams and eliminate obvious erroneous data through preliminary data verification.

[0197] The data processing and cleaning module is used to clean and preprocess the received data, including outlier detection, missing value processing and data format conversion. It combines statistical methods and machine learning algorithms to identify and process outliers, and uses interpolation or regression models to fill in missing data.

[0198] The data storage module is used for persistent storage of data. It uses a time series database to optimize read and write performance, such as lnfluxDB or TimescaleDB. These databases optimize the storage and query of time series data and support high-speed data insertion and real-time query.

[0199] The data indexing and query module uses the built-in indexing function of the database and customizes the indexing strategy to optimize specific types of queries and support complex query requirements, such as time range queries and aggregate queries;

[0200] The data security and permission management module implements TLS / SSL encrypted transmission, uses OAuth or JWT for user authentication and authorization, and database-level access control.

[0201] The data backup and recovery module is used to back up data regularly and provide a recovery mechanism when data is lost.

[0202] The data analysis support module is used to provide API and data export functions, support real-time access and batch processing of data, allow seamless integration with external analysis tools (such as MATLAB or R), and provide necessary data support and interfaces for upper-level data analysis and decision support.

[0203] These modules interact with each other through an integrated system architecture to ensure that every step of data collection, storage, processing, analysis and backup is efficient and secure. The system design also supports horizontal and vertical expansion to cope with the increase in the number of devices and the growth of data volume in the future.

[0204] Embodiment three:

[0205] See also Figure 3 Based on the device management method based on data collection described in Example 1, the real-time monitoring system designed in step S7 includes:

[0206] Data visualization module, which is used to provide real-time graphical representation of equipment operation data, such as real-time charts of temperature, pressure, and vibration parameters, to intuitively monitor equipment status and identify any abnormalities or signs of deviation from normal operating modes;

[0207] The alarm generation and management module automatically triggers alarms based on preset thresholds and the output of the analysis model. For example, if the vibration of the equipment exceeds the safety threshold, the system will immediately issue an alarm. Management functions include alarm classification, priority setting, and alarm history recording to facilitate subsequent review and analysis;

[0208] Performance indicator calculation module, used to calculate key performance indicators such as equipment efficiency, operating time and downtime;

[0209] The equipment health assessment module uses machine learning algorithms to evaluate the health status of equipment in real time and predict the remaining service life and potential failure points of the equipment;

[0210] The automatic response module controls the equipment to take preset emergency measures, such as slowing down or completely shutting down, when the detection system issues a fault warning.

[0211] At the same time, the system also provides a user-friendly interface for operators to monitor and control equipment in real time. The interface includes a dashboard, control panel and alarm log, which operators can use to quickly respond and manage equipment. The system automatically records detailed logs of all important events and system operations, including equipment status changes, alarm activities and user operations. Customized reports can also be generated for maintenance teams and management for performance review and decision support.

[0212] The beneficial effects of the present invention are:

[0213] This data acquisition-based equipment management method monitors the operating status of equipment in real time through real-time data collection and analysis, and promptly detects potential failures and performance problems. Combined with advanced data processing technologies, such as machine learning analysis models, it can learn and predict the future behavior of equipment from historical data, thereby achieving preventive maintenance and reducing the risk of equipment failure and production interruption. In addition, condition-based maintenance strategies replace traditional periodic maintenance, making maintenance work more precise and economical, and arranging maintenance activities according to the actual condition of the equipment, thereby effectively improving maintenance efficiency and the overall reliability of the equipment. This method not only improves production efficiency, but also reduces maintenance costs.

[0214] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0215] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A device management method based on data collection, characterized in that: The following steps are involved: S1. Select and configure sensors According to the working principle and key performance indicators of industrial automation equipment, select sensors to measure parameters such as temperature, pressure, and vibration, and configure the sensors. The installation location and installation method of the sensors ensure maximum data accuracy and representativeness; S2. Establish data collection protocol Develop data collection protocols, including collection frequency, data format, and communication protocol. The data collection frequency is adjusted according to the operating speed and failure rate of the equipment to capture sufficient operating data. The communication protocol used is determined to ensure timely and secure data transmission. At the same time, a communication protocol efficiency evaluation model is constructed to optimize the selection of collection frequency, data format, and communication protocol in a quantitative form. S3. Deploy data collection hardware Install data acquisition hardware, such as PLC and data acquisition cards. The data acquisition hardware is compatible with the control systems of sensors and equipment to automatically collect and forward data. At the same time, the data transmission efficiency of different data acquisition hardware is regularly analyzed, and hardware maintenance and replacement are regularly performed based on data transmission efficiency; S4. Design and application of data storage systems Design a stable and scalable data storage system using a time series database that supports fast read and write operations and data recovery capabilities; S5. Data cleaning and preprocessing Before using the collected data, the data is subjected to operations including filtering out outliers, filling in missing data and normalizing the data; S6. Deployment of analytical models Deploy analytical models applied to equipment data to identify equipment operating status and failure modes, and then learn from historical data to predict future performance and potential problems of equipment; S7. Establish a real-time monitoring system Based on the collected data and analysis models, a real-time monitoring system is established to display equipment status, alarms and performance indicators in real time; S8. Optimize maintenance strategy Utilize the collected data and analysis results to optimize maintenance strategies, and adopt condition-based maintenance strategies to replace traditional periodic maintenance; S9. Regular evaluation and optimization Regularly evaluate the efficiency and effectiveness of data collection, storage and analysis, and continuously optimize the configuration of equipment management based on equipment upgrades, the emergence of new technologies and the accumulation of operating experience.

2. The device management method based on data collection according to claim 1 is characterized in that: In step S2, the communication protocol efficiency evaluation model (CPEM) is as follows: Among them, (N) represents the network stability coefficient, (R) represents the data collection frequency, (P) represents the data packet loss rate, (L) represents the fixed network delay, (D) represents the size of the data unit, (S) represents the distance between the device and the server, and (C) represents the complexity index of the communication protocol; (λ) is a function used to adjust the impact of data size and distance on delay, and the function is expressed as: λ(D,S)=D / S 2 ; The communication protocol efficiency evaluation model is used to evaluate the efficiency of different communication protocols. According to the actual operating environment and requirements of industrial automation equipment, the model matches the most suitable protocol: Determine the acquisition frequency: By adjusting the data acquisition frequency (R), observe the (CPEM) values ​​at different frequencies, and find the optimal frequency that does not overload the storage system while ensuring data quality; Select communication protocol: Compare the (CPEM) values ​​of different communication protocols, such as Modbus and OPC UA, under the same network conditions, and select the protocol with the highest (CPEM) value to ensure timely and secure data transmission; Network optimization: By adjusting network stability (N) and solving the data packet loss problem (P), the (CPEM) value is improved, thereby optimizing the entire data transmission process through the communication protocol efficiency evaluation model.

3. The device management method based on data collection according to claim 1 is characterized in that: In step S3, when regularly analyzing the data transmission efficiency of different data acquisition hardware, a data transmission efficiency model is used to analyze the data transmission efficiency. The data transmission efficiency model is shown in the following formula: Among them, (R d ) represents the data acquisition rate, (E f ) represents the efficiency factor of the hardware, (L t ) represents the transmission delay, (P n ) represents the network congestion parameter, (γ) is the coefficient for adjusting the impact of network congestion; The data transmission efficiency model is used to analyze the data transmission efficiency of different data acquisition hardware, and the data acquisition hardware with the lowest DTEM value or DTEM value lower than the preset value is maintained or replaced.

4. The device management method based on data collection according to claim 1 is characterized in that: In step S4, the data storage system includes: The data receiving module is used to receive input data from various sensors and data acquisition hardware, use message queues to buffer the received data stream, and eliminate obvious erroneous data through preliminary data verification; Data processing and cleaning module, which is used to clean and preprocess the received data, including outlier detection, missing value processing and data format conversion, combining statistical methods and machine learning algorithms to identify and process outliers, and using interpolation or regression models to fill in missing data; The data storage module is used for persistent storage of data. It uses a time series database to optimize read and write performance, optimizes the storage and query of time series data, and supports high-speed data insertion and real-time query. The data indexing and query module uses the built-in indexing function of the database and customizes the indexing strategy to optimize specific types of queries and support complex query requirements, such as time range queries and aggregate queries; The data security and permission management module implements TLS / SSL encrypted transmission, uses 0Auth or JWT for user authentication and authorization, and database-level access control. The data backup and recovery module is used to back up data regularly and provide a recovery mechanism when data is lost. The data analysis support module is used to provide API and data export functions, support real-time access and batch processing of data, allow seamless integration with external analysis tools, and provide necessary data support and interfaces for upper-level data analysis and decision support.

5. The device management method based on data collection according to claim 1 is characterized in that: In step S5: Before filtering outliers, it is necessary to detect outliers. The detection model is as follows: Among them, (x i ) is a single data point, (μ) is the calculated data mean, (σ) is the standard deviation, (z i ) is the calculated Z-score, which is used to determine the degree of abnormality of a data point; The detection model is based on the assumption of standard normal distribution and a data point is considered anomalous if: [|from i |>θ] Where (θ) is the threshold value, which is adjusted according to the specific business scenario, for example: (θ=3.0)), indicating that points beyond three standard deviations from the mean are outliers. In the temperature or pressure readings of the equipment, the detection model is used to identify and delete abnormal readings, thereby avoiding the use of abnormal data caused by sensor errors or temporary noise in the analysis; When filling in missing data, the following formula model is used: [x′ i =x i-1 +r·(x i+1 -x i-1 )] Among them, (x′ i ) is the value after padding, (x i-1 ) and (x i+1 ) are the data points before and after the missing value, respectively, (r) is a random factor ranging from 0 to 1, which is selected according to the stability of the equipment operation to provide randomness to simulate the changes in real operating conditions; The model assumes that the data has a certain continuity and linear relationship in the time series. The missing values ​​are estimated by interpolating between the previous and next data points, and a random factor (r) is added to increase the authenticity and naturalness of the estimate.

6. The device management method based on data collection according to claim 1 is characterized in that: In step S6, the deployed analysis models include equipment failure prediction models and predictive maintenance models; The equipment failure prediction model is directly integrated into the real-time monitoring system in step S7, calculates the failure probability of the equipment in real time, and triggers an alarm when a predetermined threshold is reached; The output of the predictive maintenance model directly affects the alarm logic of the real-time monitoring system and optimizes the operation and maintenance of the equipment based on the feedback of the model; The equipment failure prediction model is as follows: [P fail (t)=σ(β0+β1x temp (t)+β2x press (t)+β3x υib (t) 2 )] in, (P fail (t)) is the probability of equipment failure at time (t), (x temp (t), x press (t), x υib (t)) are the measured values ​​of temperature, pressure and vibration, respectively, (β0, β1, β2, β3) are model parameters, which are learned through historical fault data, and (σ) is the Sigmoid function, which is used to convert linear output into probability value; Among them, the design of the model parameter (β3) makes the impact of vibration data nonlinear, which can better capture the failure risk under vibration enhancement; The predictive maintenance model is as follows: in, (RUL) is the predicted remaining useful life, (x temp (t), x prss (t)) are the current temperature and pressure measurements, (γ0, γ1, γ2) are model parameters, optimized by data-driven methods, (∈) is a small constant that prevents the denominator from being zero and increases the stability of the model; The predictive maintenance model uses inverse temperature and logarithmic pressure as input variables, and nonlinear transformation helps the model better capture the impact of complex physical processes on equipment life.

7. The device management method based on data collection according to claim 1 is characterized in that: In step S7, the real-time monitoring system includes: Data visualization module, which is used to provide real-time graphical representation of equipment operation data, such as real-time charts of temperature, pressure, and vibration parameters, to intuitively monitor equipment status and identify any abnormalities or signs of deviation from normal operating modes; The alarm generation and management module automatically triggers alarms based on preset thresholds and the output of the analysis model; Performance indicator calculation module, used to calculate key performance indicators such as equipment efficiency, operating time and downtime; The equipment health assessment module uses machine learning algorithms to evaluate the health status of equipment in real time and predict the remaining service life and potential failure points of the equipment; The automatic response module controls the equipment to take preset emergency measures, such as slowing down or completely shutting down, when the detection system issues a fault warning.

8. The device management method based on data collection according to claim 1 is characterized in that: In step S8, the condition-based maintenance strategy determines the time and method of maintenance according to the actual working condition and performance data of the equipment, rather than following a fixed maintenance schedule, and specifically includes the following process: Using the real-time monitoring system established in step S7, the key performance indicators of the equipment are continuously tracked, and the operating status of the equipment is understood in real time through the data; Set alert thresholds for each key performance indicator. The thresholds are set based on historical data analysis and equipment manufacturer specifications. When the monitored data exceeds the threshold, a maintenance signal will be triggered. Identify unusual patterns in data through analytical models; Perform necessary maintenance activities according to maintenance strategy; After maintenance, re-evaluate equipment performance to ensure the effectiveness of maintenance activities, and use performance data comparison before and after maintenance to evaluate whether the maintenance activities have achieved the expected results.

9. The device management method based on data collection according to claim 1, characterized in that: In step S9, an evaluation model is used to perform regular evaluation, and the evaluation model is as follows: P(t)=α·log(D(t))+β·S(t)+γ·L(t) Among them, (P(t)) is the system performance score at time (t), (D(t)) is the data capture rate at time (t), that is, the ratio of the amount of data successfully captured and stored to the amount of data that should be captured, (S(t)) is the data storage availability at time (t), which measures the stability and scalability of data storage, (L(t)) is the data latency rate at time (t), that is, the average time delay from data acquisition to data availability for analysis, and (α, β, γ) are weight parameters that reflect the importance of different indicators in the overall performance score.

10. The device management method based on data collection according to claim 9, characterized in that: The evaluation model periodically calculates (P(t)) to evaluate the overall performance, compares (P(t)) with the set performance benchmark to determine whether a system upgrade or optimization is needed, adjusts the values ​​of (α, β, γ) based on operating experience and the application of new technologies to reflect the impact of changes on the overall performance, such as faster data transmission protocols or more efficient storage solutions, updates model parameters to adapt to these changes, and implements specific operations based on the model results, such as increasing the data buffer area, optimizing the data collection frequency, or improving the data preprocessing process.

Images