Identification method and system based on data acquisition

Abstract

The invention is suitable for the technical field of data processing, and provides an identification method and system based on data acquisition, and the method comprises the steps: collecting detection data, and summarizing the collected data; acquiring summarized data, preprocessing the summarized data, performing feature extraction on the preprocessed data, acquiring key features in the preprocessed data through feature extraction, and comparing the key features with a key feature threshold value; if the abnormal working state is judged, querying a generation reason according to the abnormal working state, querying a processing scheme according to a specific reason, and judging the rationality of the processing scheme according to a processing result; the method comprises the following steps: preprocessing data, performing visualization processing according to the preprocessed data, predicting a subsequent operation state according to a broken line graph, comparing the broken line graph with a predicted value, and judging an alarm value according to a comparison result, thereby solving the problems that abnormal conditions in existing data cannot be queried and cannot be timely processed in data processing after data acquisition in the prior art; and early warning cannot be carried out.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and particularly relates to an identification method and system based on data collection. Background Art

[0002] With the rapid development of modern manufacturing and the improvement of equipment automation level, the health monitoring and fault diagnosis of equipment in the industrial production process become more and more important. Especially in key industries such as rail transit, metallurgy, electric power, and petrochemical industry, the long-term stable operation of equipment is directly related to production safety, economic benefits, and the sustainable development of enterprises. Therefore, the demand for real-time monitoring and warning systems for equipment is increasing day by day.

[0003] In traditional equipment monitoring systems, manual inspections and regular maintenance are mainly relied on to ensure the normal operation of equipment.

[0004] However, the existing technologies have problems that abnormal situations in existing data cannot be queried after data collection, abnormal situations cannot be processed in a timely manner, and early warnings cannot be given. Summary of the Invention

[0005] The purpose of the embodiments of the present invention is to provide an identification method based on data collection, aiming to solve the problems proposed in the third part of the background art.

[0006] The embodiments of the present invention are implemented as follows. An identification method based on data collection, the method includes:

[0007] Collect detection data, where the monitoring data includes vibration data, temperature data, current and voltage data, pressure data, and energy consumption data, and summarize the collected data;

[0008] Obtain the summarized data, preprocess the summarized data, extract features from the preprocessed data, obtain key features in the preprocessed data through feature extraction, and compare the key features with the key feature thresholds;

[0009] Obtain the comparison result. If it is determined to be an abnormal working state, obtain the cause of the abnormal state generation, query the cause according to the abnormal working state, query the treatment plan according to the specific cause, and determine the rationality of the treatment plan according to the treatment result;

[0010] Perform visualization processing based on the preprocessed data, predict the subsequent operating state according to the line chart, compare the line chart with the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0011] Preferably, the steps of obtaining the aggregated data, preprocessing the aggregated data, extracting features from the preprocessed data, obtaining key features from the preprocessed data through feature extraction, and comparing the key features with the key feature threshold specifically include:

[0012] Obtain the aggregated data, preprocess the aggregated data, where the preprocessing includes filtering and denoising, and obtain the preprocessed data;

[0013] Extract features from the preprocessed data, where the feature extraction includes time-domain analysis and frequency-domain analysis, and obtain key features from the preprocessed data through feature extraction;

[0014] Obtain the key feature threshold, where the key feature threshold is the allowable range in the normal working state, and compare the key features with the key feature threshold.

[0015] Preferably, the steps of obtaining the comparison result, if it is determined to be an abnormal working state, obtaining the cause of the abnormal state generation, querying the cause according to the abnormal working state, querying the processing solution according to the specific cause, and determining the rationality of the processing solution according to the processing result specifically include:

[0016] Obtain the comparison result, if it is determined to be an abnormal working state, where the abnormal working state is the key feature exceeding the key feature threshold, and obtain the cause of the abnormal state generation;

[0017] Query the cause according to the abnormal working state to obtain the specific cause, obtain the processing solution for the abnormal state, query the processing solution according to the specific cause, and send the processing solution to the terminal;

[0018] Obtain the processing result, and determine the rationality of the processing solution according to the processing result, where the rationality is used to evaluate the matching degree of the selected processing solution.

[0019] Preferably, the steps of performing visualization processing on the preprocessed data, predicting the subsequent operating state according to the line chart, comparing the line chart and the predicted value, determining the alarm value according to the comparison result, and sending the alarm value to the terminal specifically include:

[0020] Perform visualization processing on the preprocessed data, where the visualization processing uses a line chart, and predict the subsequent operating state according to the line chart to obtain the predicted value;

[0021] Obtain the alarm value, where the alarm value is the magnitude of the fault occurrence trend based on the healthy state benchmark, and the alarm value is set to a critical alarm value and a warning alarm value;

[0022] Compare the line chart and the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0023] Preferably, the key features are vibration, temperature, current voltage, pressure, and energy consumption.

[0024] Another object of the embodiments of the present invention is to provide an identification system based on data collection, the system comprising:

[0025] A collection module that collects detection data, the monitoring data including vibration data, temperature data, current voltage data, pressure data, and energy consumption data, and summarizes the collected data;

[0026] A feature extraction module that obtains the summarized data, preprocesses the summarized data, extracts features from the preprocessed data, obtains the key features in the preprocessed data through feature extraction, and compares the key features with the key feature threshold;

[0027] A processing solution module that obtains the comparison result, if it is determined to be an abnormal working state, obtains the cause of the abnormal state generation, queries the cause according to the abnormal working state, queries the processing solution according to the specific cause, and determines the rationality of the processing solution according to the processing result;

[0028] An early warning module that performs visualization processing on the preprocessed data, predicts the subsequent operating state according to the line chart, compares the line chart with the predicted value, determines the alarm value according to the comparison result, and sends the alarm value to the terminal.

[0029] Preferably, the feature extraction module includes:

[0030] A preprocessing unit that obtains the summarized data, preprocesses the summarized data, the preprocessing including filtering and denoising, and obtains the preprocessed data;

[0031] A feature extraction unit that extracts features from the preprocessed data, the feature extraction including time-domain analysis and frequency-domain analysis, and obtains the key features in the preprocessed data through feature extraction;

[0032] A key feature unit that obtains the key feature threshold, the key feature threshold being the allowable range of the normal working state, and compares the key features with the key feature threshold.

[0033] Preferably, the processing solution module includes:

[0034] An abnormal unit that obtains the comparison result, if it is determined to be an abnormal working state, the abnormal working state being the key feature exceeding the key feature threshold, and obtains the cause of the abnormal state generation;

[0035] A processing solution unit that queries the cause according to the abnormal working state to obtain the specific cause, obtains the processing solution for the abnormal state, queries the processing solution according to the specific cause, and sends the processing solution to the terminal;

[0036] A rationality evaluation unit obtains the processing result and determines the rationality of the processing scheme according to the processing result. The rationality is used to evaluate the matching degree of the selected processing scheme.

[0037] Preferably, the warning module includes:

[0038] A visualization unit performs visualization processing according to the preprocessed data. The visualization processing uses a line chart to predict the subsequent operating state and obtains a predicted value;

[0039] A warning determination unit obtains an alarm value. The alarm value is the magnitude of the fault occurrence trend based on the health status benchmark, and the alarm value is set as a critical alarm value and a warning alarm value;

[0040] A warning unit compares the line chart with the predicted value, determines the alarm value according to the comparison result, and sends the alarm value to the terminal.

[0041] Preferably, the key features are vibration, temperature, current and voltage, pressure, and energy consumption.

[0042] An identification method based on data collection provided by an embodiment of the present invention collects detection data, summarizes the collected data to obtain the summarized data, preprocesses the summarized data, and the preprocessing includes filtering and denoising to obtain the preprocessed data. Feature extraction is performed on the preprocessed data, and key features in the preprocessed data are obtained through feature extraction. A key feature threshold is obtained. The key feature threshold is the allowable range in the normal working state. The key feature is compared with the key feature threshold to obtain a comparison result. If it is determined to be an abnormal working state, the cause of the abnormal state is obtained, the cause is queried according to the abnormal working state to obtain the specific cause, the processing scheme for the abnormal state is obtained, the processing scheme is queried according to the specific cause and sent to the terminal, the processing result is obtained, the rationality of the processing scheme is determined according to the processing result, visualization processing is performed according to the preprocessed data, the visualization processing uses a line chart to predict the subsequent operating state to obtain a predicted value, an alarm value is obtained, the alarm value is set as a critical alarm value and a warning alarm value, the line chart is compared with the predicted value, the alarm value is determined according to the comparison result, and the alarm value is sent to the terminal, solving the problem that the existing data processing after data collection cannot query the abnormal conditions in the existing data, cannot process the abnormal conditions in time, and cannot give warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flowchart of an identification method based on data collection provided by an embodiment of the present invention;

[0044] Figure 2The flowchart of the steps for preprocessing the aggregated data, obtaining key features in the preprocessed data through feature extraction, and comparing the key features with the key feature thresholds provided by the embodiments of the present invention;

[0045] Figure 3 The flowchart of the steps for obtaining the comparison result, querying the generation reason according to the abnormal working state, and determining the rationality of the processing scheme according to the processing result provided by the embodiments of the present invention;

[0046] Figure 4 The flowchart of the steps for performing visualization processing on the preprocessed data, predicting the subsequent operating state according to the line chart, and determining the alarm value according to the comparison result provided by the embodiments of the present invention;

[0047] Figure 5 The architecture diagram of an identification system based on data acquisition provided by the embodiments of the present invention;

[0048] Figure 6 The architecture diagram of the feature extraction module provided by the embodiments of the present invention;

[0049] Figure 7 The architecture diagram of the processing scheme module provided by the embodiments of the present invention;

[0050] Figure 8 The architecture diagram of the early warning module provided by the embodiments of the present invention. Detailed implementation manners

[0051] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0052] It can be understood that the terms "first", "second", etc. used in this application can be used in this document to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish the first element from another element. For example, without departing from the scope of this application, the first xx script can be called the second xx script, and similarly, the second xx script can be called the first xx script.

[0053] As Figure 1 shown, an identification method based on data acquisition provided by the embodiments of the present invention, the method includes:

[0054] S100, collecting detection data, the monitoring data includes vibration data, temperature data, current and voltage data, pressure data and energy consumption data, and aggregating the collected data.

[0055] In this step, the detection data is collected. The collection of detection data is to ensure the normal operation of the equipment and give early warnings before potential failures occur. The collected data usually covers multiple physical quantities, which can reflect the operating status and health condition of the equipment;

[0056] The monitored data includes vibration data, temperature data, current and voltage data, pressure data, and energy consumption data. Vibration data can reflect the mechanical operating status of the equipment and is usually used to detect problems such as wear, looseness, imbalance, or misalignment of mechanical components. Temperature data is used to monitor the operating temperature of the equipment. Excessive temperature may indicate problems such as equipment overload, insufficient heat dissipation, or component damage. Current and voltage data reflect the power consumption of the equipment. Abnormal current fluctuations or voltage drops may indicate faults in the internal circuit of the equipment. Pressure data is used to monitor the pressure condition in the equipment or pipeline. Abnormal pressure changes may be problems such as pipeline rupture, valve failure, or equipment blockage. Energy consumption data helps monitor the operating efficiency of the equipment. A sudden increase in energy consumption may indicate abnormal operation of the equipment, and a decrease in energy consumption may be a signal of declining equipment performance;

[0057] The collected data needs to be summarized. It is necessary to integrate the data into a unified data platform, summarize the data, and record the timestamps of the data to ensure the integrity and traceability of the data.

[0058] S200, Obtain the summarized data, preprocess the summarized data, extract features from the preprocessed data, obtain the key features in the preprocessed data through feature extraction, and compare the key features with the key feature thresholds.

[0059] In this step, obtain the summarized data. After data collection, the data usually contains noise, etc., and needs to be preprocessed. For noise, filtering methods such as low-pass filtering or high-pass filtering are used to remove high-frequency noise or low-frequency drift;

[0060] Extract features from the preprocessed data. Extract meaningful features from the preprocessed data, which can help identify the operating status of the equipment, perform fault prediction, or optimize control, and obtain the key features in the preprocessed data through feature extraction;

[0061] Compare the key features with the key feature thresholds. The thresholds are set through historical data analysis and the definition of normal operating conditions of the equipment. These thresholds usually represent the dividing line between the healthy state and the fault state of the equipment. If the feature value exceeds the set threshold, it means that the equipment may have a fault or abnormality.

[0062] S300. Obtain the comparison result. If it is determined that the working state is abnormal, obtain the cause of the abnormal state. Query the cause according to the abnormal working state, query the treatment plan according to the specific cause, and determine the rationality of the treatment plan according to the treatment result.

[0063] In this step, obtain the comparison result. When the key feature data is within the threshold range, it indicates that the device is operating normally and is determined to be in a normal working state. When one or more key features exceed the set threshold, it indicates that the device may have a fault, deviation, or other abnormal conditions. If it is determined that the working state is abnormal;

[0064] Obtain the cause of the abnormal state. Conduct an in-depth analysis of the abnormality to find the specific cause. Query the treatment plan according to the specific cause. The treatment plan can be obtained from the device's maintenance manual and fault database. After obtaining and implementing the treatment plan, an important step is to evaluate the rationality and effectiveness of the treatment plan, and determine the rationality of the treatment plan according to the treatment result.

[0065] S400. Perform visualization processing on the preprocessed data, predict the subsequent operating state based on the line chart, compare the line chart with the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0066] In this step, perform visualization processing on the preprocessed data, display the preprocessed data in a graphical way to make it more intuitive and easy to understand, and help analysts quickly identify trends, fluctuations, and abnormalities. The visualization methods include line charts. Predict the subsequent operating state based on the line chart, and predict the future operating state of the device based on the existing data;

[0067] Compare the line chart with the predicted value to check whether the device is operating as expected. If the actual data is within the range of the predicted value, it indicates that the device is operating normally; if the actual data deviates too much from the predicted value, it indicates that the device may have an abnormality;

[0068] Determine the alarm value according to the comparison result. The alarm value is the threshold set according to the range of the device in the normal operating state. If the comparison result shows that the actual value of the device exceeds this range, an alarm needs to be triggered and the alarm value is sent to the terminal.

[0069] As Figure 2 shown, as a preferred embodiment of the present invention, the steps of obtaining the summarized data, preprocessing the summarized data, performing feature extraction on the preprocessed data, and obtaining the key features in the preprocessed data through feature extraction, and comparing the key features with the key feature threshold specifically include:

[0070] S201. Obtain the summarized data and preprocess the summarized data. The preprocessing includes filtering and denoising to obtain the preprocessed data.

[0071] In this step, the aggregated data is obtained. After obtaining the aggregated data, preprocessing is a crucial step to ensure data quality and analysis accuracy. Preprocess the aggregated data, which includes filtering and denoising. Filtering is to remove unwanted frequency components from the data, and denoising aims to remove random noise or useless signals generated by external interference in the data. Obtain the preprocessed data. After filtering and denoising, the dataset has been preliminarily cleaned and optimized, and then feature extraction and analysis can be carried out;

[0072] For example, low-pass filtering can be used to remove high-frequency noise in vibration data. If there are some sudden outliers in the vibration data, median filtering can be adopted to remove these noises.

[0073] S202. Extract features from the preprocessed data. The feature extraction includes time-domain analysis and frequency-domain analysis, and key features in the preprocessed data are obtained through feature extraction.

[0074] In this step, extract features from the preprocessed data. The feature extraction includes time-domain analysis and frequency-domain analysis. Time-domain analysis is an analysis based on the changes of data over time and is usually used to extract statistical features directly related to time changes. It reveals the change trend or volatility of the signal by calculating the statistics of the data, and reflects the average level of the data through the mean value. For vibration or temperature data, the mean value can tell us whether the working state of the device is abnormal. Frequency-domain analysis is to convert the time-domain signal into frequency components to analyze the intensity of different frequency components in the signal, which is very effective for periodic signals or systems with large fluctuations, especially vibration data in mechanical equipment;

[0075] Combining the two analysis methods helps to comprehensively understand the operating condition of the device and discover potential faults in advance;

[0076] The key features are vibration, temperature, current / voltage, pressure, and energy consumption. Vibration features are used to evaluate the mechanical health state of the device. Temperature features reflect the thermal state of the device. Abnormal temperature changes are usually early warning signals of device overload or damage. Current / voltage features are used to reflect the power consumption and load condition of the device. Abnormal current fluctuations may indicate device overload, failure, or low efficiency. Pressure features affect the working state of the device through pressure changes. Too high or too low pressure is usually an early signal of device failure. Energy consumption features are used to monitor device efficiency, and the fluctuations in energy consumption reflect the load and energy efficiency of device operation.

[0077] S203. Obtain the key feature thresholds. The key feature thresholds are the allowable ranges in the normal working state, and compare the key features with the key feature thresholds.

[0078] In this step, the key feature threshold is obtained. The key feature threshold is the allowable range in the normal working state. When the monitored key feature exceeds this range, it can be considered that the device may have a fault or anomaly, and fault diagnosis or maintenance is required;

[0079] The setting of the threshold depends on statistical analysis of the data, and combines the working conditions of the device and the historical operation situation to determine the appropriate upper and lower limit ranges, mean ± k × standard deviation: This method is used to set the threshold of the normal distribution, where k is usually a constant set according to experience, such as 2 or 3, indicating the data fluctuation within the normal working range;

[0080] For example, if the mean value of the vibration data of the device during normal operation is 0.05 m / s² and the standard deviation is 0.01 m / s², then the upper and lower limit thresholds are set as:

[0081] Upper limit threshold = 0.05 + (2 × 0.01) = 0.07 m / s²;

[0082] Lower limit threshold = 0.05 - (2 × 0.01) = 0.03 m / s²;

[0083] When the monitored vibration value exceeds the range of 0.03 to 0.07 m / s², it may indicate that there is a problem with the device.

[0084] Compare the key feature with the key feature threshold. If the data is within the threshold range, the device is operating normally. If the data exceeds the threshold range, it indicates that the device may have an anomaly or fault.

[0085] As Figure 3 shown, as a preferred embodiment of the present invention, the steps of obtaining the comparison result, if it is determined to be an abnormal working state, obtaining the cause of the abnormal state generation, querying the cause according to the abnormal working state, querying the processing solution according to the specific cause, and determining the rationality of the processing solution according to the processing result specifically include:

[0086] S301, obtain the comparison result. If it is determined to be an abnormal working state, the abnormal working state is the key feature exceeding the key feature threshold, and obtain the cause of the abnormal state generation.

[0087] In this step, obtain the comparison result. When the value of one or more key features exceeds the preset threshold, it is determined to be an abnormal working state. Once it is determined to be an abnormal working state, the next step is to obtain the cause of the abnormal state generation;

[0088] For example, excessive vibration may be related to mechanical failures of the equipment, such as imbalance and looseness. Common causes of excessive vibration may be bearing damage and motor imbalance. Excessive energy consumption may be related to overload and power system failures. The reasons for excessive energy consumption may be motor overload and reduced equipment efficiency.

[0089] S302. Query the generation cause according to the abnormal working state to obtain the specific cause, obtain the processing solution for the abnormal state, query the processing solution according to the specific cause, and send the processing solution to the terminal.

[0090] In this step, query the generation cause according to the abnormal working state. Once the equipment is determined to be in an abnormal working state, it is first necessary to determine the specific cause of the abnormality. The specific cause can be queried through data analysis and the fault database. Through data analysis, the relationship between the abnormal characteristics and the equipment health status can be analyzed in combination with historical data. Through querying the fault database, possible causes can be found by matching the current abnormal characteristics and the fault mode library.

[0091] Obtain the processing solution for the abnormal state. The processing solution is formulated based on the equipment maintenance manual or industry standards, and the operation process for specific fault causes is defined. Query the processing solution according to the specific cause. For example, for bearing wear, the common processing solution is to replace or lubricate the bearing.

[0092] After the generation of the processing solution is completed, send the processing solution to the terminal and send these solutions to the operator or maintenance personnel.

[0093] S303. Obtain the processing result and determine the rationality of the processing solution according to the processing result. The rationality is used to evaluate the matching degree of the selected processing solution.

[0094] In this step, obtain the processing result. The processing result refers to the changes in the equipment operation state after the implementation of the processing solution. These changes are real-time feedback through the monitoring system, such as whether the key characteristics of the equipment, such as vibration, temperature, and power consumption, return to the normal range.

[0095] Determine the rationality of the processing solution according to the processing result. The rationality is used to evaluate the matching degree of the selected processing solution, whether it can effectively, timely, and low-costly solve the problem.

[0096] Suppose the vibration sensor in a factory shows that a certain motor exceeds the normal vibration range. The system proposes possible fault causes (such as bearing damage) and recommends solving it by bearing replacement or balance adjustment. After the implementation of the treatment, the vibration data returns to the normal range. Then, evaluate the rationality of the plan: it matches the root cause of the problem (bearing problem), effectively solves the problem after implementation, and has low cost and less time consumption. Therefore, this plan is rated as reasonable.

[0097] Such asFigure 4 As shown, as a preferred embodiment of the present invention, the steps of performing visualization processing based on the preprocessed data, predicting the subsequent operating state according to the line chart, comparing the line chart with the predicted value, determining the alarm value according to the comparison result, and sending the alarm value to the terminal specifically include:

[0098] S401, perform visualization processing based on the preprocessed data. The visualization processing uses a line chart to predict the subsequent operating state according to the line chart and obtain a predicted value.

[0099] In this step, perform visualization processing based on the preprocessed data. The visualization processing uses a line chart to display the preprocessed data, which can help us more clearly see the operating state of the device at different time points;

[0100] Predict the subsequent operating state according to the line chart. The subsequent operating state can be predicted according to the change trend of historical data to obtain a predicted value. The possible subsequent changes can be judged by observing the change trend of historical data. For example, if the vibration intensity continues to increase and there is no obvious sign of decline, we can predict that the vibration may continue to increase at the next time point and reach the critical value of equipment failure.

[0101] S402, obtain the alarm value. The alarm value is the magnitude of the fault occurrence trend based on the healthy state benchmark, and the alarm value is set as the critical alarm value and the warning alarm value.

[0102] In this step, obtain the alarm value. The alarm value is the magnitude of the fault occurrence trend based on the healthy state benchmark. The healthy state benchmark refers to the performance index standard of the device under normal working conditions. The alarm value is set as the critical alarm value and the warning alarm value;

[0103] The warning alarm value is a threshold set near the healthy state benchmark, which is used to indicate the initial signal of a certain deviation or fault of the device. At this time, the device is still in an operable state, but monitoring and inspection are required;

[0104] The critical alarm value: It is set as the maximum deviation that the device can withstand. When the threshold is exceeded, the device may have serious faults or damages. If the operating parameters of the device reach or exceed this value, it usually means that the device needs to be immediately shut down for maintenance or repair.

[0105] S403, compare the line chart with the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0106] In this step, compare the line chart with the predicted value. By comparing the actual line chart with the predicted value, determine whether the future operating trend of the device meets the normal expectation. If the predicted value shows that the future state of the device will enter or has exceeded the set alarm threshold, the system will trigger the alarm mechanism;

[0107] Once the system determines that the device is about to enter a fault state, the system generates an alarm according to the preset alarm value and sends the alarm information to the terminal device.

[0108] As Figure 5 shown, a recognition system based on data acquisition provided by an embodiment of the present invention, the system includes:

[0109] An acquisition module 100, configured to acquire detection data, where the monitoring data includes vibration data, temperature data, current and voltage data, pressure data, and energy consumption data, and summarize the acquired data.

[0110] In this system, the acquisition module 100 acquires detection data. The detection data acquisition is to ensure the normal operation of the device and give an early warning before potential faults occur. The acquired data usually covers multiple physical quantities, and these physical quantities can reflect the operating state and health status of the device;

[0111] The monitoring data includes vibration data, temperature data, current and voltage data, pressure data, and energy consumption data. The vibration data can reflect the mechanical operating state of the device and is usually used to detect problems such as wear, looseness, imbalance, or misalignment of mechanical components. The temperature data is used to monitor the operating temperature of the device. Excessive temperature may indicate problems such as device overload, insufficient heat dissipation, or component damage. The current and voltage data reflect the power consumption of the device. Abnormal current fluctuations or voltage drops may indicate faults in the internal circuit of the device. The pressure data is used to monitor the pressure condition in the device or pipeline. Abnormal pressure changes may be problems such as pipeline rupture, valve failure, or device blockage. The energy consumption data helps monitor the operating efficiency of the device. A sudden increase in energy consumption may indicate abnormal operation of the device, and a decrease in energy consumption may be a signal of device performance degradation;

[0112] To summarize the acquired data, the data needs to be integrated into a unified data platform, the data is summarized, and the time stamp of the data is recorded to ensure the integrity and traceability of the data.

[0113] A feature extraction module 200, configured to obtain the summarized data, preprocess the summarized data, extract features from the preprocessed data, obtain the key features in the preprocessed data through feature extraction, and compare the key features with the key feature thresholds.

[0114] In this system, the feature extraction module 200 obtains the summarized data. After data acquisition, the data usually contains noise, etc., and needs to be preprocessed. For noise, filtering methods such as low-pass filtering or high-pass filtering are used to remove high-frequency noise or low-frequency drift;

[0115] Feature extraction is performed on the preprocessed data to extract meaningful features from the preprocessed data, which can help identify the operating state of the device, perform fault prediction or optimize control, and obtain the key features in the preprocessed data through feature extraction;

[0116] Compare the key features with the key feature thresholds, which are set through historical data analysis and the definition of the normal operating conditions of the device. These thresholds usually represent the dividing line between the healthy state and the fault state of the device. If the feature value exceeds the set threshold, it means that the device may have a fault or anomaly.

[0117] The processing solution module 300 is used to obtain the comparison result. If it is determined that the device is in an abnormal working state, obtain the cause of the abnormal state, query the cause according to the abnormal working state, query the processing solution according to the specific cause, and determine the rationality of the processing solution according to the processing result.

[0118] In this system, the processing solution module 300 obtains the comparison result. When the key feature data is within the threshold range, it indicates that the device is operating normally and is determined to be in a normal working state. When one or more key features exceed the set threshold, it means that the device may have a fault, deviation or other abnormal conditions. If it is determined that the device is in an abnormal working state;

[0119] Obtain the cause of the abnormal state, conduct an in-depth analysis of the anomaly to find the specific cause, query the processing solution according to the specific cause, and the processing solution can be obtained from the device's maintenance manual and fault database. After obtaining and implementing the processing solution, an important step is to evaluate the rationality and effectiveness of the processing solution, and determine the rationality of the processing solution according to the processing result.

[0120] The warning module 400 is used to perform visual processing on the preprocessed data, predict the subsequent operating state according to the line chart, compare the line chart with the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0121] In this system, the warning module 400 performs visual processing on the preprocessed data, displays the preprocessed data in a graphical way to make it more intuitive and understandable, and helps analysts quickly identify trends, fluctuations and anomalies. The visualization methods include line charts, and predict the subsequent operating state according to the line chart, and predict the future operating state of the device based on the existing data;

[0122] Compare the line chart with the predicted value to check whether the device is operating as expected. If the actual data is within the range of the predicted value, it means that the device is operating normally; if the actual data deviates too much from the predicted value, it means that the device may have an anomaly;

[0123] Determine the alarm value according to the comparison result. The alarm value is a threshold set according to the range of the device in the normal operating state. If the comparison result shows that the actual value of the device exceeds this range, an alarm needs to be triggered and the alarm value is sent to the terminal.

[0124] As Figure 6 shown, as a preferred embodiment of the present invention, the feature extraction module 200 includes:

[0125] A preprocessing unit 201 for obtaining the aggregated data, preprocessing the aggregated data, where the preprocessing includes filtering and denoising, and obtaining the preprocessed data.

[0126] In this module, the preprocessing unit 201 obtains the aggregated data. After obtaining the aggregated data, preprocessing is a key step to ensure data quality and analysis accuracy. The aggregated data is preprocessed, and the preprocessing includes filtering and denoising. Filtering is to remove unwanted frequency components from the data, and denoising aims to remove random noise or useless signals generated by external interference in the data. After filtering and denoising, the dataset has been preliminarily cleaned and optimized, and then feature extraction and analysis can be carried out;

[0127] For example, low-pass filtering can be used to remove high-frequency noise in vibration data. If there are some sudden outliers in the vibration data, median filtering can be used to remove these noises.

[0128] A feature extraction unit 202 for extracting features from the preprocessed data, where the feature extraction includes time-domain analysis and frequency-domain analysis, and key features in the preprocessed data are obtained through feature extraction.

[0129] In this module, the feature extraction unit 202 extracts features from the preprocessed data. The feature extraction includes time-domain analysis and frequency-domain analysis. Time-domain analysis is an analysis based on the change of data over time, and is usually used to extract statistical features directly related to time changes. It reveals the change trend or volatility of the signal by calculating the statistics of the data, and reflects the average level of the data through the mean. For vibration or temperature data, the mean can tell us whether the working state of the device is abnormal. Frequency-domain analysis is to convert the time-domain signal into frequency components to analyze the intensity of different frequency components in the signal, which is very effective for periodic signals or systems with large fluctuations, especially vibration data in mechanical equipment;

[0130] Combining the two analysis methods helps to comprehensively understand the operating conditions of the device and detect potential faults in advance;

[0131] The key features are vibration, temperature, current / voltage, pressure, and energy consumption. The vibration feature is used to evaluate the mechanical health status of the device. The temperature feature reflects the thermal state of the device. An abnormal temperature change is usually a warning signal of device overload or damage. The current / voltage feature is used to reflect the power consumption and load condition of the device. An abnormal current fluctuation may indicate device overload, malfunction, or low efficiency. The pressure feature affects the working state of the device through pressure changes. Excessive or too low pressure is usually an early signal of device failure. The energy consumption feature is used to monitor the device efficiency. The fluctuation of energy consumption reflects the load and energy efficiency of device operation.

[0132] The key feature unit 203 is used to obtain the key feature threshold, which is the allowable range in the normal working state, and compare the key feature with the key feature threshold.

[0133] In this module, the key feature unit 203 obtains the key feature threshold, which is the allowable range in the normal working state. When the monitored key feature exceeds this range, it can be considered that the device may have a fault or abnormality, and fault diagnosis or maintenance is required.

[0134] The setting of the threshold depends on statistical analysis of the data and combines the working conditions and historical operation of the device to determine the appropriate upper and lower limit ranges: mean ± k × standard deviation. This method is used to set the threshold of the normal distribution, where k is usually a constant set according to experience, such as 2 or 3, indicating the data fluctuation within the normal working range.

[0135] For example, if the mean value of the vibration data of the device during normal operation is 0.05 m / s² and the standard deviation is 0.01 m / s², then the upper and lower limit thresholds are set as follows:

[0136] Upper limit threshold = 0.05 + (2 × 0.01) = 0.07 m / s²;

[0137] Lower limit threshold = 0.05 - (2 × 0.01) = 0.03 m / s²;

[0138] When the monitored vibration value exceeds the range of 0.03 to 0.07 m / s², it may indicate that there is a problem with the device.

[0139] Compare the key feature with the key feature threshold. If the data is within the threshold range, the device is operating normally. If the data exceeds the threshold range, it indicates that the device may have an abnormality or fault.

[0140] As Figure 7 shown, as a preferred embodiment of the present invention, the processing solution module 300 includes:

[0141] The abnormal unit 301 is used to obtain the comparison result. If it is determined that the working state is abnormal, where the abnormal working state refers to key features exceeding the key feature threshold, the reason for the generation of the abnormal state is obtained.

[0142] In this module, the abnormal unit 301 obtains the comparison result. When the value of one or more key features exceeds the preset threshold, it is determined that the working state is abnormal. Once it is determined that the working state is abnormal, the next step is to obtain the reason for the generation of the abnormal state.

[0143] For example, excessive vibration may be related to mechanical failures of the equipment, such as imbalance and looseness. Common reasons for excessive vibration may be bearing damage and motor imbalance. Excessive energy consumption may be related to overload and power system failures. Reasons for excessive energy consumption may be motor overload and reduced equipment efficiency.

[0144] The processing solution unit 302 is used to query the reason for generation according to the abnormal working state to obtain the specific reason, obtain the processing solution for the abnormal state, query the processing solution according to the specific reason, and send the processing solution to the terminal.

[0145] In this module, the processing solution unit 302 queries the reason for generation according to the abnormal working state. Once the equipment is determined to be in an abnormal working state, it is first necessary to determine the specific reason for the abnormality. The specific reason can be queried through data analysis and the failure database. Through data analysis, the relationship between abnormal features and the equipment health status can be analyzed in combination with historical data. Through querying the failure database, possible reasons can be found by matching the current abnormal features and the failure mode library.

[0146] Obtain the processing solution for the abnormal state. The processing solution is formulated based on the equipment maintenance manual or industry standards, defining the operation process for specific failure reasons. Query the processing solution according to the specific reason. For example, for bearing wear, common processing solutions are bearing replacement or lubrication.

[0147] After the generation of the processing solution is completed, send the processing solution to the terminal and send these solutions to the operator or maintenance personnel.

[0148] The rationality evaluation unit 303 is used to obtain the processing result and determine the rationality of the processing solution according to the processing result. The rationality is used to judge the matching degree of the selected processing solution.

[0149] In this module, the rationality evaluation unit 303 obtains the processing result. The processing result refers to the changes in the equipment operation state after the implementation of the processing solution. These changes are fed back in real time through the monitoring system, such as whether key features such as the vibration, temperature, and power consumption of the equipment return to the normal range.

[0150] Determine the rationality of the processing solution based on the processing result. Rationality is used to evaluate the matching degree of the selected processing solution and whether it can effectively, promptly, and at low cost solve the problem;

[0151] Suppose the vibration sensor in a factory shows that a certain motor exceeds the normal vibration range. The system proposes possible fault causes (such as bearing damage) and suggests solving it by bearing replacement or balance adjustment. After the implementation of the treatment, the vibration data returns to the normal range. Then, evaluate the rationality of the solution: it matches the root cause of the problem (bearing problem), effectively solves the problem after implementation, and has low cost and less time consumption. Therefore, this solution is evaluated as reasonable.

[0152] As Figure 8 shown, as a preferred embodiment of the present invention, the warning module 400 includes:

[0153] A visualization unit 401 for performing visualization processing based on the preprocessed data. The visualization processing uses a line chart to predict the subsequent operating state according to the line chart and obtain a predicted value.

[0154] In this module, the visualization unit 401 performs visualization processing based on the preprocessed data. The visualization processing uses a line chart to display the preprocessed data, which can help us more clearly see the operating state of the device at different time points;

[0155] Predict the subsequent operating state according to the line chart. The subsequent operating state can be predicted according to the change trend of historical data to obtain a predicted value. The possible subsequent changes can be judged by observing the change trend of historical data. For example, if the vibration intensity continues to increase and there is no obvious sign of decline, we can predict that the vibration may continue to increase at the next time point and reach the critical value of equipment failure.

[0156] A warning determination unit 402 for obtaining an alarm value. The alarm value is the magnitude of the fault occurrence trend based on the health state reference, and the alarm value is set as a critical alarm value and a warning alarm value.

[0157] In this module, the warning determination unit 402 obtains an alarm value. The alarm value is the magnitude of the fault occurrence trend based on the health state reference. The health state reference refers to the performance index standard of the device under normal working conditions, and the alarm value is set as a critical alarm value and a warning alarm value;

[0158] The warning alarm value is a threshold set near the health state reference, which is used to indicate the initial signal of a certain deviation or fault of the device. At this time, the device is still in an operable state, but monitoring and inspection are required;

[0159] Critical alarm value: Set as the maximum deviation that the device can withstand. When the threshold is exceeded, the device may have serious failures or damages. If the operating parameters of the device reach or exceed this value, it usually indicates that the device needs to be immediately shut down for maintenance or repair.

[0160] An early warning unit 403 is used to compare the line chart with the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0161] In this module, the early warning unit 403 compares the line chart with the predicted value. By comparing the actual line chart with the predicted value, it determines whether the future operating trend of the device conforms to the normal expectation. If the predicted value shows that the future state of the device will enter or has exceeded the set alarm threshold, the system will trigger the alarm mechanism;

[0162] Once the system determines that the device will enter a faulty state, the system will generate an alarm according to the preset alarm value and send the alarm information to the terminal device.

[0163] In one embodiment, a computer device is provided. The computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented:

[0164] Collect the detection data. The monitoring data includes vibration data, temperature data, current and voltage data, pressure data, and energy consumption data, and summarize the collected data;

[0165] Obtain the summarized data, preprocess the summarized data, extract features from the preprocessed data, obtain the key features in the preprocessed data through feature extraction, and compare the key features with the key feature threshold;

[0166] Obtain the comparison result. If it is determined to be an abnormal working state, obtain the cause of the abnormal state generation, query the cause according to the abnormal working state, query the processing solution according to the specific cause, and determine the rationality of the processing solution according to the processing result;

[0167] Perform visualization processing according to the preprocessed data, predict the subsequent operating state according to the line chart, compare the line chart with the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0168] In one embodiment, a computer-readable storage medium is provided. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the processor is caused to execute the following steps:

[0169] Collect the detection data. The monitoring data includes vibration data, temperature data, current and voltage data, pressure data, and energy consumption data, and summarize the collected data;

[0170] Obtain the aggregated data, preprocess the aggregated data, extract features from the preprocessed data, obtain the key features in the preprocessed data through feature extraction, and compare the key features with the key feature thresholds;

[0171] Obtain the comparison result. If it is determined to be an abnormal working state, obtain the cause of the abnormal state generation, query the cause according to the abnormal working state, query the processing solution according to the specific cause, and determine the rationality of the processing solution according to the processing result;

[0172] Perform visualization processing based on the preprocessed data, predict the subsequent operating state according to the line chart, compare the line chart with the predicted value, determine the alarm value according to the comparison result, and send the alarm value to the terminal.

[0173] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0174] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in this application can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0175] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0176] The above embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

[0177] The above is only the preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A recognition method based on data collection, characterized in that: The method comprises: Collecting detection data, including vibration data, temperature data, current and voltage data, pressure data, and energy consumption data, and summarizing the collected data; Obtain the aggregated data, preprocess the aggregated data, perform feature extraction on the preprocessed data, obtain key features in the preprocessed data through feature extraction, and compare the key features with the key feature threshold; Obtain the comparison result. If it is determined to be an abnormal working state, obtain the cause of the abnormal state, query the cause based on the abnormal working state, query the treatment plan based on the specific cause, and determine the rationality of the treatment plan based on the treatment result; Perform visualization based on the preprocessed data, predict the subsequent operating status based on the line graph, compare the line graph and the predicted value, determine the alarm value based on the comparison result, and send the alarm value to the terminal.

2. The identification method based on data collection according to claim 1, characterized in that: The steps of obtaining the aggregated data, preprocessing the aggregated data, extracting features from the preprocessed data, obtaining key features from the preprocessed data through feature extraction, and comparing the key features with key feature thresholds specifically include: Acquire the aggregated data, perform preprocessing on the aggregated data, wherein the preprocessing includes filtering and denoising, and acquire the preprocessed data; Performing feature extraction on the preprocessed data, wherein the feature extraction includes time domain analysis and frequency domain analysis, and obtaining key features in the preprocessed data through feature extraction; A key feature threshold is obtained, where the key feature threshold is an allowable range for a normal working state, and the key feature is compared with the key feature threshold.

3. The identification method based on data collection according to claim 1, characterized in that: The steps of obtaining the comparison result, if it is determined to be an abnormal working state, obtaining the cause of the abnormal state, querying the cause according to the abnormal working state, querying the processing scheme according to the specific cause, and determining the rationality of the processing scheme according to the processing result specifically include: Obtain the comparison result, if it is determined to be an abnormal working state, the abnormal working state is a key feature that exceeds the key feature threshold, and obtain the cause of the abnormal state; Query the generation reason according to the abnormal working state, obtain the specific reason, obtain the processing plan of the abnormal state, query the processing plan according to the specific reason, and send the processing plan to the terminal; Obtain processing results, and determine the rationality of the processing solution based on the processing results. The rationality is used to judge the matching degree of the processing solution selection.

4. The identification method based on data collection according to claim 1, characterized in that: The steps of performing visualization processing according to the preprocessed data, predicting the subsequent operation status according to the line graph, comparing the line graph with the predicted value, determining the alarm value according to the comparison result, and sending the alarm value to the terminal specifically include: Performing visualization processing on the preprocessed data, wherein the visualization processing uses a line graph, and predicting subsequent operation status based on the line graph to obtain a predicted value; Obtain an alarm value, wherein the alarm value is a magnitude of a fault occurrence trend based on a healthy state, and the alarm value is set to a critical alarm value and a warning alarm value; Compare the line graph and the predicted value, determine the alarm value based on the comparison result, and send the alarm value to the terminal.

5. The identification method based on data collection according to claim 2, characterized in that: The key characteristics are vibration, temperature, current and voltage, pressure and energy consumption.

6. A recognition system based on data collection, characterized in that: The system comprises: The acquisition module collects the detection data, including vibration data, temperature data, current and voltage data, pressure data and energy consumption data, and summarizes the collected data; The feature extraction module obtains the aggregated data, preprocesses the aggregated data, extracts features from the preprocessed data, obtains key features from the preprocessed data through feature extraction, and compares the key features with the key feature thresholds; The processing scheme module obtains the comparison result. If it is determined to be an abnormal working state, the cause of the abnormal state is obtained, the cause is queried based on the abnormal working state, the processing scheme is queried based on the specific cause, and the rationality of the processing scheme is determined based on the processing result; The early warning module performs visualization based on the preprocessed data, predicts the subsequent operating status based on the line graph, compares the line graph with the predicted value, determines the alarm value based on the comparison result, and sends the alarm value to the terminal.

7. The data collection-based identification system according to claim 6, characterized in that: The feature extraction module comprises: A preprocessing unit, which obtains the aggregated data and preprocesses the aggregated data, wherein the preprocessing includes filtering and denoising, and obtains the preprocessed data; A feature extraction unit performs feature extraction on the preprocessed data, wherein the feature extraction includes time domain analysis and frequency domain analysis, and obtains key features in the preprocessed data through feature extraction; The key feature unit obtains a key feature threshold value, where the key feature threshold value is an allowable range of a normal working state, and compares the key feature with the key feature threshold value.

8. The data collection-based identification system according to claim 7, characterized in that: The processing scheme module includes: The abnormal unit obtains the comparison result, and if it is determined to be an abnormal working state, the abnormal working state is a key feature that exceeds the key feature threshold, and the cause of the abnormal state is obtained; A processing scheme unit queries and generates a cause according to the abnormal working state, obtains the specific cause, obtains a processing scheme for the abnormal state, queries the processing scheme according to the specific cause, and sends the processing scheme to the terminal; The rationality evaluation unit obtains the processing result and determines the rationality of the processing solution according to the processing result. The rationality is used to judge the matching degree of the processing solution selection.

9. The data collection-based identification system according to claim 8, characterized in that: The early warning module comprises: A visualization unit performs visualization processing according to the preprocessed data, wherein the visualization processing uses a line graph, and predicts the subsequent operation state according to the line graph to obtain a predicted value; An early warning determination unit obtains an alarm value, wherein the alarm value is a magnitude of a fault occurrence trend based on a health state, and the alarm value is set to a critical alarm value and a warning alarm value; The early warning unit compares the line graph and the predicted value, determines the alarm value according to the comparison result, and sends the alarm value to the terminal.

10. The data collection-based identification system according to claim 9, characterized in that: The key characteristics are vibration, temperature, current and voltage, pressure and energy consumption.

Images