High frequency digital acquisition method and system based on average screening point algorithm

Abstract

The present invention discloses a high-frequency digital acquisition method based on an average screening point algorithm, which relates to the technical field of data acquisition. The sampling rate is set for a sensor network, the operating environment data of a target device is recorded and an operating environment coefficient is generated. If the operating environment coefficient is higher than an operating threshold, data is collected and preprocessed at a preset sampling rate, and a data quality coefficient is generated from the quality index of the preprocessed data. If the data quality coefficient is lower than expected, the preprocessed data is optimized, the optimized data is backed up regularly, and the data backup frequency is constrained. The operating state coefficient is constructed from the change trend of the target device-related data, and the abnormal degree of the target device operation is evaluated. If the abnormal degree exceeds expectations, an optimized maintenance plan is given to the target device. By constraining the frequency of data backup, the data backup process is reasonable, and data occupation caused by excessive data backup during the data acquisition process is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of data acquisition, and in particular to a high-frequency digital acquisition method and system based on an average screening point algorithm. Background Art

[0002] The operation data collection of mechanical equipment refers to the real-time and accurate collection of various parameters and status information of mechanical equipment during operation, such as temperature, pressure, vibration, speed, current, voltage, etc., through sensors, monitoring equipment and other technical means, so as to carry out subsequent data analysis, fault diagnosis, performance evaluation and maintenance optimization, etc. This process is of great significance to ensure the safe operation of mechanical equipment, improve production efficiency and reduce maintenance costs.

[0003] In the Chinese invention patent with the authorization announcement number CN111031050B, the embodiment of the present invention provides a monitoring method for an electricity information collection system, including: collecting data from the server and equipment of the electricity information collection system based on network probe technology, the collected data including at least performance data and security data; comparing the collected data with the collection business process and server software and hardware information to analyze the causes of performance bottlenecks; and comparing the collected data with security vulnerability data to analyze security risk issues. Through the above technical solution, by integrating at least probe technology, network traffic mirroring collection technology, etc., the data of the collection system software and hardware performance status and the collection business development status are collected, monitored, and comprehensively analyzed in real time, which solves the problems of incomplete online performance analysis of the collection system, difficult to identify performance bottlenecks, and inability to judge the health of business operations, improves the trouble-free operation time of the collection system, and improves the operation stability and business carrying capacity of the collection system.

[0004] Combined with the above application and the contents of the prior art:

[0005] When the device is in operation, in order to monitor the operating status of the device, a sensor network is usually installed for the device. The sensor network is used to continuously collect various operating data of the device, and then determine whether there is a fault in the operation of the device based on whether there is an abnormality in the acquired operating data.

[0006] In the existing data collection methods, when using a sensor network matched with a device for data collection, the sensor network usually remains in operation all the time. In this case, the data collection capability of the sensor, which remains in high-intensity operation all the time, will be affected, resulting in certain deficiencies in the quality of the device-related data actually collected. At the same time, due to the lack of a data backup mechanism, the risk of loss of the collected device-related data is also high.

[0007] To this end, the present invention provides a high-frequency digital acquisition method and system based on an average screening point algorithm. Summary of the invention

[0008] 1. Technical issues to be solved

[0009] In view of the deficiencies of the prior art, the present invention provides a high-frequency digital acquisition method and system based on an average screening point algorithm, which records the operating environment data of the target device and generates an operating environment coefficient. If the operating environment coefficient is higher than the operating threshold, data acquisition and preprocessing are performed, and a data quality coefficient is generated from the quality index of the preprocessed data. If the data quality coefficient is lower than expected, the preprocessed data is optimized, the optimized data is backed up regularly, and the data backup frequency is constrained. The operating state coefficient is constructed from the change trend of the target device-related data, and the abnormal degree of the target device operation is evaluated. If the abnormal degree exceeds expectations, an optimized maintenance plan is given to the target device. By constraining the frequency of data backup, the data backup process is reasonable, thereby solving the technical problems raised in the background technology.

[0010] (II) Technical solution

[0011] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0012] A high-frequency digital acquisition method based on an average screening point algorithm includes: analyzing and obtaining similarities between different usage scenarios of a target device, generating the complexity Mst of the usage scenario, configuring a sensor network for the target device, and constraining the response time of the selected sensor according to the complexity Mst;

[0013] After setting the sampling rate for the sensor network, the operating environment data of the target device is recorded and the operating environment coefficient Hot is generated. If the operating environment coefficient Hot is higher than the operating threshold, a data collection instruction is issued to the outside.

[0014] Data is collected and preprocessed at a preset sampling rate, and a data quality coefficient Vts is generated from the quality index of the preprocessed data. If the data quality coefficient Vts is lower than expected, a data optimization instruction is issued to the outside. The data quality coefficient Vts is generated from a set of data quality indicators, wherein the accuracy Qt and the completeness Zt are linearly normalized, and then in the following manner:

[0015]

[0016] Weight coefficient, 0≤α≤1, 0≤β≤1;

[0017] Optimize the preprocessed data based on the optimization solution matched by the data optimization solution library, regularly back up the optimized data, and restrict the frequency of data backup;

[0018] The operating status coefficient Cmj is constructed based on the changing trend of the target equipment-related data. If the operating status coefficient Cmj is lower than expected, the abnormality of the target equipment operation is evaluated. If the abnormality exceeds expectations, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target equipment.

[0019] Furthermore, after determining the usage scenario of the device, a scenario description is added to it, similarity analysis is performed on different scenario descriptions and similarity data is obtained, and several similarity data obtained are aggregated to generate a scenario data set, and the complexity Mst of the usage scenario is generated from the obtained scenario data set.

[0020] Furthermore, the device to be collected data is taken as the target device, and the response time of the sensor is constrained as follows:

[0021]

[0022] Where n is the number of sensors; R i is the response time of the sensor’s i-th response, R a is the mean response time, R b is the acceptable value of response time; if the response time of the selected sensor does not meet the constraint conditions, it is replaced.

[0023] Furthermore, the operating environment data of the target device is recorded, and the acquired operating environment data is aggregated to generate an operating environment data set; the operating environment coefficient Hot of the target device is generated from the operating environment data set. If the operating environment coefficient Hot is higher than the operating threshold, this condition is used as a trigger condition for data collection, and a data collection instruction is issued to the outside.

[0024] Furthermore, after receiving the data collection instruction, a sampling period is set and real-time data collection is performed at a preset sampling rate, the collected data is obtained for preprocessing, and a preprocessed data set is obtained; data quality analysis is performed on each item of preprocessed data, corresponding data quality indicators are collected, and a data quality indicator set is generated after aggregation.

[0025] Furthermore, after receiving the data optimization instruction, feature extraction is performed on the preprocessed data, and corresponding optimization features are obtained; a number of acquired data optimization solutions are summarized to construct a data optimization solution library; and the data optimization solution library matches the corresponding optimization solution for the preprocessed data.

[0026] Further, the optimization scheme is executed to optimize the preprocessed data, the optimized data is backed up regularly, and the optimized data is backed up on the backup node that meets the constraint conditions to constrain the data backup frequency. The constraint conditions are as follows:

[0027]

[0028] Where n is the number of data backups within the data backup cycle, T ij is the time interval from the i-th data backup to the j-th data backup, T a is the average value of the time interval.

[0029] Further, a visualization tool is used to analyze the optimized data and generate a trend chart, and the similarity between the trend chart and the standard chart is calculated to obtain chart similarity data; the operating state coefficient Cmj of the target device is constructed from the obtained several chart similarity data, wherein the chart similarity Td is linearly normalized and the corresponding data value is mapped to the interval [0, 1] in the following manner:

[0030]

[0031] Among them, n is the number of graph similarities, Td i is the similarity of the i-th graph, Td a is the average value of graph similarity; Td b is the acceptable value of graph similarity, weight coefficient: 0≤F1≤1, 0≤F2≤1, and F1+F2=1;

[0032] If the operating status coefficient Cmj is lower than the status threshold, the current operating status of the target device is recorded and treated as abnormal operation, and an alarm instruction is issued to the outside.

[0033] Furthermore, the abnormal data recognition model is obtained by the neural convolution network, and the optimized data is used as input. The trained abnormal recognition model recognizes abnormalities and outputs abnormal data and corresponding abnormal nodes within the sampling period. The abnormal density Com is generated by the acquired abnormal nodes, and the time node of receiving the alarm instruction is used as the alarm node to generate the alarm density Bom.

[0034] If one of the abnormal density Com and the alarm density Bom exceeds expectations, the abnormal data within the sampling period is obtained and the corresponding abnormal features are obtained through feature extraction; and equipment maintenance optimization is used as the target word to pre-build the equipment operation optimization knowledge graph.

[0035] A high-frequency digital acquisition system based on an average screening point algorithm includes a scene analysis unit, which generates a complexity Mst of the scene after analyzing and acquiring similarities between different use scenes of a target device, configures a sensor network for the target device, and constrains the response time of the selected sensor according to the complexity Mst;

[0036] The environment analysis unit, after setting the sampling rate for the sensor network, records the operating environment data of the target device and generates the operating environment coefficient Hot. If the operating environment coefficient Hot is higher than the operating threshold, it sends a data collection instruction to the outside;

[0037] The quality analysis unit collects data and performs preprocessing at a preset sampling rate, and generates a data quality coefficient Vts based on the quality index of the preprocessed data. If the data quality coefficient Vts is lower than expected, a data optimization instruction is issued to the outside.

[0038] The data backup unit optimizes the preprocessed data according to the optimization scheme matched by the data optimization scheme library, regularly backs up the optimized data, and constrains the frequency of data backup;

[0039] The optimization unit constructs the operation status coefficient Cmj based on the change trend of the target equipment-related data. If the operation status coefficient Cmj is lower than expected, the abnormality of the target equipment operation is evaluated. If the abnormality exceeds expectations, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target equipment.

[0040] (III) Beneficial effects

[0041] The present invention provides a high-frequency digital acquisition method based on an average screening point algorithm, which has the following beneficial effects:

[0042] 1. By using scenario data to calculate the complexity Mst of the target usage scenario and evaluating the complexity of the usage scenario with the obtained complexity Mst, the coverage of data collection types can be determined, the selection of data types to be collected can be achieved, and the sensors can be screened by constraining the response time of the sensors to ensure the availability of the sensor network.

[0043] 2. After collecting the use environment data of the target device, the environmental conditions can be adjusted in time when the use environment is abnormal; the abnormality of the current use environment of the target device is evaluated according to the operating environment coefficient Hot, so that the environmental conditions are adapted to the target device to ensure the normal operation of the target device; by selecting the trigger conditions for data collection, data collection is performed when the trigger conditions are met, and the data collection frequency and working time of the corresponding sensors are reduced, which can ensure the operation status of the sensor network and the quality of subsequent data collection.

[0044] 3. Evaluate the quality of data collected in the current stage based on the acquired data quality coefficient Vts, and judge the availability of the collected data. By performing targeted optimization on data groups with lower quality, the availability of equipment-related data can be guaranteed.

[0045] 3. Match the corresponding optimization scheme for the preprocessed data through the data optimization scheme library to achieve targeted processing of low-quality data and improve the data quality of equipment-related data; back up the data after completing data preprocessing and optimization to avoid loss of the collected data and ensure data security. By restricting the frequency of data backup, the data backup process is reasonable to avoid data occupation caused by excessive data backup during the data collection process.

[0046] 4. Perform trend analysis through visual analysis tools to confirm the current operating status and operating trend of the target device. Determine the similarities and differences between the current trend image and the target image through similarity analysis. This can evaluate the current operating status of the target device and confirm whether there are any abnormalities in the current operating status. By issuing alarm instructions to the outside, possible abnormalities can be handled in a timely manner to ensure the safe operation of the target device.

[0047] 5. By analyzing the received alarm instructions and abnormal data, the abnormal density Com and alarm density Bom are obtained, so as to evaluate and judge the abnormal degree of the target equipment operation, verify whether the target equipment can continue to maintain operation, and perform targeted optimization and maintenance when it may be difficult to maintain the operation status.

[0048] 6. Based on the collected relevant data, abnormal data is identified. Taking the abnormal data as the starting point, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target equipment, which can achieve targeted maintenance of the target equipment and enable the target equipment to maintain a normal operating state. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a schematic diagram of the high-frequency digital acquisition method of the present invention;

[0050] Figure 2 It is a schematic diagram of the structure of the high-frequency digital acquisition system of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in 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.

[0052] See also Figure 1 The present invention provides a high-frequency digital acquisition method based on an average screening point algorithm, comprising:

[0053] Step 1: After analyzing and obtaining the similarities between different usage scenarios of the target device, generate the complexity Mst of the usage scenario, configure the sensor network for the target device, and constrain the response time of the selected sensor according to the complexity Mst;

[0054] The step 1 includes the following contents:

[0055] Step 101: After determining the usage scenario of the device, add a scenario description to it, perform similarity analysis on different scenario descriptions and obtain similarity data, and aggregate the obtained similarity data to generate a scenario data set;

[0056] The complexity of the usage scenario is generated from the acquired scenario data set, wherein the similarity Xs between the scenario descriptions is linearly normalized and the corresponding data values ​​are mapped to the interval [0, 1] in the following manner:

[0057]

[0058] Where n is the number of usage scenarios, Xs ij is the similarity between the i-th scene description and the j-th scene description, Xs a is the average similarity; weight coefficient: 0≤F1≤1, 0≤F2≤1, and F1+F2=1;

[0059] Step 102: The device to be collected data is taken as the target device, and a sensor network is configured for the target device, such as a voltage sensor, a current sensor, and a temperature sensor, and the response time of the sensor is constrained in the following manner:

[0060]

[0061] Where n is the number of sensors; R i is the response time of the sensor’s i-th response, R a is the mean response time, R b is the acceptable value of response time;

[0062] If the response time of the selected sensor does not meet the constraint conditions, replace it;

[0063] When using, combine the contents in steps 101 and 102:

[0064] When it is necessary to collect the operating status data of the target device, the usage scenario data of the target device is collected first. The complexity Mst of the target usage scenario is calculated by using the scenario data. The complexity of the usage scenario is evaluated by the obtained complexity Mst. The coverage of the data collection types can be determined, and the selection of the types of data to be collected can be achieved. At the same time, by constraining the response time of the sensor, the sensor can be screened to ensure the availability of the sensor network.

[0065] In the existing data collection methods, when using a sensor network matched with a device for data collection, the sensor network usually remains in operation all the time. In this case, the data collection capability of the sensor, which remains in high-intensity operation all the time, will be affected, resulting in certain deficiencies in the quality of the device-related data actually collected. At the same time, due to the lack of a data backup mechanism, the risk of loss of the collected device-related data is also high.

[0066] Step 2: After setting the sampling rate for the sensor network, record the operating environment data of the target device and generate the operating environment coefficient Hot. If the operating environment coefficient Hot is higher than the operating threshold, send a data collection instruction to the outside.

[0067] The step 2 includes the following contents:

[0068] Step 201, select a high-resolution analog-to-digital converter and configure a data transmission module for the target device, support multi-channel acquisition and synchronous sampling, facilitate the detection of multiple parameters, and set a sampling rate for the sensor network according to the signal frequency of the signal to be collected, wherein the sampling rate is more than twice the signal frequency;

[0069] When the target device is in operation, the operating environment data of the target device, such as operating temperature, operating humidity, etc., is recorded, and the acquired operating environment data is aggregated to generate an operating environment data set;

[0070] When in use, after determining the usage scenario of the target device, by collecting the usage environment data of the target device, when an abnormality occurs in the usage environment, the environmental conditions can be adjusted in time; as an optional method, after continuously acquiring several groups of data, the average screening point algorithm can be used to screen within several groups, that is, by alternately screening points with two different step sizes, the data distribution can be relatively more even, reducing the interference of abnormal data;

[0071] Step 202: Generate the operating environment coefficient Hot of the target device from the operating environment data set, wherein the ambient temperature Ht and the ambient humidity Hs are linearly normalized and the corresponding data values ​​are mapped to the interval [0, 1] in the following manner:

[0072]

[0073] Weight coefficient: 0≤k1≤1, 0≤k2≤1. The weight coefficient can be obtained by referring to the hierarchical analysis method.

[0074] Ht i is the ambient temperature of the ith monitoring node, Ht b is the acceptable value of ambient temperature, Ht a is the mean value of ambient temperature; Hs i is the ambient humidity of the i-th monitoring node, Hs b is the acceptable value of ambient humidity, Hs a is the mean value of ambient humidity;

[0075] If the operating environment coefficient Hot is higher than the operating threshold, the target device may produce an abnormality under the current operating conditions. At this time, this condition is used as a trigger condition for data collection and a data collection instruction is issued to the outside;

[0076] When using, combine the contents in steps 201 to 203:

[0077] After periodically collecting environmental condition data, the operating environment coefficient Hot is established from the collected environmental condition data. Based on the operating environment coefficient Hot, the abnormality of the current use environment of the target device is evaluated to adapt the environmental conditions to the target device, thereby ensuring that the target device can operate normally. At the same time, by selecting the trigger conditions for data collection, data collection is performed when the trigger conditions are met, thereby reducing the data collection frequency and working time of the corresponding sensors, thereby ensuring the operating status of the sensor network and the quality of subsequent data collection.

[0078] Step 3: Collect data and perform preprocessing at a preset sampling rate, and generate a data quality coefficient Vts based on the quality index of the preprocessed data. If the data quality coefficient Vts is lower than expected, send a data optimization instruction to the outside;

[0079] The step three includes the following contents:

[0080] Step 301: After receiving the data acquisition instruction, in order to ensure that the sampling time is long enough to capture the complete characteristics of the signal, set the sampling period and perform real-time data acquisition at a preset sampling rate, obtain the collected data for preprocessing, such as removing high-frequency noise and smoothing the signal, extracting the effective characteristics of the signal, etc., and obtain a preprocessed data set; perform data quality analysis on each item of preprocessed data, collect corresponding data quality indicators, such as accuracy, completeness, consistency and timeliness, etc., and generate a data quality indicator set after aggregation;

[0081] When using, after completing data collection, preprocessing through data collection can improve the quality of data and ensure feedback effect when applying the data;

[0082] Step 302: Generate a data quality coefficient Vts from the data quality indicator set, wherein the accuracy Qt and the completeness Zt are linearly normalized, and then in the following manner:

[0083]

[0084] Weight coefficient, 0≤α≤1, 0≤β≤1; the weight coefficient can be obtained by referring to the hierarchical analysis method;

[0085] The quality threshold is set in advance based on historical data and the quality management expectations for the collected data. If the data quality coefficient Vts is lower than the quality threshold, it means that the quality of the currently collected data is low and targeted optimization is required. At this time, a data optimization instruction is issued to the outside.

[0086] When used, combine the contents in steps 301 and 302:

[0087] After data quality analysis, a data quality coefficient Vts is generated. Based on the obtained data quality coefficient Vts, the quality of the data collected in the current stage is evaluated, and the availability of the collected data is judged. By performing targeted optimization on the data groups with lower quality, the availability of equipment-related data can be guaranteed.

[0088] Step 4: Optimize the preprocessed data according to the optimization solution matched by the data optimization solution library, regularly back up the optimized data, and restrict the frequency of data backup;

[0089] The step 4 includes the following contents:

[0090] Step 401: After receiving the data optimization instruction, extract the distribution state and quality state of the pre-processed data, and obtain the corresponding optimization features; through online and offline retrieval and collection, summarize the obtained data optimization solutions and build a data optimization solution library; according to the correspondence between the optimization features and the optimization solutions, the data optimization solution library matches the corresponding optimization solutions for the pre-processed data;

[0091] When in use, when the quality of the collected data related to the target device fails to meet expectations, the data optimization solution library can be used to match the corresponding optimization solution for the pre-processed data, so as to achieve targeted processing of low-quality data, thereby improving the data quality of the device-related data;

[0092] Step 402: Execute the optimization scheme to optimize the pre-processed data, and regularly back up the optimized data to ensure a redundant storage scheme and constrain the data backup frequency. The constraints are as follows:

[0093]

[0094] Where n is the number of data backups within the data backup cycle, T ij is the time interval from the i-th data backup to the j-th data backup, T a is the average value of the time interval; the optimized data is backed up on the backup node that meets the constraints;

[0095] When using, combine the contents in steps 401 and 402:

[0096] Performing a backup after completing data preprocessing and optimization can avoid the loss of the collected data and ensure data security. At the same time, by restricting the frequency of data backup, the data backup process can be reasonable to avoid data occupation caused by excessive data backup during the data collection process.

[0097] Step 5: Construct the operation status coefficient Cmj based on the change trend of the target equipment related data. If the operation status coefficient Cmj is lower than expected, evaluate the abnormality of the target equipment operation. If the abnormality exceeds expectations, provide an optimized maintenance plan for the target equipment based on the equipment operation optimization knowledge graph.

[0098] The five include the following:

[0099] Step 501: Use a visualization tool to analyze the optimized data and generate a trend chart to show the signal change trend. Preset a reference trend chart based on the use management of the target and use it as a standard chart; calculate the similarity between the trend chart and the standard chart to obtain chart similarity data;

[0100] When in use, after completing data collection, trend analysis is performed through visual analysis tools to confirm the current operating status and operating trend of the target device. When the operating status of the target device develops in a bad direction, it can be controlled and processed in a timely manner to ensure the safety of the target device's operation.

[0101] Step 502: construct the operating state coefficient Cmj of the target device from the acquired similarity data of several charts, and then monitor the operating state of the target device through high-frequency digital acquisition, wherein the chart similarity Td is linearly normalized and the corresponding data value is mapped to the interval [0, 1] in the following manner:

[0102]

[0103] Among them, n is the number of graph similarities, Td i is the similarity of the i-th graph, Td a is the average value of graph similarity; Td b is the acceptable value of graph similarity, weight coefficient: 0≤F1≤1, 0≤F2≤1, and F1+F2=1; the value of weight coefficient is consistent with the previous value;

[0104] According to historical data and the expected operation management of the target equipment, the status threshold is pre-set; if the operation status coefficient Cmj is lower than the status threshold, the current operation status of the target equipment is recorded and treated as abnormal operation. At this time, an alarm instruction is issued to the outside;

[0105] When in use, similarity analysis is used to determine the similarities and differences between the current trend image and the target image, so as to evaluate the current operating status of the target device, confirm whether there is any abnormality in the current operating status, and issue an alarm command to the outside to promptly handle the possible abnormalities, thus ensuring the safe operation of the target device.

[0106] Step 503: Obtain an abnormal data recognition model through a neural convolutional network, use the optimized data as input, identify abnormalities through the trained abnormal recognition model, output abnormal data and corresponding abnormal nodes within a sampling period, and generate abnormal density Com from the acquired abnormal nodes, where:

[0107]

[0108] Weight coefficient: 0≤K1≤1, 0≤K2≤1, and K1+K2=1. The weight coefficient can be obtained by referring to the hierarchical analysis method; m is the number of abnormal nodes, T ij is the time interval from the ith abnormal node to the jth abnormal node, T a is the average value of the time interval, T b is the acceptable value for the time interval;

[0109] At the same time, the time node when the alarm instruction is received is used as the alarm node, and the alarm density Bom is generated in the same way;

[0110] If one of the abnormal density Com and the alarm density Bom exceeds the expectation, it can be said that there is a certain degree of abnormality in the current operation of the target device. At this time, an operation optimization instruction is issued to the outside;

[0111] When in use, several device-related data are continuously collected, and when it is determined that the target device is in an abnormal operating state, the abnormal density Com and the alarm density Bom are analyzed and obtained through the received alarm instructions and the received abnormal data. The abnormal degree of the target device's operation can be evaluated and judged, and it can be verified whether the target device can continue to maintain operation, and targeted optimization and maintenance can be performed when it may be difficult to continue to maintain the operating state.

[0112] Step 504: after receiving the optimization instruction, the abnormal data within the sampling period is obtained and the corresponding abnormal features are obtained through feature extraction; the equipment maintenance optimization is used as the target word, and after deep retrieval and entity relationship building, the equipment operation optimization knowledge graph is pre-built; according to the correspondence between the abnormal features and the optimized maintenance plan, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target equipment, such as optimizing parameter settings, adjusting working status, etc.;

[0113] When using, combine the contents in steps 501 to 504:

[0114] When the target device may have an operational failure, the abnormal data is identified based on the collected relevant data. Taking the abnormal data as the starting point, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target device, which can achieve targeted maintenance of the target device and enable the target device to maintain a normal operating state.

[0115] The Analytic Hierarchy Process (AHP) is a decision-making method that decomposes decision-related elements into levels such as goals, criteria, and plans, and conducts qualitative and quantitative analysis on this basis. It is particularly suitable for dealing with target systems with hierarchical and staggered evaluation indicators, and when the target value is difficult to describe quantitatively, the AHP is an effective decision-making tool.

[0116] The core of the hierarchical analysis method is to decompose the decision problem into multiple levels to form a hierarchical structure, which usually includes the target level, the criterion level, the sub-criterion level and the solution level. By solving the eigenvector of the judgment matrix, the priority weight of each element of each level to an element of the previous level is obtained, and the final weight of each alternative solution to the total goal is recursively merged by the weighted sum method, so as to find the optimal solution.

[0117] The construction method of the knowledge graph of mechanical equipment operation optimization and maintenance can refer to the following:

[0118] The main goal of building a knowledge graph for mechanical equipment operation optimization and maintenance is to graphically represent the relevant knowledge of mechanical equipment operation, optimization, maintenance, etc., so as to facilitate quick query, understanding and application. Through the knowledge graph, the relationship between equipment, fault types and solutions, optimization strategies, etc. can be clearly displayed, thereby improving the efficiency and accuracy of equipment maintenance.

[0119] Data source selection: Collect information from a variety of data sources such as equipment manufacturer manuals, technical documents, maintenance records, operating data, expert experience, etc. Data preprocessing: Clean, transform and integrate the collected data to remove noise, redundant and inconsistent data to ensure data quality and accuracy.

[0120] Entity recognition: Use natural language processing (NLP) technology to identify named entities in text, such as equipment names, component names, fault types, etc. Entity linking: Unify the names representing the same entity in different data sources and establish link relationships between entities.

[0121] Attribute extraction: Extract attribute information related to entities from text, such as equipment model, specifications, manufacturer, etc. Relationship extraction: Identify semantic relationships between entities, such as connection relationships between devices, causal relationships between faults and causes, etc.

[0122] Entity alignment: Solve the problem of inconsistent entity names in different data sources to ensure the consistency of entities. Attribute value filling: When the same attribute has different values ​​in different data sources, make decisions based on the reliability of the data source and the actual situation and fill in the most accurate attribute value.

[0123] Data layer construction: The extracted entities, attributes and relationships are stored in the form of a graph database or graph structure to form the data layer of the knowledge graph. Model layer construction: The knowledge in the data layer is summarized and abstracted to form a conceptual hierarchy tree or ontology structure as the model layer of the knowledge graph.

[0124] Quality assessment: Conduct quality assessment on the constructed knowledge graph, including indicators such as entity coverage, attribute coverage, relationship coverage, and accuracy of entities, attributes, and relationships. Optimization and adjustment: Optimize and adjust the knowledge graph based on the evaluation results to improve the quality and accuracy of the knowledge graph.

[0125] See also Figure 2 The present invention provides a high-frequency digital acquisition system based on an average screening point algorithm, comprising:

[0126] The scenario analysis unit analyzes and obtains the similarities between different usage scenarios of the target device, generates the complexity Mst of the usage scenario, configures the sensor network for the target device, and constrains the response time of the selected sensor according to the complexity Mst;

[0127] The environment analysis unit, after setting the sampling rate for the sensor network, records the operating environment data of the target device and generates the operating environment coefficient Hot. If the operating environment coefficient Hot is higher than the operating threshold, it sends a data collection instruction to the outside;

[0128] The quality analysis unit collects data and performs preprocessing at a preset sampling rate, and generates a data quality coefficient Vts based on the quality index of the preprocessed data. If the data quality coefficient Vts is lower than expected, a data optimization instruction is issued to the outside.

[0129] The data backup unit optimizes the preprocessed data according to the optimization scheme matched by the data optimization scheme library, regularly backs up the optimized data, and constrains the frequency of data backup;

[0130] The optimization unit constructs the operation status coefficient Cmj based on the change trend of the target equipment-related data. If the operation status coefficient Cmj is lower than expected, the abnormality of the target equipment operation is evaluated. If the abnormality exceeds expectations, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target equipment.

[0131] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0132] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0133] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0134] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0135] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0136] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0137] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0138] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A high-frequency digital acquisition method based on an average screening point algorithm, characterized in that: include, After analyzing and obtaining the similarities between different usage scenarios of the target device, the complexity Mst of the usage scenario is generated, a sensor network is configured for the target device, and the response time of the selected sensor is constrained according to the complexity Mst; After setting the sampling rate for the sensor network, the operating environment data of the target device is recorded and the operating environment coefficient Hot is generated. If the operating environment coefficient Hot is higher than the operating threshold, a data collection instruction is issued to the outside. Data is collected and preprocessed at a preset sampling rate, and a data quality coefficient Vts is generated from the quality index of the preprocessed data. If the data quality coefficient Vts is lower than expected, a data optimization instruction is issued to the outside. The data quality coefficient Vts is generated from a set of data quality indicators, wherein the accuracy Qt and the completeness Zt are linearly normalized, and then in the following manner: Weight coefficient, 0≤α≤1, 0≤β≤1; Optimize the preprocessed data based on the optimization solution matched by the data optimization solution library, regularly back up the optimized data, and restrict the frequency of data backup; The operating status coefficient Cmj is constructed based on the changing trend of the target equipment-related data. If the operating status coefficient Cmj is lower than expected, the abnormality of the target equipment operation is evaluated. If the abnormality exceeds expectations, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target equipment.

2. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 1 is characterized in that: After determining the usage scenario of the device, add a scenario description to it, perform similarity analysis on different scenario descriptions and obtain similarity data, aggregate several similarity data obtained to generate a scenario data set, and generate the complexity Mst of the usage scenario from the obtained scenario data set.

3. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 2 is characterized in that: Taking the device to be collected as the target device, the response time of the sensor is constrained as follows: Where n is the number of response times; R i is the response time of the sensor’s i-th response, R a is the mean response time, R b is the acceptable value of response time; if the response time of the selected sensor does not meet the constraint conditions, it is replaced.

4. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 3 is characterized in that: Record the operating environment data of the target device, summarize the acquired operating environment data to generate an operating environment data set, and generate the operating environment coefficient Hot of the target device from the operating environment data set. If the operating environment coefficient Hot is higher than the operating threshold, use this condition as the trigger condition for data collection and issue a data collection instruction to the outside.

5. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 4 is characterized in that: After receiving the data collection instruction, set the sampling period and perform real-time data collection at a preset sampling rate, obtain the collected data for preprocessing, and obtain the preprocessed data set; Perform data quality analysis on each piece of preprocessed data, collect corresponding data quality indicators, and generate a data quality indicator set after aggregation.

6. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 5 is characterized in that: After receiving the data optimization instruction, feature extraction is performed on the preprocessed data, and the corresponding optimization features are obtained; a number of acquired data optimization solutions are summarized to build a data optimization solution library; the data optimization solution library matches the corresponding optimization solution for the preprocessed data.

7. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 6 is characterized in that: Execute the optimization plan to optimize the preprocessed data, and back up the optimized data on the backup node that meets the constraints. The constraints are as follows: Where n is the number of data backups within the data backup cycle, T ij is the time interval from the i-th data backup to the j-th data backup, T a is the average value of the time interval.

8. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 7 is characterized in that: Use visualization tools to analyze optimized data and generate trend charts, calculate the similarity between trend charts and standard charts, and obtain chart similarity data; construct the operating state coefficient Cmj of the target device from the obtained several chart similarity data; if the operating state coefficient Cmj is lower than the state threshold, record the current operating state of the target device and treat it as abnormal operation, and issue an alarm instruction to the outside; Among them, the chart similarity Td is linearly normalized, the corresponding data value is mapped to the interval [0,1], and the operating state coefficient Cmj is generated in the following way: Among them, m is the number of graph similarities, Td i is the similarity of the i-th graph, Td a is the average value of graph similarity; Td b is the acceptable value of graph similarity, weight coefficient: 0≤F1≤1, 0≤F2≤1, and F1+F2=1.

9. The high-frequency digital acquisition method based on the average sieve point algorithm according to claim 8, characterized in that: The abnormal data recognition model is obtained by the neural convolution network, and the optimized data is used as input. The trained abnormal recognition model recognizes abnormalities and outputs the abnormal data and corresponding abnormal nodes within the sampling period. The abnormal density Com is generated by the acquired abnormal nodes, and the time node when the alarm instruction is received is used as the alarm node to generate the alarm density Bom. If one of the abnormal density Com and the alarm density Bom exceeds expectations, the abnormal data within the sampling period is obtained and the corresponding abnormal features are obtained through feature extraction; and equipment maintenance optimization is used as the target word to pre-build the equipment operation optimization knowledge graph.

10. High-frequency digital acquisition system based on average screening point algorithm, characterized by: include, The scenario analysis unit analyzes and obtains the similarities between different usage scenarios of the target device, generates the complexity Mst of the usage scenario, configures the sensor network for the target device, and constrains the response time of the selected sensor according to the complexity Mst; The environment analysis unit, after setting the sampling rate for the sensor network, records the operating environment data of the target device and generates the operating environment coefficient Hot. If the operating environment coefficient Hot is higher than the operating threshold, it sends a data collection instruction to the outside; The quality analysis unit collects data and performs preprocessing at a preset sampling rate, and generates a data quality coefficient Vts based on the quality index of the preprocessed data. If the data quality coefficient Vts is lower than expected, a data optimization instruction is issued to the outside. The data backup unit optimizes the preprocessed data according to the optimization scheme matched by the data optimization scheme library, regularly backs up the optimized data, and constrains the frequency of data backup; The optimization unit constructs the operation status coefficient Cmj based on the change trend of the target equipment-related data. If the operation status coefficient Cmj is lower than expected, the abnormality of the target equipment operation is evaluated. If the abnormality exceeds expectations, the equipment operation optimization knowledge graph provides an optimized maintenance plan for the target equipment.

Images