Real-time metering monitoring and fault early warning system for a weighing instrument in an industrial internet environment

Abstract

The application discloses a weighing instrument real-time metering monitoring and fault early warning system in an industrial internet environment and belongs to the technical field of industrial data processing. The system comprises a data construction module, which acquires metering data, running state data and environment data of a weighing instrument, generates continuous monitoring data and a historical stable change range; a state analysis module, which generates a data change degree, a running stability degree and a sampling state of a current monitoring period; a sampling adjustment module, which generates hierarchical monitoring data; a degradation analysis module, which generates degradation evolution characteristics according to the hierarchical monitoring data, identifies a degradation starting boundary and generates target analysis data; a fault early warning module, which generates a metering performance change state, a fault risk result and fault early warning information; and a feedback correction module, which generates a sampling correction result, updates sampling requirements and a target sampling level. The application improves the accuracy of fault early warning by identifying metering performance gradual changes and degradation starting boundaries that cannot be explained by running conditions.

Description

Technical Field

[0001] This invention relates to the field of industrial data processing technology, and in particular to a real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment. Background Technology

[0002] Weighing instruments are measuring devices that determine the mass or related values ​​of an object by utilizing the force of gravity acting on it. They typically include a load-bearing mechanism, load cells, signal conversion devices, data processing devices, and display or output devices. They are widely used in industrial production, warehousing and logistics, transportation, and trade settlement. During long-term operation, the zero-point state, loading response, unloading recovery, and consistency of outputs from various load cells may be affected by factors such as mechanical wear, component aging, load impact, changes in ambient temperature, vibration interference, and communication abnormalities. Therefore, continuous monitoring of the weighing instrument's metrological performance and operating status is necessary to ensure the accuracy of measurement results and the reliability of equipment operation.

[0003] Existing weighing instrument monitoring systems typically use load cells, operating status sensors, and environmental sensors to collect weighing results, zero-point data, sensor output data, and environmental parameters at a pre-set fixed sampling frequency. The collected data is then uploaded to a local monitoring terminal or industrial internet platform. The monitoring terminal usually compares real-time data with preset thresholds, standard values, or historical averages. When zero-point drift, weighing error, inconsistent sensor output, or operating parameters exceeding set ranges are detected, anomaly alerts or fault alarms are generated. Some systems also perform trend analysis based on abnormal results from multiple consecutive sampling times to determine if the weighing instrument's metrological performance has deteriorated, and confirm faults through manual inspection or periodic calibration.

[0004] However, existing technologies typically employ fixed sampling periods or directly adjust sampling frequencies based on single threshold exceedances. They fail to generate tiered sampling requirements for the next monitoring period based on the degree of change in measurement data, operational stability, and the state transition relationships between adjacent monitoring periods. This makes it difficult to balance key monitoring with resource utilization when industrial internet resources are limited. Furthermore, existing technologies often directly use anomalies in measurement data as the basis for fault diagnosis, lacking analysis of the correlation between changes in operating status, environmental changes, and gradual changes in measurement performance. This makes it difficult to accurately identify degradation initiation boundaries that cannot be explained by operating conditions. In addition, existing systems typically only adjust thresholds or perform manual maintenance after a fault alarm, failing to combine the actual fault occurrence time, degradation initiation boundaries, and fault confirmation results to locate missed sampling coverage gaps or false alarm sampling verification deviations. This results in sampling strategies not being continuously corrected based on monitoring results, easily leading to delayed fault detection, false alarms, and wasted monitoring resources.

[0005] Therefore, there is an urgent need to provide a real-time weighing monitoring and fault early warning system for weighing instruments in the industrial internet environment to solve the above problems. Summary of the Invention

[0006] The technical problem this invention aims to solve is to overcome the shortcomings of existing technologies, which typically employ fixed sampling periods or directly adjust sampling frequencies based on single threshold exceedances. These technologies fail to generate tiered sampling requirements for the next monitoring period based on the degree of change in measurement data, operational stability, and the state transition relationship between adjacent monitoring periods. This makes it difficult to balance key monitoring and resource utilization when industrial internet resources are limited. Furthermore, existing technologies often use abnormal measurement data directly as the basis for fault judgment, lacking analysis of the correlation between changes in operating status, environmental changes, and gradual changes in measurement performance. This makes it difficult to accurately identify degradation initiation boundaries that cannot be explained by operating conditions. In addition, existing systems typically only adjust thresholds or perform manual maintenance after a fault alarm, failing to combine the actual fault occurrence time, degradation initiation boundaries, and fault confirmation results to locate missed sampling coverage gaps or false alarm sampling verification deviations. This results in sampling strategies not being continuously corrected based on monitoring results, easily leading to problems such as delayed fault detection, false alarms, and wasted monitoring resources. The invention provides a real-time measurement monitoring and fault early warning system for weighing instruments in an industrial internet environment.

[0007] To address the aforementioned technical problems, one technical solution adopted by this invention is: providing a real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment, comprising:

[0008] The data construction module is used to acquire the weighing instrument's measurement data, operating status data, and environmental data, and to generate continuous monitoring data and historical stable variation ranges.

[0009] The status analysis module is used to generate the degree of data change, the degree of operational stability, and the sampling status of the current monitoring period based on the continuous monitoring data and the historical stable change range.

[0010] The sampling adjustment module is used to generate sampling requirements for the next monitoring period based on the sampling status, generate a target sampling level based on the sampling requirements and resource availability, and generate graded monitoring data based on the target sampling level.

[0011] The degradation analysis module is used to generate degradation evolution characteristics based on the graded monitoring data, identify degradation initiation boundaries, and generate target analysis data based on the degradation initiation boundaries.

[0012] The fault early warning module is used to generate metering performance change status, fault risk results and fault early warning information based on the target analysis data;

[0013] The feedback correction module is used to generate a sampling correction result based on the fault warning information, subsequent metering data and fault confirmation results, combined with the deviation of the degradation start boundary positioning sampling strategy, and update the sampling requirements and target sampling level based on the sampling correction result.

[0014] The present invention is further configured such that: the data construction module generates continuous monitoring data and historical stable variation ranges, including:

[0015] Multiple monitoring objects are determined based on the measurement items of the weighing instrument. The monitoring objects include the weighing instrument as a whole, the weighing sensor, the zero point monitoring item, the loading response monitoring item, and the unloading recovery monitoring item.

[0016] When the system is running for the first time or before the target sampling level is formed, the measurement data, operating status data and environmental data of each monitored object are acquired in multiple consecutive monitoring periods according to the preset basic sampling level. The measurement data includes weighing results, zero point data, loading data, unloading data and sensor output data.

[0017] The metering data, operational status data, and environmental data are time-aligned according to their generation time to generate time-series monitoring data.

[0018] Identify missing data, duplicate data, and time-displaced data in the time-series monitoring data; continuously supplement missing data that can be supplemented; delete duplicate data; and set data status markers for missing data and time-displaced data that cannot be supplemented, thereby generating the continuous monitoring data.

[0019] Historical monitoring data generated during the normal operation of the weighing instrument is obtained, divided according to the monitoring object, and the stable change range of the corresponding measurement data of each monitoring object during the normal operation is statistically analyzed. Based on the stable change range, the historical stable change range of the corresponding monitoring object is generated.

[0020] The present invention is further configured such that: the state analysis module generates data change level, operational stability level, and sampling status for the current monitoring period, including:

[0021] The continuous monitoring data is divided into multiple monitoring data segments according to the monitoring object and monitoring period;

[0022] Determine whether the measurement data corresponding to each monitoring data segment deviates from the historical stable change range of the corresponding monitoring object, and generate the degree of data change based on the deviation magnitude, duration of deviation, direction of change and fluctuation of adjacent data;

[0023] The operational stability level is generated based on the magnitude, duration, and correlation of changes in the corresponding operational status data and environmental data for each monitoring data segment.

[0024] When the corresponding measurement data does not deviate from the historical stable change range, and the operational stability does not decrease relative to the previous monitoring period, a stable sampling state is generated;

[0025] When the corresponding measurement data deviates from the historical stable range and the number of monitoring periods that maintain the same direction of change does not reach the preset number, a trend observation status is generated.

[0026] When the number of monitoring periods in which the corresponding measurement data continuously deviates from the historical stable range and maintains the same direction of change reaches a preset number, or when the degree of data change continues to increase and the degree of operational stability decreases relative to the previous monitoring period, a degraded attention status is generated.

[0027] The sampling status for the current monitoring period is generated based on the stable sampling status, trend observation status, and degradation concern status.

[0028] The present invention is further configured such that: the sampling adjustment module generates the sampling requirements for the next monitoring period based on the sampling status and the direction of status change in adjacent monitoring periods, including:

[0029] The sampling status of the current monitoring period and the sampling status of the previous monitoring period are obtained, and the state change direction is generated based on the conversion relationship between the sampling status of the current monitoring period and the sampling status of the previous monitoring period.

[0030] When both the current monitoring period and the previous monitoring period are in a stable sampling state, and the corresponding measurement data do not show a continuous direction of change, a low-density sampling requirement is generated.

[0031] When the current monitoring period changes from a stable sampling state to a trend observation state, or when the corresponding measurement data continues the trend of change of the previous monitoring period, a trend tracking sampling requirement is generated.

[0032] When the current monitoring period is in a degraded attention state, or when it changes from a trend observation state to a degraded attention state, a key encrypted sampling requirement is generated;

[0033] When the current monitoring period changes from a degraded attention state to a trend observation state or a stable sampling state, a verification and maintenance sampling requirement is generated.

[0034] The sampling requirements for the next monitoring period are generated based on the low-density sampling requirements, trend-tracking sampling requirements, key encrypted sampling requirements, and verification and maintenance sampling requirements.

[0035] The present invention is further configured such that: the sampling adjustment module generates a target sampling level based on the sampling requirements and resource availability, including:

[0036] Acquire data transmission resources, data processing resources, and data storage resources of the Industrial Internet, and generate resource availability status based on the data transmission resources, data processing resources, and data storage resources;

[0037] The data collection priority order is determined according to the priority order of key encrypted sampling needs, verification and maintenance sampling needs, trend tracking sampling needs, and low-density sampling needs.

[0038] The low-density sampling requirement is assigned to the first sampling level, the trend tracking sampling requirement to the second sampling level, the verification and maintenance sampling requirement to the third sampling level, and the key encryption sampling requirement to the fourth sampling level. The sampling frequency of the first sampling level to the fourth sampling level increases sequentially, and the sampling interval decreases sequentially.

[0039] When the available resources cannot meet all sampling requirements, the sampling resources of the fourth and third sampling levels shall be prioritized according to the data collection priority order, and the sampling intervals corresponding to the second and first sampling levels shall be adjusted in turn.

[0040] The target sampling level for the corresponding monitoring object is generated based on the adjusted first sampling level, second sampling level, third sampling level, or fourth sampling level.

[0041] The present invention is further configured such that: the sampling adjustment module generates graded monitoring data according to the target sampling level, including:

[0042] Send sampling control commands to the weighing instrument data acquisition terminal according to the target sampling level corresponding to each monitoring object;

[0043] The data acquisition terminal is controlled to perform the corresponding sampling method according to the sampling level corresponding to the target sampling level. When the target sampling level corresponds to the first sampling level, interval sampling is performed; when it corresponds to the second sampling level, trend tracking sampling is performed; when it corresponds to the third sampling level, continuous verification sampling is performed; and when it corresponds to the fourth sampling level, key encryption sampling is performed.

[0044] Acquire sampling data generated by performing the corresponding sampling method on each monitored object;

[0045] The sampled data is associated with the corresponding monitoring object, monitoring period, sampling status, target sampling level, and data status marker to generate sampling marker data;

[0046] The sampling marker data is organized according to the monitoring object and monitoring period to generate the hierarchical monitoring data.

[0047] The present invention is further configured such that: the degradation analysis module generates degradation evolution characteristics based on the graded monitoring data, including:

[0048] The hierarchical monitoring data are divided according to the monitoring objects and monitoring periods to determine the overall change direction of the corresponding measurement data for each monitoring object;

[0049] The degree of change persistence is generated based on the number of continuous monitoring periods that maintain a consistent overall direction of change.

[0050] Determine the changes in overall change direction corresponding to changes in operational status data and / or environmental data, and generate change-related features based on change type, magnitude, and duration;

[0051] When the deviation of measurement data from the historical stable range can be explained by changes in operating status data and / or environmental data, the corresponding change is identified as a change in operating conditions.

[0052] When the measurement data continuously deviates from the historical stable range, and the changes in the operating status data and environmental data are insufficient to explain the continuous changes in the measurement data, the corresponding change is determined to be a gradual change in measurement performance.

[0053] The degradation and evolution characteristics are generated based on the overall direction of change, the degree of change, the correlation characteristics of change, the changes in operating conditions, and the gradual changes in measurement performance.

[0054] The present invention is further configured such that: the degradation analysis module identifies the degradation initiation boundary and generates target analysis data based on the degradation evolution characteristics, including:

[0055] Obtain the degradation and evolution characteristics of each monitored object and the hierarchical monitoring data formed by multiple consecutive monitoring periods;

[0056] The monitoring period in which the corresponding measurement data first deviates continuously from the corresponding historical stable range and continues to maintain the same overall direction of change is determined as the candidate starting period.

[0057] Determine whether the continuous changes in the measurement data after the candidate start time period can be explained by changes in the corresponding operating status data and / or environmental data;

[0058] When the continuous changes in the measurement data cannot be explained by changes in the corresponding operating status data and / or environmental data, the candidate starting time period is determined as the degradation starting boundary;

[0059] The system retains the preceding data within a preset monitoring period before the degradation initiation boundary, the valid change data that continuously deviates from the historical stable change range after the degradation initiation boundary, and the subsequent verification data used to verify the change trend.

[0060] Target analysis data is generated based on the aforementioned preliminary data, effective change data, and subsequent verification data.

[0061] The present invention is further configured such that: the change state, the loading response change state, the unloading recovery change state, and the sensor output consistency change state generate the measurement performance change state;

[0062] The associated monitoring objects are determined based on the physical connection relationship, data correlation relationship and the overlap of change time among the monitoring objects;

[0063] Based on the change in the metrological performance, determine the corresponding monitoring object, associated monitoring object, number of continuous monitoring periods, direction of change, and degradation initiation boundary;

[0064] When the number of consecutive monitoring periods spanned by the change in the metrological performance does not reach the preset number of degradation confirmations, and the corresponding subsequent verification data recovers to the historical stable change range, a low-risk result is generated.

[0065] When the change in the metrological performance reaches a preset number of confirmed degradations and involves only a single monitoring object, a medium-risk result is generated.

[0066] When the change in the measurement performance reaches a preset number of degradation confirmations and involves associated monitoring objects, or when the degree of data change continues to increase, a high-risk result is generated.

[0067] Based on the low-risk, medium-risk, or high-risk results, a fault warning message is generated, and the corresponding monitoring object, associated monitoring object, degradation initiation boundary, change direction, number of continuous monitoring periods, and target sampling level are written into the fault warning message.

[0068] The present invention is further configured such that: the feedback correction module generates a sampling correction result based on the fault warning information, including:

[0069] Obtain subsequent metering data and fault confirmation results for the monitoring object corresponding to the fault warning information;

[0070] When the fault confirmation result indicates the existence of a corresponding fault, the actual fault occurrence time period is determined based on the fault confirmation result, and the data continuity of each monitoring time period between the actual fault occurrence time period and the degradation initiation boundary is judged based on the data status marker. The missed sampling coverage gap is determined, and the sampling start time period and sampling duration period are determined based on the missed sampling coverage gap. A sampling requirement to increase the sampling density is generated, and the target sampling level of the corresponding monitoring object under the same sampling state and the same degradation evolution characteristics is increased based on the sampling requirement to increase the sampling density. At the same time, the sampling start time period is advanced or the sampling duration period is extended.

[0071] When the fault confirmation result indicates that there is no corresponding fault, the subsequent measurement data is used to determine whether the corresponding measurement data has recovered to the historical stable range. The cause of the false alarm is analyzed in conjunction with the operating status data and environmental data. When it is determined that the insufficient verification data is due to the sampling strategy, the corresponding monitoring period is determined as the false alarm sampling verification deviation. The sampling requirement for restoring normal sampling is generated based on the false alarm sampling verification deviation. The sampling start period, sampling duration and target sampling level are re-determined based on the sampling requirement for restoring normal sampling. At the same time, the target sampling level of the corresponding monitoring object under the same operating condition changes is reduced or the sampling duration is shortened.

[0072] Based on the adjusted sampling requirements, target sampling level, sampling start time, and sampling duration, a sampling correction result is generated, and the sampling requirements and target sampling level of the corresponding monitoring object are updated according to the sampling correction result, and then sent to the sampling adjustment module.

[0073] The beneficial effects of this invention are as follows:

[0074] 1. This invention generates hierarchical sampling requirements by sampling status and the direction of status change, and determines the target sampling level by combining industrial Internet resources, thereby improving the data collection capability and monitoring resource utilization efficiency during key abnormal periods;

[0075] 2. This invention analyzes the correlation between changes in measurement data, operating status data, and environmental data to identify gradual changes and degradation initiation boundaries of measurement performance that cannot be explained by operating conditions, thereby improving the accuracy of fault early warning.

[0076] 3. This invention locates the sampling coverage gaps and false alarm verification deviations by confirming the fault results, and corrects the sampling requirements, sampling level and sampling period to reduce the number of missed, false and duplicate invalid samples. Attached Figure Description

[0077] Figure 1 This is a system flowchart of the present invention;

[0078] Figure 2 This is a flowchart of the degradation initiation boundary identification process of the present invention. Detailed Implementation

[0079] The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby providing a clearer and more explicit definition of the scope of protection of the present invention.

[0080] Please see Figure 1 - Figure 2 A real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment includes:

[0081] The data construction module is used to acquire the weighing instrument's measurement data, operating status data, and environmental data, and to generate continuous monitoring data and historical stable variation ranges.

[0082] The status analysis module is used to generate the degree of data change, the degree of operational stability, and the sampling status of the current monitoring period based on the continuous monitoring data and the historical stable change range.

[0083] The sampling adjustment module is used to generate sampling requirements for the next monitoring period based on the sampling status, generate a target sampling level based on the sampling requirements and resource availability, and generate graded monitoring data based on the target sampling level.

[0084] The degradation analysis module is used to generate degradation evolution characteristics based on the graded monitoring data, identify degradation initiation boundaries, and generate target analysis data based on the degradation initiation boundaries.

[0085] The fault early warning module is used to generate metering performance change status, fault risk results and fault early warning information based on the target analysis data;

[0086] The feedback correction module is used to generate a sampling correction result based on the fault warning information, subsequent metering data and fault confirmation results, combined with the deviation of the degradation start boundary positioning sampling strategy, and update the sampling requirements and target sampling level based on the sampling correction result.

[0087] In a specific embodiment of the present invention, the data construction module generates continuous monitoring data and historical stable change range in the following specific implementation method.

[0088] First, multiple monitoring objects are determined based on the weighing items of the weighing instrument. In this embodiment, the monitoring objects include the weighing instrument as a whole, the load cell, zero-point monitoring items, loading response monitoring items, and unloading recovery monitoring items. The weighing instrument as a whole reflects its overall weighing performance, the load cell reflects the operating status of the weighing components, and the zero-point monitoring items, loading response monitoring items, and unloading recovery monitoring items reflect changes in the weighing performance of the instrument during no-load, loading, and unloading processes, respectively. The data construction module establishes corresponding data acquisition objects for each monitoring object to ensure that the subsequently acquired data corresponds one-to-one with the corresponding monitoring object.

[0089] Subsequently, during the initial system operation or before the target sampling level is established, the data construction module continuously acquires measurement data, operational status data, and environmental data for each monitored object across multiple consecutive monitoring periods, according to a pre-set basic sampling level. The measurement data includes weighing results, zero-point data, loading data, unloading data, and sensor output data. The data construction module acquires the corresponding measurement data, operational status data, and environmental data for each monitored object separately, ensuring continuous data acquisition across multiple consecutive monitoring periods to form a complete data acquisition process.

[0090] After acquiring the data corresponding to each monitoring object, the data construction module aligns the measurement data, operational status data, and environmental data according to their generation time. This ensures that the measurement data, operational status data, and environmental data of the corresponding monitoring objects within the same monitoring period maintain a consistent chronological order, and generates time-series monitoring data for the corresponding monitoring objects. The generated time-series monitoring data is arranged continuously in chronological order according to the monitoring period to ensure data continuity and temporal consistency for the same monitoring object.

[0091] After generating time-series monitoring data, the data construction module identifies missing, duplicate, and time-displaced data within the data. For missing data that can be supplemented, continuous supplementation is performed; for duplicate data, deletion is performed; for missing or time-displaced data that cannot be supplemented, corresponding data status flags are set and retained, ensuring that the generated continuous monitoring data simultaneously includes the data status information of the corresponding monitored object, allowing the subsequent status analysis module to identify the completeness and validity of the corresponding data.

[0092] After generating continuous monitoring data, the data construction module acquires historical monitoring data generated during the normal operation of the weighing instrument and divides it according to the monitored objects, so that each monitored object corresponds to its own historical monitoring data. Subsequently, it statistically analyzes the stable variation range of the measurement data corresponding to each monitored object during normal operation and generates the historical stable variation range for the corresponding monitored object based on the stable variation range. Different monitored objects correspond to their own historical stable variation ranges, serving as a reference for subsequent judgments on whether the corresponding measurement data deviates from the normal variation state.

[0093] For example, over multiple consecutive monitoring periods, the data construction module continuously acquires metering data, operational status data, and environmental data corresponding to a specific monitoring object, and performs time alignment, continuity supplementation, and data status marking processing to generate corresponding continuous monitoring data. Subsequently, based on the historical monitoring data generated during the normal operation of the monitoring object, it statistically analyzes the corresponding stable variation intervals and generates the historical stable variation range corresponding to the monitoring object. Finally, the data construction module sends the continuous monitoring data and historical stable variation range to the status analysis module, so that the status analysis module can generate the degree of data change, operational stability, and sampling status of the current monitoring period based on the continuous monitoring data and historical stable variation range.

[0094] In a specific embodiment of the present invention, the specific implementation method for the state analysis module to generate the degree of data change, the degree of operational stability, and the sampling status of the current monitoring period is as follows.

[0095] First, the status analysis module receives continuous monitoring data and historical stable variation ranges from the data construction module. Then, the status analysis module divides the continuous monitoring data into multiple monitoring data segments according to the monitoring object and monitoring period, so that each monitoring data segment corresponds to continuous monitoring data generated by a single monitoring object within a specific monitoring period. Each monitoring data segment corresponds to its respective monitoring object and monitoring period, providing a unified data foundation for subsequent status analysis.

[0096] Subsequently, the status analysis module determines whether the measurement data corresponding to each monitoring data segment deviates from the historical stable variation range of the corresponding monitored object. Specifically, the status analysis module compares the measurement data corresponding to each monitoring data segment with the historical stable variation range of the corresponding monitored object, and generates the degree of data change for the corresponding monitoring data segment by combining the deviation magnitude, deviation duration, change direction, and fluctuation of adjacent data. Among them, the deviation magnitude reflects the degree to which the measurement data deviates from the historical stable variation range, the deviation duration reflects the duration of continuous deviation, the change direction reflects the trend of measurement data change, and the fluctuation of adjacent data reflects the continuity of change between continuous monitoring data, thereby comprehensively generating the degree of data change for the corresponding monitoring data segment.

[0097] Meanwhile, the status analysis module generates the operational stability level based on the magnitude, duration, and correlation of changes in the corresponding operational status data and environmental data for each monitoring data segment. The status analysis module comprehensively analyzes the changes in operational status data and environmental data within the corresponding monitoring period to determine the operational stability level of the currently monitored object, ensuring that the operational stability level reflects the overall operational status of the corresponding monitored object within the current monitoring period.

[0098] After obtaining the degree of data change and operational stability, the status analysis module generates the sampling status for the current monitoring period based on whether the corresponding measurement data deviates from the historical stable change range and the change in operational stability. When the corresponding measurement data does not deviate from the historical stable change range and the operational stability has not decreased compared to the previous monitoring period, the status analysis module generates a stable sampling status. When the corresponding measurement data deviates from the historical stable change range and the number of monitoring periods that maintain the same direction of change continuously does not reach a preset number, the status analysis module generates a trend observation status. When the corresponding measurement data continuously deviates from the historical stable change range and the number of monitoring periods that maintain the same direction of change continuously reaches a preset number, or when the degree of data change continues to increase and the operational stability decreases compared to the previous monitoring period, the status analysis module generates a degradation concern status.

[0099] For example, if, during multiple consecutive monitoring periods, the measurement data corresponding to a certain monitoring object continuously deviates from its historical stable range while maintaining the same direction of change, and the corresponding operational stability decreases compared to the previous monitoring period, the status analysis module generates a degradation concern status based on the above analysis results. If the corresponding measurement data only deviates for a short time and the number of monitoring periods maintaining the same direction of change does not reach the preset number, the status analysis module generates a trend observation status. If the corresponding measurement data remains within its historical stable range and the operational stability does not decrease, the status analysis module generates a stable sampling status.

[0100] Finally, the status analysis module sends the generated data change level, operational stability level, and sampling status of the current monitoring period to the sampling adjustment module, so that the sampling adjustment module can generate the sampling requirements for the next monitoring period based on the sampling status and the direction of status change in adjacent monitoring periods, thereby realizing the data connection between status analysis and sampling adjustment.

[0101] In a specific embodiment of the present invention, the sampling adjustment module generates the sampling requirements for the next monitoring period based on the sampling status and the direction of status change in adjacent monitoring periods as follows:

[0102] First, the sampling adjustment module receives data from the state analysis module regarding the degree of data change, operational stability, and sampling status for the current monitoring period. Then, the sampling adjustment module acquires the sampling status for the current monitoring period and the sampling status for the previous monitoring period, and generates a state change direction based on the transition relationship between the current and previous sampling statuses. The sampling status includes stable sampling status, trend observation status, and degradation concern status. The state change direction reflects the change process of the sampling status between adjacent monitoring periods, providing a basis for generating sampling requirements for the next monitoring period.

[0103] Subsequently, the sampling adjustment module generates sampling requirements for the next monitoring period based on the direction of state change and the corresponding changes in measurement data. When both the current and previous monitoring periods are in a stable sampling state, and the corresponding measurement data does not show a continuous direction of change, the sampling adjustment module generates low-density sampling requirements. When the current monitoring period changes from a stable sampling state to a trend observation state, or when the corresponding measurement data continues the direction of change of the previous monitoring period, the sampling adjustment module generates trend-tracking sampling requirements. When the current monitoring period is in a degraded attention state, or changes from a trend observation state to a degraded attention state, the sampling adjustment module generates key encrypted sampling requirements. When the current monitoring period changes from a degraded attention state to a trend observation state or a stable sampling state, the sampling adjustment module generates verification and maintenance sampling requirements.

[0104] Among them, low-density sampling is used for monitoring objects whose measurement data remain stable and do not show a continuous trend of change; trend tracking sampling is used for monitoring objects whose measurement data show a continuous trend of change; key encrypted sampling is used for monitoring objects whose measurement data changes continuously and whose sampling status enters a degraded state of concern; and verification and maintenance sampling is used for monitoring objects whose sampling status has changed from a degraded state of concern to a trend observation state or a stable sampling state, so as to maintain the continuous acquisition of subsequent monitoring data.

[0105] For example, if, during multiple consecutive monitoring periods, the sampling status of a monitored object changes from a stable sampling status to a trend observation status, while the corresponding measurement data maintains the same direction of change as in the previous monitoring period, the sampling adjustment module generates trend tracking sampling requirements based on the corresponding state change direction. When the corresponding sampling status further changes to a degraded attention status in subsequent monitoring periods, the sampling adjustment module generates key encrypted sampling requirements. When the corresponding sampling status changes from a degraded attention status to a trend observation status or a stable sampling status, the sampling adjustment module generates verification and maintenance sampling requirements to ensure the continuity of subsequent monitoring data for the corresponding monitored object.

[0106] Finally, the sampling adjustment module generates the sampling requirements for the next monitoring period based on the generated low-density sampling requirements, trend tracking sampling requirements, key encryption sampling requirements, and verification and maintenance sampling requirements. The sampling requirements for the next monitoring period are then used to generate the target sampling level based on the resource availability status, thereby achieving a continuous connection between the sampling status, the direction of status change, and the sampling requirements.

[0107] In a specific embodiment of the present invention, the sampling adjustment module generates the target sampling level according to the sampling requirements and resource availability status in the following specific implementation method.

[0108] First, the sampling adjustment module receives the sampling requirements for the next monitoring period generated by the previous processing step. Then, the sampling adjustment module acquires the current data transmission resources, data processing resources, and data storage resources of the industrial internet, and generates a resource availability status based on these resources. The resource availability status reflects the comprehensive resource situation capable of supporting data transmission, data processing, and data storage in the current industrial internet environment, providing a basis for the subsequent generation of the target sampling level.

[0109] Subsequently, the sampling adjustment module determines the data collection priority order for each monitored object based on sampling requirements and resource availability. Specifically, the key encryption sampling requirement corresponds to the highest priority, the verification and maintenance sampling requirement corresponds to the second highest priority, the trend tracking sampling requirement corresponds to the next highest priority, and the low-density sampling requirement corresponds to the lowest priority. This ensures that data collection for monitored objects with different sampling requirements can be allocated resources according to the preset priority order.

[0110] Next, the sampling adjustment module establishes a correspondence between sampling requirements and sampling levels, assigning the low-density sampling requirement to the first sampling level, the trend-tracking sampling requirement to the second sampling level, the verification and maintenance sampling requirement to the third sampling level, and the key encryption sampling requirement to the fourth sampling level. The sampling frequency increases sequentially from the first to the fourth sampling level, while the sampling interval decreases sequentially to adapt to the data acquisition requirements corresponding to different sampling needs.

[0111] After establishing the correspondence, the sampling adjustment module determines whether the resource availability status meets all sampling requirements. When the resource availability status can meet all sampling requirements, the sampling adjustment module generates a target sampling level according to the sampling level corresponding to the sampling requirement; when the resource availability status is insufficient to meet all sampling requirements, the sampling adjustment module reallocates resources according to the data acquisition priority order, prioritizing the sampling level corresponding to the sampling requirement with higher priority, and adjusting the sampling level corresponding to the remaining sampling requirements according to the resource availability status, so that each monitored object can complete the corresponding data acquisition under the current resource conditions.

[0112] For example, during multiple consecutive monitoring periods, the sampling adjustment module continuously receives sampling requests from different monitoring objects and generates corresponding target sampling levels based on the current resource availability. When resource availability can meet all sampling requests, each monitoring object maintains the sampling level matched to its corresponding sampling request. When resource availability changes and is insufficient to meet all sampling requests, the sampling adjustment module adjusts the target sampling level of the corresponding monitoring object according to a preset data acquisition priority order to ensure that monitoring objects with higher priority complete data acquisition first.

[0113] Finally, the sampling adjustment module generates a target sampling level for the corresponding monitoring object based on the adjusted first sampling level, second sampling level, third sampling level, or fourth sampling level, and uses the target sampling level to send sampling control commands in the future, thereby realizing the continuous connection between sampling requirements, resource availability status, and target sampling level.

[0114] In a specific embodiment of the present invention, the sampling adjustment module generates graded monitoring data according to the target sampling level in the following specific implementation method.

[0115] First, the sampling adjustment module obtains the target sampling level for the corresponding monitoring object and sends sampling control commands according to the target sampling level for each monitoring object. These sampling control commands are used to control the corresponding monitoring object to perform data acquisition according to the target sampling level, ensuring that different monitoring objects complete the corresponding data acquisition process based on their respective target sampling levels.

[0116] Subsequently, the sampling adjustment module controls the corresponding sampling method for the monitored object according to the target sampling level. Specifically, when the target sampling level corresponds to the first sampling level, interval sampling is performed; when the target sampling level corresponds to the second sampling level, trend tracking sampling is performed; when the target sampling level corresponds to the third sampling level, continuous verification sampling is performed; and when the target sampling level corresponds to the fourth sampling level, focused and encrypted sampling is performed. Different sampling methods correspond to different data acquisition frequencies to adapt to the data acquisition requirements of different monitored objects under different sampling needs.

[0117] Next, the sampling adjustment module acquires the sampling data generated by the corresponding sampling method. This sampling data includes metering data, operational status data, and environmental data generated by the corresponding monitored object. The sampling adjustment module receives the sampling data generated by the corresponding sampling method and uses it as the data basis for subsequently generating sampling marker data.

[0118] Subsequently, the sampling adjustment module associates the sampled data with the corresponding monitoring object, monitoring period, sampling status, target sampling level, and data status marker to generate sampling marker data. Specifically, the monitoring object identifies the data source of the sampled data, the monitoring period identifies the monitoring time to which the sampled data belongs, the sampling status identifies the current sampling status of the corresponding monitoring object, the target sampling level identifies the sampling level of the corresponding sampled data, and the data status marker identifies the status information of the corresponding sampled data. This provides each sampled data with unified data identification information, facilitating subsequent retrieval and analysis.

[0119] For example, during multiple consecutive monitoring periods, the sampling adjustment module continuously acquires sampling data according to the target sampling level of the corresponding monitoring object, and associates it with the corresponding monitoring object, monitoring period, sampling status, target sampling level and data status mark to form corresponding sampling mark data, so that the sampling data formed by different monitoring objects in different monitoring periods have corresponding data identification information.

[0120] Finally, the sampling adjustment module organizes the sampling marker data according to the monitoring object and monitoring period to generate hierarchical monitoring data. Specifically, the sampling marker data corresponding to the same monitoring object is organized according to the monitoring period, while different monitoring objects generate corresponding hierarchical monitoring data. This hierarchical monitoring data retains the data collection results corresponding to different target sampling levels and maintains the correspondence between the monitoring object and the monitoring period. The sampling adjustment module sends the generated hierarchical monitoring data to the degradation analysis module, which then generates degradation evolution characteristics based on the hierarchical monitoring data.

[0121] In a specific embodiment of the present invention, the degradation analysis module generates degradation evolution characteristics based on the hierarchical monitoring data in the following specific implementation method.

[0122] First, the degradation analysis module receives the tiered monitoring data sent by the sampling and adjustment module, and organizes the tiered monitoring data according to the monitoring object and monitoring period to form a data sequence for the corresponding monitoring object. Then, the degradation analysis module analyzes the changes in the measurement data over the continuous monitoring period based on the data sequence of the corresponding monitoring object, determining the overall direction of change of the corresponding measurement data. The overall direction of change is used to characterize the overall trend of change of the measurement data over the continuous monitoring period, including continuous increase, continuous decrease, or remaining stable.

[0123] Subsequently, the degradation analysis module generates a degree of change persistence based on whether the overall direction of change remains consistent over continuous monitoring periods. The degree of change persistence is low when the overall direction of change remains consistent only over a few monitoring periods; and high when the overall direction of change remains consistent over multiple consecutive monitoring periods. This degree of change persistence reflects the continuity of the overall direction of change and serves as the basis for subsequent analysis of the causes of changes in the measurement data.

[0124] Next, the degradation analysis module acquires the operational status data and / or environmental data corresponding to the overall direction of change, analyzes the changes in the corresponding operational status data and / or environmental data within the relevant monitoring period, and generates change correlation features based on the change type, magnitude, and duration. These change correlation features describe the correspondence between changes in measurement data and changes in operational status data and / or environmental data, providing a basis for subsequent determination of the causes of changes in measurement data.

[0125] Subsequently, the degradation analysis module determines the reasons for the deviation of the measurement data from the historical stable range based on the change correlation characteristics. When the deviation of the measurement data from the historical stable range can be explained by changes in operating status data and / or environmental data, the corresponding change is identified as a change in operating conditions; when the measurement data continuously deviates from the historical stable range, and the changes in operating status data and environmental data are insufficient to explain the continuous change in the measurement data, the corresponding change is identified as a gradual change in measurement performance.

[0126] For example, over multiple consecutive monitoring periods, the degradation analysis module continuously analyzes the changes in the measurement data of the corresponding monitored object, and simultaneously analyzes the changes in the corresponding operating status data and environmental data. When the changes in measurement data can be explained by changes in operating status data and / or environmental data, the corresponding change is determined to be a change in operating conditions; when the measurement data maintains the same direction of change and no corresponding changes occur in the operating status data and environmental data that can explain the changes, the corresponding change is determined to be a gradual change in measurement performance.

[0127] Finally, the degradation analysis module generates degradation evolution features based on the overall change direction, duration of change, correlation characteristics of change, changes in operating conditions, and gradual changes in measurement performance. These degradation evolution features characterize the performance change process of the corresponding monitored object during continuous monitoring periods and serve as data for subsequent identification of degradation initiation boundaries. The degradation analysis module sends these degradation evolution features to the subsequent processing module for identification of degradation initiation boundaries.

[0128] In a specific embodiment of the present invention, the degradation analysis module identifies the degradation initiation boundary and generates target analysis data based on the degradation evolution characteristics as follows.

[0129] First, the degradation analysis module acquires the degradation evolution characteristics and hierarchical monitoring data corresponding to each monitored object, and determines the changes in the corresponding measurement data within the continuous monitoring period according to the monitored object and monitoring period. The degradation evolution characteristics include the overall direction of change, the degree of change duration, and the correlation characteristics of change, used to characterize the change process of the measurement data of the corresponding monitored object.

[0130] Subsequently, based on the degradation evolution characteristics and hierarchical monitoring data, the degradation analysis module determines the monitoring period in which the corresponding measurement data first continuously deviates from the historical stable range and maintains the same overall direction of change, and identifies this monitoring period as a candidate starting period. Specifically, when the corresponding measurement data begins to deviate from the historical stable range in a certain monitoring period and maintains the same overall direction of change in subsequent continuous monitoring periods, the monitoring period in which the first continuous deviation occurs is identified as the candidate starting period.

[0131] Next, the degradation analysis module acquires the operational status data and / or environmental data corresponding to the candidate start time period and its subsequent monitoring periods, and determines whether the continuous changes in the measurement data after the candidate start time period can be explained by the changes in the corresponding operational status data and / or environmental data. When the continuous changes in the measurement data can be explained by the changes in the operational status data and / or environmental data, the corresponding changes are identified as changes in operational conditions, and the candidate start time period is not identified as the degradation initiation boundary. When the measurement data continuously deviates from the historical stable change range, and the changes in the corresponding operational status data and environmental data are insufficient to explain the continuous changes in the measurement data, the corresponding changes are identified as gradual changes in measurement performance, and the candidate start time period is identified as the degradation initiation boundary.

[0132] Subsequently, the degradation analysis module, based on the degradation initiation boundary, retains from the graded monitoring data the preceding data within a preset number of monitoring periods before the degradation initiation boundary, the valid change data after the degradation initiation boundary that continuously deviates from the historical stable change range, and subsequent verification data used to verify the corresponding change trend. The preceding data reflects the measurement state before the degradation initiation boundary, the valid change data reflects the continuous change process of the measurement data after the degradation initiation boundary, and the subsequent verification data verifies whether the continuous change maintains the corresponding overall change direction.

[0133] Finally, the degradation analysis module correlates and organizes the preceding data, effective change data, and subsequent verification data according to the monitoring object, monitoring period, and degradation initiation boundary to generate target analysis data. This target analysis data covers the measurement data before the degradation initiation boundary, during the degradation change stage, and during the subsequent verification stage, and is sent to the fault early warning module so that the fault early warning module can generate measurement performance change status and fault risk results based on the target analysis data.

[0134] In a specific embodiment of the present invention, the fault early warning module generates the metering performance change status, fault risk results and fault early warning information based on the target analysis data as follows.

[0135] First, the fault early warning module receives the target analysis data sent by the degradation analysis module and extracts the preceding data, valid change data, and subsequent verification data from the target analysis data. Based on the target analysis data corresponding to different monitoring objects, the fault early warning module determines the zero-point change state, loading response change state, unloading recovery change state, and sensor output consistency change state, and generates a metrological performance change state based on these states. The metrological performance change state characterizes the metrological performance changes of different monitoring objects of the weighing instrument during a continuous monitoring period.

[0136] Subsequently, the fault early warning module determines the associated monitoring objects based on the physical connections, data correlations, and overlapping change times among the monitored objects. Specifically, when different monitoring objects have physical connections, their corresponding data show the same or corresponding changes within a continuous monitoring period, or the corresponding changes occur within the same monitoring period, the corresponding monitoring object is identified as an associated monitoring object.

[0137] Next, the fault early warning module determines the corresponding monitoring object, associated monitoring object, number of continuous monitoring periods, direction of change, and degradation initiation boundary based on the metrological performance change status and target analysis data. The number of continuous monitoring periods reflects the continuity of the metrological performance change status, the direction of change reflects the overall direction of change in the corresponding metrological data, and the degradation initiation boundary characterizes the monitoring period in which the gradual change in metrological performance begins.

[0138] Subsequently, the fault early warning module compares the number of continuous monitoring periods with the preset degradation confirmation number and generates a fault risk result based on the determination result of the associated monitoring objects. A low-risk result is generated when the number of continuous monitoring periods does not reach the preset degradation confirmation number; a medium-risk result is generated when the number of continuous monitoring periods reaches the preset degradation confirmation number and the change in metering performance only involves the corresponding monitoring object; and a high-risk result is generated when the number of continuous monitoring periods reaches the preset degradation confirmation number and the change in metering performance involves the associated monitoring objects.

[0139] Finally, the fault warning module generates fault warning information based on the low-risk, medium-risk, or high-risk results. This fault warning information includes the corresponding monitoring object, associated monitoring objects, degradation initiation boundary, direction of change, number of continuous monitoring periods, and target sampling level, used to characterize the fault risk of the corresponding monitoring object. The fault warning module sends the fault warning information to the feedback correction module, which then combines subsequent measurement data and fault confirmation results to generate sampling correction results.

[0140] In a specific embodiment of the present invention, the feedback correction module generates sampling correction results based on the fault warning information in the following specific implementation method.

[0141] First, the feedback correction module receives fault warning information sent by the fault warning module and obtains subsequent metering data and fault confirmation results for the monitored object corresponding to the fault warning information. The fault confirmation results are used to characterize whether a confirmed fault exists in the corresponding monitored object. The feedback correction module obtains the corresponding monitored object, degradation initiation boundary, and target sampling level from the fault warning information as the basis for sampling strategy deviation analysis.

[0142] Subsequently, the feedback correction module determines whether the corresponding monitored object has a fault based on the fault confirmation result.

[0143] When the fault confirmation result indicates the existence of a corresponding fault, the feedback correction module determines the actual fault occurrence time based on the fault confirmation result, and determines whether there is a monitoring period not fully covered by the current sampling strategy based on the time relationship between the actual fault occurrence time and the degradation initiation boundary. The feedback correction module further determines whether there are missing sampling data, insufficient effective sampling data, or situations where the sampling period does not cover the measurement data change process based on the data status markers within the corresponding monitoring period, and determines the underreported sampling coverage gap accordingly.

[0144] Subsequently, the feedback correction module determines the sampling start time and sampling duration based on the missed sampling coverage gap, generates a sampling requirement to increase the sampling density, and raises the target sampling level of the corresponding monitoring object according to the increased sampling density requirement. Specifically, by advancing the sampling start time and / or extending the sampling duration, subsequent monitoring processes can obtain tiered monitoring data covering the corresponding measurement data change process.

[0145] When the fault confirmation result indicates that there is no corresponding fault, the feedback correction module determines whether the corresponding measurement data has recovered to the historical stable range based on the subsequent measurement data. When the corresponding measurement data has recovered to the historical stable range, and the existing sampling data is insufficient to verify the recovery process, the feedback correction module identifies the corresponding monitoring period as a false alarm sampling verification deviation.

[0146] Subsequently, the feedback correction module generates a sampling requirement to restore normal sampling based on the false alarm sampling verification deviation, and redetermines the sampling start time, sampling duration, and target sampling level based on the restored normal sampling requirements. Specifically, by lowering the target sampling level and / or shortening the sampling duration, the corresponding monitored object is restored to a sampling method adapted to the current sampling state.

[0147] Finally, the feedback correction module generates a sampling correction result based on the adjusted sampling requirements, target sampling level, sampling start time, and sampling duration, and sends the sampling correction result to the sampling adjustment module. The sampling adjustment module updates the sampling requirements and target sampling level of the corresponding monitored object based on the sampling correction result, so that subsequent monitoring periods perform data collection according to the corrected target sampling level, thereby forming a closed-loop processing process of fault warning, fault confirmation, sampling strategy deviation location, and sampling strategy correction.

[0148] Example

[0149] A weighing instrument in an industrial internet environment operates continuously for 20 monitoring periods. The system determines the weighing instrument as a whole, the weighing sensor, the zero-point monitoring item, the loading response monitoring item, and the unloading recovery monitoring item as monitoring objects based on the instrument's measurement items. During the initial system run, measurement data, operational status data, and environmental data are collected for each monitoring object according to the basic sampling level. The weighing results are 1000.0, 1000.1, 1000.0, 1000.2, 1000.4, 1000.6, 1000.8, 1001.0, 1001.2, and 1001.5 kg respectively; the zero-point data are 0.00, 0.01, 0.02, 0.03, 0.05, 0.07, 0.09, 0.12, 0.15, and 0.18 kg respectively; loading data, unloading data, and sensor output data are collected simultaneously; and operational status data and environmental data are simultaneously generated to form data for 20 consecutive monitoring periods. The data construction module aligns the data according to its generation time, identifies and deletes one duplicate data entry, identifies one missing data entry that cannot be filled in and sets a data status flag, generating continuous monitoring data. Subsequently, historical monitoring data is statistically analyzed, yielding a historical stable variation range of ±0.05 kg for the zero-point monitoring item. The status analysis module divides the monitoring data into segments according to the monitoring object and monitoring period. It calculates that the zero-point data for monitoring periods 1 to 4 do not deviate from the historical stable variation range, generating a stable sampling state. For monitoring periods 5 to 7, the zero-point data deviates from 0.00 kg to 0.05 kg, 0.07 kg, and 0.09 kg respectively, with 3 monitoring periods maintaining the same direction of change, falling short of the preset number of 5, generating a trend observation state. For monitoring periods 8 to 12, the zero-point data reaches 0.12 kg, 0.15 kg, 0.18 kg, 0.22 kg, and 0.25 kg respectively, with 5 monitoring periods maintaining the same direction of change. Simultaneously, the degree of data change continues to increase, and the operational stability decreases compared to the previous monitoring period, generating a degradation concern state. The sampling adjustment module generates sampling requirements for the next monitoring period based on the direction of state change. The stable sampling state corresponds to low-density sampling requirements, the trend observation state corresponds to trend tracking sampling requirements, and the degradation attention state corresponds to key encrypted sampling requirements. The data transmission resources, data processing resources, and data storage resources of the Industrial Internet meet all sampling requirements. Therefore, the first sampling level, the second sampling level, and the fourth sampling level are generated as target sampling levels. Interval sampling, trend tracking sampling, and key encrypted sampling are performed according to the target sampling levels to form sampling data. The sampling data is associated with the monitoring object, monitoring period, sampling state, target sampling level, and data status marker to generate sampling marker data. Finally, the hierarchical monitoring data is organized according to the monitoring object and monitoring period.The degradation analysis module determined that the overall trend of change in the zero-point data was a continuous increase based on the graded monitoring data, with 8 monitoring periods maintaining consistency, indicating a high degree of change persistence. Analysis of the operational status data and environmental data revealed that the change amplitude remained stable, generating change correlation features. Since the measurement data continuously deviated from the historical stable change range and the changes in the operational status data and environmental data were insufficient to explain the continuous change in the measurement data, it was determined to be a gradual change in measurement performance, and degradation evolution features were generated. The 8th monitoring period was further identified as a candidate starting period. It was determined that the continuous change in measurement data after the 8th monitoring period could not be explained by the operational status data and environmental data, so the 8th monitoring period was determined as the degradation starting boundary. The 5th to 7th monitoring periods were retained as preliminary data, the 8th to 16th monitoring periods as effective change data, and the 17th to 20th monitoring periods as subsequent verification data, and target analysis data were generated. The fault early warning module generates zero-point change status, loading response change status, unloading recovery change status, and sensor output consistency change status based on target analysis data, and comprehensively generates the metrological performance change status. Based on physical connection relationships, data correlation relationships, and overlap of change times, the weighing sensor is identified as the associated monitoring object. Based on the metrological performance change status, the corresponding monitoring object, associated monitoring object, number of continuous monitoring periods of 9, change direction of continuous increase, and degradation start boundary of the 8th monitoring period are determined. Since the number of continuous monitoring periods reaches the preset degradation confirmation number of 5 and involves associated monitoring objects, a high-risk result is generated, and the corresponding monitoring object, associated monitoring object, degradation start boundary, change direction, number of continuous monitoring periods, and target sampling level are written into the fault early warning information. The feedback correction module obtains the subsequent metering data and fault confirmation results of the monitoring object corresponding to the fault warning information. The fault confirmation results show that there is a corresponding fault, and the actual fault occurred during the 7th monitoring period. According to the data status mark, there is a data continuity deficiency of 1 monitoring period between the 7th monitoring period and the degradation start boundary. The missed sampling coverage gap is identified, and the sampling start period is adjusted to the 6th monitoring period and the sampling duration is adjusted from 6 monitoring periods to 10 monitoring periods. A sampling requirement to increase the sampling density is generated, and the target sampling level is increased to the fourth sampling level. Finally, the sampling correction result is generated, and the sampling requirement and target sampling level are updated according to the sampling correction result and sent to the sampling adjustment module to execute the data collection for the next monitoring cycle.

[0150] This embodiment can dynamically generate target sampling levels based on sampling status, accurately identify degradation initiation boundaries, generate fault risk results, and revise sampling requirements and target sampling levels based on fault confirmation results, thereby improving the accuracy of fault early warning and reducing missed reports, false reports, and waste of monitoring resources.

[0151] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment, characterized by: include: The data construction module is used to acquire the weighing instrument's measurement data, operating status data, and environmental data, and to generate continuous monitoring data and historical stable variation ranges. The status analysis module is used to generate the degree of data change, the degree of operational stability, and the sampling status of the current monitoring period based on the continuous monitoring data and the historical stable change range. The sampling adjustment module is used to generate sampling requirements for the next monitoring period based on the sampling status, generate a target sampling level based on the sampling requirements and resource availability, and generate graded monitoring data based on the target sampling level. The degradation analysis module is used to generate degradation evolution characteristics based on the graded monitoring data, identify degradation initiation boundaries, and generate target analysis data based on the degradation initiation boundaries. The fault early warning module is used to generate metering performance change status, fault risk results and fault early warning information based on the target analysis data; The feedback correction module is used to generate a sampling correction result based on the fault warning information, subsequent metering data and fault confirmation results, combined with the deviation of the degradation start boundary positioning sampling strategy, and update the sampling requirements and target sampling level based on the sampling correction result.

2. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 1, characterized in that: The data construction module generates continuous monitoring data and historical stable variation ranges, including: Multiple monitoring objects are determined based on the measurement items of the weighing instrument. The monitoring objects include the weighing instrument as a whole, the weighing sensor, the zero point monitoring item, the loading response monitoring item, and the unloading recovery monitoring item. When the system is running for the first time or before the target sampling level is formed, the measurement data, operating status data and environmental data of each monitored object are acquired in multiple consecutive monitoring periods according to the preset basic sampling level. The measurement data includes weighing results, zero point data, loading data, unloading data and sensor output data. The metering data, operational status data, and environmental data are time-aligned according to their generation time to generate time-series monitoring data. Identify missing data, duplicate data, and time-displaced data in the time-series monitoring data; continuously supplement missing data that can be supplemented; delete duplicate data; and set data status markers for missing data and time-displaced data that cannot be supplemented, thereby generating the continuous monitoring data. Historical monitoring data generated during the normal operation of the weighing instrument is obtained, divided according to the monitoring object, and the stable change range of the corresponding measurement data of each monitoring object during the normal operation is statistically analyzed. Based on the stable change range, the historical stable change range of the corresponding monitoring object is generated.

3. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 2, characterized in that: The status analysis module generates data change level, operational stability level, and sampling status for the current monitoring period, including: The continuous monitoring data is divided into multiple monitoring data segments according to the monitoring object and monitoring period; Determine whether the measurement data corresponding to each monitoring data segment deviates from the historical stable change range of the corresponding monitoring object, and generate the degree of data change based on the deviation magnitude, duration of deviation, direction of change and fluctuation of adjacent data; The operational stability level is generated based on the magnitude, duration, and correlation of changes in the corresponding operational status data and environmental data for each monitoring data segment. When the corresponding measurement data does not deviate from the historical stable change range, and the operational stability does not decrease relative to the previous monitoring period, a stable sampling state is generated; When the corresponding measurement data deviates from the historical stable range and the number of monitoring periods that maintain the same direction of change does not reach the preset number, a trend observation status is generated. When the number of monitoring periods in which the corresponding measurement data continuously deviates from the historical stable range and maintains the same direction of change reaches a preset number, or when the degree of data change continues to increase and the degree of operational stability decreases relative to the previous monitoring period, a degraded attention status is generated. The sampling status for the current monitoring period is generated based on the stable sampling status, trend observation status, and degradation concern status.

4. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 3, characterized in that: The sampling adjustment module generates the sampling requirements for the next monitoring period based on the sampling status and the direction of status changes in adjacent monitoring periods, including: The sampling status of the current monitoring period and the sampling status of the previous monitoring period are obtained, and the state change direction is generated based on the conversion relationship between the sampling status of the current monitoring period and the sampling status of the previous monitoring period. When both the current monitoring period and the previous monitoring period are in a stable sampling state, and the corresponding measurement data do not show a continuous direction of change, a low-density sampling requirement is generated. When the current monitoring period changes from a stable sampling state to a trend observation state, or when the corresponding measurement data continues the trend of change of the previous monitoring period, a trend tracking sampling requirement is generated. When the current monitoring period is in a degraded attention state, or when it changes from a trend observation state to a degraded attention state, a key encrypted sampling requirement is generated; When the current monitoring period changes from a degraded attention state to a trend observation state or a stable sampling state, a verification and maintenance sampling requirement is generated. The sampling requirements for the next monitoring period are generated based on the low-density sampling requirements, trend-tracking sampling requirements, key encrypted sampling requirements, and verification and maintenance sampling requirements.

5. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 4, characterized in that: The sampling adjustment module generates a target sampling level based on the sampling requirements and resource availability, including: Acquire data transmission resources, data processing resources, and data storage resources of the Industrial Internet, and generate resource availability status based on the data transmission resources, data processing resources, and data storage resources; The data collection priority order is determined according to the priority order of key encrypted sampling needs, verification and maintenance sampling needs, trend tracking sampling needs, and low-density sampling needs. The low-density sampling requirement is assigned to the first sampling level, the trend tracking sampling requirement to the second sampling level, the verification and maintenance sampling requirement to the third sampling level, and the key encryption sampling requirement to the fourth sampling level. The sampling frequency of the first sampling level to the fourth sampling level increases sequentially, and the sampling interval decreases sequentially. When the available resources cannot meet all sampling requirements, the sampling resources of the fourth and third sampling levels shall be prioritized according to the data collection priority order, and the sampling intervals corresponding to the second and first sampling levels shall be adjusted in turn. The target sampling level for the corresponding monitoring object is generated based on the adjusted first sampling level, second sampling level, third sampling level, or fourth sampling level.

6. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 5, characterized in that: The sampling adjustment module generates graded monitoring data based on the target sampling level, including: Send sampling control commands to the weighing instrument data acquisition terminal according to the target sampling level corresponding to each monitoring object; The data acquisition terminal is controlled to perform the corresponding sampling method according to the sampling level corresponding to the target sampling level. When the target sampling level corresponds to the first sampling level, interval sampling is performed; when it corresponds to the second sampling level, trend tracking sampling is performed; when it corresponds to the third sampling level, continuous verification sampling is performed; and when it corresponds to the fourth sampling level, key encryption sampling is performed. Acquire sampling data generated by performing the corresponding sampling method on each monitored object; The sampled data is associated with the corresponding monitoring object, monitoring period, sampling status, target sampling level, and data status marker to generate sampling marker data; The sampling marker data is organized according to the monitoring object and monitoring period to generate the hierarchical monitoring data.

7. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 6, characterized in that: The degradation analysis module generates degradation evolution characteristics based on the hierarchical monitoring data, including: The hierarchical monitoring data are divided according to the monitoring objects and monitoring periods to determine the overall change direction of the corresponding measurement data for each monitoring object; The degree of change persistence is generated based on the number of continuous monitoring periods that maintain a consistent overall direction of change. Determine the changes in overall change direction corresponding to changes in operational status data and / or environmental data, and generate change-related features based on change type, magnitude, and duration; When the deviation of measurement data from the historical stable range can be explained by changes in operating status data and / or environmental data, the corresponding change is identified as a change in operating conditions. When the measurement data continuously deviates from the historical stable range, and the changes in the operating status data and environmental data are insufficient to explain the continuous changes in the measurement data, the corresponding change is determined to be a gradual change in measurement performance. The degradation and evolution characteristics are generated based on the overall direction of change, the degree of change, the correlation characteristics of change, the changes in operating conditions, and the gradual changes in measurement performance.

8. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 7, characterized in that: The degradation analysis module identifies the degradation initiation boundary based on the degradation evolution characteristics and generates target analysis data, including: Obtain the degradation and evolution characteristics of each monitored object and the hierarchical monitoring data formed by multiple consecutive monitoring periods; The monitoring period in which the corresponding measurement data first deviates continuously from the corresponding historical stable range and continues to maintain the same overall direction of change is determined as the candidate starting period. Determine whether the continuous changes in the measurement data after the candidate start time period can be explained by changes in the corresponding operating status data and / or environmental data; When the continuous changes in the measurement data cannot be explained by changes in the corresponding operating status data and / or environmental data, the candidate starting time period is determined as the degradation starting boundary; The system retains the preceding data within a preset monitoring period before the degradation initiation boundary, the valid change data that continuously deviates from the historical stable change range after the degradation initiation boundary, and the subsequent verification data used to verify the change trend. Target analysis data is generated based on the aforementioned preliminary data, effective change data, and subsequent verification data.

9. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 8, characterized in that: The change states, loading response change states, unloading recovery change states, and sensor output consistency change states are used to generate the measurement performance change states. The associated monitoring objects are determined based on the physical connection relationship, data correlation relationship and the overlap of change time among the monitoring objects; Based on the change in the metrological performance, determine the corresponding monitoring object, associated monitoring object, number of continuous monitoring periods, direction of change, and degradation initiation boundary; When the number of consecutive monitoring periods spanned by the change in the metrological performance does not reach the preset number of degradation confirmations, and the corresponding subsequent verification data recovers to the historical stable change range, a low-risk result is generated. When the change in the metrological performance reaches a preset number of confirmed degradations and involves only a single monitoring object, a medium-risk result is generated. When the change in the measurement performance reaches a preset number of degradation confirmations and involves associated monitoring objects, or when the degree of data change continues to increase, a high-risk result is generated. Based on the low-risk, medium-risk, or high-risk results, a fault warning message is generated, and the corresponding monitoring object, associated monitoring object, degradation initiation boundary, change direction, number of continuous monitoring periods, and target sampling level are written into the fault warning message.

10. The real-time weighing monitoring and fault early warning system for weighing instruments in an industrial internet environment according to claim 9, characterized in that: The feedback correction module generates a sampling correction result based on the fault warning information, including: Obtain subsequent metering data and fault confirmation results for the monitoring object corresponding to the fault warning information; When the fault confirmation result indicates the existence of a corresponding fault, the actual fault occurrence time period is determined based on the fault confirmation result, and the data continuity of each monitoring time period between the actual fault occurrence time period and the degradation initiation boundary is judged based on the data status marker. The missed sampling coverage gap is determined, and the sampling start time period and sampling duration period are determined based on the missed sampling coverage gap. A sampling requirement to increase the sampling density is generated, and the target sampling level of the corresponding monitoring object under the same sampling state and the same degradation evolution characteristics is increased based on the sampling requirement to increase the sampling density. At the same time, the sampling start time period is advanced or the sampling duration period is extended. When the fault confirmation result indicates that there is no corresponding fault, the subsequent measurement data is used to determine whether the corresponding measurement data has recovered to the historical stable range. The cause of the false alarm is analyzed in conjunction with the operating status data and environmental data. When it is determined that the insufficient verification data is due to the sampling strategy, the corresponding monitoring period is determined as the false alarm sampling verification deviation. The sampling requirement for restoring normal sampling is generated based on the false alarm sampling verification deviation. The sampling start period, sampling duration and target sampling level are re-determined based on the sampling requirement for restoring normal sampling. At the same time, the target sampling level of the corresponding monitoring object under the same operating condition changes is reduced or the sampling duration is shortened. Based on the adjusted sampling requirements, target sampling level, sampling start time, and sampling duration, a sampling correction result is generated, and the sampling requirements and target sampling level of the corresponding monitoring object are updated according to the sampling correction result, and then sent to the sampling adjustment module.

Images