Error detection data processing method and system based on mine data sensor

Abstract

The application discloses a kind of error detection data processing method and system based on mining data sensor, belong to mine safety monitoring technical field, this method includes: the real-time monitoring data of multiple mining data sensors in mine monitoring system is collected, the sensor at least includes two types same or physical quantity related sensor;Real-time data is time-synchronized and is handled, including timestamp correction and interpolation alignment, form unified time series data;According to the horizontal comparison of same type sensor and the pre-established different physical quantity physical correlation model, the consistency of uniform sequence is checked;Based on the statistical distribution characteristics of historical stable operation condition data, the dynamic error threshold of each sensor is calculated and periodically updated;The application improves the reliability of mine sensor data through multi-source fusion, dynamic threshold and trend classification, reduces false alarm and misses, and is convenient for targeted maintenance.

Description

Technical Field

[0001] This application relates to the field of mine safety monitoring technology, specifically to an error detection data processing method and system based on mining data sensors. Background Technology

[0002] Mining, a pillar industry of my country's energy supply, plays a vital role in promoting economic and social development. However, the mining environment is complex and ever-changing, with underground operations facing various safety risks such as gas explosions, fires, water hazards, and rock bursts, directly threatening miners' lives and production stability. Statistics show that coal mine accidents still occur frequently in my country, indicating a severe safety situation. To ensure efficient and safe mine operation, real-time monitoring of underground environmental parameters, equipment status, and geological changes has become an industry consensus. By deploying a monitoring system composed of multiple sensors (such as temperature, pressure, gas concentration, and vibration sensors), key data can be collected, providing a basis for hazard identification and emergency response.

[0003] Existing mine safety monitoring systems primarily rely on various types of sensors for data acquisition and anomaly detection through threshold comparisons or simple alarm mechanisms. For example, traditional systems often use catalytic combustion or electrochemical gas sensors combined with fixed thresholds to monitor gas concentration, supplemented by vibration and temperature sensors to monitor equipment operating status. Some systems have introduced wireless communication or IoT technologies to achieve data transmission and centralized display, supporting linkage with personnel positioning or ventilation systems. Furthermore, some upgrade and renovation solutions are promoting laser methane sensors to improve anti-interference capabilities and exploring multi-parameter fusion to enhance monitoring coverage.

[0004] However, existing mine monitoring systems still have significant shortcomings in sensor error handling. First, sensors are prone to zero-point drift, abrupt changes, or random noise in harsh underground environments such as humidity, high dust levels, and electromagnetic interference, leading to increased data deviation. Traditional systems often use fixed thresholds, which cannot adapt to changing operating conditions, resulting in frequent false alarms or missed alarms. Second, data from multiple sensors is often difficult to fuse effectively due to differences in sampling frequencies or time synchronization. Consistency verification is mainly limited to comparing single parameter thresholds, lacking in-depth utilization of horizontal comparisons between similar sensors and the correlations between different physical quantities. Furthermore, error types are difficult to classify precisely, failing to distinguish between drift, abrupt changes, or noise, resulting in poor maintenance targeting and increased manual intervention burden. These problems reduce the overall reliability of the system and affect safety production decisions. Summary of the Invention

[0005] This application provides a method and system for error detection data processing based on mining data sensors. Through time synchronization, multi-source consistency verification, dynamic threshold calculation, and error trend classification, it achieves accurate error detection and classification of sensor data, which can improve the reliability and accuracy of mine monitoring data and effectively solve the problems of poor adaptability of fixed thresholds, insufficient consistency verification, and difficulty in distinguishing error types in the background technology.

[0006] To achieve the above objectives, this application provides the following technical solution: a method for processing error detection data based on a mining data sensor, comprising the following steps:

[0007] S1. Collect real-time monitoring data from multiple mining data sensors in the mine monitoring system, wherein the mining data sensors include at least two types of sensors that are the same or have related physical quantities.

[0008] S2. Perform time synchronization processing on real-time monitoring data, and perform timestamp correction and time alignment on sensor data with different sampling frequencies to form unified time series data;

[0009] S3. Based on the horizontal data comparison between similar sensors and the pre-established physical correlation model between different physical quantities, perform consistency verification on the unified time series data.

[0010] S4. Based on the statistical distribution characteristics of sensor data in the historical stable operating data of the mine monitoring system, calculate the dynamic error threshold corresponding to each mine data sensor.

[0011] S5. Compare the consistency verification result with the corresponding dynamic error threshold, and determine the error of the sensing data that exceeds the dynamic error threshold.

[0012] S6. Based on the magnitude of the deviation of the sensing data from the reference value and the trend characteristics of its change over time, the sensing data that are determined to have errors are classified into zero-point drift type errors, sudden change anomaly type errors, or random noise type errors, and the sensing data are output after being marked with errors.

[0013] Furthermore, the mining data sensor includes at least two of the following: a temperature sensor, a pressure sensor, a displacement sensor, a vibration sensor, or a gas concentration sensor.

[0014] Furthermore, the time synchronization process includes synchronization processing based on timestamp correction and sampling frequency alignment processing based on interpolation.

[0015] Furthermore, the consistency check in step S3 includes:

[0016] Perform differential analysis on the monitoring data of similar sensors to determine whether the changes in their values ​​are within the acceptable range of consistency determined based on historical stable operating data;

[0017] Furthermore, based on the physical correlation model between different physical quantities, the monitoring data of relevant sensors are subjected to correlation consistency verification.

[0018] Furthermore, the dynamic error threshold is periodically updated based on the mean, variance, or fluctuation range of the sensor data in the historical stable operating condition data.

[0019] Furthermore, the zero-point drift error is determined by the continuous deviation of the sensing data from the reference value over a relatively long period of time; the abrupt change error is determined by the sudden change in amplitude of the sensing data within a short period of time; and the random noise error is determined by the high-frequency fluctuation characteristics of the sensing data.

[0020] The error detection data processing system based on mining data sensors includes a data acquisition module, a time synchronization module, a consistency verification module, a dynamic threshold calculation module, an error judgment module, and a data output module.

[0021] The data acquisition module is used to collect real-time monitoring data from multiple mining data sensors in the mine monitoring system;

[0022] The time synchronization module is used to perform time synchronization processing on real-time monitoring data to form unified time series data;

[0023] The consistency verification module performs consistency verification on unified time series data based on the horizontal data comparison between sensors of the same type and the physical correlation model between different physical quantities.

[0024] The dynamic threshold calculation module calculates the dynamic error threshold corresponding to each mining data sensor based on the statistical distribution characteristics of sensor data in historical stable operating condition data.

[0025] The error determination module is used to compare the consistency verification results with the dynamic error threshold and to determine and classify the errors in the sensor data.

[0026] The data output module is used to mark and output the sensor data that is determined to have errors.

[0027] The output of the time synchronization module is connected to the input of the consistency verification module, the output of the consistency verification module is connected to the dynamic threshold calculation module and the error determination module, and the output of the error determination module is connected to the data output module.

[0028] Furthermore, the error determination module classifies and determines the error type based on the magnitude of the deviation of the sensing data from the reference value and the trend characteristics of its change over time.

[0029] Compared with the prior art, the beneficial effects of this application are:

[0030] 1. By performing time synchronization processing and unified sequence alignment on real-time monitoring data, combined with horizontal comparison of similar sensors and consistency verification of physical correlation models of different physical quantities, the accuracy of multi-source data fusion is improved, and false error judgments caused by time misalignment or correlation deviation are reduced.

[0031] 2. The dynamic error threshold is calculated based on the statistical distribution characteristics of historical stable operating data and is updated periodically. This allows the threshold to adapt to changes in mine operating conditions, reducing the false alarm rate and false negative rate caused by fixed thresholds and improving the robustness of error detection.

[0032] 3. Based on the deviation magnitude and time trend characteristics, errors are classified into zero-point drift type, sudden change anomaly type and random noise type, and the output is marked to facilitate targeted maintenance and fault tracing, shorten the sensor maintenance response time, and improve the overall reliability of the mine monitoring system and the level of safety production assurance. Attached Figure Description

[0033] Figure 1 This is a flowchart of the method steps in this application.

[0034] Figure 2 This is a schematic diagram of the system flow of this application. Detailed Implementation

[0035] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0036] Please see Figure 1-2 This application provides the following technical solutions:

[0037] A method and system for error detection and data processing based on mining data sensors are proposed. This method can be applied to mine safety monitoring scenarios to detect and process errors generated by mining data sensors under complex working conditions.

[0038] The specific execution process of the method is as follows:

[0039] Step S1: Multi-source mining data acquisition; In step S1, real-time monitoring data from multiple mining data sensors are acquired from the mine monitoring system. These sensors cover at least two categories of the same type or related physical quantities.

[0040] Specifically, mining data sensors can be selected from at least two combinations of temperature sensors, pressure sensors, displacement sensors, vibration sensors, or gas concentration sensors to monitor environmental parameters in underground mines. For example, two DS18B20 digital temperature sensors and one BMP280 pressure sensor can be deployed at key locations underground to collect and monitor ambient temperature and atmospheric pressure.

[0041] Data acquisition can be achieved through a data acquisition device, which includes a signal interface unit and an analog-to-digital converter (ADC) unit (such as an ADS1115 chip). The ADC unit has a resolution of at least 16 bits. During acquisition, each sensor outputs monitoring data according to a preset sampling period. The sampling period for different types of sensors can be set according to monitoring requirements (e.g., a temperature sensor acquires data every 5 seconds, and a pressure sensor acquires data every 1 second). Each acquired data point includes at least the monitored value, the corresponding sensor identification information, and an initial timestamp. Sensor data can be transmitted to the data processing unit via an industrial communication protocol to ensure the reliability and integrity of data transmission.

[0042] Step S2: Time synchronization and sequence alignment; In step S2, the collected real-time monitoring data is processed for time synchronization to generate time series data under a unified time reference.

[0043] Time synchronization processing includes two sub-processes: timestamp correction and time alignment. Timestamp correction can be achieved based on an external reference clock, such as obtaining a reference time through a global positioning system or network time synchronization mechanism, and calculating the deviation between the original timestamp of each data point and the reference time to correct the original timestamp.

[0044] For sensor data with different sampling frequencies, further time alignment processing is performed. First, a uniform time sampling interval is determined. Then, data sequences with sampling frequencies higher than the defined interval are downsampled, and data sequences with sampling frequencies lower than the defined interval are interpolated. In one exemplary implementation, linear interpolation can be used to pad the low-frequency data, thereby forming a time-aligned data matrix, where each row corresponds to the same time point and each column corresponds to different sensor channels.

[0045] Step S3: Consistency verification process; In step S3, the consistency of the unified time series data is verified based on the horizontal data comparison between sensors of the same type and the pre-built physical correlation model between different physical quantities.

[0046] The consistency verification includes two parts: horizontal consistency verification and associated consistency verification.

[0047] In lateral consistency verification, difference analysis is performed on data collected by similar sensors at the same time point to obtain a difference sequence. It is then determined whether this difference sequence falls within a permissible consistency range determined based on historical stable operating data. This permissible range can be determined based on statistical results from historical stable operating periods, for example, by constructing an interval range based on the mean and standard deviation of the difference sequence. When the difference result exceeds the permissible range, the data at the corresponding time point is deemed to have lateral inconsistencies.

[0048] In correlation consistency verification, relevant sensor data are verified based on a physical correlation model between different physical quantities. For example, a linear or nonlinear correlation model between temperature and displacement can be established to predict the expected value of one physical quantity based on another, and the residual between the actual measured value and the predicted value is calculated. When the residual exceeds the allowable range determined based on historical data statistics, it is determined that there is a correlation inconsistency at that point in time.

[0049] The output of the consistency check can be represented as the consistency status information corresponding to the time series, which can be used for subsequent error determination.

[0050] Step S4: Dynamic error threshold calculation; In step S4, based on the historical stable operating condition data of the mine monitoring system, the dynamic error threshold corresponding to each mine data sensor is calculated, and the threshold is periodically updated.

[0051] Historical stable operating condition data can be extracted from the monitoring database, and the data is selected from continuous operating periods without triggering abnormal alarms. For each sensor channel, statistical distribution parameters, including mean, variance, or data fluctuation range, are calculated based on historical data, and the corresponding error threshold interval is determined accordingly.

[0052] In one exemplary implementation, a threshold range can be constructed based on the mean and variance of historical data to cover the vast majority of normal operating conditions. To adapt to changes in environmental conditions, the dynamic error threshold can be updated using a sliding time window approach, with the update cycle set according to actual application requirements.

[0053] Step S5: Error determination processing; In step S5, the consistency verification result obtained in step S3 is compared point by point with the dynamic error threshold obtained in step S4, and the sensing data that exceeds the dynamic error threshold and has consistency anomalies is determined to have errors.

[0054] Specifically, for each data point in a unified time series, if the consistency check result indicates inconsistency and the deviation of the data point from the historical benchmark exceeds the corresponding dynamic error threshold, then the data point is marked as error data. The error determination results can form an error label sequence of the same length as the original data sequence.

[0055] Step S6: Error type classification and labeling output; In step S6, for the sensor data that is determined to have errors, the error type is classified according to its deviation from the reference value and the trend characteristics of its change over time, and the error data is labeled and output.

[0056] In one alternative implementation, if the sensing data deviates from the reference value continuously over a long time window and the rate of change is slow, the error is determined to be a zero-point drift error; if the sensing data experiences a sudden change in amplitude within a short period of time, the error is determined to be a sudden change anomaly error; if the sensing data exhibits high-frequency fluctuation characteristics, the error is determined to be a random noise error.

[0057] In this embodiment, the trend analysis and frequency domain analysis methods are only exemplary implementations; other equivalent time domain or frequency domain analysis methods can also be used to classify error types. After classification, corresponding type tags are added to the error data, and the tagged data is output to an external monitoring terminal or storage device through a data output interface.

[0058] Based on the above method, this application also provides an error detection data processing system based on a mining data sensor, including a data acquisition module, a time synchronization module, a consistency verification module, a dynamic threshold calculation module, an error judgment module, and a data output module.

[0059] The data acquisition module is used to implement the multi-source mining data acquisition function described in step S1; the time synchronization module is used to implement the time synchronization and sequence alignment function described in step S2; the consistency verification module is used to implement the consistency verification process described in step S3; the dynamic threshold calculation module is used to implement the dynamic error threshold calculation and update described in step S4; the error judgment module is used to implement the error judgment and classification process described in steps S5 and S6; and the data output module is used to implement the marking and output of error data.

[0060] The data connections between modules are as follows: the output of the time synchronization module is connected to the input of the consistency verification module; the output of the consistency verification module is connected to both the dynamic threshold calculation module and the error determination module; and the output of the error determination module is connected to the data output module. The system can be controlled by an embedded processor or a programmable controller, and data interaction between modules can be achieved through shared storage or a message mechanism.

[0061] It is worth noting that in this embodiment, the sensor sampling interval should be controlled within the range of 1 to 100 seconds. The core chip of the control switch group can be a PLC microcontroller, specifically the Siemens S7-200. The mining data sensors can be freely configured according to the actual application scenario. It is recommended to use intrinsically safe mining temperature sensors (such as models integrating PT100 platinum resistance), intrinsically safe mining pressure sensors (such as the GPD60 or GPD25 series), intrinsically safe mining vibration sensors (such as the GBD20 or HZ-892A series), intrinsically safe mining displacement sensors, or intrinsically safe mining gas concentration sensors (such as the CH-D3 series catalytic combustion methane sensor). The data acquisition device can be an analog-to-digital converter with a resolution of not less than 16 bits (such as the ADS1115 chip). The control switch group (or microcontroller) controls the operation of each module using methods commonly used in existing technologies, such as timed interrupt-driven data processing loops and data transmission via RS485 or Modbus protocols.

[0062] Although embodiments of this application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for processing error detection data based on mining data sensors, characterized in that, Includes the following steps: S1. Collect real-time monitoring data from multiple mining data sensors in the mine monitoring system, wherein the mining data sensors include at least two types of sensors that are the same or have related physical quantities. S2. Perform time synchronization processing on real-time monitoring data, and perform timestamp correction and time alignment on sensor data with different sampling frequencies to form unified time series data; S3. Based on the horizontal data comparison between similar sensors and the pre-established physical correlation model between different physical quantities, perform consistency verification on the unified time series data. S4. Based on the statistical distribution characteristics of sensor data in the historical stable operating data of the mine monitoring system, calculate the dynamic error threshold corresponding to each mine data sensor. S5. Compare the consistency verification result with the corresponding dynamic error threshold, and determine the error of the sensing data that exceeds the dynamic error threshold. S6. Based on the magnitude of the deviation of the sensing data from the reference value and the trend characteristics of its change over time, the sensing data that are determined to have errors are classified into zero-point drift type errors, sudden change anomaly type errors, or random noise type errors, and the sensing data are output after being marked with errors.

2. The error detection data processing method based on a mining data sensor according to claim 1, characterized in that: The mining data sensor includes at least two of the following: a temperature sensor, a pressure sensor, a displacement sensor, a vibration sensor, or a gas concentration sensor.

3. The error detection data processing method based on a mining data sensor according to claim 1, characterized in that: The time synchronization process includes synchronization processing based on timestamp correction and sampling frequency alignment processing based on interpolation.

4. The error detection data processing method based on a mining data sensor according to claim 1, characterized in that: The consistency check in step S3 includes: Perform differential analysis on the monitoring data of similar sensors to determine whether the changes in their values ​​are within the acceptable range of consistency determined based on historical stable operating data; Furthermore, based on the physical correlation model between different physical quantities, the monitoring data of relevant sensors are subjected to correlation consistency verification.

5. The error detection data processing method based on a mining data sensor according to claim 1, characterized in that: The dynamic error threshold is periodically updated based on the mean, variance, or fluctuation range of the sensor data in historical stable operating condition data.

6. The error detection data processing method based on a mining data sensor according to claim 1, characterized in that: The zero-point drift error is determined by the continuous deviation of the sensor data from the reference value over a relatively long period of time, while the sudden change error is determined by the sudden change in amplitude of the sensor data within a short period of time. Random noise errors are determined by the high-frequency fluctuation characteristics of sensor data.

7. An error detection data processing system based on mining data sensors, characterized in that, It includes a data acquisition module, a time synchronization module, a consistency verification module, a dynamic threshold calculation module, an error judgment module, and a data output module; The data acquisition module is used to collect real-time monitoring data from multiple mining data sensors in the mine monitoring system; The time synchronization module is used to perform time synchronization processing on real-time monitoring data to form unified time series data; The consistency verification module performs consistency verification on unified time series data based on the horizontal data comparison between sensors of the same type and the physical correlation model between different physical quantities. The dynamic threshold calculation module calculates the dynamic error threshold corresponding to each mining data sensor based on the statistical distribution characteristics of sensor data in historical stable operating condition data. The error determination module is used to compare the consistency verification results with the dynamic error threshold and to determine and classify the errors in the sensor data. The data output module is used to mark and output the sensor data that is determined to have errors. The output of the time synchronization module is connected to the input of the consistency verification module, the output of the consistency verification module is connected to the dynamic threshold calculation module and the error determination module, and the output of the error determination module is connected to the data output module.

8. The error detection data processing system based on a mining data sensor according to claim 7, characterized in that: The error determination module classifies and determines the error type based on the magnitude of the deviation of the sensing data from the reference value and the trend characteristics of its change over time.

Images