Novel mine pressure monitoring method based on wireless sensor network and intelligent data analysis

Abstract

The invention relates to the technical field of mine pressure monitoring, and discloses a novel mine pressure monitoring method based on a wireless sensor network and intelligent data analysis, and the method comprises the steps: collecting mine pressure data in a mine environment through the wireless sensor network; transmitting the encrypted data of the mine pressure data to a local data processing position through wireless communication; decrypting and analyzing the encrypted data at the local data processing position to obtain an analysis result of the encrypted data; performing mine pressure anomaly analysis on the analysis result to obtain a mine pressure risk level of the analysis result; performing early warning strategy matching on the mine pressure risk grade to obtain graded early warning information of the mine environment; by optimizing wireless sensor network deployment and a data acquisition process, mine pressure data can be efficiently acquired, timeliness and pertinence of mine pressure early warning are effectively improved, and powerful guarantee is provided for mine safety prevention and control.

Description

Technical Field

[0001] This invention relates to the field of mine pressure monitoring technology, and in particular to a novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis. Background Technology

[0002] In traditional mine pressure monitoring systems, data acquisition mainly relies on wired sensors. These systems require numerous cables to connect sensors to data acquisition equipment, resulting in complex wiring processes and high installation and maintenance costs. Furthermore, wired sensors have poor adaptability to the complex environments of mines, such as humid, corrosive, or high-temperature conditions. Cable damage is common, leading to frequent system failures, increasing maintenance difficulty and costs, and making it difficult to consistently meet the needs of mine pressure monitoring.

[0003] Traditional systems rely on cable links for data transmission, making them susceptible to environmental interference and prone to data loss or transmission delays, severely reducing data accuracy and reliability. Furthermore, data analysis is mostly conducted in a central control room, where limitations in data transmission efficiency and processing capacity prevent real-time processing and monitoring of mine pressure data, hindering timely identification of anomalies. In addition, traditional systems lack effective data encryption and security protection mechanisms, leaving data vulnerable to leakage and tampering during transmission, further compromising the security and effectiveness of mine pressure monitoring. Summary of the Invention

[0004] This invention provides a novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis, the method comprising:

[0006] S1. Collect mine pressure data in the mine environment through a wireless sensor network;

[0007] S2. The encrypted data of the mine pressure data is transmitted to the local data processing location via wireless communication;

[0008] S3. Decrypt and analyze the encrypted data at the local data processing location to obtain the analysis result of the encrypted data;

[0009] S4. Perform anomaly analysis on the analysis results to obtain the mine pressure risk level of the analysis results;

[0010] S5. Match the early warning strategy to the mine pressure risk level to obtain the graded early warning information of the mine environment.

[0011] In a preferred embodiment, the step of collecting mine pressure data in the mine environment via a wireless sensor network includes:

[0012] The wireless sensor nodes deployed in the mine environment are networked to obtain the optimized network topology of the wireless sensor network;

[0013] Coordinated data acquisition is performed on the optimized network topology to obtain the mine pressure data of the wireless sensor network.

[0014] In a preferred embodiment, the step of networking the wireless sensor nodes deployed in the mine environment to obtain an optimized network topology for the wireless sensor network includes:

[0015] Dynamic route discovery is performed on the wireless sensor nodes to obtain the initial routing table of the wireless sensor nodes;

[0016] The initial routing table is evaluated for communication link quality to obtain the link quality parameters of the initial routing table;

[0017] The link quality parameters and the wireless sensor nodes are optimized to obtain the optimized network topology of the wireless sensor network.

[0018] In a preferred embodiment, the step of collaboratively acquiring data from the optimized network topology to obtain the mining pressure data of the wireless sensor network includes:

[0019] A time-series data acquisition plan is performed on the optimized network topology to obtain a node acquisition time-series table for the optimized network topology;

[0020] Distributed acquisition is triggered on the node acquisition time sequence table to obtain the original data packets of the node acquisition time sequence table;

[0021] The original data packets are fused to obtain the mining pressure data of the wireless sensor network.

[0022] In a preferred embodiment, transmitting the encrypted data of the mine pressure data to a local data processing location via wireless communication includes:

[0023] The mine pressure data is encrypted to obtain an encrypted data stream of the mine pressure data;

[0024] The encrypted data stream is encapsulated into data packets, and the encrypted data stream is transmitted via data packets;

[0025] The data packets are transmitted in a distributed manner through multiple communication sub-channels, and then sent to the local data processing location.

[0026] In a preferred embodiment, the step of decrypting and analyzing the encrypted data at the local data processing location to obtain the analysis result of the encrypted data includes:

[0027] The encrypted data is decrypted to obtain the decrypted data stream of the encrypted data;

[0028] Perform data integrity verification on the decrypted data stream to obtain verification passed data for the decrypted data stream;

[0029] The data quality of the verified data is evaluated to obtain the data quality evaluation results of the verified data.

[0030] Based on the data quality assessment results, the verification data is reconstructed to obtain the analysis results of the encrypted data.

[0031] The data reconstruction process uses the following reconstruction formula for calculation:

[0032]

[0033] In the formula, This represents the reconstructed data vector. This represents the data vector that passed the verification. This represents the statistical feature vector of the data vector that passed the verification. This represents the original data weighting coefficients that are dynamically adjusted based on the data quality assessment results. This represents the statistical feature weighting coefficients that are dynamically adjusted based on the data quality assessment results.

[0034] In a preferred embodiment, the step of performing mine pressure anomaly analysis on the analysis results to obtain the mine pressure risk level of the analysis results includes:

[0035] Multi-dimensional feature extraction is performed on the analysis results to obtain the mine pressure feature vector;

[0036] Anomaly pattern recognition is performed on the aforementioned mine pressure feature vector to obtain anomaly pattern recognition results;

[0037] The risk level of the anomaly pattern identification results is assessed by a risk probability calculation model to obtain the mine pressure risk level of the analysis results.

[0038] The risk probability calculation model is as follows:

[0039]

[0040] In the formula, Indicates the probability of risk. The first element in the ore pressure feature vector represents the... Each feature component Indicates the relationship with the first Preset weight coefficients corresponding to each feature component This represents the total number of characteristic components.

[0041] In a preferred embodiment, the step of assessing the risk level of the abnormal pattern recognition results using a risk probability calculation model to obtain the mine pressure risk level of the analysis results includes:

[0042] The abnormal pattern recognition result is subjected to pattern classification processing to obtain the abnormal pattern classification result of the abnormal pattern recognition result;

[0043] The risk of the abnormal pattern classification result is determined to obtain the risk probability of the abnormal pattern classification result.

[0044] The risk probability is mapped to a risk level to obtain the mine pressure risk level of the analysis results.

[0045] In a preferred embodiment, the step of matching early warning strategies for the mine pressure risk level to obtain graded early warning information for the mine pressure risk level includes:

[0046] The mine pressure risk level is queried from the early warning strategy database to obtain the initial early warning strategy for the mine environment;

[0047] The initial early warning strategy is configured with strategy parameters to obtain a personalized early warning strategy for the mine environment;

[0048] The personalized early warning strategy is converted into early warning information to obtain the graded early warning information of the mine environment.

[0049] In a preferred embodiment, the step of converting the personalized early warning strategy into early warning information to obtain the graded early warning information of the mine environment includes:

[0050] The initial warning content of the personalized warning strategy is adapted to channel characteristics to obtain the warning content items of the mine environment;

[0051] The warning content items are sorted by information priority to obtain the sorted warning content of the mine environment;

[0052] The sorted warning content is encapsulated in a warning format to obtain a standardized warning data packet of the mine environment;

[0053] The standardized early warning data packets are distributed through distribution channels to obtain the graded early warning information for the mine environment.

[0054] Beneficial effects

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] 1. This technology optimizes the deployment and data acquisition process of wireless sensor networks, enabling efficient acquisition of mine pressure data. By optimizing the network topology and collaborative data acquisition of wireless sensor nodes, combined with time-series planning and distributed triggering mechanisms, the accuracy and completeness of mine pressure data acquisition are improved. Simultaneously, decryption analysis and data reconstruction are performed at the local data processing location. A reconstruction algorithm with dynamically adjusted weight coefficients ensures the quality of the decrypted data, providing reliable data support for subsequent mine pressure analysis and significantly improving the effectiveness of mine pressure monitoring data.

[0057] 2. This technology excels in mine pressure risk identification and early warning response. Through multi-dimensional feature extraction and abnormal pattern recognition, combined with a risk probability calculation model, it achieves accurate assessment of mine pressure risk levels. Then, based on the risk level, it matches personalized early warning strategies, completing early warning content adaptation, priority ranking, and standardized encapsulation. This not only quickly locates potential mine pressure risks but also generates clear and standardized tiered early warning information, effectively improving the timeliness and relevance of mine pressure early warnings and providing strong support for mine safety control. Attached Figure Description

[0058] Figure 1 This is a flowchart illustrating a novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis, provided in an embodiment of the present invention.

[0059] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0060] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0061] This application provides a novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis. The executing entity of this novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0062] Reference Figure 1 The diagram shown is a flowchart illustrating a novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis according to an embodiment of the present invention. In this embodiment, the novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis includes:

[0063] S1. Collect mine pressure data in the mine environment through a wireless sensor network.

[0064] In this embodiment of the invention, the step of collecting mine pressure data in the mine environment via a wireless sensor network includes:

[0065] The wireless sensor nodes deployed in the mine environment are networked to obtain the optimized network topology of the wireless sensor network;

[0066] Coordinated data acquisition is performed on the optimized network topology to obtain the mine pressure data of the wireless sensor network.

[0067] The process of networking the wireless sensor nodes deployed in the mine environment to obtain the optimized network topology of the wireless sensor network includes:

[0068] Dynamic route discovery is performed on the wireless sensor nodes to obtain the initial routing table of the wireless sensor nodes;

[0069] The initial routing table is evaluated for communication link quality to obtain the link quality parameters of the initial routing table;

[0070] The link quality parameters and the wireless sensor nodes are optimized to obtain the optimized network topology of the wireless sensor network.

[0071] The step of collaboratively acquiring data from the optimized network topology to obtain the mining pressure data of the wireless sensor network includes:

[0072] A time-series data acquisition plan is performed on the optimized network topology to obtain a node acquisition time-series table for the optimized network topology;

[0073] Distributed acquisition is triggered on the node acquisition time sequence table to obtain the original data packets of the node acquisition time sequence table;

[0074] The original data packets are fused to obtain the mining pressure data of the wireless sensor network.

[0075] Specifically, the mine environment is first divided into multiple fixed areas, and a preset number of wireless sensor nodes are deployed in each area. Then, each node sends a signal to the surrounding nodes and records the information of the nodes that can receive the signal. Based on the signal strength and distance between nodes, invalid connections with weak signals and long distances are eliminated, and valid connections with stable signals and appropriate distances are retained. Finally, the network structure is constructed according to the valid connection relationships to obtain the optimized network topology of the wireless sensor network.

[0076] Furthermore, based on the optimized network topology, a node is designated as the coordinating node. The coordinating node sends a synchronous acquisition command to all sensor nodes. After receiving the command, each sensor node simultaneously starts the mine pressure data acquisition module to collect the mine pressure data at its location. After the acquisition is completed, each node sends the data to the coordinating node in a preset order. The coordinating node receives the mine pressure data from all nodes and integrates it to obtain the mine pressure data of the wireless sensor network.

[0077] Specifically, a wireless sensor node is selected as the initial routing initiating node. This node sends a route probing signal to all surrounding sensor nodes. After receiving the signal, the surrounding nodes record the location information of the initiating node and forward the probing signal to the surrounding nodes that have not received the signal before, until all nodes have received the signal. Each node organizes the information of the source node of the received signal in the order of receipt to form an initial routing table for wireless sensor nodes containing the communication paths from itself to other nodes.

[0078] Furthermore, extract the two nodes corresponding to each communication link from the initial routing table, and have these two nodes send test data packets of a fixed size to each other for a preset time. Calculate the ratio of the number of successfully received data packets to the total number of data packets sent during this period, as well as the average time from data packet transmission to reception. Use these two values ​​as the link quality parameters of the communication link to obtain the link quality parameters of the initial routing table.

[0079] Furthermore, based on the link quality parameters, high-quality links with a high ratio of successfully received data packets and a short average time are selected. At the same time, it is checked whether all nodes are connected through high-quality links. If there are unconnected nodes, nodes around them that can establish high-quality links are searched and connections are added. All non-high-quality links are removed. The resulting network structure is the optimized network topology of the wireless sensor network.

[0080] Specifically, based on optimizing the network topology, the number and location distribution of all wireless sensor nodes are counted, adjacent nodes are divided into the same group, and each group is assigned a fixed collection time period that does not overlap. At the same time, the specific collection time and duration of each node within the collection time period of its group are set. This information is organized in order of node number to form a node collection time sequence table for optimizing the network topology.

[0081] Furthermore, the coordinating node reads the node acquisition timing table and sends a wireless signal containing acquisition instructions to the node at the corresponding acquisition time. After receiving the signal, the node immediately starts the mine pressure detection module and records the mine pressure change data according to the set duration. After the acquisition is completed, the data is combined with the node number and acquisition time information to generate a raw data packet and temporarily stored.

[0082] Furthermore, the coordinating node receives the raw data packets from all nodes, first checks whether the node number and acquisition time of each data packet are consistent with the node acquisition time sequence table, removes inconsistent data packets, then compares the mine pressure data of adjacent nodes in the same time period, retains the data with deviation within a reasonable range, calculates the average value of these data, and organizes all average values ​​according to time order and location distribution to obtain the mine pressure data of the wireless sensor network.

[0083] In summary, by building an initial routing table through dynamic route discovery, selecting high-quality links through link quality assessment, and then optimizing the topology, invalid connections with weak signals and high transmission delays in the complex environment of mines can be eliminated, while stable links are retained. This solves the problems of complex wiring and susceptibility to damage from humid / corrosive environments in traditional wired sensors, making wireless sensor networks adaptable to the mine environment, reducing data acquisition interruptions caused by link failures, and improving network stability.

[0084] In summary, timing planning avoids node acquisition conflicts, distributed triggering ensures synchronous acquisition by multiple nodes, and reduces data time deviation; data fusion integrates raw data from multiple nodes, removes anomalies and fills in missing data, improving the accuracy and completeness of mining pressure data, and avoiding the problems of fragmented and large errors in traditional data acquisition.

[0085] In summary, it can provide reliable data support for subsequent encrypted transmission, decryption analysis, and risk assessment of mine pressure data, thereby improving the effectiveness and efficiency of mine pressure monitoring from the source of data collection.

[0086] S2. The encrypted data of the mining pressure data is transmitted to the local data processing location via wireless communication.

[0087] In this embodiment of the invention, transmitting the encrypted data of the mine pressure data to the local data processing location via wireless communication includes:

[0088] The mine pressure data is encrypted to obtain an encrypted data stream of the mine pressure data;

[0089] The encrypted data stream is encapsulated into data packets, and the encrypted data stream is transmitted via data packets;

[0090] The data packets are transmitted in a distributed manner through multiple communication sub-channels, and then sent to the local data processing location.

[0091] Specifically, after acquiring the mining pressure data, the data is divided into several data blocks of a fixed length, each data block is given a unique serial number, and each data block is then converted using a preset encryption key to make the content of the data block into a character sequence that cannot be directly recognized. Finally, all the encrypted data blocks are concatenated in the order of the serial numbers to form an encrypted data stream of mining pressure data.

[0092] Furthermore, in accordance with the format specified in the wireless communication protocol, a start identifier and data length information are added before the encrypted data stream, and a checksum is added after the encrypted data stream. The checksum is obtained by performing calculations on all characters in the encrypted data stream according to fixed rules. The overall data structure after adding the information is determined as the transmission data packet of the encrypted data stream.

[0093] Furthermore, the transmitted data packet is divided into multiple data segments of equal length, and each segment is marked with a corresponding position number. At the same time, multiple communication sub-channels in the wireless sensor network are activated, each sub-channel corresponding to a different communication frequency. Each data segment is sent through a different sub-channel. After receiving the data segments transmitted by all sub-channels, the local data processing location reassembles them according to the position number to restore the complete transmitted data packet.

[0094] In summary, traditional mine pressure monitoring systems lack effective encryption mechanisms, making data susceptible to leakage and tampering. This method transforms mine pressure data into an encrypted data stream through encryption processing (such as the AES 256-bit encryption algorithm mentioned in the technical disclosure), blocking security risks in data transmission from the source, ensuring the confidentiality of sensitive mine monitoring data, and meeting the needs of secure communication.

[0095] In summary, by encapsulating encrypted data streams with data packets (adding start identifiers, checksums, etc.), the integrity of the data can be verified after transmission using the checksum. This avoids data loss or damage caused by signal interference in the complex environment of mines, solves the problem of unstable data transmission in traditional wired transmission, and provides a complete data foundation for subsequent local decryption and analysis.

[0096] In summary, the use of multiple communication sub-channels for distributed transmission avoids the problems of single-channel congestion or the impact of the mine environment (such as humidity and electromagnetic interference), speeds up transmission, ensures that encrypted data is delivered to the local processing location efficiently, avoids the delay of traditional centralized transmission, and creates conditions for subsequent real-time data processing and rapid assessment of mine pressure risks.

[0097] S3. Decrypt and analyze the encrypted data at the local data processing location to obtain the analysis results of the encrypted data.

[0098] In this embodiment of the invention, the step of decrypting and analyzing the encrypted data at the local data processing location to obtain the analysis result of the encrypted data includes:

[0099] The encrypted data is decrypted to obtain the decrypted data stream of the encrypted data;

[0100] Perform data integrity verification on the decrypted data stream to obtain verification passed data for the decrypted data stream;

[0101] The data quality of the verified data is evaluated to obtain the data quality evaluation results of the verified data.

[0102] Based on the data quality assessment results, the verification data is reconstructed to obtain the analysis results of the encrypted data.

[0103] The data reconstruction process uses the following reconstruction formula for calculation:

[0104]

[0105] In the formula, This represents the reconstructed data vector. This represents the data vector that passed the verification. This represents the statistical feature vector of the data vector that passed the verification. This represents the original data weighting coefficients that are dynamically adjusted based on the data quality assessment results. This represents the statistical feature weighting coefficients that are dynamically adjusted based on the data quality assessment results.

[0106] Specifically, after obtaining the encrypted data, the segmentation markers in the data are identified, and the encrypted data is split into multiple encrypted data blocks. Each encrypted data block is reverse-converted using the key corresponding to the encryption, so that each data block is restored to the original recognizable character sequence. Then, the data blocks are rearranged and concatenated according to their unique sequence numbers to form the decrypted data stream of the encrypted data.

[0107] Furthermore, the verification information in the decrypted data stream is extracted, and all data characters in the decrypted data stream are recalculated according to the rules agreed upon before encryption to obtain a new verification result. The new verification result is compared with the extracted verification information. If the two are completely consistent, the decrypted data stream is determined to be complete and valid, and it is identified as the verified data of the decrypted data stream.

[0108] Furthermore, the total amount of mine pressure data contained in the statistical verification data is checked to see if there is any missing time point or location information. The proportion of missing data in the total data volume is calculated. At the same time, the mine pressure data at adjacent time points are compared, and the number of numerical abrupt changes is recorded. Based on whether the missing proportion and the number of abrupt changes are within the preset range, a data quality assessment result of the verified data is formed.

[0109] Furthermore, if the data quality assessment results show that the validated data is without missing data and without abnormal mutations, it is directly used as the analysis result; if there is a small amount of missing data, the average value of the adjacent valid data is used to fill the missing positions; if there is abnormal mutation data, the mutation value is replaced with the smooth transition value of the adjacent data. The complete data set formed after processing is the analysis result of the encrypted data.

[0110] Specifically, The verified data is derived from the data integrity verification of the decrypted data stream. The verified data is organized into an ordered vector form according to the collection time sequence and node number, which is the verified data vector.

[0111] Furthermore, It is obtained by statistical analysis of the verified data vector. The specific process is as follows: calculate the average value of all data in the verified data vector, count the difference in change between adjacent data, extract the maximum and minimum values ​​in the data, and arrange these statistically obtained values ​​in a preset order to form the statistical feature vector of the verified data vector.

[0112] Furthermore, and The source is the data quality assessment results. The specific method for determining this is as follows: first, check the percentage of missing data and the number of abnormal mutations in the data quality assessment results. If the percentage of missing data is low and the number of mutations is few, it indicates high data quality. In this case, [the data quality will be assessed]. Set to a larger value. Set it to a smaller value; if the missing percentage is high or the number of mutations is high, it indicates low data quality, and in this case, set it to a smaller value. Set to a smaller value. Set to a larger value, and and The sum of their values ​​is always 1.

[0113] Furthermore, the meaning of this formula is that it multiplies the validated data vector and its corresponding statistical feature vector by weighting coefficients determined based on the data quality assessment results. and Then, the two product results are added together to obtain a reconstructed data vector that can take into account both the validity of the original data and the rationality of the statistical characteristics. This enables the optimization and adjustment of the verified data, and makes up for the impact of possible minor anomalies or missing data in the original data.

[0114] Furthermore, the formula shows a trend where: the higher the data quality, the better. The larger the value of , the better. The smaller the value of , the greater the weight of the validated data vector in the calculation, the smaller the difference between the reconstructed data vector and the validated data vector, and the closer it is to the original validated data; when the data quality assessment results show that the data quality is low, The smaller the value of , the better. The larger the value of , the greater the weight of the statistical feature vector in the calculation, the smaller the difference between the reconstructed data vector and the statistical feature vector, and the more it relies on statistical features to correct the original data.

[0115] In summary, by first decrypting the encrypted data to restore it into an analyzable decrypted data stream, and then performing integrity verification (such as checksum comparison), data damage or tampering caused by interference from the mining environment (such as electromagnetic or humid environments) during transmission can be eliminated. This solves the problems of easy data loss and lack of authenticity in traditional systems, providing a reliable data foundation for subsequent analysis.

[0116] In summary, data quality assessment can identify and verify issues such as missing data and abrupt changes, avoiding the direct use of low-quality data that could negatively impact analysis results; this, combined with the dynamic weighting reconstruction formula ( and Adjustments based on quality), when data quality is high, emphasis is placed on the original data ( When the quality is low, rely on statistical characteristics ( This feature completes and corrects data, effectively compensating for data flaws and making the final analysis results more accurate, thus avoiding misjudgments caused by poor data quality in traditional data analysis.

[0117] In summary, the local processing location completes the decryption and analysis without transmitting the data to a remote central control room, reducing the transmission delay of traditional remote processing. Furthermore, the reconstructed high-quality data can be directly used for subsequent mine pressure risk level assessment, laying the foundation for rapid identification of mine pressure anomalies and timely early warning, and improving the real-time performance and efficiency of mine pressure monitoring.

[0118] S4. Perform anomaly analysis on the analysis results to obtain the mine pressure risk level of the analysis results.

[0119] In this embodiment of the invention, the step of performing mine pressure anomaly analysis on the analysis results to obtain the mine pressure risk level of the analysis results includes:

[0120] Multi-dimensional feature extraction is performed on the analysis results to obtain the mine pressure feature vector;

[0121] Anomaly pattern recognition is performed on the aforementioned mine pressure feature vector to obtain anomaly pattern recognition results;

[0122] The risk level of the anomaly pattern identification results is assessed by a risk probability calculation model to obtain the mine pressure risk level of the analysis results.

[0123] The risk probability calculation model is as follows:

[0124]

[0125] In the formula, Indicates the probability of risk. The first element in the ore pressure feature vector represents the... Each feature component Indicates the relationship with the first Preset weight coefficients corresponding to each feature component This represents the total number of characteristic components.

[0126] The step of assessing the risk level of the abnormal pattern recognition results using a risk probability calculation model to obtain the mine pressure risk level of the analysis results includes:

[0127] The abnormal pattern recognition result is subjected to pattern classification processing to obtain the abnormal pattern classification result of the abnormal pattern recognition result;

[0128] The risk of the abnormal pattern classification result is determined to obtain the risk probability of the abnormal pattern classification result.

[0129] The risk probability is mapped to a risk level to obtain the mine pressure risk level of the analysis results.

[0130] Specifically, features in three dimensions—time, space, and trend—are extracted from the analysis results. In the time dimension, the mine pressure value at each moment is extracted at fixed intervals. In the spatial dimension, the average mine pressure value of each region is calculated according to the distribution area of ​​sensor nodes. In the trend dimension, the difference between the mine pressure values ​​at two consecutive moments is calculated. These extracted features are arranged in the order of time, space, and trend to form a mine pressure feature vector.

[0131] Furthermore, the mine pressure feature vector is compared with the preset normal mine pressure feature pattern. The normal pattern includes the time feature range, spatial feature range and trend feature range when there are no anomalies in history. Each feature in the mine pressure feature vector is checked one by one to see if it is within the corresponding normal range. If a feature exceeds the range, it is marked as an anomaly. The number of abnormal features and their dimensions are counted and summarized to form the anomaly pattern recognition result.

[0132] Furthermore, the risk probability calculation model presets basic risk values ​​corresponding to different anomaly patterns. Based on the number of abnormal features in the anomaly pattern identification results, a fixed value is added to the basic risk value for each additional abnormal feature. At the same time, depending on the extent to which the abnormal feature exceeds the normal range, the larger the extent, the more additional value is added. The final value is compared with the preset risk level thresholds (low, medium, and high). The threshold range in which the value falls corresponds to the mine pressure risk level, which is the mine pressure risk level of the analysis result.

[0133] Specifically, the abnormal feature information contained in the abnormal pattern recognition results is first sorted out, including the dimension in which the abnormality exists (time dimension, spatial dimension, and trend dimension) and the number of abnormal features in each dimension. Three types of abnormal pattern classification standards are preset: single-dimensional single abnormality (only one abnormal feature exists in one dimension), single-dimensional multiple abnormality (only two or more abnormal features exist in one dimension), and multi-dimensional abnormality (abnormal features exist in two or more dimensions). The number of abnormal dimensions and the number of abnormalities in each dimension in the abnormal pattern recognition results are compared with the classification standards one by one. The abnormal pattern is classified into the corresponding category according to the standard, thus obtaining the abnormal pattern classification result of the abnormal pattern recognition results.

[0134] Furthermore, a fixed correspondence table between the abnormal pattern classification results and the risk probability is preset, in which a single-dimensional single abnormality corresponds to a risk probability of 30%, a single-dimensional multiple abnormality corresponds to a risk probability of 50%, and a multi-dimensional abnormality corresponds to a risk probability of 80%. Based on the obtained abnormal pattern classification results, the risk probability value matching the classification is found in the correspondence table and directly determined as the risk probability of the abnormal pattern classification result.

[0135] Furthermore, a preset risk level mapping rule is established: a risk probability of 30% corresponds to a low risk level, 50% corresponds to a medium risk level, and 80% corresponds to a high risk level. The determined risk probability is matched with this rule to find the risk level corresponding to the risk probability. This risk level is the mine pressure risk level in the analysis result.

[0136] Specifically, Derived from the mine pressure feature vector obtained after multi-dimensional feature extraction of the analysis results, each independent feature data is selected sequentially from the mine pressure feature vector in the extraction order, and each feature data is the first element in the mine pressure feature vector. Each feature component.

[0137] Specifically, These are fixed coefficients pre-set based on the importance of the characteristic components of the mining pressure. For characteristic components that have a high degree of influence on mining pressure anomalies, larger coefficients are set. For characteristic components with low impact on mine pressure anomalies, a smaller value is set. And each All are corresponding to the first eigenvector in the ore pressure feature vector. Each feature component corresponds one-to-one.

[0138] Specifically, It is the total number obtained by counting all feature components in the mine pressure feature vector, that is, how many independent feature data are contained in the mine pressure feature vector. Just for how much.

[0139] Furthermore, this calculation process involves taking each of the following elements from the mine pressure feature vector: Each feature component and its corresponding preset weight coefficient By multiplying them separately, we obtain the weighted value corresponding to each feature component. Then, we sum up the weighted values ​​of all feature components. The sum obtained is the risk probability. In this way, different feature components affecting mine pressure risk can be assigned different weights according to their importance, thus obtaining a risk probability that is more in line with the actual situation.

[0140] Furthermore, when all preset weight coefficients When fixed, one or more of the first elements in the ore pressure feature vector are... The larger the value of a feature component, the closer that feature component is to its corresponding... The larger the weighted value obtained by multiplication, the higher the probability of risk after accumulating all weighted values; when all the values ​​in the mining pressure feature vector are... When the values ​​of each feature component are fixed, the preset weight coefficients are used. The larger the feature component, the more significant its numerical change has on the risk probability. Overall, as the overall value of the feature components in the mining pressure feature vector increases, the risk probability will gradually increase, and vice versa.

[0141] In summary, by extracting features from multiple dimensions (such as time, space, and trend dimensions) to construct mine pressure feature vectors, compared with traditional single-dimensional data analysis, it can more comprehensively cover mine pressure change information and avoid anomaly omissions caused by incomplete information. Furthermore, by comparing with preset normal patterns to achieve anomaly pattern recognition, it can accurately locate abnormal features beyond the range, solve the problems of traditional manual judgment relying on experience, being prone to errors, and being lagging behind, and improve the accuracy of anomaly identification.

[0142] In summary, the risk probability calculation model uses preset weighting coefficients (set according to the degree of influence of features on mine pressure safety) to weight and sum the feature components, avoiding the problem of traditional extensive risk assessment ignoring the differences in the importance of features, and making the risk probability calculation more in line with the actual mine safety needs. Combined with pattern classification (such as single-dimensional single anomaly, multi-dimensional anomaly) to determine the risk probability, the risk assessment basis is further refined, making the results more valuable for reference.

[0143] In summary, mapping risk probabilities to specific mine pressure risk levels (low, medium, and high) allows staff to quickly and intuitively grasp the degree of mine pressure risk, avoiding the problems of traditional early warning systems lacking clear levels and chaotic response measures. This provides a clear basis for matching personalized early warning strategies. Furthermore, the entire process is completed locally, without relying on a remote central control room, solving the delay problem of traditional remote analysis, improving the real-time output of risk levels, and buying time for mine safety control.

[0144] S5. Match the early warning strategy to the mine pressure risk level to obtain the graded early warning information of the mine environment.

[0145] In this embodiment of the invention, the step of matching the early warning strategy for the mine pressure risk level to obtain the graded early warning information for the mine pressure risk level includes:

[0146] The mine pressure risk level is queried from the early warning strategy database to obtain the initial early warning strategy for the mine environment;

[0147] The initial early warning strategy is configured with strategy parameters to obtain a personalized early warning strategy for the mine environment;

[0148] The personalized early warning strategy is converted into early warning information to obtain the graded early warning information of the mine environment.

[0149] The process of converting the personalized early warning strategy into graded early warning information for the mine environment includes:

[0150] The initial warning content of the personalized warning strategy is adapted to channel characteristics to obtain the warning content items of the mine environment;

[0151] The warning content items are sorted by information priority to obtain the sorted warning content of the mine environment;

[0152] The sorted warning content is encapsulated in a warning format to obtain a standardized warning data packet of the mine environment;

[0153] The standardized early warning data packets are distributed through distribution channels to obtain the graded early warning information for the mine environment.

[0154] Specifically, after obtaining the mine pressure risk level, a preset early warning strategy library is invoked. This strategy library stores corresponding early warning rules according to low, medium and high risk levels. The rules for each level include early warning methods, coverage, response requirements and other content. The matching rule is searched in the library according to the mine pressure risk level. The found rule is the initial early warning strategy for the mine environment.

[0155] Furthermore, specific information about the mine is collected, including the location of underground workers, the distribution area of ​​major equipment, and real-time communication status. Based on this information, the parameters in the initial early warning strategy are adjusted, such as limiting the early warning coverage to areas where personnel are actually working and adjusting the early warning frequency to an interval that is compatible with the current communication load. The adjusted strategy is the personalized early warning strategy for the mine environment.

[0156] Furthermore, the analysis of the early warning channels involved in the personalized early warning strategy, including the underground broadcasting system, the surface monitoring center display screen, and the mobile terminals of management personnel, clarifies the information carrying characteristics of each channel. For example, the broadcasting system is suitable for brief voice information, while the mobile terminal is suitable for information combining text and images. Based on these characteristics, the expression form and length of the initial early warning content are adjusted so that it can be effectively transmitted in the corresponding channel. The processed content is the early warning content item for the mine environment.

[0157] Furthermore, priorities are set according to the urgency of the warning items. Items directly related to personnel safety are set to the highest priority, those related to equipment operation safety are set to the medium priority, and those that only reflect fluctuations in environmental parameters are set to the low priority. All warning items are arranged in descending order, and the resulting sequence is the sorted warning content of the mine environment.

[0158] Furthermore, the sorted warning content is organized according to a preset standard format, which includes fixed fields: warning level, release time, affected area, core content, and response suggestions. The sorted content is then filled into the corresponding fields in sequence to ensure that the information in each field is complete and the format is consistent. The resulting structured data set is the standardized warning data package for the mine environment.

[0159] Furthermore, the distribution channels are determined according to the mine pressure risk level. Standardized early warning data packets for high-risk levels are simultaneously distributed to the underground broadcasting system, the ground monitoring center display screen, and all management personnel's mobile terminals. Data packets for medium-risk levels are distributed to the ground monitoring center display screen and management personnel's mobile terminals. Data packets for low-risk levels are distributed only to management personnel's mobile terminals. The early warning information formed after the distribution is completed and adapted to each channel is the graded early warning information for the mine environment.

[0160] In summary, the initial strategy is first obtained by querying the early warning strategy library based on the mine pressure risk level. Then, personalized parameters are configured in combination with the actual mine scenario (such as the location of workers, equipment distribution, and communication status). This avoids the problem of fixed early warning strategies in traditional systems that are out of touch with on-site needs, making the early warning strategy more in line with the real-time situation of the mine and providing a foundation for accurate early warning.

[0161] In summary, the channel characteristic adaptation adjusts the form of early warning content according to the carrying characteristics of different channels such as underground broadcasts and mobile terminals (e.g., short voice for broadcasts and text and graphics for terminals), avoiding the transmission failure of traditional information due to channel incompatibility; the information priority sorting puts personnel safety-related content first to prevent information confusion, help staff quickly capture key risks, and improve response efficiency.

[0162] In summary, standardized encapsulation unifies the warning format, making it easier for different terminals to parse and reducing format compatibility issues. Distributing channels according to risk level (such as multi-channel coverage for high-risk and targeted distribution for low-risk) ensures comprehensive emergency warnings while avoiding resource waste. It also leverages the real-time advantages of local processing, solves the delay problem of traditional remote warnings, and buys time for mine safety control.

[0163] In the several embodiments provided by this invention, it should be understood that the disclosed method can be implemented in other ways.

[0164] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0165] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, and technology that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis, characterized in that, The method includes: S1. Collect mine pressure data in the mine environment through a wireless sensor network; S2. The encrypted data of the mine pressure data is transmitted to the local data processing location via wireless communication; S3. Decrypt and analyze the encrypted data at the local data processing location to obtain the analysis result of the encrypted data; S4. Perform anomaly analysis on the analysis results to obtain the mine pressure risk level of the analysis results; S5. Match the early warning strategy to the mine pressure risk level to obtain the graded early warning information of the mine environment.

2. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 1, characterized in that, The method of collecting mine pressure data in the mine environment via a wireless sensor network includes: The wireless sensor nodes deployed in the mine environment are networked to obtain the optimized network topology of the wireless sensor network; Coordinated data acquisition is performed on the optimized network topology to obtain the mine pressure data of the wireless sensor network.

3. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 2, characterized in that, The process of networking the wireless sensor nodes deployed in the mine environment to obtain the optimized network topology of the wireless sensor network includes: Dynamic route discovery is performed on the wireless sensor nodes to obtain the initial routing table of the wireless sensor nodes; The initial routing table is evaluated for communication link quality to obtain the link quality parameters of the initial routing table; The link quality parameters and the wireless sensor nodes are optimized to obtain the optimized network topology of the wireless sensor network.

4. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 3, characterized in that, The step of collaboratively acquiring data from the optimized network topology to obtain the mining pressure data of the wireless sensor network includes: A time-series data acquisition plan is performed on the optimized network topology to obtain a node acquisition time-series table for the optimized network topology; Distributed acquisition is triggered on the node acquisition time sequence table to obtain the original data packets of the node acquisition time sequence table; The original data packets are fused to obtain the mining pressure data of the wireless sensor network.

5. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 1, characterized in that, The step of transmitting the encrypted data of the mine pressure data to the local data processing location via wireless communication includes: The mine pressure data is encrypted to obtain an encrypted data stream of the mine pressure data; The encrypted data stream is encapsulated into data packets, and the encrypted data stream is transmitted via data packets; The data packets are transmitted in a distributed manner through multiple communication sub-channels, and then sent to the local data processing location.

6. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 1, characterized in that, The step of decrypting and analyzing the encrypted data at the local data processing location to obtain the analysis result of the encrypted data includes: The encrypted data is decrypted to obtain the decrypted data stream of the encrypted data; Perform data integrity verification on the decrypted data stream to obtain verification passed data for the decrypted data stream; The data quality of the verified data is evaluated to obtain the data quality evaluation results of the verified data. Based on the data quality assessment results, the verification data is reconstructed to obtain the analysis results of the encrypted data. The data reconstruction process uses the following reconstruction formula for calculation: In the formula, This represents the reconstructed data vector. This represents the data vector that passed the verification. This represents the statistical feature vector of the data vector that passed the verification. This represents the original data weighting coefficients that are dynamically adjusted based on the data quality assessment results. This represents the statistical feature weighting coefficients that are dynamically adjusted based on the data quality assessment results.

7. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 1, characterized in that, The step of performing mine pressure anomaly analysis on the analysis results to obtain the mine pressure risk level of the analysis results includes: Multi-dimensional feature extraction is performed on the analysis results to obtain the mine pressure feature vector; Anomaly pattern recognition is performed on the aforementioned mine pressure feature vector to obtain anomaly pattern recognition results; The risk level of the anomaly pattern identification results is assessed by a risk probability calculation model to obtain the mine pressure risk level of the analysis results. The risk probability calculation model is as follows: In the formula, Indicates the probability of risk. The first element in the ore pressure feature vector represents the... Each feature component Indicates the relationship with the first Preset weight coefficients corresponding to each feature component This represents the total number of characteristic components.

8. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 7, characterized in that, The step of assessing the risk level of the abnormal pattern recognition results using a risk probability calculation model to obtain the mine pressure risk level of the analysis results includes: The abnormal pattern recognition result is subjected to pattern classification processing to obtain the abnormal pattern classification result of the abnormal pattern recognition result; The risk of the abnormal pattern classification result is determined to obtain the risk probability of the abnormal pattern classification result. The risk probability is mapped to a risk level to obtain the mine pressure risk level of the analysis results.

9. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 1, characterized in that, The process of matching early warning strategies for the mine pressure risk level to obtain graded early warning information for the mine pressure risk level includes: The mine pressure risk level is queried from the early warning strategy database to obtain the initial early warning strategy for the mine environment; The initial early warning strategy is configured with strategy parameters to obtain a personalized early warning strategy for the mine environment; The personalized early warning strategy is converted into early warning information to obtain the graded early warning information of the mine environment.

10. The novel mine pressure monitoring method based on wireless sensor networks and intelligent data analysis as described in claim 9, characterized in that, The process of converting the personalized early warning strategy into graded early warning information for the mine environment includes: The initial warning content of the personalized warning strategy is adapted to channel characteristics to obtain the warning content items of the mine environment; The warning content items are sorted by information priority to obtain the sorted warning content of the mine environment; The sorted warning content is encapsulated in a warning format to obtain a standardized warning data packet of the mine environment; The standardized early warning data packets are distributed through distribution channels to obtain the graded early warning information for the mine environment.

Images