A mine tunnel wireless link blind area detection method based on cooperative broadcasting

Abstract

The application discloses a mine tunnel wireless link blind area detection method based on cooperative broadcasting, and relates to the technical field of mine tunnel monitoring. The specific scheme is as follows: a wireless transceiver device transmits and receives distance measurement messages to measure the adjacent distance, accumulates the longitudinal coordinates section by section, divides virtual sections, extracts wireless propagation statistical features based on a received signal strength sequence, matches the features with a multipath propagation feature fingerprint library to obtain a propagation cause pre-classification result, determines a potential transmission blind area according to the result, controls associated wireless transceiver devices to transmit at a preset transmission power level and construct a link response curve, substitutes wireless propagation statistical features corresponding to a critical transmission power level into a mine tunnel waveguide propagation model to solve a propagation attenuation distance, superimposes the longitudinal coordinates of a transmission end to obtain blind area boundary coordinates, and outputs a blind area detection result after verification. Through the application, intelligent identification and accurate positioning of a blind area are realized, and the connectivity and monitoring continuity of a mine wireless sensor network are ensured.

Description

Technical Field

[0001] This application relates to the field of mine roadway monitoring technology, and more specifically, to a method for detecting blind spots in mine roadway wireless links based on cooperative broadcasting. Background Technology

[0002] In the field of mine roadway environmental monitoring, wireless sensor networks are often deployed in mobile work areas such as longwall faces and tunneling heads to monitor gas concentration, temperature, humidity, and equipment status. The roadway length, cross-section, and internal facilities (such as large mining equipment, temporary supports, and material storage points) in these areas are constantly changing, requiring frequent adjustments to the monitoring network topology as operations progress. This results in the temporary and mobile nature of wireless transceiver deployment. These transceivers are typically powered by built-in batteries, and their installation location within the roadway is limited by the roadway wall conditions. Deployment spacing and positional accuracy cannot be perfectly aligned with ideal plans, leading to communication blind spots in certain areas due to excessive spacing between transceivers, obstructions, or abrupt changes in roadway structure. Reliably detecting these wireless communication blind spots, which affect the continuity of data acquisition, in such complex and dynamically changing environments is a key requirement for ensuring the effective operation of mine safety monitoring systems.

[0003] The common approach in existing technologies is to pre-plan and permanently deploy several wireless sensor nodes within the tunnel, maintaining network connectivity through periodic communication between the nodes. When communication is interrupted in a certain area due to obstruction or excessive distance, the central node infers the existence of blind spots by monitoring changes in the network topology (such as neighbor list updates or routing table anomalies). This method primarily relies on a pre-fixed, stable network structure and a connectivity maintenance mechanism between nodes.

[0004] However, existing technologies have significant shortcomings in temporary operational scenarios where nodes need to be deployed flexibly and their spacing is not fixed. First, methods relying on topology changes for detection suffer from unstable signal propagation when node spacing is large or at abrupt structural changes in the roadway (such as intersections or bends) with severe multipath interference. This leads to false alarms or missed alarms based on connectivity status, and makes it difficult to accurately distinguish between communication quality degradation caused by environmental fading and genuine communication outages. Second, these methods cannot precisely quantify the specific location and boundary range of blind zones in the longitudinal direction of the roadway, only providing a vague conclusion that "blind zones exist," which is detrimental to subsequent targeted compensation or network optimization. Therefore, there is an urgent need for a detection method that can adapt to dynamic node deployment, accurately locate blind zone boundaries, and identify the causes of blind zones to improve the reliability of temporary monitoring networks in mines. Summary of the Invention

[0005] In view of this, the purpose of this application is to provide a method for detecting blind spots in mine roadway wireless links based on cooperative broadcasting, which can realize intelligent identification and accurate positioning of various blind spot causes, and effectively ensure the connectivity and monitoring continuity of the mine wireless sensor network.

[0006] According to a first aspect of this application, a method for detecting blind spots in wireless links in mine roadways based on cooperative broadcasting is provided, comprising: Several transceiver devices are deployed in the mine roadway. The wireless transceiver devices send ranging messages to each other to measure the distance between adjacent wireless transceiver devices. Taking the wireless transceiver device at the entrance of the roadway as the origin, the distance is accumulated segment by segment according to the arrangement order along the longitudinal direction of the roadway to obtain the longitudinal coordinate of each wireless transceiver device. The roadway space between two wireless transceiver devices with adjacent longitudinal coordinates is defined as a virtual segment. Each wireless transceiver receives probe messages transmitted by adjacent wireless transceivers, forms a sequence of received signal strength, and extracts variance, skewness, kurtosis and fading distribution type based on a preset sliding time window to obtain the statistical characteristics of wireless propagation. The wireless propagation statistical features are matched with the pre-established multipath propagation feature fingerprint database to obtain the propagation cause pre-classification results of each wireless transceiver. When the propagation cause pre-classification results of the wireless transceivers at both ends of the virtual segment are both abnormal propagation causes, or the distance between the two wireless transceivers is greater than the preset transmission distance threshold, the virtual segment is determined to be a potential transmission blind zone. For potential transmission blind spots, control the wireless transceiver devices associated with the potential transmission blind spot to transmit detection messages at a preset transmission power level. The receiving end counts the reception success rate at different transmission power levels and constructs a link response curve between the transmission power level and the average reception success rate. The critical transmit power level is determined based on the link response curve. The wireless propagation statistical characteristics corresponding to the critical transmit power level are substituted into the waveguide propagation model of the mine roadway to calculate the propagation attenuation distance of the detection message. The coordinates of the transmission blind zone boundary are obtained by combining the longitudinal coordinates of the transmitter. When the boundary coordinates of the transmission blind zone meet the preset physical constraints corresponding to the pre-classification results of the propagation cause, the output includes the boundary coordinates of the transmission blind zone and the cause of the blind zone.

[0007] In some embodiments, the distance between adjacent wireless transceivers is measured based on the propagation time of the ranging message, including: each wireless transceiver sends a ranging message to an adjacent wireless transceiver and records the sending time; the adjacent wireless transceivers return an acknowledgment message after a preset acknowledgment delay; the wireless transceiver that sent the ranging message records the time of receiving the acknowledgment message; the round-trip propagation time of the ranging message is calculated based on the sending time, the receiving time, and the preset acknowledgment delay; and the propagation time after deducting the acknowledgment delay is halved and multiplied by the electromagnetic wave propagation speed to obtain the distance between adjacent wireless transceivers. The distance between adjacent wireless transceivers is measured multiple times. When the deviation of a certain measurement result from the average of the other measurement results exceeds the preset ranging deviation threshold, the measurement result is discarded, and the average of the remaining measurement results is taken as the corresponding adjacent distance. The longitudinal coordinates of each wireless transceiver are obtained by accumulating the distances between adjacent devices along the longitudinal direction of the roadway.

[0008] In some embodiments, the fading distribution type is determined by fitting the received signal strength sequence with the Rayleigh distribution model and the Rice distribution model respectively, obtaining the corresponding fitting errors, and determining the distribution model with the smaller fitting error as the fading distribution type corresponding to the received signal strength sequence.

[0009] In some embodiments, the multipath propagation feature fingerprint database is established by: simulating multiple typical propagation conditions in a mine roadway, collecting the received signal strength sequence under each typical propagation condition and extracting the corresponding wireless propagation statistical features, and associating and storing the propagation causes of each typical propagation condition with the corresponding wireless propagation statistical features to form the multipath propagation feature fingerprint database. The similarity between the wireless propagation statistical features to be matched and each fingerprint pattern in the multipath propagation feature fingerprint database is calculated, and the propagation cause corresponding to the fingerprint pattern with the highest similarity is used as the propagation cause pre-classification result of the corresponding wireless transceiver. The causes of the various typical propagation conditions include normal propagation causes and abnormal propagation causes. Abnormal propagation causes represent the types of causes for which objectively non-existent propagation conditions do not exist.

[0010] In some embodiments, constructing the link response curve includes: selecting the two ends of the wireless transceiver device of the virtual segment under test where the potential transmission blind zone is located, and the adjacent wireless transceiver device located outside the two ends of the wireless transceiver device as the measurement association device; The wireless transceiver devices at both ends of the virtual segment under test are used as transmitters in sequence, and the remaining measurement and association devices are used as receivers. The transmitting end increases the transmission power step by step according to the preset transmission power level and transmits detection messages. The receiving end records the reception success rate at each transmission power level and extracts the corresponding wireless propagation statistical features. The average reception success rate of each receiver is used as the average reception success rate under the corresponding transmission power level, and the link response curve is plotted with the transmission power level as the horizontal axis and the average reception success rate as the vertical axis.

[0011] In some embodiments, when the transmitter increases its transmission power step by step according to a preset transmission power level and transmits a detection message, the preset transmission power levels are arranged in order from low to high, and the transmitter increases its transmission power step by step from the lowest transmission power level. At each transmission power level, the receiver counts the number of probe messages successfully received within a preset reception time, and uses the ratio of the number of successfully received probe messages to the total number of probe messages transmitted at that transmission power level as the reception success rate of the receiver at that transmission power level.

[0012] In some embodiments, substituting the wireless propagation statistical characteristics corresponding to the critical transmit power level into the mine tunnel waveguide propagation model to calculate the propagation attenuation distance of the probe message includes: setting the mine tunnel waveguide propagation model as a path loss model characterizing the attenuation law of the received signal strength with the propagation distance, and making the model parameters of the path loss model correspond to different fading distribution types. Based on the fading distribution type in the wireless propagation statistical characteristics corresponding to the critical transmit power level, select the corresponding model parameters; Substituting the transmission power corresponding to the critical transmission power level and the preset receiving sensitivity threshold into the path loss model, the propagation attenuation distance of the probe message is obtained.

[0013] In some embodiments, the preset physical constraints are set as follows: based on the preset roadway facility distribution information, cause reference coordinates are set for each abnormal propagation cause, wherein the cause reference coordinates for the cause of metal equipment obstruction are the coordinates of the metal equipment placement location, the cause reference coordinates for the cause of roadway intersection are the coordinates of the intersection center, the cause reference coordinates for the cause of roadway bend are the coordinates of the bend start point, and the cause reference coordinates for the cause of exceeding the transmit / receive spacing limit are the coordinates of the midpoint of the virtual segment to be tested where the potential transmission blind zone is located. The preset physical constraint is set as follows: the deviation between the boundary coordinates of the transmission blind zone and the causal reference coordinates corresponding to the propagation cause pre-classification result does not exceed the preset position deviation threshold.

[0014] In some embodiments, when the wireless transceivers at both ends of the virtual segment under test alternately transmit probe messages as transmitters, the propagation attenuation distance corresponding to each transmission is calculated according to the corresponding link response curves, and the propagation attenuation distance corresponding to each transmission is superimposed with the longitudinal coordinates of the corresponding transmitter along the longitudinal direction of the roadway to obtain the candidate boundary coordinates corresponding to each transmission. When the deviation between two candidate boundary coordinates does not exceed a preset boundary deviation threshold, the average value of the two candidate boundary coordinates is used as the boundary coordinate of the transmission blind zone. When the deviation between two candidate boundary coordinates exceeds a preset boundary deviation threshold, the two candidate boundary coordinates are checked to see if they meet the preset physical constraints. If both candidate boundary coordinates meet the preset physical constraints, the one with the smaller deviation from the causal reference coordinate is taken as the transmission blind zone boundary coordinate. If only one candidate boundary coordinate meets the preset physical constraints, that candidate boundary coordinate is taken as the transmission blind zone boundary coordinate. If neither candidate boundary coordinate meets the preset physical constraints, the average of the two candidate boundary coordinates is taken as the transmission blind zone boundary coordinate, and a check mark is added to the blind zone detection result.

[0015] In some embodiments, after outputting the blind zone detection result, the method further includes: re-acquiring the received signal strength sequence and extracting wireless propagation statistical features according to a preset retesting cycle, and matching the re-extracted wireless propagation statistical features with the multipath propagation feature fingerprint database to obtain the corresponding propagation cause retesting result. When the propagation cause retest results of the wireless transceivers at both ends of the virtual segment under test are inconsistent with the propagation cause pre-classification results of the previous retest cycle, the transmission power classification detection and link response curve construction steps are re-executed, and the transmission blind zone boundary coordinates are recalculated based on the updated link response curve to update the blind zone detection results of the corresponding virtual segment under test.

[0016] One embodiment of the above application has the following advantages or beneficial effects: This application solves the technical problem of communication blind spots and difficulty in accurate positioning in wireless sensor networks in complex mine roadways, providing a reliable means for blind spot detection and cause identification for mine safety monitoring networks, significantly enhancing the environmental adaptability and early warning capabilities of the monitoring network, and ensuring the continuity of underground communication and data acquisition.

[0017] This application uses wireless transceivers to collaboratively locate and divide virtual segments, and extracts multidimensional wireless propagation statistical features of received signal strength and performs cause matching with a pre-built fingerprint database to achieve pre-classification and identification of the causes of blind spots. On this basis, multiple wireless transceivers are used to perform collaborative broadcast scanning at preset transmission power levels and plot link response curves. By finding the critical transmission power point on the curve, potential transmission blind spots can be triggered and located more accurately, effectively overcoming the shortcomings of high misjudgment rate and ambiguous positioning in existing technologies.

[0018] This application addresses the aforementioned technical challenges by combining a cooperative broadcasting mechanism with roadway spatial modeling and multipath fingerprint matching. It divides the roadway space into virtual segments with longitudinal coordinates, establishing a spatial binding relationship between blind zone detection and the roadway's physical structure, overcoming the positioning ambiguity caused by the lack of spatial reference in traditional methods. It pre-classifies the causes of wireless transceiver device failures using a multipath propagation feature fingerprint database, transforming blind zone cause identification from post-investigation to pre-judgment, avoiding the confusion caused by single measurement methods. Through the transmit power-link response curve mechanism of cooperative broadcasting, it transforms connectivity criticality determination into a stable transmit power level critical criterion, eliminating the interference of instantaneous signal fluctuations on the determination results. By jointly verifying the mine roadway waveguide propagation model with physical constraints, it ensures that the boundary coordinates of the transmission blind zone simultaneously satisfy the consistency constraints of electromagnetic propagation laws and roadway facility distribution, achieving a leap from empirical estimation to physical model-driven blind zone boundary positioning. Therefore, this application improves the accuracy of blind zone detection while endowing the detection results with clear physical interpretability and spatial traceability.

[0019] Other effects of the above-mentioned alternative methods will be described below in conjunction with specific embodiments. Attached Figure Description

[0020] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of this application. Wherein: Figure 1 This is a flowchart illustrating the wireless link blind spot detection method for mine roadways based on cooperative broadcasting provided in the embodiments of this application; Figure 2 This is a schematic diagram of the pre-classification results of the propagation causes of wireless transceivers at both ends of each virtual segment, provided in an embodiment of this application. Figure 3 This is a schematic diagram illustrating the variation of average reception success rate of cooperative broadcasting with transmission power level, provided in an embodiment of this application. Figure 4 This is a schematic diagram illustrating the attenuation of received signal strength with propagation distance, provided in an embodiment of this application. Detailed Implementation

[0021] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0022] To facilitate understanding of this application, the embodiments of this application will be briefly described below: This application proposes a wireless sensor network for use in mine roadways. Multiple wireless transceivers are deployed at equal intervals along the longitudinal direction of a transport roadway, maintaining network connectivity through wireless communication. The roadway presents complex conditions such as metal equipment obstructions, roadway junctions, and bends. The system needs to detect communication blind spots caused by these factors in real time to ensure the continuity of mine safety monitoring. This solution first completes self-positioning and divides virtual segments by exchanging ranging messages between the wireless transceivers. Then, through periodic broadcasting and receiving signals, it analyzes the statistical characteristics of wireless propagation to preliminarily identify the causes and locations of potential blind spots.

[0023] The complex environment of mine tunnels significantly affects wireless signal propagation due to multipath effects and obstacles, easily creating coverage blind spots and threatening the reliability of monitoring networks. This application utilizes a collaborative broadcast mechanism between wireless transceivers to extract statistical characteristics of wireless propagation for preliminary screening, and combines this with a waveguide propagation model of mine tunnels for precise boundary calculation. This enables intelligent identification and accurate location of various blind spot causes, effectively ensuring the connectivity and monitoring continuity of the mine wireless sensor network.

[0024] See Figure 1 , Figure 1 This is a flowchart illustrating the wireless link blind spot detection method for mine roadways based on cooperative broadcasting provided in this application embodiment. Figure 1 The execution subject of the method shown can be a combination of software and / or hardware, specifically, it can be one or more of various types of terminals, hardware systems, cloud computing, etc.

[0025] Figure 1 The method for detecting blind spots in a mine roadway wireless link based on cooperative broadcasting includes steps S1 to S6, as detailed below: S1: Several transceiver devices are deployed in the mine roadway. The wireless transceiver devices send ranging messages to each other to measure the distance between adjacent wireless transceiver devices. Taking the wireless transceiver device at the entrance of the roadway as the origin, the distance is accumulated segment by segment according to the arrangement order along the longitudinal direction of the roadway to obtain the longitudinal coordinate of each wireless transceiver device. The roadway space between two wireless transceiver devices with adjacent longitudinal coordinates is determined as a virtual segment.

[0026] Specifically, the adjacent spacing refers to the spatial distance between two adjacent wireless transceivers along the longitudinal direction of the tunnel. It is used to characterize the actual deployment interval of the wireless transceivers within the tunnel, providing a basic measurement for subsequently constructing the longitudinal coordinates of the wireless transceivers. Since the underground environment cannot receive satellite positioning signals, and tunnel design drawings cannot accurately reflect the actual installation positions of the wireless transceivers, this step utilizes the known and constant propagation speed of electromagnetic waves. The adjacent spacing is measured based on the propagation time of the ranging messages exchanged between the wireless transceivers. For example, the spacing can be calculated by recording the round-trip propagation time of the messages and subtracting the fixed response delay of the wireless transceivers. Taking the wireless transceiver at the tunnel entrance as the origin, the adjacent spacings of each segment are accumulated according to the arrangement order of the wireless transceivers along the longitudinal direction of the tunnel, thus transforming the relative distance relationship between the wireless transceivers into absolute longitudinal coordinates in a unified coordinate system. A virtual segment, with two adjacent wireless transceivers in the longitudinal coordinates as the boundary, is used to discretize the continuous tunnel space into several independently identifiable monitoring units, providing a spatial index for subsequent segment-by-segment screening of potential transmission blind spots. This step transforms the discrete ranging results between wireless transceivers into a continuous spatial reference covering the entire tunnel, providing a precise location reference for all subsequent detections.

[0027] S2: Each wireless transceiver receives probe messages transmitted by adjacent wireless transceivers, forms a sequence of received signal strength, and extracts variance, skewness, kurtosis and fading distribution type based on a preset sliding time window to obtain the statistical characteristics of wireless propagation.

[0028] Specifically, the received signal strength sequence refers to the sequence of signal strength values ​​recorded by each wireless transceiver device according to the reception time when it periodically receives probe messages broadcast by adjacent wireless transceivers. This sequence is used to characterize the fluctuation process of the wireless channel between wireless transceivers over time, providing raw data for subsequent extraction of statistical features of wireless propagation. Because the impact mechanisms on signal propagation differ depending on typical working conditions such as metal equipment obstruction, roadway intersections, and roadway bends within mine roadways, the statistical forms of received signal strength vary significantly. Therefore, this step uses a preset sliding time window to extract the received signal strength sequence and calculates the variance, skewness, kurtosis, and fading distribution type of the sequence within the window. Variance characterizes the dispersion of signal strength around the mean; skewness characterizes the asymmetry of the distribution; kurtosis characterizes the sharpness and tail thickness of the distribution; and the fading distribution type characterizes whether a dominant line-of-sight path exists in the channel. For example, the Rayleigh distribution corresponds to a severe multipath environment without a line-of-sight path, while the Rice distribution corresponds to an environment with a dominant line-of-sight path. This step transforms the raw fluctuating received signal strength data into a set of stable and comparable wireless propagation statistical features, overcoming the limitation of relying solely on the average signal strength for judgment, and providing feature input for subsequent pre-classification of propagation causes.

[0029] S3: Match the wireless propagation statistical features with the pre-established multipath propagation feature fingerprint database to obtain the propagation cause pre-classification results of each wireless transceiver. When the propagation cause pre-classification results of the wireless transceivers at both ends of the virtual segment are both abnormal propagation causes, or the distance between the two wireless transceivers is greater than the preset transmission distance threshold, the virtual segment is determined to be a potential transmission blind zone.

[0030] Specifically, the multipath propagation feature fingerprint database is a pre-established database of correlations between typical operating conditions and their wireless propagation statistical characteristics. It provides a comparison benchmark for online cause identification. The fingerprint patterns are constructed by collecting received signal strength sequences under various typical operating conditions, generating wireless propagation statistical characteristics, and then associating and storing the operating condition causes with these statistical characteristics. The pre-classification results of propagation causes include normal propagation causes and abnormal propagation causes. Normal propagation causes indicate that the wireless channel has good propagation conditions, while abnormal propagation causes indicate cause types where propagation conditions objectively do not exist, such as metal equipment obstruction, roadway intersections, and roadway bends. During online detection, the wireless propagation statistical characteristics of the wireless transceiver are matched one by one with the fingerprint patterns in the fingerprint database. The cause corresponding to the fingerprint pattern with the highest similarity is taken as the pre-classification result of the propagation cause for that wireless transceiver. Since a single malfunction of a wireless transceiver can affect the communication quality of adjacent links, virtual segments are used as the determination unit. When the propagation cause pre-classification results of the wireless transceivers at both ends of a virtual segment are both abnormal propagation causes, or the corresponding adjacent distance exceeds a preset transmission distance threshold, the virtual segment is determined to be a potential transmission blind spot. Through this step, abstract wireless propagation statistical characteristics are mapped to specific physical cause labels, realizing automated preliminary screening of potential transmission blind spots and indicating the target area for subsequent cooperative broadcast detection.

[0031] S4: For potential transmission blind spots, control the wireless transceiver devices associated with the potential transmission blind spot to transmit probe messages at a preset transmission power level. The receiving end will count the reception success rate at different transmission power levels and construct a link response curve between the transmission power level and the average reception success rate.

[0032] Specifically, the virtual segment identified as a potential transmission blind zone is designated as the virtual segment under test. Measurement association devices refer to the two ends of the virtual segment under test and adjacent wireless transceivers outside these two ends. The adjacent wireless transceivers are introduced because they can provide reference receiving points from outside the blind zone's influence range, providing a comparative basis for determining whether the link interruption originates from the blind zone. During cooperative broadcasting, the two ends of the wireless transceivers take turns acting as transmitters, while the remaining measurement association devices act as receivers, probing link connectivity from two directions of the virtual segment under test. The transmitters gradually increase their transmission power according to preset transmission power levels and broadcast probe messages. The receivers count the number of successfully received probe messages within a preset reception duration and extract the corresponding wireless propagation statistical characteristics at each transmission power level. The average reception success rate is taken as the average reception success rate. Since the signal attenuation within potential transmission dead zones cannot be directly measured using static received signal strength sequences, this step involves actively changing the transmit power to obtain a quantitative response relationship between link connectivity and transmit power. A link response curve is then plotted with transmit power level on the horizontal axis and average reception success rate on the vertical axis. This step transforms qualitative suspicions about potential transmission dead zones into quantitatively analyzable experimental data on transmit power and connectivity, providing a direct basis for calculating the boundaries of transmission dead zones.

[0033] S5: Determine the critical transmit power level based on the link response curve, substitute the wireless propagation statistical characteristics corresponding to the critical transmit power level into the mine tunnel waveguide propagation model, calculate the propagation attenuation distance of the detection message, and obtain the boundary coordinates of the transmission blind zone by combining the longitudinal coordinates of the transmitter.

[0034] Specifically, the critical transmit power level refers to the transmit power level at which the average reception success rate of the first consecutive preset number of transmit power levels (e.g., 3) appears below a preset connectivity threshold when searching along the link response curve from high to low transmit power levels. It characterizes the minimum power boundary required for the transmitter to maintain reliable connectivity. The mine tunnel waveguide propagation model is a path loss model characterizing the attenuation of received signal strength with propagation distance. Since mine tunnels are confined spaces, electromagnetic wave propagation exhibits waveguide and multipath effects, and free-space models cannot accurately describe the attenuation law. Therefore, this step selects model parameters based on the fading distribution type in the wireless propagation statistical characteristics corresponding to the critical transmit power level, and calculates the distance at which the probe message can just reliably propagate, i.e., the propagation attenuation distance. By superimposing the propagation attenuation distance and the longitudinal coordinates of the measurement association device for transmitting probe messages along the tunnel longitudinally away from the transmitter, the critical state of link connectivity can be mapped to a specific location in the tunnel space. This step achieves a quantitative transformation from power and connectivity curves to specific spatial coordinates, overcoming the limitation that relying solely on connectivity status cannot locate blind zone boundaries.

[0035] S6: When the boundary coordinates of the transmission blind zone meet the preset physical constraints corresponding to the pre-classification results of the propagation cause, output the blind zone detection results including the boundary coordinates of the transmission blind zone and the cause of the blind zone.

[0036] Specifically, the preset physical constraints refer to the spatial constraints set for each anomaly propagation cause based on preset roadway facility distribution information. These constraints are used to verify whether the calculated transmission blind zone boundary coordinates match the actual physical environment. Since the actual blind zone boundary is necessarily closely related to the physical location of the cause—for example, the blind zone boundary corresponding to metal equipment obstruction should be located near the location of the metal equipment, and the blind zone boundary corresponding to roadway intersection should be located near the center of the intersection—this step sets cause reference coordinates for each anomaly propagation cause. When the deviation between the transmission blind zone boundary coordinates and the cause reference coordinates corresponding to the propagation cause pre-classification results does not exceed a preset position deviation threshold, the transmission blind zone boundary coordinates are deemed to meet the preset physical constraints. If the calculation results deviate significantly from the cause reference coordinates, it indicates that the model parameters were improperly selected or the propagation cause pre-classification was incorrect, and the detection results should not be directly output. Through this step, prior knowledge of roadway facilities is introduced to cross-validate the calculation results, significantly reducing the false alarm rate. The output includes the transmission blind zone boundary coordinates and the blind zone cause, completing the decision-making closed loop from mathematical calculation to physical reality.

[0037] This embodiment, through the synergistic effect of the above six steps, upgrades the blind zone detection of mine wireless sensor networks from a crude mode relying on manual inspection and simple connectivity judgment to an automated and precise detection system based on spatial modeling, statistical identification, and quantitative calculation. It constructs a digital spatial benchmark for the roadway through ranging and positioning and virtual segmentation; achieves intelligent identification of causes and initial screening of potential transmission blind zones through wireless propagation statistical feature extraction and fingerprint database matching; actively detects link connectivity boundaries through power scanning collaborative broadcasting; and outputs reliable transmission blind zone boundary coordinates through mine roadway waveguide propagation model calculation and physical constraint verification. These four interconnected steps construct a complete closed-loop logical framework of "spatial modeling - feature identification - active detection - calculation and verification." This solution has the ability to identify blind zone causes and risks in advance, such as metal equipment obstruction, roadway intersections, roadway bends, and exceeding transmit / receive distance limits. It can output decision results including transmission blind zone boundary coordinates and blind zone causes, and dynamically track the evolution of blind zones through periodic retesting. This achieves the integration of wireless channel statistical characteristics and roadway geometric spatial information, as well as the integration of passive network monitoring and active detection mechanism of wireless transceivers, providing technical support for the continuous and reliable operation of the mine safety monitoring network.

[0038] This embodiment elaborates on the location of wireless transceivers and the division of virtual segments.

[0039] The location of wireless transceivers serves as the spatial reference for all subsequent detection stages. The core of blind spot detection in mine roadways using wireless sensor networks lies in accurately identifying the boundary locations of areas with severe communication attenuation. To achieve this, the physical location of each wireless transceiver along the longitudinal direction of the roadway must be known. Since satellite positioning signals cannot be received underground, and roadway design drawings often fail to reflect the actual installation locations of the wireless transceivers, it is necessary to utilize the direct communication capabilities between the wireless transceivers. By measuring signal propagation time, the distance between the wireless transceivers can be calculated, thereby constructing a one-dimensional coordinate system along the longitudinal direction of the roadway.

[0040] Specifically, the objective mechanism for measuring the distance between adjacent devices lies in the fact that the propagation speed of electromagnetic waves is known and constant. By measuring the total time it takes for a ranging message to travel back and forth between two wireless transceivers, and subtracting the fixed preset response delay of the adjacent wireless transceivers, the actual flight time of the electromagnetic waves can be calculated, thus determining the spatial distance between the wireless transceivers. Each wireless transceiver sends a ranging message to its adjacent wireless transceiver and records the transmission time. Adjacent wireless transceivers within a preset response delay Afterwards, a response message is sent, and each wireless transceiver records the time when it is received. Let the round-trip propagation time of the ranging message be... Then the adjacent spacing The formula is expressed as: Where d is the adjacent spacing in meters (m); c is the propagation speed of electromagnetic waves in the equivalent waveguide environment of a mine roadway, which is taken in this embodiment. m / s; This is the total time from sending the ranging message to receiving the response message, i.e. The unit is seconds (s); To account for the fixed processing and waiting delay between adjacent wireless transceivers receiving a ranging message and sending a response message, this embodiment defaults to 0.1ms, i.e. The preset response delay is a fixed, known value that is precisely deducted from the calculation. Therefore, its value does not affect the ranging accuracy. It is set to 0.1ms to allow sufficient and stable time for the wireless transceiver to receive, process, parse the protocol stack, and prepare the response, so as to avoid ranging errors caused by processing jitter.

[0041] Multiple measurements are performed on the same adjacent distance. When the deviation of a measurement result from the average of the remaining measurement results exceeds a preset ranging deviation threshold, that measurement result is discarded, and the average of the remaining measurement results is used as the adjacent distance. In this embodiment, the preset ranging deviation threshold can be set to 0.3m. This threshold is set based on error analysis of the ranging system in the complex electromagnetic environment of a mine: the standard deviation of a single ranging measurement under typical working conditions is approximately 0.1m, and according to the 3σ principle, the judgment limit for gross errors is approximately 0.3m. If the threshold is set too small, normal measurement fluctuations will be mistakenly judged as outliers and discarded, reducing the amount of effective data; if it is set too large, gross errors caused by sudden interference or message conflicts cannot be identified. 0.3m represents a compromise between the two.

[0042] For example, consider the measurement of the distance between wireless transceiver n2 and wireless transceiver n3. Assume three measurements were performed, with total round-trip propagation times of 100132.4 ns, 100131.6 ns, and 100385.2 ns, respectively. Subtract the preset response delay of 1 × 10⁻⁶. 5 After ns, the actual round-trip flight times of the electromagnetic wave are 132.4ns, 131.6ns, and 385.2ns, respectively. Substituting these values ​​into the above formula, the first measurement results are obtained. Similarly, the second and third measurement results were 19.74m and 57.78m, respectively. The deviation of the third measurement result from the average of the other two measurement results (19.80m) was 37.98m, exceeding the preset distance deviation threshold of 0.3m, and was therefore discarded. The mutual deviation of the other two measurement results was checked and found to be 0.12m, which did not exceed the threshold. Finally, the average of 19.86m and 19.74m, 19.80m, was used as the adjacent distance between wireless transceiver n2 and wireless transceiver n3.

[0043] After obtaining the adjacent spacing, taking the wireless transceiver at the entrance of the tunnel (denoted as n1) as the origin, and following the longitudinal arrangement of the wireless transceivers along the tunnel (n1→n2→...→n9), the adjacent spacings are accumulated segment by segment along the longitudinal direction of the tunnel. The accumulated result for each wireless transceiver is used as the longitudinal coordinate of that wireless transceiver. Mathematically, the longitudinal coordinate of the k-th wireless transceiver... Represented as: ,in, Let be the vertical coordinate of the kth wireless transceiver, in meters (m). The distance between adjacent segments in the i-th segment is the distance between wireless transceiver devices ni and ni+1, expressed in meters (m). The summation starts from wireless transceiver device n1 and includes the vertical coordinates of wireless transceiver device n1. For example, the vertical coordinate of the wireless transceiver n4 =20.10+19.80+20.15=60.05m.

[0044] Finally, the tunnel space between two adjacent wireless transceivers with vertical coordinates is defined as a virtual segment. The virtual segment is bounded by the vertical coordinates of the two adjacent wireless transceivers, representing the monitoring area of ​​that tunnel segment. For example, the tunnel space between wireless transceivers n2 and n3 is virtual segment VS2, with a corresponding coordinate range of [20.10m, 39.90m]. To verify the correctness of the wireless transceiver positioning and virtual segment division, this embodiment provides the measurement results of each adjacent distance and the calculation results of the vertical coordinates of each wireless transceiver, as shown in Table 1.

[0045] Table 1: Calculation Results of Ranging Values ​​and Vertical Coordinates of Wireless Transceiver Device

[0046] Table 1 shows that the measured values ​​of each adjacent spacing are between 19.70m and 20.30m, with a deviation of no more than 0.30m from the designed deployment spacing of 20m. This indicates that the ranging and outlier removal mechanism can effectively suppress measurement errors in the mine environment. The longitudinal coordinates of each wireless transceiver are obtained by accumulating adjacent spacings segment by segment. For example, the longitudinal coordinate of wireless transceiver n5 is 80.00m, and the longitudinal coordinate of wireless transceiver n9 is 159.80m, which is compatible with the total length of the 160m roadway. Between two wireless transceivers with adjacent longitudinal coordinates, eight virtual segments are formed sequentially from VS1 to VS8, each with a clear coordinate boundary. The above results verify the accuracy of the method of accumulating adjacent spacings segment by segment to obtain the longitudinal coordinates based on the wireless transceiver at the roadway entrance as the origin, and dividing the virtual segments accordingly. This provides a clear and accurate spatial index for subsequent blind spot determination based on virtual segments.

[0047] This embodiment employs a ranging and coordinate accumulation method based on round-trip propagation time because it directly utilizes the physical characteristic that the propagation speed of electromagnetic waves is known and constant. The ranging results are significantly less affected by multipath fading than those based on signal strength, providing reliable absolute distance information. If ranging relies on received signal strength, the received signal strength is severely affected by shadowing fading and multipath effects, resulting in significant errors in the non-uniform channels of mines, making it impossible to provide a reliable geometric benchmark for blind zone boundary calculation. Through precise wireless transceiver positioning and virtual segment division, blind zone detection can be anchored from fuzzy connectivity judgments to specific physical locations within the tunnel, avoiding blind zone misjudgments and boundary positioning deviations caused by positioning ambiguity. This establishes a quantifiable spatial coordinate system for subsequent wireless propagation statistical feature analysis, potential transmission blind zone determination, and blind zone boundary calculation.

[0048] This embodiment elaborates on the extraction of statistical features of wireless propagation and the determination of fading type.

[0049] Specifically, the objective mechanism for extracting the wireless propagation statistical characteristics and fading distribution types of the received signal strength sequence lies in the fact that the multipath effect intensity and fluctuation characteristics of the wireless channel within the mine roadway directly shape the statistical form of the received signal strength. When a dominant line-of-sight path exists between the transceivers, the envelope of the received signal strength follows a Ricean distribution with relatively gentle fluctuations. When the line-of-sight path is blocked by metal equipment, or when the signal undergoes multiple reflections and scatterings at roadway junctions and bends, the envelope of the received signal strength follows a Rayleigh distribution, with frequent deep fading, significantly increased variance and kurtosis, and significantly negative skewness. Simply relying on the mean of the received signal strength cannot distinguish between these two fundamentally different channel states; therefore, it is necessary to extract a complete set of wireless propagation statistical characteristics and determine the fading distribution type.

[0050] Specifically, in this embodiment, for the signal strength sequence of the probe messages received by each wireless transceiver device, a window containing K sampling points is first extracted using a preset sliding time window (e.g., set to 10s). Then, the mean, variance, skewness, and kurtosis of the sequence within the window are calculated. Next, the empirical distribution of the sequence within the window is fitted with the Rayleigh distribution model and the Rice distribution model, respectively, and the fitting error is compared to determine the fading distribution type. The sliding time window repeats the above process as time progresses, obtaining the dynamic wireless propagation statistical characteristics of each wireless transceiver device. Let the current sliding time window contain K received signal strength sampling values, denoted as RSSI(1), RSSI(2), ..., RSSI(K). First, the mean of the received signal strength sequence within the window is calculated, and its formula is expressed as: ,in, is the arithmetic mean of the received signal strength sequence within the current sliding time window, in dBm; K is the number of sampling points within the sliding time window. The value of K is determined by the length of the sliding time window and the reception period of the probe message. For example, when the sliding time window is 10s and the reception period is 10ms, K is 1000 to ensure the stability of the statistical estimation.

[0051] Subsequently, the variance of the received signal strength sequence was calculated. Its formula is expressed as: ,in, The sample variance of the received signal strength sequence, in dBm. 2 The variance represents the degree of dispersion of the signal strength around its mean; the larger the variance, the more severe the signal fluctuation. RSSI(k) is the received signal strength value of the kth sampling point within the window, in dBm. The denominator is K-1 to obtain an unbiased estimate of the overall variance.

[0052] Next, the skewness γ of the received signal strength sequence is calculated, and its formula is expressed as: , where γ is the sample skewness, which is dimensionless and characterizes the asymmetry of the distribution of the received signal intensity sequence; The aforementioned sample variance; It is a very small constant, for example, it can be set to 10. -6 This is used to prevent numerical calculation anomalies where the denominator is zero when the variance is zero. A skewness less than 0 indicates that the distribution is left-skewed, meaning that there are many deep fading sampling points in the sequence that are much lower than the mean, which is a typical characteristic of multipath obstruction channels; a skewness close to 0 indicates that the distribution is approximately symmetrical.

[0053] Then, the kurtosis κ of the received signal strength sequence is calculated, and its formula is expressed as: , where κ is the sample kurtosis, which is dimensionless and characterizes the sharpness and tail thickness of the distribution; subtracting 3 in the formula is to make the kurtosis of the normal distribution 0. A kurtosis greater than 0 indicates that the distribution is sharper and has thicker tails than the normal distribution, that is, the probability of extreme deep fading is higher than that expected by the normal distribution.

[0054] After obtaining the aforementioned wireless propagation statistical characteristics, the fading distribution type is determined. Specifically, the empirical histogram distribution of the received signal strength sequence within the window is fitted with the theoretical probability density functions of the Rayleigh and Rice distribution models, respectively. The least squares method is used for fitting, with the root mean square error as the fitting error. Let the fitting error with the Rayleigh distribution model be... The fitting error with the Rice distribution model is ,when If the received signal strength sequence is in the Rayleigh distribution, it is determined to be in the Rice distribution; otherwise, it is determined to be in the Rice distribution.

[0055] To verify the effectiveness of wireless propagation statistical feature extraction and fading distribution type determination, this embodiment presents the calculation results and determination results of wireless propagation statistical features for each wireless transceiver device, as shown in Table 2.

[0056] Table 2: Statistical Characteristics of Wireless Propagation and Determination Results of Fading Distribution Types for Various Wireless Transceivers

[0057] As shown in Table 2, the variances of wireless transceivers n1, n4, and n7 are all within 2.1 dBm. 2 Up to 2.3dBm 2 Between these values, the skewness does not exceed 0.22, the kurtosis is close to 0, and the fitting error of the Rice distribution model is smaller than that of the Rayleigh distribution model, thus they are determined to be Rice distributions. This indicates that the channels in which these wireless transceivers are located have a dominant line-of-sight path, weak multipath effects, and good communication conditions. The variance of wireless transceivers n2, n3, n5, n6, n8, and n9 significantly increases to 3.9 dBm. 2 Up to 4.8dBm 2The skewness of all samples was below -1.0, and the kurtosis was above 2.2. Furthermore, the Rayleigh distribution model showed a smaller fitting error, leading to the identification of a Rayleigh distribution. This indicates that these wireless transceivers frequently experienced deep fading in the received signals, exhibiting a significantly left-skewed distribution with sharp peaks and thick tails, suggesting severe multipath interference or obstruction. The variance of wireless transceiver n5 was 4.8 dBm. 2 The most prominent characteristics are skewness (-1.42) and kurtosis (3.12), which are consistent with the physical fact that multipath scattering is strongest at the junction of the tunnel. The above-mentioned statistical characteristics of wireless propagation and the determination results of fading distribution type verify the effectiveness of characterizing channel state through variance, skewness, kurtosis and fading distribution type, and provide reliable feature input for subsequent pre-classification of propagation causes.

[0058] By introducing variance, skewness, kurtosis, and fading distribution types to construct multidimensional wireless propagation statistical features, we can quantitatively characterize the fluctuation characteristics and multipath effect intensity of wireless channels, avoiding the one-sidedness of relying solely on the mean of a single signal strength. This provides an accurate feature basis for fingerprint database matching and potential transmission blind zone determination.

[0059] This embodiment elaborates on the pre-construction method of the multipath propagation feature fingerprint database and the pre-classification of propagation causes.

[0060] Specifically, the multipath propagation fingerprint database is a database that stores the statistical characteristics and causal relationships of wireless propagation under typical working conditions in mine roadways. The objective mechanism for its pre-construction lies in the fact that the impact mechanisms of typical working conditions such as metal equipment obstruction, roadway intersections, roadway bends, and exceeding the transmit / receive distance limit on wireless signal propagation are different, and the corresponding wireless propagation statistical characteristics are significantly different. However, it is difficult to distinguish these causes based on a single indicator of received signal strength. Therefore, it is necessary to construct a fingerprint database by simulating various working conditions to collect multidimensional wireless propagation statistical characteristics as a benchmark for online pre-classification of propagation causes.

[0061] The pre-construction process in this embodiment is as follows: First, five typical working conditions are simulated in the mine roadway: normal working condition, metal equipment obstruction working condition, roadway intersection working condition, roadway turning working condition, and transceiver spacing exceeding the limit working condition. Then, under each typical working condition, the wireless transceiver device is controlled to periodically broadcast detection messages, and the receiving end continuously records the received signal strength sequence and extracts the sequence with a preset sliding time window. Next, the variance, skewness, and kurtosis of the extracted sequence are calculated, and the fading distribution type is determined to generate the corresponding wireless propagation statistical features. Finally, the causes of each typical working condition are associated with the generated wireless propagation statistical features and stored as a fingerprint pattern to form a multipath propagation feature fingerprint database.

[0062] For example, when simulating the working condition of metal equipment obstruction, metal obstacles are placed in the tunnel. The signal between the wireless transceivers undergoes strong reflection and diffraction. The collected received signal strength sequence shows high variance, left skew, sharp peaks and thick tails, and Rayleigh fading characteristics. This wireless propagation statistical characteristic is associated with the abnormal propagation cause of "metal equipment obstruction" and stored as a fingerprint pattern. Under normal working conditions, line-of-sight propagation is dominant, and the wireless propagation statistical characteristic shows low variance, approximately symmetry, and Ricean fading characteristics. This is associated with the "normal propagation cause" and stored.

[0063] During online detection, for each wireless transceiver device to be classified, its wireless propagation statistical features are matched one by one with fingerprint patterns in the multipath propagation feature fingerprint database. Let F be the wireless propagation statistical feature vector of the wireless transceiver device to be classified, and let the feature vector of the i-th fingerprint pattern in the fingerprint database be... Then the similarity The formula for calculating cosine similarity is as follows: ,in, is the similarity between the wireless propagation statistical features of the wireless transceiver device to be classified and the i-th fingerprint pattern, with a value range of [-1, 1]. The larger the value, the more similar the two are; F is the wireless propagation statistical feature vector of the wireless transceiver device to be classified, which is composed of variance, skewness, kurtosis and fading distribution type encoding. Let be the feature vector of the i-th fingerprint pattern in the fingerprint database. The cause corresponding to the fingerprint pattern with the highest similarity is used as the pre-classification result of the propagation cause of the wireless transceiver device. This matching process only requires vector inner product operation, with low computational overhead, making it suitable for distributed implementation of wireless transceivers in resource-constrained mines. This embodiment uses predefined wireless propagation statistical features to construct the fingerprint database because the physical meaning of variance, skewness, kurtosis, and fading distribution type is clear and highly interpretable; if machine learning is used to automatically extract features, the model training and inference overhead is large, and the decision-making process is difficult to interpret, making it unsuitable for mine safety monitoring scenarios.

[0064] The pre-classification results of propagation causes include normal propagation causes and abnormal propagation causes. Normal operating conditions correspond to normal propagation causes, indicating that the wireless channel has good propagation conditions. Abnormal propagation causes correspond to conditions such as metal equipment obstruction, roadway intersections, roadway bends, and exceeding the transmit / receive distance limit. Abnormal propagation causes represent the types of causes where propagation conditions objectively do not exist. It should be noted that environmental interference such as humidity and dust, which only weaken signal strength, does not constitute abnormal propagation causes. Only situations where reliable propagation conditions objectively do not exist, such as physical obstruction, abrupt changes in roadway structure, and excessive distance, are considered abnormal propagation causes. During the virtual segment determination, when the pre-classification results of propagation causes for both ends of the virtual segment's wireless transceiver are abnormal propagation causes, or when the corresponding adjacent distance exceeds a preset transmission distance threshold (e.g., set to 30m), the virtual segment is determined to be a potential transmission dead zone.

[0065] See Figure 2 , Figure 2 This is a schematic diagram of the pre-classification results of the propagation causes of wireless transceivers at both ends of each virtual segment provided in this embodiment. Figure 2 The horizontal axis represents virtual segment numbers VS1 to VS8, and the vertical axis represents the cause category code. Code 1 indicates normal propagation cause, code 2 indicates metal equipment obstruction, code 3 indicates roadway intersection, and code 4 indicates roadway bend. The bar chart contains two series, representing the pre-classification results of propagation causes for the wireless transceiver with the smaller number (wireless transceiver A) and the wireless transceiver with the larger number (wireless transceiver B) at both ends of each virtual segment. It should be noted that the cause category code also includes code 5 (transmitter-receiver spacing exceeding the limit). In this embodiment, the pre-classification results of propagation causes for each wireless transceiver do not involve this cause. Figure 2 No bar corresponding to code 5 appeared in the table.

[0066] Depend on Figure 2 It can be seen that the cause categories of the wireless transceivers at both ends of virtual segment VS1 are 1 and 2, virtual segment VS3 is 2 and 1, virtual segment VS4 is 1 and 3, virtual segment VS6 is 3 and 1, and virtual segment VS7 is 1 and 4. The pre-classification results of the propagation causes of the wireless transceivers at both ends of these five virtual segments are not both abnormal propagation causes, and the corresponding adjacent distances do not exceed the preset transmission distance threshold of 30m. Therefore, none of them are judged as potential transmission blind spots. The cause category of the wireless transceivers at both ends of virtual segment VS2 is 2, that is, the wireless transceivers at both ends are abnormal propagation caused by metal equipment obstruction. The cause category of the wireless transceivers at both ends of virtual segment VS5 is 3, which is abnormal propagation caused by the intersection of the roadway. The cause category of the wireless transceivers at both ends of virtual segment VS8 is 4, which is abnormal propagation caused by the bend of the roadway. Therefore, VS2, VS5, and VS8 are judged as potential transmission blind spots. Figure 2 The judgment results are causally consistent with the wireless propagation statistical characteristics and fading distribution types of each wireless transceiver in Table 2: the virtual segment where both ends of the wireless transceiver exhibit Rayleigh fading and significantly abnormal wireless propagation statistical characteristics is precisely the virtual segment that was judged as a potential transmission blind zone, which verifies the effectiveness of the fingerprint database construction and matching method and provides reliable input for the potential transmission blind zone judgment logic.

[0067] By constructing a multipath propagation feature fingerprint database containing five typical operating conditions and using similarity matching to pre-classify propagation causes, we can comprehensively utilize the multidimensional features of signal fading to distinguish the causes of different blind zones. This avoids the one-sidedness and misjudgment of relying solely on signal strength or a single indicator, thus providing an accurate and automated basis for determining the causes of potential transmission blind zones and improving the robustness of the detection method.

[0068] This embodiment provides a detailed explanation of cooperative broadcasting and link response curve plotting.

[0069] Specifically, the objective mechanism of cooperative broadcasting is that the signal propagation conditions in potential transmission blind zones are poor, and the precise boundary of the blind zone cannot be determined solely based on the static received signal strength sequence. However, by controlling the measurement and correlation device to gradually increase the transmission power for active detection, a quantitative response relationship of link connectivity as a function of transmission power can be obtained. The critical characteristics on the link response curve directly correspond to the spatial boundary of the blind zone.

[0070] In this embodiment, the selection of measurement association devices includes the two ends of the virtual segment under test and the adjacent wireless transceivers outside the two ends. For example, if the two ends of the potential transmission blind zone VS2 are n2 and n3, and the adjacent wireless transceivers outside the two ends are n1 and n4, then the measurement association devices are n1, n2, n3, and n4. The adjacent wireless transceivers are introduced because they can evaluate the connectivity boundary of the link from outside the influence range of the blind zone, improving the robustness of the boundary calculation.

[0071] During cooperative broadcasting, the wireless transceivers at both ends take turns acting as transmitters, while the remaining measurement and correlation devices act as receivers. For example, initially, n2 acts as the transmitter, and n1, n3, and n4 act as receivers; subsequently, n3 acts as the transmitter, and n1, n2, and n4 act as receivers. Alternating transmissions probe the communication link from both directions of the virtual segment, which can characterize the asymmetry of blind zone propagation in both directions, providing a data foundation for subsequent bidirectional candidate boundary fusion.

[0072] The transmitting end increases its transmission power incrementally according to preset transmission power levels and broadcasts probe messages. The preset transmission power levels are arranged in ascending order, for example, -10dBm, -5dBm, 0dBm, 5dBm, 10dBm, 15dBm, 20dBm, in 5dBm increments. The transmitting end broadcasts probe messages, increasing its transmission power incrementally from the lowest transmission power level. The receiving end counts the number of successfully received probe messages within a preset reception duration (e.g., 1 second) and extracts the corresponding wireless propagation statistical characteristics for each transmission power level. Let's assume that at a certain transmission power level, the total number of probe messages broadcast by the transmitting end is... The number of probe messages successfully received by a certain receiving end is The reception success rate of the receiving end at this transmission power level is... The formula is expressed as: ,in, The reception success rate is dimensionless and its value ranges from [0, 1]. This represents the number of probe messages successfully received within a preset reception time. This represents the total number of detection messages broadcast at this transmission power level, determined by the detection message broadcast frequency and the preset reception duration.

[0073] To comprehensively characterize the link connectivity along this transmission direction, the average reception success rate is the average success rate of reception at each receiver. Let the number of receivers be... The average reception success rate The formula is expressed as: ,in, The average reception success rate is dimensionless. The number of receivers; Let be the reception success rate of the i-th receiver at the same transmit power level.

[0074] After completing the scan of all transmit power levels, plot the link response curve with transmit power level on the horizontal axis and average reception success rate on the vertical axis. The lower limit of the preset transmit power level is lower than the minimum reliable transmit power of the wireless transceiver, and the upper limit is not lower than the maximum transmit power of the wireless transceiver; the step value is a trade-off between scan resolution and test time; the preset reception duration is a trade-off between statistical stability and real-time performance; the preset connectivity threshold (e.g., can be set to 0.8) is set according to the reliability requirements of data transmission for mine safety monitoring.

[0075] See Figure 3 , Figure 3 This is a schematic diagram illustrating how the average reception success rate of cooperative broadcasting varies with the transmission power level, as provided in this embodiment. Figure 3 The horizontal axis represents the transmit power level, and the vertical axis represents the average reception success rate. The solid and dashed lines represent the link response curves when the wireless transceivers n2 and n3 at both ends of the potential transmission blind zone VS2 act as transmitters, respectively. The horizontal dotted line represents the preset connectivity threshold of 0.8. Figure 3It can be seen that the link response curve of transmitter n2 has an average reception success rate of approximately 0.15 at -10dBm, which monotonically increases with the increase of the transmission power level, reaching 0.72 at 10dBm, still below the preset connectivity threshold of 0.8. It rises to 0.85 at 15dBm, exceeding 0.8 for the first time. Looking from high to low transmission power levels, the average reception success rate at 10dBm, 5dBm, and 0dBm all drops below 0.8. Therefore, the critical transmission power level for n2 is 10dBm. The link response curve of transmitter n3 is generally higher than that of n2. At 10dBm, the average reception success rate is 0.83, already above 0.8. However, the average reception success rates at 5dBm, 0dBm, and -5dBm are 0.69, 0.52, and 0.36 respectively, all falling below 0.8. Therefore, the critical transmission power level for n3 is 5dBm. The difference in the response curves of the two links indicates that there is an asymmetry in link quality in the two propagation directions for the same potential transmission dead zone. Figure 3 This is completely consistent with the critical transmit power levels recorded in Table 3, verifying that the cooperative broadcasting mechanism can accurately capture the power critical point where connectivity undergoes a qualitative change.

[0076] To verify the results of potential transmission blind zone determination and cooperative broadcast response, this embodiment presents the determination results of each virtual segment and the cooperative broadcast response results of potential transmission blind zones, as shown in Table 3.

[0077] Table 3: Results of Virtual Segment Determination and Cooperative Broadcast Response for Potential Transmission Dead Zones

[0078] As shown in Table 3, the pre-classification results of the propagation causes of the wireless transceivers at both ends of virtual segments VS1, VS3, VS4, VS6, and VS7 are not simultaneously classified as abnormal propagation causes, and the adjacent distances are all between 19.70m and 20.30m, which does not exceed the preset transmission distance threshold of 30m. Therefore, they are not identified as potential transmission blind zones and no cooperative broadcasting is required. The wireless transceivers at both ends of virtual segments VS2, VS5, and VS8 are all classified as abnormal propagation causes and are identified as potential transmission blind zones. Cooperative broadcasting in two transmission directions is completed respectively. For VS2, the critical transmission power levels of transmitters n2 and n3 are 10dBm and 5dBm, respectively, and the average reception success rates at the critical times are 0.72 and 0.69, respectively. For VS5, the critical transmission power level of transmitters n5 and n6 is 10dBm. For VS8, the critical transmission power levels of transmitters n8 and n9 are 10dBm and 0dBm, respectively. The above results verify the rationality of the selection of measurement correlation devices, alternating transmission and power step-by-step scanning design, and show that cooperative broadcasting can provide quantitative boundary calculation inputs in two directions for each potential transmission blind zone.

[0079] By selecting measurement correlation devices to conduct bidirectional, multi-transmission power level coordinated broadcasting and plotting link response curves, it is possible to quantitatively detect the pattern of connectivity variation with transmission power in potential transmission blind zones, accurately capture power critical points, and avoid the problem of failing to locate blind zone boundaries by relying solely on static received signal strength sequences. This provides accurate and robust input parameters for blind zone boundary calculation.

[0080] This embodiment elaborates on the blind zone boundary calculation and physical constraint verification.

[0081] Specifically, the waveguide propagation model in mine roadways is a path loss model that characterizes the attenuation of received signal strength with propagation distance. Mine roadways are confined spaces, and electromagnetic wave propagation exhibits waveguide and multipath effects. Free-space models cannot accurately describe the attenuation behavior; therefore, this embodiment employs a parameterized waveguide propagation model for mine roadways. The model parameters are set corresponding to the fading distribution type, and the model is expressed as follows: Where L(d) is the average path loss at propagation distance d, in dB; L0 is the reference path loss at reference distance d0, in dB; d0 is the reference distance, taken as 1m in this embodiment; n is the path loss exponent, dimensionless; ξ is the log-normal shadowing fading margin, taken as its expected value of 0dB when calculating the propagation attenuation distance. The model parameters n and L0 are related to the characteristics of the tunnel, calibrated based on measured tunnel data for the 2.4GHz band, and set accordingly to the fading distribution type: when the fading distribution type is Ricean distribution, n is 1.8 and L0 is 40dB; when the fading distribution type is Rayleigh distribution, n is 2.8 and L0 is 40dB. An excessively large path loss exponent will underestimate the propagation attenuation distance, while an excessively small one will overestimate it; therefore, it must be set accordingly to the fading distribution type.

[0082] The critical transmit power level has been determined above. Let the preset receiver sensitivity threshold be... In the critical state, the probe message can propagate to the blind zone boundary with a reliability not lower than the preset connectivity threshold. The propagation attenuation distance D is the propagation distance corresponding to the received power attenuating to the preset receive sensitivity threshold. Assume an effective link margin. = Let the path loss at propagation distance D equal the effective link margin and substitute it into the model formula. The formula for solving the propagation attenuation distance is expressed as: Where D is the propagation attenuation distance in meters (m); d0 is the reference distance, taken as 1m; L0 represents the effective link margin in dB; L0 represents the baseline path loss in dB; and n represents the path loss exponent. The preset receiver sensitivity threshold can be set to -60 dBm in this embodiment. It should be noted that this threshold is not the limit of the RF chip's physical sensitivity, but rather the minimum received power required to ensure a reception success rate not lower than the preset connectivity threshold of 0.8. It is calibrated based on the RF specifications and packet error rate requirements of the wireless transceiver and corresponds to the preset connectivity threshold on the link response curve, ensuring consistency between the calculated critical transmit power level and propagation attenuation distance.

[0083] For example, for the potential transmission dead zone VS2 (the propagation cause pre-classification result is metal device blockage, the fading distribution type is Rayleigh distribution, corresponding to n=2.8, L0=40dB), the critical transmit power level in the n2 direction of the transmitter is 10dBm, and the effective link margin is... Substituting into the above equation, we get: Similarly, the critical transmit power level in the n3 direction of the transmitter is 5dBm, and the effective link margin is 65dB. Substituting these values ​​into the above formula, we get: After obtaining the propagation attenuation distance, the propagation attenuation distance is superimposed on the longitudinal coordinates of the measurement association device for transmitting the detection message along the longitudinal direction of the roadway away from the transmitting end. The longitudinal coordinate of the transmitting end n2 is 20.10m, the transmission direction is along the positive longitudinal direction of the roadway, and the candidate boundary coordinates are 20.10m + 11.8m = 31.9m; ​​the longitudinal coordinate of the transmitting end n3 is 39.90m, the transmission direction is along the negative longitudinal direction of the roadway, and the candidate boundary coordinates are 39.90m - 7.8m = 32.1m.

[0084] See Figure 4 , Figure 4 This is a schematic diagram illustrating the attenuation of received signal strength with propagation distance provided in this embodiment. Figure 4 The horizontal axis represents the propagation distance, and the vertical axis represents the received power. The solid and dashed lines represent the attenuation curves of received power with propagation distance when the transmit power is 10dBm and 5dBm, respectively. The horizontal dotted line represents the preset receive sensitivity threshold of -60dBm. Figure 4 It can be seen that the attenuation curve with a transmission power of 10dBm intersects the threshold line at a propagation distance of 11.8m, which corresponds to the propagation attenuation distance in the n2 direction of the transmitting end. This is superimposed with the longitudinal coordinate of n2, 20.10m, along the positive longitudinal direction of the roadway to obtain the candidate boundary coordinates of 31.9m. The attenuation curve with a transmission power of 5dBm intersects the threshold line at a propagation distance of 7.8m, which corresponds to the propagation attenuation distance in the n3 direction of the transmitting end. This is superimposed with the longitudinal coordinate of n3, 39.90m, along the negative longitudinal direction of the roadway to obtain the candidate boundary coordinates of 32.1m. Figure 4Consistent with the data in Table 3, the former explains why the signal attenuates to the point of being unreliable at that distance, while the latter shows how cooperative broadcasting responds to and captures the critical transmit power level. Both fully present the end-to-end mechanism from power scanning to boundary calculation.

[0085] Physical constraint verification is then performed. Based on the preset roadway facility distribution information, causal reference coordinates are set for each abnormal propagation cause. Specifically, the causal reference coordinates for metal equipment obstruction are the coordinates of the metal equipment's location; for roadway intersections, the reference coordinates are the center coordinates of the intersection; for roadway bends, the reference coordinates are the coordinates of the bend's starting point; and for excessive transmit / receive spacing, the reference coordinates are the coordinates of the midpoint of the virtual segment containing the potential transmission blind zone. The preset physical constraint condition is set as follows: the deviation between the boundary coordinates of the transmission blind zone and the causal reference coordinates corresponding to the pre-classification results of the propagation causes does not exceed a preset position deviation threshold. For example, the preset position deviation threshold can be set to 3m. This threshold is set based on the mapping accuracy of the roadway facility distribution information and the statistical analysis of the solution error of the mine roadway waveguide propagation model. The core idea of ​​physical constraint verification is that the actual blind zone boundary is necessarily closely related to the physical location of the cause. If the solution result deviates significantly from the causal reference coordinates, it indicates that the model parameters are improperly selected or the pre-classification of the propagation causes is incorrect. This verification can eliminate physically unreasonable solution results.

[0086] In engineering implementation, the 10n in the denominator of the solution formula will not be zero when n is greater than 0. If an extreme calibration case of n being 0 occurs, a very small constant ε can be added to the denominator to ensure numerical stability. The origin reference coordinate depends on the pre-mapped information on the distribution of roadway facilities. If the information on the distribution of roadway facilities is missing, it can be set by engineers based on on-site surveys and the data source can be noted.

[0087] A deterministic propagation model based on ray tracing can also be used to calculate the propagation attenuation distance, but ray tracing has extremely high computational complexity, making it difficult to meet the requirements of online real-time detection. This embodiment uses a parameterized mine roadway waveguide propagation model because the model parameters can be calibrated offline, and the online calculation only requires a single exponentiation operation, achieving a balance between computational complexity and calculation accuracy. By introducing the mine roadway waveguide propagation model to convert the critical transmit power level into propagation attenuation distance, and combining it with causal reference coordinates for physical constraint verification, the critical power characteristics of the link layer can be accurately calculated into boundary coordinates in the roadway space, eliminating physically impossible calculation results, thereby improving the accuracy and reliability of blind zone boundary positioning.

[0088] This embodiment elaborates on the bidirectional broadcast candidate boundary fusion and retest update.

[0089] Specifically, the objective mechanism of bidirectional solution fusion is that the solution of a unidirectional propagation path may be affected by local occlusion, non-uniform multipath and hardware transceiver asymmetry, resulting in deviations in the candidate boundary coordinates of the solution in a single direction. However, the wireless transceiver devices at both ends of the virtual segment alternately act as transmitters to perform bidirectional solution, which can independently estimate the same blind zone boundary from two directions. By fusion or verification, the accuracy and robustness of boundary positioning can be improved.

[0090] For the fusion process, the deviation between the two candidate boundary coordinates is first calculated. Let the candidate boundary coordinates calculated by wireless transceiver A as the transmitter be... The candidate boundary coordinates calculated by wireless transceiver B as the transmitter are: The deviation between the coordinates of the two candidate boundary lines The formula is expressed as: ,in, The absolute deviation between the coordinates of the two candidate boundary lines is expressed in meters (m). , These are the candidate boundary coordinates calculated for the two transmission directions, in meters (m). A preset boundary deviation threshold is used to measure the consistency of the bidirectional calculation results. For example, it can be set to 2m. This threshold is set based on the theoretical error range of the bidirectional mine roadway waveguide propagation model, taking approximately twice the standard deviation. This allows for tolerance of reasonable measurement errors while identifying significant inconsistencies.

[0091] When the deviation between the two candidate boundary coordinates does not exceed the preset boundary deviation threshold, it indicates that the bidirectional solution results are highly consistent. The average value of the two candidate boundary coordinates is used as the boundary coordinate of the transmission blind zone, expressed by the formula: For example, for virtual segment VS2, the candidate boundary coordinates calculated by transmitter n2 are 31.9m, and the candidate boundary coordinates calculated by transmitter n3 are 32.1m, with an absolute deviation. Since the deviation is less than the preset boundary deviation threshold of 2m, the average value is taken. This serves as the boundary coordinate of the transmission blind zone in VS2. It is evident that fusing the bidirectional solution results can smooth out the random errors of the unidirectional solution and improve the stability of boundary positioning.

[0092] When the deviation between two candidate boundary coordinates exceeds the preset boundary deviation threshold, it indicates a significant discrepancy in the bidirectional solution results. In this case, the two candidate boundary coordinates are checked to ensure they meet the preset physical constraints: the absolute deviation between each candidate boundary coordinate and the corresponding causal reference coordinate in the propagation cause pre-classification result is calculated. Candidate boundary coordinates with a deviation not exceeding the preset position deviation threshold (3m) meet the preset physical constraints. If both candidate boundary coordinates meet the preset physical constraints, the one with the smaller deviation from the causal reference coordinate is taken as the transmission blind zone boundary coordinate. If only one candidate boundary coordinate meets the preset physical constraints, that candidate boundary coordinate is taken as the transmission blind zone boundary coordinate. If neither candidate boundary coordinate meets the preset physical constraints, the average of the two candidate boundary coordinates is taken as the transmission blind zone boundary coordinate, and a "to be verified" mark is added to the blind zone detection result to prompt maintenance personnel for on-site verification. For example, for virtual segment VS5, the deviation between the two candidate boundary coordinates 91.8m and 88.5m is 3.3m, which exceeds 2m. The deviations between the two and the reference coordinates for the formation of the roadway intersection 90.0m are 1.8m and 1.5m, respectively, both of which meet the preset physical constraints. Therefore, the smaller deviation, 88.5m, is taken as the transmission blind zone boundary coordinate of VS5. For virtual segment VS8, the deviation between the two candidate boundary coordinates 151.8m and 154.6m is 2.8m, which exceeds 2m. The deviations between the two and the reference coordinates for the formation of the roadway bend 150.0m are 1.8m and 4.6m, respectively. Only 151.8m meets the preset physical constraints. Therefore, 151.8m is taken as the transmission blind zone boundary coordinate of VS8.

[0093] To verify the effectiveness of bidirectional solution fusion and physical constraint verification, this embodiment presents the bidirectional solution and verification results for each potential transmission blind zone, as shown in Table 4.

[0094] Table 4: Results of Two-Way Calculation and Physical Constraint Verification of Transmission Dead Zone Boundary Coordinates

[0095] As shown in Table 4, the deviation between the two candidate boundary coordinates of virtual segment VS2 (31.9m and 32.1m) is 0.2m, which does not exceed the preset boundary deviation threshold of 2m. The average value of 32.0m is taken as the boundary coordinate of the transmission blind zone. Its deviation from the reference coordinate of 32.0m (the cause of metal equipment obstruction) is 0.0m, satisfying the preset physical constraint conditions. The deviation between the two candidate boundary coordinates of virtual segment VS5 is 3.3m, exceeding the threshold. After physical constraint verification, both meet the conditions. The smaller deviation from the intersection center coordinate of 90.0m, 88.5m, is taken. The deviation between the two candidate boundary coordinates of virtual segment VS8 is 2.8m, exceeding the threshold. After verification, only 151.8m meets the conditions; therefore, 151.8m is taken as the boundary coordinate of the transmission blind zone. The final boundary coordinates of the three potential transmission blind zones all deviate from their respective cause reference coordinates by no more than 1.8m, indicating that bidirectional solution fusion and physical constraint verification can effectively handle various situations of bidirectional solution consistency and divergence, verifying the correctness and completeness of the candidate boundary fusion rules.

[0096] After outputting the blind zone detection results, this embodiment also re-collects the received signal strength sequence of each wireless transceiver device according to a preset retesting period (e.g., 30 minutes) and extracts wireless propagation statistical features. The re-extracted wireless propagation statistical features are matched one by one with the multipath propagation feature fingerprint database to obtain the propagation cause retesting results of each wireless transceiver device. When the propagation cause retesting results of the wireless transceivers at both ends of the virtual segment under test are inconsistent with the propagation cause pre-classification results of the previous retesting period, it indicates that the channel conditions of the virtual segment under test have changed. The transmit power graded detection and link response curve construction are re-executed, that is, the measurement association device of the virtual segment under test is triggered to increase the transmit power step by step according to the preset transmit power level and broadcast the detection message. The transmission blind zone boundary coordinates are recalculated according to the updated link response curve, and the blind zone detection results of the virtual segment under test are updated. The preset retest cycle is a compromise between tracking timeliness and the power consumption of the wireless transceiver: the typical time scale for equipment movement and changes in operating conditions in mine roadways is tens of minutes. The 30-minute retest cycle can capture the evolution of blind spots in a timely manner without excessively consuming the power of the wireless transceiver due to frequent collaborative broadcasts.

[0097] By using bidirectional candidate boundary fusion and physical constraint verification in branch decision-making, random errors can be smoothed out, directional solution deviations can be identified and eliminated, and unreliable results can be marked for verification, thus avoiding the problem of direct output of unidirectional solution deviations. Through periodic retesting and updating, the dynamic changes of roadway conditions can be tracked, thereby ensuring the accuracy and timeliness of blind spot detection results throughout the entire life cycle.

[0098] To more clearly illustrate the technical solution and effects of this application, the following section uses a wireless sensor network deployed in a mine transport roadway as an example to provide a detailed explanation of the mine roadway wireless link blind zone detection method based on cooperative broadcasting.

[0099] Specifically, the scenario is set in a mine transport roadway with a total length of 160m, width of 5m, and height of 3.5m. Nine wireless transceivers, numbered n1 to n9, are deployed at equal intervals along the longitudinal direction of the roadway, with a designed spacing of 20m. Typical operating conditions exist within the roadway: a metal device forms an obstruction approximately 32m between n2 and n3; a roadway fork occurs approximately 90m between n5 and n6; and a roadway bend begins approximately 150m between n8 and n9. The wireless transceivers use 2.4GHz wireless communication modules with an adjustable transmission power range of -10dBm to 20dBm. The key parameters involved in this embodiment are set as follows: preset response delay 0.1ms; preset ranging deviation threshold 0.3m; preset sliding time window 10s; detection message reception period 10ms; preset transmission distance threshold 30m; preset transmission power level sequence of -10dBm to 20dBm, step 5dBm; preset reception duration 1s; preset connectivity threshold 0.8; preset reception sensitivity threshold -60dBm; preset position deviation threshold 3m; preset boundary deviation threshold 2m; preset quantity 3; preset retest period 30min.

[0100] First, the positioning of wireless transceivers and the division of virtual segments are performed. Each wireless transceiver exchanges ranging messages with its neighboring wireless transceivers. The round-trip propagation time is calculated based on the transmission time, reception time, and preset acknowledgment delay. After deducting the preset acknowledgment delay and taking half of it, the result is divided by the electromagnetic wave propagation speed 3 × 10⁻⁶. 8 Multiplying by m / s gives the adjacent spacing. For example, the total round-trip propagation time obtained from three measurements between wireless transceivers n2 and n3 are 100132.4ns, 100131.6ns, and 100385.2ns, respectively. Subtracting 1×10 5 After calculating the preset response delay of ns, the ranging values ​​were 19.86m, 19.74m, and 57.78m. Among them, the deviation of 57.78m from the average of the other two values ​​of 19.80m was 37.98m, which exceeded the preset ranging deviation threshold of 0.3m and was therefore discarded. The final adjacent spacing of this segment was 19.80m. After completing all measurements, with the wireless transceiver n1 at the entrance of the tunnel as the origin, the adjacent spacing of each segment was accumulated to obtain the longitudinal coordinates of each wireless transceiver. The longitudinal coordinates of wireless transceivers n2 to n9 were 20.10m, 39.90m, 60.05m, 80.00m, 100.30m, 120.00m, 140.00m, and 159.80m, respectively. Virtual segments VS1 to VS8 were formed between two wireless transceivers with adjacent longitudinal coordinates.

[0101] Subsequently, each wireless transceiver performs wireless propagation statistical feature extraction. Each wireless transceiver receives probe messages broadcast by neighboring wireless transceivers at 10ms intervals, constructs a received signal strength sequence, and after truncating it with a 10s sliding time window, calculates the variance, skewness, and kurtosis according to the aforementioned formulas. These are then fitted with Rayleigh and Rice distribution models to determine the fading distribution type. For example, the received signal strength sequence of wireless transceiver n2 yields a variance of 4.3dBm. 2 The skewness is -1.24 and the kurtosis is 2.61. Its fitting error with the Rayleigh distribution model (0.013) is less than the fitting error with the Rice distribution model (0.048), therefore it is determined to be a Rayleigh distribution. The calculation and determination results of the wireless propagation statistical characteristics of each wireless transceiver are shown in Table 2. The wireless propagation statistical characteristics of wireless transceivers n2, n3, n5, n6, n8, and n9 are significantly abnormal and all exhibit Rayleigh distributions, indicating that they are in a channel environment with significant multipath effects or severe obstruction.

[0102] Next, propagation cause pre-classification and potential transmission blind zone determination are performed. The wireless propagation statistical characteristics of each wireless transceiver are matched one by one with fingerprint patterns in a pre-built multipath propagation feature fingerprint database. The cause corresponding to the fingerprint pattern with the highest similarity is taken as the propagation cause pre-classification result. For example, wireless transceivers n2 and n3 have the highest similarity to the fingerprint pattern of metal equipment obstruction; wireless transceivers n5 and n6 have the highest similarity to the fingerprint pattern of road intersections; wireless transceivers n8 and n9 have the highest similarity to the fingerprint pattern of road bends; and the propagation cause pre-classification result for wireless transceivers n1, n4, and n7 is normal propagation cause. Virtual segment-by-virtual segment analysis: For VS2, both wireless transceivers n2 and n3 have abnormal propagation caused by metal equipment obstruction; for VS5, both wireless transceivers n5 and n6 have abnormal propagation caused by tunnel intersections; for VS8, both wireless transceivers n8 and n9 have abnormal propagation caused by tunnel bends. All are identified as potential transmission blind spots. For the remaining virtual segments, the wireless transceivers at both ends do not have simultaneous abnormal propagation causes, and the adjacent distances do not exceed 30m; therefore, they are not identified as potential transmission blind spots. The analysis results are as follows: Figure 2 As shown in Table 3.

[0103] Next, cooperative broadcasting and link response curve plotting are performed for potential transmission blind spots. Taking VS2 as an example, the measurement associated devices are n1, n2, n3, and n4. First, n2 acts as the transmitter, and n1, n3, and n4 act as the receivers. Starting from -10dBm, n2 broadcasts probe messages with its transmission power increasing in 5dBm increments. The receivers count the number of successfully received probe messages within 1 second and calculate the reception success rate. For example, at 10dBm, the reception success rates of n1, n3, and n4 are 0.68, 0.75, and 0.73, respectively, with an average reception success rate of 0.72. Subsequently, n3 acts as the transmitter, and the above process is repeated. Two link response curves are plotted with the transmission power level on the horizontal axis and the average reception success rate on the vertical axis, as shown below. Figure 3 As shown.

[0104] Subsequently, blind zone boundary calculation was performed. On the link response curve in the n2 direction, the average reception success rate at 15dBm was 0.85, while the three consecutive transmit power levels of 10dBm, 5dBm, and 0dBm were 0.72, 0.61, and 0.44 respectively, all falling below the preset connectivity threshold of 0.8. Therefore, the critical transmit power level was 10dBm. The fading distribution type of VS2 is Rayleigh distribution, corresponding to model parameters n=2.8, L0=40dB, and an effective link margin of 70dB. Substituting these values ​​into the calculation formula yielded a propagation attenuation distance of approximately 11.8m. This distance was superimposed with the longitudinal coordinate of n2 (20.10m) along the positive longitudinal direction of the channel to obtain the candidate boundary coordinates of 31.9m. Similarly, the critical transmit power level in the n3 direction was 5dBm, with a propagation attenuation distance of approximately 7.8m. This distance was superimposed with the longitudinal coordinate of n3 (39.90m) along the negative longitudinal direction of the channel to obtain the candidate boundary coordinates of 32.1m. The attenuation curve and calculation process are as follows: Figure 4 As shown.

[0105] Finally, candidate boundary fusion, physical constraint verification, and result output are performed. The deviation between the two candidate boundary coordinates of VS2 (31.9m and 32.1m) is 0.2m, which does not exceed the preset boundary deviation threshold of 2m. The average value of 32.0m is taken as the boundary coordinate of the transmission blind zone. According to the preset roadway facility distribution information, the reference coordinate for the cause of metal equipment obstruction is 32.0m. The deviation between the boundary coordinate of the transmission blind zone and the reference coordinate is 0.0m, which does not exceed the preset position deviation threshold of 3m. The preset physical constraint conditions are met, and the blind zone detection result is output: Virtual segment VS2 has a blind zone, the boundary coordinate of the transmission blind zone is 32.0m, and the cause of the blind zone is metal equipment obstruction. Similarly, VS5 outputs the boundary coordinate of the transmission blind zone as 88.5m, and the cause of the blind zone is a roadway intersection. VS8 outputs the boundary coordinate of the transmission blind zone as 151.8m, and the cause of the blind zone is a roadway bend. Subsequently, the system regenerates the wireless propagation statistical features and performs retesting pre-classification according to a preset retesting cycle of 30 minutes. When the retesting result of the propagation cause of any virtual segment changes, the system triggers the virtual segment to re-execute cooperative broadcasting and boundary calculation, and dynamically updates the blind zone detection results.

[0106] As can be seen from the operation process of the above specific scenarios, the method of this embodiment can rely on the cooperative broadcasting mechanism between wireless transceivers to autonomously complete the construction of spatial reference, channel status identification, screening of potential transmission blind spots, precise location of blind spot boundaries and determination of causes in the mine roadway environment where wireless transceivers are deployed. The output blind spot detection results are highly consistent with the actual physical locations of metal equipment, intersections, turns and other facilities in the roadway, providing accurate decision-making basis for blind spot compensation and topology optimization of the mine safety monitoring network.

[0107] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0108] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0109] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for detecting blind spots in wireless links in mine roadways based on cooperative broadcasting, characterized in that, include: Several transceiver devices are deployed in the mine roadway. The wireless transceiver devices send ranging messages to each other to measure the distance between adjacent wireless transceiver devices. Taking the wireless transceiver device at the entrance of the roadway as the origin, the distance is accumulated segment by segment according to the arrangement order along the longitudinal direction of the roadway to obtain the longitudinal coordinate of each wireless transceiver device. The roadway space between two wireless transceiver devices with adjacent longitudinal coordinates is defined as a virtual segment. Each wireless transceiver receives probe messages transmitted by adjacent wireless transceivers, forms a sequence of received signal strength, and extracts variance, skewness, kurtosis and fading distribution type based on a preset sliding time window to obtain the statistical characteristics of wireless propagation. The wireless propagation statistical features are matched with the pre-established multipath propagation feature fingerprint database to obtain the propagation cause pre-classification results of each wireless transceiver. When the propagation cause pre-classification results of the wireless transceivers at both ends of the virtual segment are both abnormal propagation causes, or the distance between the two wireless transceivers is greater than the preset transmission distance threshold, the virtual segment is determined to be a potential transmission blind zone. For potential transmission blind spots, control the wireless transceiver devices associated with the potential transmission blind spot to transmit detection messages at a preset transmission power level. The receiving end counts the reception success rate at different transmission power levels and constructs a link response curve between the transmission power level and the average reception success rate. The critical transmit power level is determined based on the link response curve. The wireless propagation statistical characteristics corresponding to the critical transmit power level are substituted into the waveguide propagation model of the mine roadway to calculate the propagation attenuation distance of the detection message. The coordinates of the transmission blind zone boundary are obtained by combining the longitudinal coordinates of the transmitter. When the boundary coordinates of the transmission blind zone meet the preset physical constraints corresponding to the pre-classification results of the propagation cause, the output includes the boundary coordinates of the transmission blind zone and the cause of the blind zone.

2. The method according to claim 1, characterized in that, The distance between adjacent wireless transceivers is measured based on the propagation time of the ranging message, including: Each wireless transceiver sends a ranging message to its neighboring wireless transceiver and records the sending time. The neighboring wireless transceiver returns a response message after a preset response delay. The wireless transceiver that sent the ranging message records the time when it receives the response message. The round-trip propagation time of the ranging message is calculated based on the sending time, the receiving time, and the preset response delay. The propagation time after deducting the response delay is halved and multiplied by the electromagnetic wave propagation speed to obtain the distance between the neighboring wireless transceivers. The distance between adjacent wireless transceivers is measured multiple times. When the deviation of a certain measurement result from the average of the other measurement results exceeds the preset ranging deviation threshold, the measurement result is discarded, and the average of the remaining measurement results is taken as the corresponding adjacent distance. The longitudinal coordinates of each wireless transceiver are obtained by accumulating the distances between adjacent devices along the longitudinal direction of the roadway.

3. The method according to claim 1, characterized in that, The fading distribution type is determined by fitting the received signal strength sequence with the Rayleigh distribution model and the Rice distribution model respectively, obtaining the corresponding fitting errors, and determining the distribution model with the smaller fitting error as the fading distribution type corresponding to the received signal strength sequence.

4. The method according to claim 1, characterized in that, The method for establishing the multipath propagation feature fingerprint database is as follows: Multiple typical propagation conditions are simulated in the mine roadway. The received signal strength sequence is collected under each typical propagation condition and the corresponding wireless propagation statistical features are extracted. The propagation causes of each typical propagation condition are associated with and stored with the corresponding wireless propagation statistical features to form the multipath propagation feature fingerprint database. The similarity between the wireless propagation statistical features to be matched and each fingerprint pattern in the multipath propagation feature fingerprint database is calculated, and the propagation cause corresponding to the fingerprint pattern with the highest similarity is used as the propagation cause pre-classification result of the corresponding wireless transceiver. The causes of the various typical propagation conditions include normal propagation causes and abnormal propagation causes. Abnormal propagation causes represent the types of causes for which objectively non-existent propagation conditions do not exist.

5. The method according to claim 1, characterized in that, The construction of the link response curve includes: The two ends of the wireless transceiver device in the virtual segment to be tested where the potential transmission blind zone is located, and the adjacent wireless transceiver device located outside the two ends of the wireless transceiver device are selected as the measurement association device. The wireless transceiver devices at both ends of the virtual segment under test are used as transmitters in sequence, and the remaining measurement and association devices are used as receivers. The transmitting end increases the transmission power step by step according to the preset transmission power level and transmits detection messages. The receiving end records the reception success rate at each transmission power level and extracts the corresponding wireless propagation statistical features. The average reception success rate of each receiver is used as the average reception success rate under the corresponding transmission power level, and the link response curve is plotted with the transmission power level as the horizontal axis and the average reception success rate as the vertical axis.

6. The method according to claim 5, characterized in that, When the transmitting end increases the transmission power step by step according to the preset transmission power level and transmits the detection message, the preset transmission power level is arranged in order from low to high, and the transmitting end increases the transmission power step by step from the lowest transmission power level. At each transmission power level, the receiver counts the number of probe messages successfully received within a preset reception time, and uses the ratio of the number of successfully received probe messages to the total number of probe messages transmitted at that transmission power level as the reception success rate of the receiver at that transmission power level.

7. The method according to claim 1, characterized in that, Substituting the wireless propagation statistical characteristics corresponding to the critical transmit power level into the waveguide propagation model of the mine roadway, the propagation attenuation distance of the detection message is calculated, including: The waveguide propagation model of the mine roadway is set as a path loss model that characterizes the attenuation of the received signal strength with the propagation distance, and the model parameters of the path loss model are made to correspond to different fading distribution types. Based on the fading distribution type in the wireless propagation statistical characteristics corresponding to the critical transmit power level, select the corresponding model parameters; Substituting the transmission power corresponding to the critical transmission power level and the preset receiving sensitivity threshold into the path loss model, the propagation attenuation distance of the probe message is obtained.

8. The method according to claim 1, characterized in that, The preset physical constraints are set as follows: Based on the preset roadway facility distribution information, cause reference coordinates are set for each abnormal propagation cause. Among them, the cause reference coordinates for the cause of metal equipment obstruction are the coordinates of the metal equipment placement location, the cause reference coordinates for the cause of roadway intersection are the coordinates of the intersection center, the cause reference coordinates for the cause of roadway bend are the coordinates of the bend start point, and the cause reference coordinates for the cause of exceeding the transmit / receive spacing limit are the coordinates of the midpoint of the virtual segment to be tested where the potential transmission blind zone is located. The preset physical constraint is set as follows: the deviation between the boundary coordinates of the transmission blind zone and the causal reference coordinates corresponding to the propagation cause pre-classification result does not exceed the preset position deviation threshold.

9. The method according to claim 8, characterized in that, When the wireless transceivers at both ends of the virtual segment under test alternately transmit probe messages as transmitters, the propagation attenuation distance corresponding to each transmission is calculated according to the corresponding link response curves. The propagation attenuation distance corresponding to each transmission is then superimposed with the longitudinal coordinates of the corresponding transmitter along the longitudinal direction of the tunnel to obtain the candidate boundary coordinates corresponding to each transmission. When the deviation between two candidate boundary coordinates does not exceed a preset boundary deviation threshold, the average value of the two candidate boundary coordinates is used as the boundary coordinate of the transmission blind zone. When the deviation between two candidate boundary coordinates exceeds a preset boundary deviation threshold, the two candidate boundary coordinates are checked to see if they meet the preset physical constraints. If both candidate boundary coordinates meet the preset physical constraints, the one with the smaller deviation from the causal reference coordinate is taken as the transmission blind zone boundary coordinate. If only one candidate boundary coordinate meets the preset physical constraints, that candidate boundary coordinate is taken as the transmission blind zone boundary coordinate. If neither candidate boundary coordinate meets the preset physical constraints, the average of the two candidate boundary coordinates is taken as the transmission blind zone boundary coordinate, and a check mark is added to the blind zone detection result.

10. The method according to any one of claims 1 to 9, characterized in that, After outputting the blind zone detection results, it also includes: The received signal strength sequence is re-acquired according to the preset retesting cycle and the wireless propagation statistical features are extracted. The re-extracted wireless propagation statistical features are matched with the multipath propagation feature fingerprint database to obtain the corresponding propagation cause retesting results. When the propagation cause retest results of the wireless transceivers at both ends of the virtual segment are inconsistent with the propagation cause preclassification results of the previous retest period, the transmission power classification detection and link response curve construction steps are re-executed, and the transmission blind zone boundary coordinates are recalculated based on the updated link response curve to update the blind zone detection results of the corresponding virtual segment.

Images