Layered transmission and abnormal supplement transmission method and system of tunnel geological exploration while-drilling data

Abstract

The present application relates to the technical field of data transmission, and discloses a layered transmission and abnormal retransmission method and system for tunnel geological exploration while drilling data, comprising: obtaining a while-drilling data frame covering a comprehensive rock breaking ability, extracting a cutting lag amount and correcting a measured point formation depth therefrom; then combining the measured point formation depth and the cutting lag amount to extract an abnormal fluctuation component, calculating a transmission priority and a bandwidth allocation rate of different attribute data groups; taking a single drill rod span as a cache segment, generating a depth data packet with a chain hash value, and scheduling packet sending according to a sending weight value; after packet sending, comparing observed and predicted link state quantization transmission interference, calculating retransmission priority for unconfirmed packets to trigger retransmission; and at the receiving end, interpolating and aligning received packets based on the measured point formation depth to reorganize a continuous geological profile. The present application effectively avoids spatial semantic dislocation of while-drilling data in the transmission and retransmission process, and improves the spatial matching degree of geological prediction.

Description

Technical Field

[0001] This invention relates to the field of data transmission technology, and more specifically, to a method and system for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling. Background Technology

[0002] Horizontal directional drilling (WDD) for tunnel exploration typically involves long-distance exploration boreholes along the tunnel axis. Drilling data, including drilling speed, mud pressure, and bottom hole measurements, are widely used for surrounding rock identification and advanced geological forecasting. In conventional WDD operations, the drill pipe, joints, and borehole mud together form the information transmission channel. When the drill bit enters a fault fracture zone or water-rich zone, drilling pressure, mud flow field, and borehole wall contact conditions change drastically. This often leads to geological anomalies, accompanied by worsening mud pulsation and channel attenuation, resulting in the continuous loss of valuable geological data during this period. Furthermore, due to gravity and drill pipe eccentricity, the transport of cuttings within the borehole not only exists in a suspended state but also forms a slowly moving cuttings bed along the borehole bottom. This asymmetric flow field distribution of mud and cuttings results in a significant and dynamically changing retention deviation between the actual formation depth generated by WDD parameters and the corresponding data measured on the surface. Existing layered transmission and automatic retransmission protocols mostly rely on fixed time windows and arrival order for packet retransmission, without considering the spatial calibration errors caused by the aforementioned delay bias. Once the system performs a retransmission, the retransmitted data is often directly attached to the current drill bit depth, rather than the actual stratigraphic source segment from which the data originated. This leads to spatial semantic misalignment after data reconstruction, resulting in spatial-level deviations in the location and assessment of adverse geological bodies ahead. Summary of the Invention

[0003] This invention provides a method and system for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling, which solves the technical problems mentioned in the background art.

[0004] This invention provides the following technical solution: Firstly, a method for layered transmission and anomaly retransmission of tunnel geological exploration data while drilling is applied to a horizontal directional drilling system comprising a drilling rig end, a measuring end, and a receiving end, including: Drilling parameters and data while drilling are collected, packaged into data frames containing comprehensive rock breaking energy according to time windows, and the data frames while drilling are divided into data groups with different attributes. Extract the cuttings hysteresis of the drilling data frame to eliminate the physical delay interference caused by drill pipe eccentricity and annular cuttings bed, and correct the current hole depth based on the cuttings hysteresis to obtain the formation depth at the measuring point; For each of the data groups, the abnormal fluctuation characteristics are extracted by combining the formation depth of the measuring point and the cuttings hysteresis, and the transmission priority and bandwidth allocation rate of each data group are calculated. Using the length of a single drill pipe as the physical buffer boundary, the drilling data frame is spatially mapped according to the formation depth of the measuring point to generate a depth data packet with an anti-tampering hash value; The channel adaptability of the depth data packet is evaluated using the transmission priority and the cuttings lag, a transmission weight value is obtained, and a scheduled packet transmission is performed to the receiving end accordingly. After the packet is sent, the measured link state is compared with the predicted link state to quantify the transmission interference. The retransmission priority of the unacknowledged lost packets is calculated in combination with the rock cuttings lag to trigger the retransmission queue. At the receiving end, spatial interpolation and alignment are performed on the received depth data packets based on the stratum depth of the measuring point, and the resulting continuous geological profile with out-of-order interference is reconstructed.

[0005] Secondly, the layered transmission and anomaly retransmission system for tunnel geological exploration data during drilling includes the drilling rig end, the measurement end, and the receiving end; The drilling rig end is responsible for data frame encapsulation, cuttings hysteresis extraction and measurement point formation depth correction. The drilling rig end also performs priority calculation functions; The drilling rig end also undertakes the function of data packet generation; The drilling rig end also undertakes scheduling and contracting functions; The drilling rig end also undertakes the function of retransmission queue management; The measuring end is responsible for synchronous acquisition of drilling parameters and clock synchronization. The receiving end is responsible for data packet reception, spatial interpolation alignment, hash integrity verification, and geological profile reconstruction and archiving.

[0006] The beneficial effects of this invention are as follows: This invention overcomes the limitations of conventional drilling data communication that treats transmission interference and geological disturbances in isolation. By introducing cuttings hysteresis to correct the borehole depth, it transforms the data retention delay in the mud circulation channel into a quantifiable basis for spatial compensation and dynamically adjusts the bandwidth allocation of data at different levels. At the same time, it uses the span of a single drill pipe entering the well as the buffer benchmark for data archiving, so that even after experiencing channel fading, packet loss, and delayed retransmission, the ground receiver can still seamlessly stitch and spatially align the data according to the true original stratum source depth. This effectively avoids spatial semantic misalignment errors in surrounding rock identification and the location of adverse geological bodies, and improves the spatial restoration of long-distance tunnel exploration data. Attached Figure Description

[0007] Figure 1 This is a flowchart of the method and system for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling, as described in this invention. Detailed Implementation

[0008] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0009] like Figure 1 As shown, a layered transmission and anomaly retransmission method and system for tunnel geological exploration data during drilling is applied to a horizontal directional drilling system comprising a drilling rig end, a measuring end, and a receiving end, including: Drilling parameters and data while drilling are collected, packaged into data frames containing comprehensive rock breaking energy according to time windows, and the data frames while drilling are divided into data groups with different attributes. Extract the cuttings hysteresis of the drilling data frame to eliminate the physical delay interference caused by drill pipe eccentricity and annular cuttings bed, and correct the current hole depth based on the cuttings hysteresis to obtain the formation depth at the measuring point; For each of the data groups, the abnormal fluctuation characteristics are extracted by combining the formation depth of the measuring point and the cuttings hysteresis, and the transmission priority and bandwidth allocation rate of each data group are calculated. Using the length of a single drill pipe as the physical buffer boundary, the drilling data frame is spatially mapped according to the formation depth of the measuring point to generate a depth data packet with an anti-tampering hash value; The channel adaptability of the depth data packet is evaluated using the transmission priority and the cuttings lag, a transmission weight value is obtained, and a scheduled packet transmission is performed to the receiving end accordingly. After the packet is sent, the measured link state is compared with the predicted link state to quantify the transmission interference. The retransmission priority of the unacknowledged lost packets is calculated in combination with the rock cuttings lag to trigger the retransmission queue. At the receiving end, spatial interpolation and alignment are performed on the received depth data packets based on the stratum depth of the measuring point, and the resulting continuous geological profile with out-of-order interference is reconstructed.

[0010] This embodiment is applied to a horizontal directional drilling system comprising a drilling rig, a measurement rig, and a receiving rig. The drilling rig is responsible for data frame encapsulation, priority calculation, data packet generation, scheduling packet transmission, and retransmission queue management. The measurement rig is responsible for synchronous acquisition of drilling parameters and clock synchronization. The receiving rig is responsible for data packet reception, spatial interpolation alignment, hash integrity verification, and geological profile reconstruction and archiving. All acquisition devices and computing nodes within the system are clock synchronized via Network Time Protocol (NTP), with synchronization errors controlled within 10% of the sampling interval. All acquired data corresponds to the same synchronization clock stamp, ensuring the temporal consistency of the data.

[0011] Drilling data is divided into four groups based on attributes: key measurement data, rig status data, auxiliary data, and communication status data. The key measurement data group includes formation physical parameters and borehole environmental parameters; the rig status data group includes rig operating condition parameters and equipment operating status parameters; the auxiliary data group includes equipment calibration parameters and environmental correction parameters; and the communication status group includes channel real-time parameters and link transmission parameters. All data groups maintain a consistent acquisition frequency and sampling interval. All acquired data undergoes a preprocessing process, which employs the following preprocessing method: The criteria remove outliers, linear interpolation fills in missing values, and moving average filtering suppresses random noise. The sliding window length is set to 5 to 15 sampling points, which can be dynamically adjusted according to drilling conditions.

[0012] To suit different computational scenarios, zero-prevention constants with corresponding dimensions are set, namely, zero-prevention constants for length dimensions. The range of values ​​is to Time dimension prevents zero constant The range of values ​​is to Pressure dimension is zero constant The range of values ​​is to ; Dimensionless zero constant The range of values ​​is to Information dimension: zero constant The range of values ​​is to All zero-prevention constants are used to eliminate zero singularities in the corresponding computational scenarios and do not contribute to the numerical values ​​of core physical quantities.

[0013] To eliminate drilling statistical errors caused solely by time accumulation, the time difference between the current and previous sampling moments is calculated as the sampling interval. The calculation formula is as follows: ; in, This is the synchronization clock stamp for the current sampling time, in seconds. This is the synchronization clock stamp of the previous sampling time, in seconds; This represents the sampling interval corresponding to the current sampling time, in seconds (s). The basic range of the sampling interval is 1 ms to 1 s, which can be adjusted according to the dynamic changes in drilling conditions. When the drilling speed fluctuation exceeds a preset threshold, the sampling interval is reduced to increase the sampling density; when the drilling state is stable, the sampling interval is increased to reduce the computational load.

[0014] Extract drilling speed, axial feed force, torque, rotational speed, mud pressure, and mud flow rate to characterize the combined energy output of mechanical rock breaking and hydraulic rock clearing. Simultaneously calculate the hole depth at the current sampling moment using the following formula: ; in, The drilling speed at the current sampling moment is expressed in m / s. The depth of the hole at the previous sampling time, in meters; This represents the hole depth at the current sampling time, in meters (m). Among the parameters acquired synchronously, The axial feed force at the current sampling moment, in N; The torque at the current sampling time, in N. m; The rotational speed at the current sampling time is expressed in r / min. The mud pressure at the current sampling time is expressed in Pa. The mud flow rate at the current sampling time is expressed in m³. / s. All parameters are synchronously acquired by the corresponding sensor at the measurement end, and the acquisition clock is consistent with the synchronization clock stamp.

[0015] By calculating the cumulative work done by the combined energy output within the sampling interval, and introducing a length-dimension zero-prevention constant divided by the difference in borehole depth between the current and previous moments to eliminate zero-depth singularity interference, the comprehensive rock-breaking energy per unit depth is obtained. The calculation formula is as follows: ; in, This represents the total rock-breaking energy per unit depth at the current sampling time, expressed in J / m. In the numerator of the formula... This represents the mechanical rock-breaking power corresponding to the axial feed force, in W. This represents the rotary rock-breaking power corresponding to torque and rotational speed, in W. This represents the hydraulic rock-clearing power corresponding to mud pressure and flow rate, in W. The sum of the three power values ​​and the sampling interval are also considered. Multiplying these components yields the total cumulative work done by mechanical rock breaking and hydraulic rock clearing within the sampling interval, expressed in J. The denominator is the sum of the absolute value of the hole depth difference within the current sampling interval and the zero-constant in length dimension, thus normalizing the total work done to energy per unit depth and avoiding the singularity of zero denominator in zero-depth conditions.

[0016] The current time, borehole depth, comprehensive rock-breaking energy, key measurement data, and drilling rig status data are combined and encapsulated into a drilling-while-drilling data frame to provide a physically consistent data benchmark. The data frame structure is as follows: ; in, The drilling data frame encapsulated at the current sampling moment; This is the key measurement data set at the current sampling time; This is the drilling rig status data set at the current sampling time; This is the auxiliary data set for the current sampling time; This represents the communication status group at the current sampling time. All parameters within the drilling data frame correspond to the same synchronization clock stamp. This ensures the temporal and physical origin of the data.

[0017] To quantify the cross-sectional change caused by drill pipe eccentricity, the theoretical annular area is calculated from the borehole diameter and drill pipe outer diameter using the following formula: ; in, The theoretical annular area at the current sampling time, in meters. ; The borehole diameter at the current sampling time is in meters. It is obtained by combining the nominal diameter of the drill bit with the borehole diameter expansion rate, which is calibrated from historical drilling data under the same formation conditions. The outer diameter of the drill pipe at the current sampling time, in meters, is determined by the nominal structural parameters of the drill pipe used in the well. The theoretical annulus area is the annular cross-sectional area between the borehole inner wall and the drill pipe outer wall, and is the main flow channel for mud carrying rock back up.

[0018] To eliminate the base pressure interference caused by mud circulation, the flow field smoothness is calculated by combining the residual of the measured pressure deviating from the cuttings-free reference pressure with the pressure dimensionless constant. The calculation formula is as follows: ; in, represents the flow field smoothness at the current sampling time, which is a dimensionless parameter with a value range of (0,1]. The cuttings-free reference pressure at the current sampling time, in Pa, is obtained through mud pressure calibration under the same drilling parameters and cuttings-free circulation. The calibration process is completed during the mud circulation stage before drilling, and the mud flow rate and rotation speed parameters are consistent with the actual drilling conditions. Flow field unobstructedness is used to characterize the degree of blockage of the mud flow field by the cuttings bed in the annulus. The greater the thickness of the cuttings bed, the greater the deviation between the measured mud pressure and the cuttings-free reference pressure, and the smaller the flow field unobstructedness value.

[0019] Combining mud flow rate, theoretical annular area, and flow field smoothness, the equivalent upward return velocity of mud carrying rock is calculated using the following formula: ; in, This represents the equivalent upward velocity of the mud carrying rock at the current sampling time, expressed in m / s. The equivalent upward velocity is the actual average upward velocity of the mud-rock mixture within the annulus after considering the impact of rock cuttings bed blockage, providing a fundamental parameter for calculating the rock cuttings upward velocity time.

[0020] By integrating the equivalent upward return velocity backwards over time to the matching annular path length, the upward return of cuttings is determined to characterize the flow field retention effect. The calculation formula is as follows: ; in, It is an equivalent time series function with upward reversal. The loop path length at the current sampling time, in meters, is the same as the current hole depth. The values ​​are consistent; The cuttings return time at the current sampling moment is expressed in seconds (s), representing the time required for cuttings to rise from the bottom of the borehole with the drilling mud to the surface measurement end. The cuttings return time is numerically solved using the fourth-order Runge-Kutta method, with the integration step size set to 1 / 10 of the sampling interval. The iteration termination condition is that the relative error between the integration result and the annular path length is less than 0.1%. During the solution process, the dynamic update of the annular path length is simultaneously locked to avoid integration boundary shifts caused by changes in drilling progress.

[0021] The cuttings hysteresis is obtained by integrating the drilling speed over the time it takes for the cuttings to return to the surface. The calculation formula is as follows: ; in, The time series function of drilling speed; The cuttings lag at the current sampling time, expressed in meters (m), represents the depth to which the drill bit has advanced during the time it takes for the cuttings to return from the bottom of the borehole to the surface. It characterizes the spatial deviation between the formation information carried by the cuttings and the current drill bit position. The cuttings lag is calculated using the trapezoidal numerical integration method, with all sampling points within the corresponding time window as integration nodes to ensure the accuracy of the integration results.

[0022] The cuttings lag is subtracted from the current borehole depth to remove spatial delay interference and reconstruct the true geological origin of the stratigraphic depth at the measuring point. The calculation formula is as follows: ; in, The depth of the formation at the current sampling time is expressed in meters (m), representing the actual formation source depth corresponding to the drilling data. To address the coupling relationship between cuttings hysteresis, cuttings uptake time, and borehole depth, a fixed-point iterative method is employed for decoupling. The initial value for iteration is the convergence result from the previous sampling time, and the iteration terminates when the iterative change in the formation depth at the measuring point is less than [a certain value]. If the iteration count exceeds the preset upper limit and the convergence is still not achieved, the cuttings lag at the previous sampling time is used for correction, and an abnormal working condition alarm is triggered.

[0023] To eliminate the influence of differences in physical dimensions and numerical magnitudes between different data sets on the calculation results, min-max normalization was performed on the measured values ​​of each data set, mapping all parameters to the dimensionless interval [0,1]. The calculation formula s is: ; Among them, superscript It serves as a unique identifier for the data group, corresponding to the four attribute data groups divided within the drilling data frame; For the first The measured values ​​of each data set at the current sampling time; and The first The theoretical minimum and maximum values ​​of each data set under the corresponding formation conditions are obtained by calibration based on industry standards and historical drilling data. For the first The normalized measured values ​​of each data set at the current sampling time are dimensionless parameters.

[0024] To capture data anomalies caused by geological mutations, the normalized measured values ​​of each data set are subtracted from the predicted values ​​arranged according to the stratigraphic depth of the measurement point, and the innovation error is calculated using the following formula: ; in, For the first Each data set is located at the current measurement point at the current stratigraphic depth. The corresponding normalized predicted value is a dimensionless parameter; For the first The innovation error of each data set at the current sampling time is a dimensionless parameter that characterizes the degree of deviation between the normalized measured value and the predicted value. The predicted value is calculated using a linear Kalman filter algorithm, and the state equation is set as follows: The observation equation is set as follows ,in Let be the system state vector at time k. For process noise, To observe noise, the process noise covariance and the observation noise covariance are obtained from historical data calibration. When there is no historical data in the initial stage of drilling, the calibration data under the same formation conditions are used as the cold start initial value for the predicted value. The Kalman filter model is updated synchronously with the measured data of each sampling point. When the cumulative footage of the formation depth at the measuring point exceeds the length of a single drill pipe, a global optimization update of the model parameters is performed.

[0025] Combining online covariance extraction with the prediction bias value representing the degree of anomaly, the calculation formula is as follows: ; in, For the first The prediction deviation of each data set at the current sampling time is a dimensionless parameter. The larger the value, the higher the degree of data anomaly. For the first The online covariance matrix of the information error for each data set is a dimensionless matrix. It is updated in real time using a sliding window method with a forgetting factor. The sliding window length is set to 20 to 100 sampling points, and the forgetting factor ranges from 0.95 to 0.99. It is an identity matrix with the same dimensions as the covariance matrix.

[0026] The data generation rate is assessed using historical sampling intervals, and a real-time factor characterizing the urgency of data is generated by combining a time-dimension zero constant. The calculation formula is as follows: ; in, For the first The real-time factor of each data set at the current sampling time is a dimensionless parameter; The function for calculating the harmonic mean; For the first Each data set is a historical sampling interval sequence from the initial sampling time to the current sampling time, including zero values ​​and values ​​exceeding the specified interval. Outliers in the threshold are replaced with the median of the sequence. The real-time factor is used to characterize the rate of change in data generation; when the current sampling interval is less than the historical harmonic mean, the real-time factor is greater than 1, indicating a higher urgency of data timeliness.

[0027] To enhance the system's sensitivity to geological boundaries, phase sensitivity is calculated using the normalized measured values ​​and the derivative of cuttings hysteresis with respect to borehole depth. The calculation formula is as follows: ; in, For the first The phase sensitivity of each data set at the current sampling time is a dimensionless parameter. For the first The normalized measured values ​​of each data set are the absolute values ​​of the first derivative of the formation depth at the measurement point. These are dimensionless parameters and are calculated using the central difference method for non-equal interval sequences. The difference step size matches the sampling interval of the formation depth at the measurement point. The absolute value of the first derivative of the cuttings hysteresis with respect to the current borehole depth is a dimensionless parameter, calculated using the same central difference method. Phase sensitivity increases with the gradient of geological parameter changes and the gradient of cuttings hysteresis changes, enhancing the system's ability to identify and respond to stratigraphic interfaces and geological abrupt change zones.

[0028] The transmission priority of each data group is calculated by multiplying and fusing the prediction deviation, real-time factor, and phase sensitivity. The calculation formula is as follows: ; in, For the first The transmission priority of each data set at the current sampling time is a dimensionless parameter; a larger value indicates a higher transmission priority. The transmission priority considers the degree of data anomaly, timeliness, and geological boundary sensitivity, achieving a quantitative classification of the importance of data transmission for different attributes. The transmission priority is updated synchronously with the calculation results at each sampling time. The updated results only apply to newly generated data packets; the priority of data packets already in the waiting queue remains unchanged.

[0029] The bandwidth allocation rate is determined by weighting each group's transmission priority within the total global priority, ensuring that high-anomaly characteristics occupy channels preferentially. The calculation formula is as follows: ; in, This represents the total available channel bandwidth at the current sampling time, in bits per second. It is updated based on real-time channel detection results, and the update cycle is consistent with the drilling cycle of a single drill rod. This represents the sum of the transmission priorities of all data groups at the current sampling time. For the first The bandwidth allocated to each data group at the current sampling time, in bits per second. During the bandwidth allocation process, the peak bandwidth limit for high-priority data is set to 80% of the total available channel bandwidth, while the minimum guaranteed bandwidth for low-priority data is set to 5% of the total available channel bandwidth. This is to prevent high-priority data from excessively occupying channel resources, which could lead to low-priority data being completely unable to be transmitted.

[0030] To eliminate the sensitivity of fixed-time buffers to drilling progress fluctuations, the effective length of the drill pipe is accumulated, and the physical span of a single drill pipe is established as the buffer segment boundary. The calculation formula is as follows: ; Where j is the number of the drill pipe segment, which is a positive integer; For the first The effective length of the drill pipe inserted into the well, in meters, is determined by the nominal structural parameters of the drill pipe. The boundary depth of the j-th cache segment, in meters, represents the cumulative effective length of the first j drill pipes. The physical span of a single drill pipe is used as the cache segment boundary, directly binding the data cache to the physical drilling process. This avoids misalignment between the cache segment and the formation position caused by drilling speed fluctuations. The cache segment boundary is updated synchronously with each drill pipe entering the well.

[0031] By comparing the formation depth at the measuring point with the boundary of the buffer segment, the drill pipe segment number to which the current data belongs is determined. The calculation formula is as follows: ; Where j is the drill pipe segment number to which the data belongs at the current sampling time. By comparing the formation depth of the measuring point with the boundary depth of each buffer segment, the data is mapped to the depth range of the corresponding drill pipe segment, thus binding the data with the physical location of the formation. When the formation depth of the measuring point exceeds the current maximum segment boundary, a new drill pipe segment is created.

[0032] The relative borehole depth is calculated by combining the relative offset of the measuring point within the current segment boundary with a length dimension zero constant, in order to accurately locate the data source segment. The calculation formula is as follows: ; in, is the relative hole depth of the data at the current sampling time, which is a dimensionless parameter with a value range of [0,1); The lower boundary depth of the current segment, in meters; This represents the upper boundary depth of the current segment, in meters. The relative hole depth indicates the relative position of the data within its assigned drill pipe segment, enabling precise location of the data source segment.

[0033] The normalized measured value, the formation depth at the measurement point, the transmission priority, and the relative borehole depth are jointly encoded into a characteristic load, with the following structure: ; in, For the first The characteristic payloads of each data set at the current sampling time. The characteristic payloads integrate the normalized measured values, spatial location, transmission priority, and relative position information of the data, providing core data content for data packet encapsulation. The encoding format adopts binary encoding to ensure the compactness of data transmission.

[0034] The hash value of the previous packet, the drill pipe segment number, and the characteristic load are concatenated to generate a tamper-proof hash value. The calculation formula is as follows: ; in, This is the tamper-proof hash value of the current data packet; The hash value of the previous data packet within the same drill pipe segment; For the first The packet sequence number of each data group within the current segment is a positive integer; The SHA-256 cryptographic hash algorithm is used. Input parameters are concatenated sequentially according to the formula, using UTF-8 encoding, and the padding rules follow the SHA-256 algorithm standard. Each drill segment corresponds to an independent chained hash chain, with the initial hash value of the first data packet within the segment... The hash chain is generated by concatenating the drill pipe number, entry time, and drill bit number of the segment. The hash chain extends sequentially as data packets are generated within the segment, thus achieving anti-tampering and sequence integrity verification of the data packets.

[0035] This encapsulates a deep data packet with the following structure: ; in, This is the encapsulated depth data packet. Packet sequence number. Data packets are segmented and independently encoded. Packet numbers within each segment start from 1 and increment sequentially. When a new segment is created, the packet number is reset to 1. The maximum packet number is set to 65535; exceeding this value triggers segment splitting. The depth data packet contains spatial location information, attribution identifier, priority information, core payload, and integrity check hash value, thus binding the data to the formation depth.

[0036] Obtain the queuing dwell time and packet length for any depth packet in the pending queue. The queuing dwell time is... For data packets The time elapsed from entering the queue to the current scheduling time, in seconds; data packet length For data packets The length is expressed in bytes. The pending queue uses a first-in, first-out (FIFO) queuing rule, and also has a timeout cleanup mechanism. When a data packet's queuing time exceeds the preset maximum retention time, it is removed from the pending queue. The maximum retention time is set to twice the drilling cycle of a single drill pipe. The pending queue is sorted in two levels according to the drill pipe segment number and transmission priority to ensure that data packets within the same segment enter the scheduling stage in order.

[0037] To balance data importance and queue congestion, the congestion factor is obtained by dividing the queuing dwell time by the sum of the average queuing time and a time-diminishing constant. Here, the average queuing time... The congestion factor is the arithmetic mean of the queuing dwell time of all data packets in the queue at the current scheduling time, expressed in seconds. The calculation of the congestion factor incorporates the ratio of queuing dwell time to average queuing time, characterizing the degree of data packet retention in the queue. The longer the retention time, the larger the congestion factor, and the higher the urgency of sending the data packet.

[0038] Multiply the congestion factor and transmission priority by the derivative of the cuttings hysteresis with respect to borehole depth, scale it, and then divide it by the sum of the data packet length and the information dimension of the zero constant. This process removes the excessive channel congestion caused by long packets, yielding the transmission weight value. The calculation formula is as follows: ; in, For data packets The sending weight value is a dimensionless parameter; the larger the value, the higher the sending priority of the data packet. For data packets The transmission priority of the corresponding data group is a dimensionless parameter; For data packets The absolute value of the derivative of the cuttings hysteresis with respect to the borehole depth is a dimensionless parameter. The congestion factor is a dimensionless parameter. The transmission weight value is positively correlated with transmission priority, congestion factor, and cuttings hysteresis gradient, and negatively correlated with data packet length. This ensures that high-priority data is sent first, while avoiding excessive consumption of channel resources by long data packets, thus balancing transmission efficiency and transmission fairness.

[0039] The queue of items to be sent is reordered in descending order of their sending weight values ​​to generate a sending sequence. Sending sequence This is the sequence of all data packets in the queue to be sent at the current scheduling moment, sorted from largest to smallest by their sending weight value. When multiple data packets have the same sending weight value, they are sorted a second time by ascending drill pipe segment number and then by packet sequence number to ensure that the sending order of data packets within the same segment is consistent with the generation order.

[0040] Under the constraint of available channel capacity, high-weight packets are extracted sequentially according to the transmission sequence to form a transmission set and pushed outward, so as to achieve disturbance-resistant packet transmission scheduling. The calculation formula is as follows: ; in, This is the set of transmissions at the current scheduling moment; This represents the available channel capacity within the current scheduling period, in bytes. Data packet lengths are incremented sequentially according to the transmission sequence until the incremented value approaches but does not exceed the available channel capacity. The corresponding data packets form a transmission set and are pushed to the receiver. During the packet transmission scheduling process, 20% of the available channel capacity is reserved for the retransmission queue to ensure the retransmission of lost packets. Retransmission data packets have a higher transmission priority than data packets in the normal transmission queue with the same weight.

[0041] The measured link state, encompassing latency jitter and packet loss gaps, is obtained. The predicted link state is then derived using a mapping model based on mud parameters and cuttings hysteresis. The calculation formula is as follows: ; in, The measured link state at the current moment is a multi-dimensional vector containing parameters in four dimensions: channel one-way delay, delay jitter, packet loss rate, and available bandwidth. The predicted link state at the current moment is a vector with the same dimension as the measured link state; The model employs a three-layer backpropagation (BP) neural network. The input layer contains 5 neurons, corresponding to the 5 input parameters in the formula; the hidden layer contains 10 neurons, using the ReLU activation function; and the output layer contains 4 neurons, corresponding to the four dimensions of the predicted link state, using the Sigmoid activation function. The training dataset consists of input parameters synchronously collected during historical drilling and corresponding measured link state data, with a minimum of 10,000 datasets. The labeled data is clock-synchronized with the input data. The loss function is the mean squared error function, and the optimizer is the Adam optimizer with a learning rate of 0.001 and 1000 iterations. After training, the model's inference latency is no more than 1ms. During the cold start phase, pre-trained weights under the same drilling conditions are used. Incremental training is performed every 100m of drilling progress to update the model weights.

[0042] Min-max normalization was performed on all dimensional parameters of the measured and predicted link states to eliminate dimensional differences and map them to the dimensionless interval [0,1]. The normalization formula was consistent with the normalization formula for the measured values ​​of the data set, and the maximum and minimum values ​​of each dimensional parameter were used. By comparing the normalized residuals of the measured and predicted link states, a dimensionless zero-prevention constant was introduced to suppress data noise, and the anomaly intensity characterizing the degree of channel degradation was obtained. The calculation formula is as follows: ; in, The current channel anomaly intensity is represented by a dimensionless parameter. A larger value indicates a greater deviation of the channel state from the predicted value and a more severe channel degradation. This is the normalized measured link state vector; This is the normalized predicted link state vector; The online covariance matrix of the link-state prediction residual is a dimensionless matrix, which is updated in real time using a sliding window method with a forgetting factor. It is an identity matrix with the same dimensions as the covariance matrix.

[0043] To avoid unnecessary retransmission of outdated data, a freshness factor is extracted based on the queuing time of lost packets. The calculation formula is as follows: ; in, For lost packets The freshness factor is a dimensionless parameter. The freshness factor increases as the queueing time of lost packets increases, representing the freshness of the data packets. When the queuing time exceeds the maximum retention time, the freshness factor is set to 0, prohibiting entry into the retransmission queue and preventing stale data that has been stuck for a long time from occupying retransmission channel resources.

[0044] To correct for spatiotemporal migration errors in retransmission, the cuttings lag of the lost packet is compared with the average delay distance of its segment. A spatial correction is generated by combining this with a length-dimension zero constant. The calculation formula is as follows: ; in, For lost packets The corresponding spatial correction is a dimensionless parameter; For lost packets The corresponding cuttings hysteresis, in meters; For lost packets The spatial correction is the arithmetic mean of the cuttings lag of all data packets within the corresponding drill pipe segment, expressed in meters (m). It is used to correct spatial positioning errors in retransmitted data. The greater the deviation between the cuttings lag of a lost packet and the average value of its segment, the larger the spatial correction and the higher the retransmission priority, thus avoiding spatial semantic misalignment in the retransmitted data.

[0045] The retransmission priority is calculated based on the combined transmission priority, anomaly intensity, spatial correction amount, and freshness factor to trigger retransmission. The calculation formula is as follows: ; in, For lost packets The retransmission priority is a dimensionless parameter; a higher value indicates a higher urgency for retransmission. The retransmission queue is arranged in descending order of priority. When the retransmission priority exceeds a preset trigger threshold, the data packet enters the retransmission queue. The base value of the trigger threshold is 10, which can be dynamically adjusted according to the real-time channel status. The more severe the channel degradation, the higher the trigger threshold, and only high-priority data packets trigger retransmission. The maximum number of retransmissions for retransmission data packets is set to 3. Data packets that have not received acknowledgment after exceeding the retransmission limit are removed from the retransmission queue and marked as missing data in the interpolation processing stage at the receiving end.

[0046] At the receiving end, successfully arriving depth data packets are collected according to the drill pipe segment number to form a segmented receiving pool, expressed as: ; in, This is the segment receiving pool corresponding to the j-th drill pipe segment. The receiving end classifies and collects all successfully received depth data packets according to the drill pipe segment number. Data packets within the same segment are stored in the corresponding segment receiving pool. At the same time, hash integrity checks are performed on the received data packets. Data packets that fail the check are discarded, triggering a retransmission request for the corresponding packet.

[0047] The discrete distribution error of the measurement point depth in each packet within the segmented receiving pool is used to define the smoothing bandwidth. The calculation formula is as follows: ; in, This represents the smoothing bandwidth corresponding to the j-th segment receiving pool, in meters. The smoothing bandwidth is the arithmetic mean of the depth of the measurement points for all data packets within the j-th segment's receiving pool, expressed in meters. It is calculated from the standard deviation of the depth of the measurement points within the segment, adapting to the spatial dispersion of the data; the higher the data dispersion, the larger the smoothing bandwidth.

[0048] To eliminate temporal fault interference caused by packet loss and out-of-order delivery, a smoothing mapping function is used to project discrete packet data into a continuous stratigraphic depth coordinate system. Spatial interpolation and alignment are then performed to restore the geological trend, resulting in a continuous geological profile. The calculation formula is as follows: ; in, These are depth coordinates within a continuous stratigraphic depth coordinate system, in meters (m). For the j-th segment in the receiving pool, the first segment is... A set of valid data packets for each data group; For data packets Inner Measured values ​​for each data set; This is the Gaussian kernel smoothing mapping function, with the specific expression as follows: ; For the j-th segment, the first... Each data set corresponds to a continuous geological profile, with depth coordinates. The data is a continuous function. When splicing geological profiles from adjacent drill pipe segments, a weighted smoothing transition is performed on the profile data within a 1m radius on both sides of the segment boundary. The weights change linearly with the distance from the segment boundary to eliminate numerical faults at the segment boundary. The reliability verification rule for the interpolation results is as follows: when the data packet reception rate within a segment is lower than 80%, emergency retransmission of missing data packets for the corresponding segment is triggered; when the reception rate is lower than 50%, the segment profile is marked as low reliability and is prohibited from being directly used for geological prediction.

[0049] To identify missing hash sets, the hash set sent by the drilling rig is compared with the hash set received by the receiving end. The calculation formula is as follows: ; in, The set of sending hashes, which consists of the hash values ​​of all data packets within the j-th segment of the drilling rig; The receive hash set is composed of the hash values ​​of all valid data packets in the j-th segment receive pool of the receiving end; This represents the missing hash set corresponding to the j-th segment, i.e., the hash values ​​in the sent hash set that do not appear in the received hash set, corresponding to the data packets lost by the receiving end. The sent hash set is sent by the drilling rig to the receiving end through an independent control channel after the corresponding drill pipe segment is completely inserted into the well. The sending period is consistent with the update period of the drill pipe segment. If the receiving end does not receive the sent hash set for the corresponding segment, it triggers a retransmission request for the hash set.

[0050] When the missing hash set is empty, the closed, tamper-proof hash value and continuous geological profiles are stored and archived. The archived dataset structure is as follows: ; in, For the first The archived datasets corresponding to each segment; For the sequence of cuttings hysteresis within the j-th segment; , , These are the continuous geological profiles corresponding to the key measurement data, drilling rig status data, and auxiliary data within the j-th segment, respectively. This represents the closed-chain hash root value of the data packets within the j-th segment. When the missing hash set is empty, it means that all data packets within that segment have been successfully received, and the data is complete and without any missing data. The corresponding continuous geological profiles and chained hash values ​​are packaged together to form an archived dataset, which is then stored in the database, completing the entire closed-loop process of data transmission and processing.

[0051] The applicable operating conditions for this implementation are as follows: hole depth range of 50m to 5000m, drilling speed range of 0.1m / h to 100m / h, available channel bandwidth range of 1kbit / s to 10Mbit / s, and mud flow rate range of 0.1m³ / h. / min to 10m / min. The parameter adaptation rules for different operating conditions are as follows: the greater the hole depth, the smaller the sampling interval and the longer the sliding window length; the greater the fluctuation of drilling speed, the smaller the sampling interval and the higher the Kalman filter model update frequency; the lower the available channel bandwidth, the higher the retransmission trigger threshold and the lower the minimum guaranteed bandwidth for low-priority data.

[0052] The abnormal operating condition handling mechanism includes: when sensor data is missing, the linear interpolation result of the first 3 valid sampling points is used to fill the gap; when more than 5 sampling points are missing consecutively, a sensor abnormality alarm is triggered; when a singularity in the operation is triggered, the valid operation result of the previous sampling time is used to replace it, and the abnormal event is recorded at the same time; when communication is interrupted, the drilling rig caches all generated data packets according to the drill rod segment, and after communication is restored, the cached data packets are resent in the order of segment number from smallest to largest and transmission priority from highest to lowest; when hash verification fails consecutively, the full data packet of the corresponding segment is retransmitted, and channel security verification is started at the same time.

[0053] The quantitative verification method for the spatial matching degree of geological prediction is to use the real stratigraphic profile obtained by borehole core sampling as a benchmark to calculate the depth positioning error between the continuous geological profile generated by this scheme and the real stratigraphic profile. The formula for calculating the depth positioning error is the arithmetic mean of the absolute values ​​of the depth difference of the profile feature points. At the same time, the correlation coefficient of the profile data is calculated. The closer the correlation coefficient is to 1, the smaller the depth positioning error, which represents the higher the spatial matching degree.

[0054] The forgetting factor is the weight decay coefficient of historical data on the current covariance matrix. The closer the value is to 1, the higher the influence weight of historical data. It is used to adapt to slowly changing drilling conditions and balance the smoothness of calculation results with real-time response capability. The sliding window length is the effective length of the historical data sequence participating in the current operation, used to limit the time / depth range of the operation and balance the operation's noise resistance with its response speed to sudden changes in operating conditions; The covariance between process noise and observation noise is a quantified value of the statistical characteristics of system state prediction error and sensor measurement error in Kalman filtering. The magnitude of the value corresponds to the degree of confidence in the corresponding error. The lower the value, the higher the confidence weight of the corresponding data. The number of neurons in the hidden layer of a neural network is a dimension of the model's ability to fit the nonlinear relationship between input parameters and output link states. The number matches the nonlinear complexity of the input and output, balancing the model's fitting accuracy and the risk of overfitting. The maximum number of retransmissions is a critical threshold for balancing data packet transmission reliability and channel resource usage, to prevent the infinite retransmission of a single lost packet from crowding out the transmission bandwidth of high-priority data. The maximum dwell time is the effective time window for geological forecasting of drilling data. The formation information corresponding to data packets that exceed this window has lost the timeliness of advance forecasting and no longer needs to occupy channel resources for transmission. The enlargement rate is the ratio of the actual borehole diameter to the nominal diameter of the drill bit. It quantifies the degree of borehole enlargement caused by borehole wall collapse and wear, and is obtained by statistical calibration of historical drilling data under the same formation conditions. The normalization upper and lower limits are the theoretical extreme value ranges of the corresponding parameters under the target formation conditions. They are used to map physical parameters of different dimensions and magnitudes to a unified dimensionless range, ensuring the legality and comparability of cross-parameter calculations.

[0055] The outlier removal process is as follows: For a single dataset, a sliding window of 20 consecutive sampling points is used. The arithmetic mean and standard deviation of the data within the window are calculated. Outliers exceeding the mean are removed. Values ​​that are more than one standard deviation in length are considered outliers and are replaced with the median within the window, with the process performed by sliding the sampling point one by one. The fixed-point iterative decoupling operation process is as follows: take the current measured borehole depth as the initial value of the iteration, calculate the cuttings return time, cuttings lag, and measuring point formation depth in sequence, then correct the annular path length with the updated measuring point formation depth, and repeat the iteration process until the difference between the measuring point formation depths of two iterations is less than the convergence threshold, or the number of iterations reaches the preset upper limit. The chain hash chain construction process is as follows: After each drill pipe is fully inserted into the well, an independent hash chain is created for the corresponding segment. The UTF-8 encoding of the drill pipe number, insertion time, and drill bit number is concatenated and then SHA-256 operation is performed to generate the initial hash value of the segment. For each data packet generated for this segment, the hash value of the previous packet is concatenated with the parameters of the current packet in a fixed order and then SHA-256 operation is performed to generate the hash value of the current packet, forming an immutable chain structure. The Kalman filter cold start operation process is as follows: before drilling, obtain the industry standard parameter range of the target formation, take the midpoint of the range as the initial state vector of the filter, take 1 / 10 of the range of the process noise covariance, take twice the nominal accuracy of the sensor for the observation noise covariance, and complete the cold start initialization; after every 5 valid sampling points are obtained, update the filter covariance matrix once. The incremental training process for neural networks is as follows: after each 100m drilling advance, the input parameters collected synchronously within that section and the measured link state data are extracted to form an incremental training set. The current model weights are used as the initial values, the learning rate is set to 1 / 10 of the initial training, and the incremental training is completed after 100 iterations. At the same time, the historical optimal weights are retained as backups to avoid model performance degradation. The process of splicing adjacent segment profiles is as follows: For the boundary of two adjacent segments, take the profile data within a 1m range. With the segment boundary as the zero point, the points closer to the zero point have higher weights. The weights change linearly from 1 to 0. The values ​​within the overlapping range are weighted and averaged to achieve a smooth transition between the two profiles and eliminate numerical faults at the boundary. The communication interruption buffering and retransmission operation process is as follows: if no reception confirmation is received for three consecutive packet transmission cycles, it is determined that the communication is interrupted. The drilling rig stops real-time packet transmission and stores the newly generated data packets in local non-volatile storage according to the drill pipe segment number and data group priority. After the channel is restored, the unacknowledged data packets before the interruption are retransmitted first, and then the buffered data is retransmitted in the order of segment number from small to large and priority from high to low within the same segment. The real-time packet transmission requests of newly generated data packets are processed synchronously.

[0056] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. A method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling, applied to a horizontal directional drilling system comprising a drilling rig end, a measuring end, and a receiving end, characterized in that... include: Drilling parameters and data while drilling are collected, packaged into data frames containing comprehensive rock breaking energy according to time windows, and the data frames while drilling are divided into data groups with different attributes. Extract the cuttings hysteresis of the drilling data frame to eliminate the physical delay interference caused by drill pipe eccentricity and annular cuttings bed, and correct the current hole depth based on the cuttings hysteresis to obtain the formation depth at the measuring point; For each of the data groups, the abnormal fluctuation characteristics are extracted by combining the formation depth of the measuring point and the cuttings hysteresis, and the transmission priority and bandwidth allocation rate of each data group are calculated. Using the length of a single drill pipe as the physical buffer boundary, the drilling data frame is spatially mapped according to the formation depth of the measuring point to generate a depth data packet with an anti-tampering hash value; The channel adaptability of the depth data packet is evaluated using the transmission priority and the cuttings lag, a transmission weight value is obtained, and a scheduled packet transmission is performed to the receiving end accordingly. After the packet is sent, the measured link state is compared with the predicted link state to quantify the transmission interference. The retransmission priority of the unacknowledged lost packets is calculated in combination with the rock cuttings lag to trigger the retransmission queue. At the receiving end, spatial interpolation and alignment are performed on the received depth data packets based on the stratum depth of the measuring point, and the resulting continuous geological profile with out-of-order interference is reconstructed.

2. The method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling according to claim 1, characterized in that, The data frames encapsulated by time window and containing comprehensive rock-breaking energy during drilling include: To eliminate drilling statistical errors caused solely by time accumulation, the time difference between the current and previous sampling moments is calculated as the sampling interval. Extract drilling speed, axial feed force, torque, rotational speed, mud pressure, and mud flow rate to characterize the combined energy output of mechanical rock breaking and hydraulic rock clearing; The combined rock-breaking energy per unit depth is obtained by calculating the cumulative work done by the combined energy output within the sampling interval and introducing a zero-prevention constant divided by the difference in hole depth between the current and previous moments to eliminate the interference of zero-depth singularity. The current time, borehole depth, the comprehensive rock-breaking energy, key measurement data, and drilling rig status data are combined and encapsulated into the drilling data frame to provide a physically homogeneous data benchmark.

3. The method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling according to claim 2, characterized in that, The step of extracting the cuttings hysteresis of the drilling data frame and correcting the current borehole depth based on the cuttings hysteresis to obtain the formation depth at the measuring point includes: To quantify the cross-sectional change caused by drill pipe eccentricity, the theoretical annular area is calculated from the borehole diameter and the outer diameter of the drill pipe; To eliminate the interference of base pressure caused by mud circulation, the flow field smoothness is calculated by combining the residual of the measured pressure deviating from the rock cuttings-free reference pressure with the aforementioned zero constant. By combining the mud flow rate, the theoretical annular area, and the flow field smoothness, the equivalent upward return velocity of the mud carrying rock is calculated. Integrate the equivalent upward return velocity in reverse time to the matching annular path length to determine the upward return of rock cuttings, so as to characterize the flow field retention effect; The drilling speed is integrated over the time it takes for the cuttings to return to the surface to obtain the cuttings hysteresis. This hysteresis is then subtracted from the current borehole depth to remove spatial delay interference and restore the true geological origin of the stratum depth at the measuring point.

4. The method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling according to claim 3, characterized in that, The calculation of the transmission priority and bandwidth allocation rate for each data group includes: To capture data anomalies caused by geological mutations, the measured values ​​of each data group are subtracted from the predicted values ​​arranged according to the stratum depth of the measuring point, and the prediction deviation value representing the degree of anomaly is extracted by combining online covariance. The data generation rate is evaluated using historical sampling intervals, and a real-time factor characterizing the urgency of data timeliness is generated in combination with the aforementioned zero constant. To enhance the system's sensitivity to geological boundaries, phase sensitivity is calculated using the measured values ​​and the derivative of the cuttings hysteresis with respect to borehole depth. The transmission priority of each data group is calculated by multiplying and fusing the prediction deviation value, the real-time factor, and the phase sensitivity. The bandwidth allocation rate is determined by using the proportion of each group's transmission priority in the total global priority as the weight, so as to ensure that high-abnormality features occupy the channel first.

5. The method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling according to claim 4, characterized in that, The generation of deep data packets with tamper-proof hash values ​​includes: To eliminate the sensitivity of fixed-time buffers to drilling progress fluctuations, the effective length of the drill pipe is accumulated, and the physical span of a single drill pipe is established as the buffer segment boundary. By comparing the formation depth at the measuring point with the boundary of the buffer segment, the drill pipe segment number to which the current data belongs is determined; The relative borehole depth is calculated by combining the relative offset of the formation depth at the measuring point within the current segment boundary with the zero-constant to accurately locate the data source segment; The measured value, the formation depth of the measuring point, the transmission priority, and the relative borehole depth are jointly encoded into a characteristic load; The tamper-proof hash value is generated by concatenating the hash value of the previous packet, the drill pipe segment number, and the characteristic load, and then encapsulating the depth data packet.

6. The method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling according to claim 5, characterized in that, The calculation of the transmission weight value and the execution of scheduled packet transmission include: Obtain the queuing dwell time and packet length of any of the aforementioned depth packets in the pending queue; To balance data importance and queue congestion, the congestion factor is obtained by dividing the queue dwell time by the sum of the average queue time and the zero-prevention constant. Multiply the congestion factor, the transmission priority, and the derivative of the cuttings hysteresis with respect to the borehole depth by the product and scale, then divide by the sum of the data packet length and the anti-zero constant to remove the excessive crowding interference of long packets on the channel and obtain the transmission weight value. The queue to be sent is reordered in descending order according to the sending weight values ​​to generate a sending sequence; Under the constraint of available channel capacity, high-weight packets are sequentially extracted according to the transmission sequence to form a transmission set and pushed outward, so as to achieve anti-disturbance scheduling of packet transmission.

7. The method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling according to claim 6, characterized in that, The calculation of the retransmission priority for unacknowledged lost packets includes: The measured link state, which includes latency jitter and packet loss gaps, is obtained, and the predicted link state is derived by using mud parameters and cuttings hysteresis through a mapping model. By comparing the residuals of the measured link state and the predicted link state, the zero-prevention constant is introduced to suppress data noise, and the abnormal intensity characterizing the degree of channel degradation is obtained. To avoid unnecessary retransmission of outdated data, a freshness factor is extracted during the queuing and dwell time of the lost packets. To correct the spatiotemporal offset error of the retransmission, the cuttings lag of the lost packet is compared with the average delay distance of the segment to which it belongs, and a spatial correction amount is generated in combination with the zero constant. The retransmission priority is calculated by combining the transmission priority, the anomaly intensity, the spatial correction amount, and the freshness factor to trigger retransmission.

8. The method for layered transmission and anomaly retransmission of tunnel geological exploration data during drilling according to claim 7, characterized in that, The process of performing spatial interpolation and alignment on the received depth data packets based on the stratigraphic depth of the measuring point to reconstruct a continuous geological profile free from disordered interference includes: At the receiving end, the successfully arriving depth data packets are collected according to the drill pipe segment number to form a segment receiving pool; The discrete distribution error of the stratum depth of the measurement point in each packet within the segmented receiving pool is statistically analyzed to define the smoothing bandwidth; To eliminate temporal fault interference caused by packet loss and out-of-order delivery, a smoothing mapping function is used to project discrete packet data into a continuous stratigraphic depth coordinate system. Spatial interpolation and alignment are then performed to restore the geological trend, resulting in the continuous geological profile. Compare the hash set sent by the drilling rig with the hash set received by the receiving end to identify the missing hash set; When the missing hash set is empty, the closed anti-tampering hash value and the continuous geological profile are stored and archived in the database.

9. A layered transmission and anomaly retransmission system for tunnel geological exploration data during drilling, comprising the layered transmission and anomaly retransmission method for tunnel geological exploration data during drilling as described in any one of claims 1-8, characterized in that, Includes the drilling rig end, the measuring end, and the receiving end; The drilling rig end is responsible for data frame encapsulation, cuttings hysteresis extraction and measurement point formation depth correction. The drilling rig end also performs priority calculation functions; The drilling rig end also undertakes the function of data packet generation; The drilling rig end also undertakes scheduling and contracting functions; The drilling rig end also undertakes the function of retransmission queue management; The measuring end is responsible for synchronous acquisition of drilling parameters and clock synchronization. The receiving end is responsible for data packet reception, spatial interpolation alignment, hash integrity verification, and geological profile reconstruction and archiving.

Images