A high-speed logging communication method and system for dual-mode FSK demodulation

Abstract

The present application provides a high-speed logging communication method and system for dual-mode FSK demodulation, which relates to the field of communication technology. After receiving and preprocessing the analog FSK modulation signal, signal granularity discrimination is performed to dynamically update the internal recognition mode; the preprocessing result is sent to the first and second demodulation modules; after obtaining the output results of the modules, the internal recognition mode is used to fuse the output results, and a preliminary fusion decision and an adjacent frequency point set are established; based on the preliminary fusion decision and the adjacent frequency point set, soft decision trigger analysis is performed. If soft decision compensation is triggered, the adjacent frequency point set is used to update the preliminary fusion decision and then high-speed logging communication is carried out. The present application solves the technical problems in the prior art that due to the synchronization method in the demodulation process relying on a fixed recognition mode, the anti-interference ability is insufficient in the strong noise environment underground, the communication reliability decreases, and the error rate is high, improves the demodulation accuracy of the communication signal, and significantly improves the overall communication quality and stability.

Description

Technical Field

[0001] The present application relates to the field of communication technologies, and in particular, to a high-speed logging communication method and system for dual-mode FSK demodulation. Background Art

[0002] With the continuous improvement of the requirements for real-time performance and data volume in oil and gas exploration and development, high-speed logging communication technology has become an important means of information transmission between the underground and the ground. Among them, the method of using frequency shift keying (FSK) modulation and demodulation has become the mainstream technical route in current downhole logging communication due to its simple implementation and good system stability. However, the existing FSK demodulation methods have obvious deficiencies. Since the synchronization and demodulation processes rely on fixed templates, in the complex underground electromagnetic environment, they are easily affected by factors such as strong noise interference and frequency drift, with limited anti-interference ability, difficult to adapt to changes in channel characteristics in a timely manner, easily leading to a decrease in communication reliability and an increase in the error rate, and affecting the accurate transmission of logging data. Summary of the Invention

[0003] The present application provides a high-speed logging communication method and system for dual-mode FSK demodulation, which solves the technical problem that the existing technology has insufficient anti-interference ability in the strong noise environment underground due to the synchronization method in the FSK demodulation process relying on a fixed recognition mode, resulting in a decrease in communication reliability and a high error rate, and achieves the technical effect of improving the demodulation accuracy of high-speed logging communication signals, thereby significantly improving the overall communication quality and stability.

[0004] In view of the above problems, on the one hand, the present application provides a high-speed logging communication method for dual-mode FSK demodulation, and the method includes: receiving an analog FSK modulation signal of an underground communication link, and performing signal granularity discrimination on the preprocessed result after preprocessing the analog FSK modulation signal; dynamically updating the internal recognition mode of the dual-mode FSK by using the signal granularity discrimination result; respectively sending the preprocessed result to a first demodulation module and a second demodulation module of the dual-mode FSK, where the first demodulation module is a main frequency-energy demodulation module, and the second demodulation module is a local spectrum dynamic feature extraction module; respectively obtaining the output results corresponding to the first demodulation module and the second demodulation module, fusing the output results by using the internal recognition mode to establish a preliminary fusion decision; performing adjacent frequency point search on the preliminary fusion decision to establish an adjacent frequency point set; performing soft decision trigger analysis based on the preliminary fusion decision and the adjacent frequency point set, and if soft decision compensation is triggered, updating the preliminary fusion decision by using the adjacent frequency point set; and performing high-speed logging communication by using the updated preliminary fusion decision.

[0005] On the other hand, the present application also provides a high-speed logging communication system for dual-mode FSK demodulation. The system includes: a signal granularity discrimination module, configured to receive an analog FSK modulation signal of a downhole communication link, and perform signal granularity discrimination on a preprocessing result after preprocessing the analog FSK modulation signal; an identification mode update module, configured to dynamically update an internal identification mode of the dual-mode FSK by using the signal granularity discrimination result; a result sending module, configured to respectively send the preprocessing result to a first demodulation module and a second demodulation module of the dual-mode FSK, wherein the first demodulation module is a main frequency-energy demodulation module, and the second demodulation module is a local spectrum dynamic feature extraction module; a result fusion module, configured to respectively obtain output results corresponding to the first demodulation module and the second demodulation module, fuse the output results by using the internal identification mode, and establish a preliminary fusion decision; an adjacent frequency point search module, configured to perform adjacent frequency point search on the preliminary fusion decision to establish an adjacent frequency point set; a trigger analysis module, configured to perform soft decision trigger analysis based on the preliminary fusion decision and the adjacent frequency point set, and if soft decision compensation is triggered, update the preliminary fusion decision by using the adjacent frequency point set; and a communication execution module, configured to perform high-speed logging communication by using the updated preliminary fusion decision.

[0006] One or more technical solutions provided in the present application have at least the following beneficial effects:

[0007] By receiving and preprocessing the analog FSK modulation signal of the downhole communication link, the standardization processing of the original FSK signal is completed, and then the detail change degree of the signal is analyzed through signal granularity discrimination, providing a basic basis for dynamically adapting the demodulation mode subsequently. Dynamically updating the internal identification mode of the dual-mode FSK by using the signal granularity discrimination result can dynamically adjust the internal mode relied on by demodulation according to the current signal characteristics, improve the adaptive ability to channel changes, and thus enhance the anti-interference performance. The preprocessing result is respectively sent to the first demodulation module and the second demodulation module of the dual-mode FSK. Through a dual-path processing mechanism, on the one hand, coarse demodulation is performed based on the overall frequency and energy, and on the other hand, weak changes are captured through local spectrum feature extraction, forming a complementarity to improve the accuracy and robustness of demodulation. The output results corresponding to the first demodulation module and the second demodulation module are respectively obtained, and the demodulation results of the dual paths are fused by using the internal identification mode to establish a preliminary fusion decision, improving the demodulation reliability. Adjacent frequency point search is performed on the preliminary fusion decision to establish an adjacent frequency point set, constructing an auxiliary reference set for subsequent soft decision compensation. Soft decision trigger analysis is performed based on the preliminary fusion decision and the adjacent frequency point set. If soft decision compensation is triggered, the preliminary fusion decision is fine-tuned or corrected by using the adjacent frequency point set, thereby significantly reducing the demodulation error probability and further enhancing the anti-noise ability. Data restoration and communication are performed by using the updated preliminary fusion decision to ensure stable and reliable communication in a high-speed logging environment and achieve system-level performance improvement.

[0008] In summary, the present application realizes the dynamic perception of the FSK signal characteristics in the complex downhole environment by introducing signal granularity discrimination, and updates the internal recognition mode in real time based on the granularity analysis results, significantly improving the adaptability of the demodulation process to the channel state changes; adopts a dual-mode demodulation structure of main frequency-energy demodulation and local spectrum dynamic feature extraction, and combines the internal recognition mode to perform fusion decision on the dual-mode output, which not only improves the accuracy of signal recognition, but also enhances the overall anti-interference ability; by constructing an adjacent frequency point set and triggering soft decision compensation when there is uncertainty in the fusion preliminary decision, further corrects potential demodulation deviations and effectively reduces the bit error rate. Overall, the present application effectively enhances the accuracy and reliability of signal demodulation. Especially in the complex downhole environment, it can significantly reduce the bit error rate, improve the communication quality, and meet the actual requirements of high-speed logging for high-reliability data transmission.

[0009] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present application are specifically given below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 It is a schematic flow chart of a high-speed logging communication method for dual-mode FSK demodulation provided by an embodiment of the present application.

[0011] Figure 2 It is a schematic flow chart of item-by-item verification backtracking based on a verification chain in a high-speed logging communication method for dual-mode FSK demodulation provided by an embodiment of the present application.

[0012] Figure 3 It is a schematic structural diagram of a high-speed logging communication system for dual-mode FSK demodulation provided by an embodiment of the present application.

[0013] Description of the reference numerals: signal granularity discrimination module 10, recognition mode update module 20, result sending module 30, result fusion module 40, adjacent frequency point search module 50, trigger analysis module 60, communication execution module 70. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] The embodiments of the present application provide a high-speed logging communication method and system for dual-mode FSK demodulation. By dynamically updating the internal recognition mode and combining the dual-mode feature fusion decision, it solves the technical problems in the prior art that due to the synchronous method in the FSK demodulation process relying on a fixed recognition mode, the anti-interference ability is insufficient in the strong noise environment downhole, resulting in a decrease in communication reliability and a high bit error rate, and achieves the technical effect of improving the demodulation accuracy of high-speed logging communication signals, thereby significantly improving the overall communication quality and stability.

[0015] Embodiment 1, as Figure 1 shown, the embodiment of the present application provides a high-speed logging communication method for dual-mode FSK demodulation, and the method includes:

[0016] Step S100: Receive the analog FSK modulation signal of the downhole communication link, and after preprocessing the analog FSK modulation signal, perform signal granularity discrimination on the preprocessing result.

[0017] Specifically, the downhole communication link refers to the signal transmission channel between the downhole instrument and the ground control center, which is used to transmit information such as logging data. The analog FSK modulation signal is an analog signal generated by using the frequency shift keying (FSK) modulation method, and different data bits are represented by changing the frequency of the signal. The granularity of the signal, that is, the fineness of the signal or the severity of the characteristic change.

[0018] Receive the analog FSK modulation signal through the downhole communication link. Taking oil drilling logging as an example, the geological data (such as resistivity, porosity, etc.) collected by the downhole instrument in different formation sections will be converted into an analog FSK modulation signal and sent to the ground system. Then, perform preprocessing operations such as filtering, noise suppression, and amplitude normalization on the analog FSK modulation signal, and then perform signal granularity discrimination on the preprocessed signal. By analyzing the change trend of the signal main frequency, the stability of the energy distribution, and the short-time spectrum expansion situation, determine whether the signal belongs to the coarse-grained stable change state or the fine-grained rapid change state, so as to provide a basis for subsequent dynamic demodulation mode selection. For example, when the frequency of the signal changes slowly and the amplitude is large, it is determined as a coarse-grained signal; on the contrary, when the signal frequency changes frequently and the amplitude is small, it is determined as a fine-grained signal.

[0019] Through this step, basic cleaning and granularity classification can be quickly completed when the original signal is received, providing a targeted mode selection basis for the subsequent demodulation process, and effectively improving the adaptability and robustness of the overall demodulation process.

[0020] Step S200: Dynamically update the internal recognition mode of the dual-mode FSK by using the signal granularity discrimination result.

[0021] Specifically, the internal recognition mode is the mode or strategy used inside the dual-mode FSK demodulation system to identify and process signals. Different modes correspond to different demodulation parameters and algorithm configurations to adapt to signals with different characteristics. According to the signal granularity discrimination result obtained in step S100, the internal recognition mode of the dual-mode FSK is updated in real time and dynamically. If it is discriminated as coarse granularity, a demodulation strategy with a long time window and low change sensitivity is enabled to improve the demodulation speed and stability. If it is discriminated as fine granularity, a high-confidence and short-time window tracking mode is enabled to improve the perception ability of subtle changes and ensure that the subsequent demodulation module can extract signal features and perform bit decisions according to the optimal mode.

[0022] By dynamically updating the internal recognition mode in this step, it is possible to adapt to the complex and changing downhole signal environment, improve the demodulation accuracy and reduce the probability of misrecognition, ensuring that the subsequent demodulation module operates in the best working state.

[0023] Step S300: Send the preprocessing results to the first demodulation module and the second demodulation module of the dual-mode FSK respectively, where the first demodulation module is the main frequency-energy demodulation module, and the second demodulation module is the local spectrum dynamic feature extraction module.

[0024] Specifically, the preprocessing results obtained in step S100 are sent to the two demodulation modules simultaneously. The first demodulation module is the main frequency-energy demodulation module, which focuses on extracting the main frequency of the signal and its energy distribution characteristics, and makes a preliminary bit judgment through the energy peak of the main frequency point. The second demodulation module is the local spectrum dynamic feature extraction module, which is responsible for extracting the dynamic change trends in the local spectrum of the signal, such as frequency offset, local energy diffusion and other characteristics, to assist in detecting potential signal drift or weak variation conditions. The two operate in parallel and output independent preliminary analysis results respectively.

[0025] In this step, the signal is demodulated from different angles by two different demodulation modules, so that more comprehensive signal information can be obtained, and the accuracy and reliability of demodulation are improved.

[0026] Step S400: Obtain the output results corresponding to the first demodulation module and the second demodulation module respectively, and fuse the output results using the internal recognition mode to establish a fused preliminary decision.

[0027] Specifically, the fused preliminary decision is a preliminary decision result obtained by fusing the output results of two demodulation modules according to certain rules, which is used to represent the bit sequence represented by the signal. The main frequency-energy decision result from the first demodulation module and the local spectral dynamic feature extraction result from the second demodulation module are received respectively. According to the internal recognition mode updated in step S200, mechanisms such as weighted fusion, feature complementarity, or confidence screening are used to fuse these two output results. The bit judgments with consistent main frequencies and supported dynamic features are preferentially retained. At the same time, for the discrepant information, the set fusion rules are applied for decision-making, and finally the fused preliminary decision is generated, providing a basis for further optimization.

[0028] In this step, by deeply fusing the information of the dual modules, not only the stability of the main frequency energy is utilized, but also the sensitivity of the local frequency change is taken into account, greatly improving the accuracy and anti-interference ability of the preliminary bit decision, and providing a basis for subsequent further optimization and compensation.

[0029] Step S500: Search for adjacent frequency points for the fused preliminary decision and establish an adjacent frequency point set.

[0030] Specifically, the adjacent frequency point set is a set of frequency points adjacent to the frequency points in the preliminary decision result in the frequency domain. These frequency points may contain secondary components of the useful signal or frequency point information of the interference signal, which can assist in verifying and optimizing the preliminary decision. Taking the central frequency point corresponding to the fused preliminary decision as a reference, within a certain frequency range around it, energy scanning and dynamic feature analysis are carried out to find adjacent frequency points with potential signal components, and effective frequency points are screened according to conditions such as energy threshold and frequency shift trend consistency. Finally, a set containing several adjacent frequency points is formed to assist in subsequent soft decision compensation analysis.

[0031] By establishing the adjacent frequency point set, the frequency information that cannot be fully covered by the preliminary decision can be effectively captured, increasing the redundancy and flexibility during soft decision-making, and improving the ability to recover weak signals under complex channel conditions.

[0032] Step S600: Based on the fused preliminary decision and the adjacent frequency point set, conduct soft decision trigger analysis. If soft decision compensation is triggered, update the fused preliminary decision using the adjacent frequency point set.

[0033] Specifically, soft decision trigger analysis refers to judging whether soft decision compensation is required for the preliminary decision result according to certain rules and conditions, that is, whether to use the information of adjacent frequency point sets to correct the preliminary decision to improve the accuracy of the decision. Soft decision compensation refers to adjusting and optimizing the preliminary decision result by using the frequency point information in the adjacent frequency point sets to reduce misjudgment and improve the reliability of demodulation. Based on the fusion of the preliminary decision and the adjacent frequency point sets, soft decision trigger analysis is performed. First, it is evaluated whether the confidence level of the preliminary decision meets the set threshold. If it is lower than the threshold or there is a decision conflict, the soft decision compensation process is triggered. By referring to the energy level, dynamic consistency, and correlation index with the main decision frequency point of each frequency point in the adjacent frequency point sets, the decision result is re-corrected after comprehensive weighting, and finally, a fusion preliminary decision with a higher confidence level is updated.

[0034] By setting up a soft decision mechanism, it is possible to automatically perform decision correction and optimization in the case of large fluctuations in channel quality or strong interference, greatly improving the adaptability and fault tolerance to the complex downhole environment and increasing the accuracy of the final logging communication data.

[0035] Step S700: Perform high-speed logging communication using the updated fusion preliminary decision.

[0036] Specifically, based on the fusion preliminary decision result updated by soft decision compensation, actual high-speed logging communication is performed. The bit stream obtained by the decision is transmitted back to the ground system in real time. At the same time, the transmission rate and coding parameters are dynamically adjusted according to the communication link state to ensure stable and low-bit-error-rate data transmission in the high-interference and high-attenuation downhole environment and support the real-time monitoring and control requirements of downhole operations. For example, the bit sequence is converted into specific geological parameter values (such as the resistivity value of the formation, etc.) or other instrument status information. Then, through the downhole communication link, according to a certain communication protocol (such as the DSL protocol), these data information is transmitted to the ground control center at high speed. At the ground control center, the received data is further processed and analyzed for guiding drilling operations, geological analysis, etc.

[0037] By using the updated high-confidence decision result for data transmission, the bit error rate and the number of retransmissions in the logging communication process can be significantly reduced, the effective payload capacity and real-time performance of the communication link can be improved, and the accurate and efficient transmission of downhole measurement and control data can be guaranteed.

[0038] Furthermore, step S600 includes:

[0039] Step S610: If soft decision compensation is not triggered, establish a preliminary candidate bit stream based on the fusion preliminary decision.

[0040] Step S620: Perform consistency check on the preliminary candidate bit stream using a sliding window.

[0041] Step S630: If the consistency check fails, trigger strong decision compensation, use the strong decision compensation to perform backtracking correction on the preliminary fusion decision, and use the backtracking correction result for high-speed logging communication.

[0042] Specifically, the preliminary candidate bitstream is a set of preliminary bit sequences generated according to the preliminary fusion decision result. It has not undergone in-depth correction but already has a communication structure and will be used as the object for subsequent further verification and processing. If the soft decision compensation is not triggered, that is, it is considered that the preliminary fusion decision already has basic credibility, then directly construct the preliminary candidate bitstream and further enter the sliding window consistency check process. During the check process, redundant detection and forward and backward logical consistency comparison are performed on local segments of the bitstream according to the set window length. For example, set the window length to 4 bits, place the window at the starting position of the preliminary candidate bitstream, and check whether these 4 bits conform to the preset consistency rules, such as whether they conform to specific coding rules (such as Manchester coding rules) or whether they meet requirements such as data continuity and regularity. Then the window slides forward by one bit position and continues the check, and so on until the entire preliminary candidate bitstream is checked. If the check passes, it is determined that the bitstream is in an acceptable state and enters the high-speed logging communication stage; if the consistency check fails, trigger the strong decision compensation mechanism, backtrack the decision basis in the previous multiple sampling periods, re-evaluate the spectrum structure and decision chain of each node, eliminate the original decisions with low confidence, and re-fit the most likely bit sequence according to the frequency point intensity trend, so as to achieve backtracking correction of the preliminary fusion decision. Finally, use the corrected result as the new bitstream for communication.

[0043] For example, during a certain communication process, the preliminary fusion decision generated the bitstream "1011001". When performing the sliding window consistency check (window size is 4), it is found that the "1100" segment appears with frequent inconsistent support features, triggering strong decision compensation; by backtracking the spectrum data 5 cycles ago, it is found that the original energy of the "1" part in "1100" is lower than the threshold, so it is corrected to "1000" through backtracking correction, and finally the reliable bitstream "1000001" is obtained and used for actual communication transmission.

[0044] The above steps introduce two-level mechanisms of soft decision compensation and strong decision backtracking, which can adaptively adjust the decision path according to different degrees of uncertainty, ensure the overall consistency and accuracy of the data stream while avoiding excessive correction, and significantly improve the robustness of bitstream construction and communication reliability in the high-interference downhole environment, providing a solid guarantee for subsequent efficient logging data transmission.

[0045] Further, step S630 includes:

[0046] Step S631: During the backtracking verification process, backtrack multiple data nodes to establish a verification chain, which includes a timing verification chain and a signal verification chain.

[0047] Step S632: Starting from the node that triggers the strong decision, perform item-by-item verification backtracking based on the verification chain to establish the backtracking correction result.

[0048] Specifically, when the consistency verification fails and triggers the strong decision compensation, first start the backtracking verification mechanism, backtrack multiple historical data nodes forward, and establish two types of verification chains during the backtracking process, namely, the timing verification chain and the signal verification chain. Among them, the timing verification chain is used to verify the time logic continuity between nodes to ensure that the bit transition conforms to the timing jump rule unique to FSK communication. The signal verification chain is used to detect the coherence of the spectral characteristics changes of each node, such as main frequency drift, energy mutation, etc. For the establishment of the timing verification chain, it is necessary to connect the relevant historical data nodes in chronological order according to the time sequence requirements of data transmission, and determine the relationship rules in time between each node, such as a certain node should appear within a specific time after another node, etc. For the establishment of the signal verification chain, it is necessary to connect the historical data nodes related to the signal characteristics and determine the relationship rules that the signal frequency, energy and other characteristics should satisfy between these nodes.

[0049] Subsequently, starting from the current node that triggers the strong decision, perform item-by-item verification backtracking based on the established verification chain above, that is, perform time consistency check and spectral feature consistency check on each backtracked node respectively. If a deviation is found, it will be corrected according to the support degree and change trend of adjacent nodes. Through continuous multi-node verification, a backtracking correction result is finally formed. The corrected decision result will be used as a new basis for the bit stream to participate in subsequent high-speed logging communication. For example, in a section of logging communication, it is detected that there is a local anomaly in the bit stream "10011010". After triggering the strong decision, backtrack the previous 10 data nodes. After establishing the timing verification chain, it is found that there is an unreasonable timing jump (the time span is abnormally small) between the 4th node and the 5th node. At the same time, on the signal verification chain, there is an unexpected shift in the main frequency of the spectrum of the 3rd to 5th nodes. According to the double verification results of timing and signal, adjust the bit decision of the 4th node, change the node originally determined as "1" to "0", and regenerate the backtracking correction bit stream "10001010", which greatly improves the data reliability.

[0050] The above steps, by introducing systematic verification chain backtracking during the strong decision compensation process, can not only comprehensively identify and correct isolated error points, but also achieve continuous correction according to the dual characteristics of timing and spectrum, significantly improve the overall correct rate of bit decision in the logging communication link, enhance the anti-interference ability and signal integrity, and greatly reduce the misjudgment rate caused by short-term anomalies, thereby ensuring the stable and reliable transmission of high-speed logging data.

[0051] Further, as Figure 2 shown, step S632 includes:

[0052] Step S632-1: Activate the timing check sub-channel, and use the timing check sub-channel to perform time consistency check based on the timing check chain to generate a first focus check result.

[0053] Step S632-2: Activate the spectrum check sub-channel, and use the spectrum check sub-channel to perform spectrum feature consistency check based on the signal check chain to generate a second focus check result.

[0054] Step S632-3: Perform double-check verification on the first focus check result and the second focus check result to complete item-by-item check backtracking.

[0055] Specifically, the timing check sub-channel is a functional module for performing time consistency check based on the timing check chain, focusing on the rationality of node time arrangement and jump. The spectrum check sub-channel is a functional module for performing spectrum feature consistency check based on the signal check chain, focusing on the rationality of feature changes such as node frequency and energy.

[0056] To complete the item-by-item check backtracking of the fusion preliminary decision, first activate the timing check sub-channel. For each data node in the timing check chain, perform consistency check according to the time evolution law of the sampling moment and the decision result. If phenomena such as abnormal time span and abnormal bit jump are detected, mark them as the first focus points and output the first focus check result. For example, during the logging data transmission process, each data node has a corresponding timestamp. Normally, the time interval between two adjacent data nodes should be ΔT. If it is actually detected that a certain interval deviates significantly from ΔT, such as exceeding 1.5ΔT or being less than 0.5ΔT, it is determined that the time consistency check at this point fails. At the same time, the rationality of the data change rate over time will also be checked. For example, if the change rate of a certain geological parameter suddenly exceeds the physical possibility range in a short time, it is also regarded as a time consistency problem.

[0057] Next, activate the spectrum parity sub-channel, and analyze the evolution of spectrum features such as the main frequency, bandwidth, and energy node by node based on the signal parity chain. If abnormalities such as feature mutations and continuity breaks are detected, mark them as the second focus points and output the second focus parity results. Taking the analog FSK modulation signal in downhole communication as an example, the normal signal spectrum should have a stable main frequency position, a reasonable spectrum width, and a specific spectrum shape. Perform a fast Fourier transform (FFT) on each traced data node to extract its spectrum information. For example, check whether the main frequency point fluctuates within the allowed frequency offset range and whether the spectrum energy distribution conforms to the preset template. If it is found that the main frequency of a certain data node has a too large offset or abnormal sidelobe growth appears in the spectrum, it is determined that the spectrum feature parity at that point fails.

[0058] When performing double parity verification on the first focus parity result and the second focus parity result, only when both point to the same problem or corroborate each other, is it determined that there is an error in this node and backtracking correction is performed; otherwise, the original judgment is maintained. Thus, the entire step-by-step parity backtracking process is completed. For example, in the analysis of a certain logging communication data, for the 10 traced nodes, the timing parity sub-channel detected a consistency abnormality (the first focus parity result) with a time span shorter than the normal jump period between the 6th node and the 7th node, and at the same time, the spectrum parity sub-channel found that the main frequency drift speed of the 6th node exceeded the normal change threshold (the second focus parity result); in the double parity verification, these two abnormal results complement and corroborate each other in terms of node positioning and abnormal type, so it is confirmed that the judgment of the 6th node is incorrect and corrected. On the contrary, for the 3rd node, only weak energy abnormalities were detected by spectrum parity, but no timing problems were found by timing parity. After double parity verification, it is considered that this abnormality is not sufficient to trigger correction, so the original judgment is maintained.

[0059] The above steps set up dedicated timing parity sub-channels and spectrum parity sub-channels to independently carry out consistency detection hierarchically and by feature, and finally comprehensively determine the error nodes through double cross-verification, greatly improving the accuracy and robustness of error recognition, avoiding the risk of false correction that may be brought by a single parity method, and at the same time strengthening the perception and repair ability of tiny abnormalities in complex interference environments, thereby ensuring the highly reliable and stable data quality during high-speed logging communication.

[0060] Further, step S300 includes:

[0061] Step S310: After extracting the frequency-domain information of the preprocessing result using the first demodulation module, perform a fast Fourier transform.

[0062] Step S320: Perform energy detection on the output of the fast Fourier transform to determine the main frequency point with the maximum energy.

[0063] Step S330: Compare the main frequency point with a preset frequency threshold to complete the preliminary bit decision.

[0064] Step S340: Perform a confidence level identification of the preliminary bit decision based on the energy value corresponding to the main frequency point.

[0065] Step S350: Perform smoothing of the preliminary bit decision based on the confidence level identification based on the energy-time window, and output the output result of the first demodulation module.

[0066] Specifically, the main frequency point refers to the frequency component with the maximum energy, which usually represents the carrier frequency corresponding to the current bit in signal modulation. The confidence level identification is an index characterizing the credibility of the preliminary bit decision, which is calculated based on the frequency energy intensity. The energy-time window is a sliding analysis region jointly restricted in the energy and time dimensions, which defines the energy change law within a time range and is used to smooth the decision error caused by short-term jitter.

[0067] The first demodulation module extracts the frequency domain information of the input preprocessing result, performs fast Fourier transform (FFT) processing on it to obtain the spectrum distribution of the signal, then performs energy detection on the spectrum curve output by the fast Fourier transform, searches for and determines the frequency point with the strongest energy, which is the main frequency point. Then, compare the extracted main frequency point with a preset frequency threshold to determine whether the main frequency belongs to the high-frequency or low-frequency region, and complete the preliminary bit decision. For example, low frequency corresponds to bit 1 and high frequency corresponds to bit 0. Further, generate a confidence level identification according to the magnitude of the energy value corresponding to the main frequency point. The stronger the energy, the higher the confidence level. The confidence level, as an important reference for subsequent processing, can be achieved by pre-establishing a mapping relationship. For example, divide the energy value into different intervals, and each interval corresponds to a confidence level value.

[0068] Finally, based on the energy-timing window and combined with the confidence flag, smooth processing is performed on the preliminary bit decision. The main purpose is to correct the isolated decision jumps caused by channel disturbances in a short period of time, and finally output the stable demodulation result of the first demodulation module. An example of the smooth processing of the preliminary bit decision is as follows: First, define a sliding window with a fixed length, that is, the energy-timing window. The window length is generally set to 5 to 9 bit decision periods, and can be dynamically adjusted according to the system delay and channel stability. The corresponding preliminary bit decision value, main frequency energy value, and confidence flag of each period are recorded in the window. In the current sliding window, count the bit decision value that appears the most times, and record it as the main decision value. If the occurrence ratio of the main decision value exceeds the set threshold (such as more than 80%), it is considered that the overall decision of the window is stable, and the main decision value is output as the smooth result. If the threshold is not exceeded, enter the abnormal discrimination. If there is only a single or a small number (generally one or two) of bit decisions in the window that are inconsistent with the main decision value, and the confidence levels corresponding to these abnormal points are significantly low (lower than the preset confidence threshold, for example, 60%), it is considered an isolated jump. The bit decision value of the detected isolated jump point is directly replaced with the main decision value, and at the same time, its confidence level is updated to the average confidence level in the window. After correction, recalculate the final decision output of the entire window to ensure continuity and smoothness. After each window smooth processing is completed, the window slides forward by one bit period, and continue to perform the same detection and correction on the data in the new window to form a continuous streaming smooth decision. If multiple windows continuously fail to form high consistency (such as the consistency discrimination fails for 3 consecutive windows), the strong decision compensation process is triggered for further backtracking correction.

[0069] For example, when processing a downhole FSK signal, after processing by the fast Fourier transform, it is found that the main frequency point frequency is 17 kHz and the energy density is 95%. If the preset frequency threshold is set to 15 kHz, then this main frequency point is judged as bit 0 and marked as high confidence according to the energy intensity of 95%. If the decision results are consistent in the subsequent 5 consecutive sampling periods, it is directly output. If the main frequency energy drops and the bit decision jumps in a certain sampling period, the energy-timing window smoothing algorithm is used for correction to maintain the decision continuity and avoid bit errors caused by short-term interference.

[0070] The above steps accurately extract the frequency domain information by using the fast Fourier transform, perform bit decision and confidence evaluation based on the main frequency point with the maximum energy, and at the same time combine the energy-timing window for decision smoothing, effectively improving the accuracy and stability of demodulation. Especially in the high-noise downhole environment, it can significantly reduce the bit error rate and enhance the data transmission reliability of the overall high-speed logging communication system.

[0071] Furthermore, step S300 further includes:

[0072] Step S360: Perform adaptive time-window segmentation on the preprocessing result to establish a time-window set.

[0073] Step S370: Extract local spectra for each time-window set.

[0074] Step S380: Calculate the spectral change feature between adjacent time-windows. The spectral change feature includes main frequency offset, energy distribution diffusion feature, energy distribution contraction feature, and local noise rise rate feature.

[0075] Step S390: Combine the spectral change features to construct a local feature vector and perform bit decision to establish the output result of the second demodulation module.

[0076] Specifically, adaptive time-window segmentation means dynamically determining the time-window length according to signal characteristics (such as energy change, timing markers, etc.), and dividing the preprocessing result into several local time-window regions to facilitate subsequent local extraction and analysis of spectral features. First, perform preliminary energy analysis on the preprocessed analog FSK signal. According to the set starting threshold, use an adaptively adjusted time-window mechanism to divide the signal into multiple local time-windows. If the current energy level is stable, the time-window maintains the default length (such as 100 μs); if it is detected that the energy change rate exceeds the preset threshold, dynamically shorten the time-window length (such as shortening to 50 μs) to improve the resolution. Finally, a time-window set covering the entire signal is obtained, and the lengths of each time-window in the time-window set are not exactly the same, providing a basis for subsequent local spectral feature extraction.

[0077] Then, inside each established time-window, perform a fast Fourier transform (FFT) to convert the time-domain signal into a frequency-domain representation. Extract the main frequency point position and spectral energy distribution pattern of each time-window, and record the change trend index of the local spectrum. To suppress spectral leakage and improve the accuracy of spectral features, apply windowing processing (such as Hanning window or Blackman window) to each time-window during the extraction process.

[0078] Next, compare the spectral data extracted from two adjacent time-windows and calculate change feature indexes such as main frequency offset, energy distribution diffusion feature, energy distribution contraction feature, and local noise rise rate feature. Among them, the main frequency offset refers to the difference between the main frequencies of two adjacent time-windows, indicating the degree of change of the signal's main frequency; the energy distribution diffusion feature is represented by the increase in spectral width, reflecting the change in the degree of energy dispersion on the spectrum; the energy distribution contraction feature is represented by the decrease in spectral width, reflecting the change in the degree of energy concentration on the spectrum, which is opposite to the energy distribution diffusion feature; the local noise rise rate feature is represented by the growth rate of the background energy in the non-main frequency region, used to measure the change speed of noise in adjacent time-windows.

[0079] Taking two adjacent time windows as an example, assume that the main frequency of the previous time window is f1 and the main frequency of the next time window is f2. Then the main frequency offset is |f2 - f1|. For the energy distribution diffusion feature, it can be calculated by comparing the spectral energy distributions of the two time windows. For example, calculate the energy concentration degree of the previous time window (such as the proportion of energy concentrated in a certain narrow band) and the energy concentration degree of the next time window. If the energy concentration degree of the next time window decreases, it is considered that the energy distribution has a diffusion trend, otherwise it has a contraction trend. The local noise rising rate feature can be obtained by calculating the ratio of the difference in noise intensity between adjacent time windows to the time interval. For example, in a certain frequency band, the noise intensity of the previous time window is N1, the noise intensity of the next time window is N2, and the time interval is Δt. Then the local noise rising rate feature is (N2 - N1) / Δt. During the feature calculation process, normalization processing is adopted to eliminate the interference of the absolute energy magnitude on the bit decision, ensuring that the feature reflects the relative change trend rather than the signal amplitude deviation.

[0080] Finally, arrange the main frequency offset, energy diffusion feature, energy contraction feature, and local noise rise rate feature calculated for each time window in a fixed order to form a four-dimensional local feature vector. Subsequently, use the trained pattern recognition model (such as the K-nearest neighbor algorithm, support vector machine, or lightweight neural network, etc.) to perform classification on each feature vector and determine the bit value corresponding to the current time window. Combine the judgment results of each time window and output the demodulated bit stream of the second demodulation module. An example of the pattern recognition model training process is as follows: Collect a large number of preprocessed analog FSK signal data samples and accurately label each segment of the signal according to the known bit stream. Each sample corresponds to a specific bit label (such as "0" or "1"). Process each segment of the collected signal according to the actual demodulation process, perform adaptive time window segmentation, local spectrum extraction, and calculate four indicators: main frequency offset, energy diffusion feature, energy contraction feature, and local noise rise rate, to form a four-dimensional local feature vector. Randomly divide all the feature vector data into a training set (such as 80%) and a validation set (such as 20%), ensuring that both the training set and the validation set are representative under different signal-to-noise ratios and different signal conditions. For the K-nearest neighbor algorithm: Determine the optimal number of neighbors k value. The training process is to store the training set feature vectors and their labels in the model, and when classifying, judge based on the nearest neighbor rule. For the support vector machine: Select an appropriate kernel function (such as linear kernel, radial basis kernel RBF), perform hyperplane segmentation training, and maximize the class interval. For the lightweight neural network: Design a small neural network with one or two hidden layers, use the cross-entropy loss function, and use the Adam optimizer for iterative training to adjust the network weights to minimize the training error. Use the validation set to evaluate the accuracy, recall rate, and other indicators of the trained model. If the performance is not ideal, adjust the feature selection method, model parameters (such as k value, kernel function type, network layer number, etc.) or increase the training data sample size for retraining and optimization. Save the trained and validated model for classifying and judging the real-time feature vectors during the subsequent online demodulation process.

[0081] Further, step S200 includes:

[0082] Step S210: When the signal granularity discrimination result is the first granularity discrimination result, activate the long-time window smoothing mode and update the long-time window smoothing mode to the internal recognition mode of dual-mode FSK, where the first granularity discrimination result is the coarse granularity discrimination result.

[0083] Step S220: When the signal granularity discrimination result is the second granularity discrimination result, activate the high-confidence tracking mode and update the high-confidence tracking mode to the internal recognition mode of dual-mode FSK, where the second granularity discrimination result is the fine granularity discrimination result.

[0084] Specifically, the signal granularity discrimination result refers to the degree of granularity of the signal in the time domain and frequency domain determined by analyzing the characteristics of the preprocessed analog FSK signal, which is divided into coarse-grained discrimination results (i.e., the first granularity discrimination result) and fine-grained discrimination results (i.e., the second granularity discrimination result).

[0085] The mode update judgment is performed according to the signal granularity discrimination result obtained in step S100. When the discrimination result is the first granularity discrimination result (i.e., the coarse-grained discrimination result), it is considered that the current downhole communication link signal has a large scale variation characteristic and poor stability in a short time, so the long-time window smoothing mode is activated and the mode is updated to the internal recognition mode of the dual-mode FSK. In the long-time window smoothing mode, a larger energy smoothing window is used and the time domain cumulative weight is increased to suppress the interference of random noise on spectrum extraction and bit judgment, thereby improving the demodulation stability. For example, for the energy detection of the signal, the energy values of multiple sampling points in the time window are averaged to obtain a smoothed energy value for subsequent main frequency point determination and bit judgment. This smoothing process can reduce the energy fluctuations caused by short-term noise or interference and improve the stability of the demodulation.

[0086] When the discrimination result is the second granularity discrimination result (i.e., fine-grained discrimination result), it is considered that the current signal has good short-term stability and high local feature discernibility. At this time, the high-confidence tracking mode is activated and updated to the internal recognition mode of dual-mode FSK. In the high-confidence tracking mode, the energy smoothing window length is reduced, the sensitivity of feature changes is increased, and a dynamic confidence calculation mechanism is adopted to enhance the response capability to subtle main frequency drift and local feature disturbances, thereby improving the real-time judgment accuracy under high-speed communication.

[0087] For example, when the energy of the received downhole signal fluctuates violently and the main frequency is unstable in a short period of time, the coarse-grained discrimination result is output after granularity discrimination, and the mode is switched to long-time window smoothing to extract signal characteristics under an energy smoothing window of 200μs. When the signal shows good continuity, slow changes in the main frequency and low local noise, the fine-grained discrimination result is output, and then the mode is switched to high-confidence tracking mode, shortening the energy smoothing window to 50μs, and enabling dynamic confidence weighted judgment to capture the changes of each bit in detail.

[0088] During the demodulation process of the simulated FSK modulation signal, the first demodulation module and the second demodulation module always work in parallel, while the signal granularity discrimination result determines the weight allocation mechanism and the confidence fusion method. At a coarse granularity, the output result of the first demodulation module is mainly used, and the output result of the second demodulation module is used for confidence correction or low-weight repeated discrimination. At a fine granularity, the local spectral dynamic features extracted by the second demodulation module are mainly used, and the first demodulation module provides a candidate support and a redundant error correction channel. The output of each demodulated bit is based on the output results of the first demodulation module and the second demodulation module to jointly give a preliminary fusion decision after confidence weighting fusion.

[0089] By dynamically updating the internal recognition mode of the dual-mode FSK according to the signal granularity discrimination result, the above steps can adaptively adjust the demodulation strategy according to the actual signal state of the communication link, maintain high stability in poor channel conditions, improve the demodulation sensitivity and rate in good channel conditions, and significantly enhance the robustness and overall transmission quality of high-speed logging communication.

[0090] Further, step S100 includes:

[0091] Step S110: Obtain the working condition characteristics of the front-end working condition, and adaptively match the noise suppression parameters based on the working condition characteristics.

[0092] Step S120: Perform noise suppression preprocessing on the simulated FSK modulation signal by using the adaptively matched noise suppression parameters.

[0093] Specifically, the front-end working condition refers to the current physical environment state during downhole logging communication, including relevant environmental parameters affecting signal quality such as geological layer type, well depth, lithology, well fluid viscosity, conductivity, vibration intensity, etc. The working condition characteristics are environmental quantization indexes obtained through sensor collection or input from the regulation end, such as noise background intensity, signal attenuation rate, instantaneous signal-to-noise ratio, etc., which are used as the basis for matching noise suppression parameters. First, obtain the real-time or quasi-real-time information of the front-end working condition, which is collected by the downhole sensing module or transmitted and obtained from the ground system. After extracting the working condition characteristics from the front-end working condition, compare them with the predefined working condition-parameter mapping model, and determine the currently applicable noise suppression parameters according to the matching result, including key parameters such as filter bandwidth, cut-off frequency, noise reduction weight, suppression gain threshold, etc. These parameters will then be used for noise suppression processing.

[0094] Execute multi - level noise suppression on the simulated FSK signal according to the selected parameter configuration. Common processes include: Band - pass filtering: Retain the FSK modulation frequency range (such as 2 kHz - 5 kHz) and remove out - of - band interference; Spectral subtraction for noise reduction: Estimate the noise spectrum and subtract it from the signal spectrum to increase the proportion of the net signal; Adaptive gain adjustment: Automatically adjust the signal intensity to avoid loss of features due to too low signal amplitude; Anti - mutation smoothing processing: Suppress abnormal energy values caused by sudden shocks or instantaneous spikes.

[0095] The above steps achieve robust adaptation to different downhole communication environments by obtaining working condition characteristics and performing adaptive noise suppression, effectively suppressing background noise and environmental interference, improving the quality of signal pre - processing, enhancing the accuracy of subsequent signal granularity discrimination and demodulation, and providing a solid foundation for dual - mode FSK communication.

[0096] Further, step S700 includes:

[0097] Step S710: Establish an early - warning monitoring index set, which includes a signal strength change index, a main frequency drift rate index, a bit error rate index, and an abnormal confidence distribution index.

[0098] Step S720: Use the early - warning monitoring index set to perform early - warning trigger analysis on the simulated FSK modulation signal and report an early - warning signal.

[0099] Specifically, according to the communication stability requirements of the simulated FSK modulation signal, establish a multi - dimensional early - warning monitoring index set including a signal strength change index, a main frequency drift rate index, a bit error rate index, and an abnormal confidence distribution index. Among them, the signal strength change index is used to measure the fluctuation of the signal strength and reflects the stability of the signal during transmission. The main frequency drift rate index focuses on the change rate of the main frequency of the signal over time and reflects the frequency stability of the signal. If the main frequency drift rate is too high, it means that the signal source is unstable or affected by external interference. The bit error rate index refers to the probability of incorrect bits occurring during the transmission of digital signals. It directly reflects the reliability of signal transmission and is one of the key indicators for evaluating communication quality. The abnormal confidence distribution index is used to monitor the distribution of the confidence of each bit decision during signal demodulation. The normal confidence distribution usually has a certain regularity, and an abnormal distribution may indicate potential problems during the demodulation process, such as increased noise interference or signal feature changes.

[0100] Examples of calculating each warning and monitoring indicator are as follows: For the signal strength change indicator, energy detection needs to be performed on the input analog FSK signal. In each sliding window, calculate: the maximum energy value Emax, the minimum energy value Emin, and the root mean square energy Erms; define the energy fluctuation ΔE = Emax - Emin. When ΔE exceeds the set fluctuation threshold, or Erms shows an abnormal decrease (such as a decrease exceeding the set percentage), it is determined that the signal strength is abnormal, and a signal strength change warning is triggered.

[0101] For the main frequency drift rate indicator, continuously extract the main frequency point f main (t) of each time slice based on the fast Fourier transform (FFT). Calculate the change amount of the main frequency between adjacent time slices. The drift amount = ∣f main (t + Δt) - f main (t)∣. Then, calculate the main frequency drift rate per unit time: drift rate = drift amount / Δt. When the drift rate exceeds the set threshold T f (such as 1 kHz / ms), it is determined that the main frequency is unstable, and a main frequency drift warning is triggered.

[0102] For the bit error rate indicator, in the sliding window, count the comparison between the bit decision result and the reference bit stream (or through self-consistent backtracking) to obtain the number of bit errors N err . At the same time, record the total number of decision bits N total , and calculate the bit error rate: BER = N err / N total . When the BER value continuously exceeds the set threshold (such as 1% or a dynamic adaptive threshold), a bit error rate anomaly warning is triggered.

[0103] For the confidence distribution anomaly indicator, at each bit decision, extract the corresponding confidence value C(t) (such as energy difference, discrimination probability, etc.). In the sliding window, count the confidence distribution, including the mean μ C , the standard deviation σ C , and the proportion of low-confidence samples (such as the proportion of samples below a certain threshold). Then check the following abnormal situations: the confidence mean decreases by more than the threshold (such as more than 20% decrease), the confidence standard deviation increases abnormally (such as more than 15%), and the proportion of low-confidence samples surges (such as more than 30%). When any one of the abnormalities occurs, a confidence anomaly warning is triggered.

[0104] The above indicators are continuously monitored and analyzed in real time, and early warning trigger analysis is performed based on the monitoring and analysis results. Each early warning indicator can be determined independently, or comprehensive early warning rules can be set. For example, "any single indicator is abnormal three times in a row" or "two indicators are abnormal at the same time" triggers an overall early warning and reports an early warning signal. The early warning signal is used to indicate that the current communication quality is deteriorating and the potential risk of failure is increasing, and supports subsequent communication link switching, parameter adaptive adjustment or manual intervention operations. In addition, the time when the early warning occurs, the specific values of related indicators and other information can be recorded for subsequent troubleshooting and analysis.

[0105] By triggering analysis and issuing warning signals, relevant personnel can be reminded in time to pay attention to the status of the communication link and take corresponding solutions, which will help improve the reliability and stability of high-speed logging communications, reduce the risk of logging data loss or errors due to signal quality problems, and ensure the smooth progress of logging operations.

[0106] In summary, the high-speed well logging communication method with dual-mode FSK demodulation provided in the embodiment of the present application has the following beneficial effects:

[0107] Embodiments of the present application achieve precise demodulation and real-time warning of simulated FSK modulation signals under strong noise interference and complex working conditions through a multi-stage and multi-modal parallel processing architecture. First, the front-end working condition characteristics are obtained and noise adaptive parameter matching is performed to achieve the dynamic adaptability of signal preprocessing. Then, signal granularity discrimination is introduced to dynamically perceive the characteristics of FSK signals in the complex downhole environment. According to the results of signal granularity discrimination, the long-time window smoothing mode under coarse granularity or the high-confidence tracking mode under fine granularity is respectively activated. In both modes, a dual-mode demodulation structure of main frequency-energy demodulation and local spectrum dynamic feature extraction runs in parallel. The first demodulation module (main frequency-energy demodulation module) quickly extracts the main frequency point and energy through fast Fourier analysis, performs confidence weighting and bit decision smoothing, and outputs the first demodulation result; while the second demodulation module (local spectrum dynamic feature extraction module) generates dynamic feature vectors such as main frequency offset, energy diffusion / shrinkage, and noise rise rate through adaptive time window division, local spectrum feature extraction, and change amount calculation, and outputs the second demodulation result after performing bit decision. The dual-mode output is fused and judged by combining the internal recognition mode, and the bit selection mechanism is dynamically adjusted by combining the noise level, granularity category, and confidence, which improves the stability under anti-interference and variable environments, improves the accuracy of signal recognition, and enhances the overall anti-interference ability. On the basis of the fusion decision, to cope with the short-term recognition instability caused by spectrum offset, an adjacent frequency point set is further constructed to assist in judging whether to trigger the soft decision compensation mechanism. If the soft decision compensation is triggered, the fusion preliminary decision is updated using the adjacent frequency point set. If the soft decision compensation is not triggered, it enters the consistency verification and backtracking correction stage, and time series sub-channels and spectrum sub-channels are introduced for focus comparison and correction verification. On this basis, an early warning monitoring index set centered on signal intensity change, main frequency drift rate, bit error rate, and confidence distribution is established, and through sliding window analysis and dynamic threshold judgment, fast early warning and response to demodulation anomalies are realized.

[0108] Overall, the embodiments of the present application effectively enhance the accuracy and reliability of signal demodulation. Especially in the complex downhole environment, it can significantly reduce the bit error rate, improve the communication quality, and meet the actual requirements of high-speed logging for high-reliability data transmission.

[0109] Embodiment 2, as Figure 3 shown, based on the same inventive concept as the foregoing Embodiment 1, the embodiments of the present application provide a high-speed logging communication system for dual-mode FSK demodulation, and the system includes:

[0110] A signal granularity discrimination module 10, configured to receive an analog FSK modulation signal of a downhole communication link, and perform signal granularity discrimination on the preprocessing result after preprocessing the analog FSK modulation signal.

[0111] An identification mode update module 20 is configured to dynamically update the internal identification mode of the dual-mode FSK by using the signal granularity discrimination result.

[0112] A result sending module 30 is configured to send the preprocessing result to a first demodulation module and a second demodulation module of the dual-mode FSK respectively, wherein the first demodulation module is a main frequency-energy demodulation module, and the second demodulation module is a local spectrum dynamic feature extraction module.

[0113] A result fusion module 40 is configured to respectively obtain the output results corresponding to the first demodulation module and the second demodulation module, fuse the output results by using the internal identification mode, and establish a preliminary fusion decision.

[0114] An adjacent frequency point search module 50 is configured to search for adjacent frequency points for the preliminary fusion decision and establish an adjacent frequency point set.

[0115] A trigger analysis module 60 is configured to perform soft decision trigger analysis based on the preliminary fusion decision and the adjacent frequency point set. If soft decision compensation is triggered, the preliminary fusion decision is updated by using the adjacent frequency point set.

[0116] A communication execution module 70 is configured to perform high-speed logging communication by using the updated preliminary fusion decision.

[0117] Further, the trigger analysis module 60 in the embodiment of the present application is further configured to perform the following steps:

[0118] If soft decision compensation is not triggered, a preliminary candidate bit stream is established based on the preliminary fusion decision; the consistency check of the preliminary candidate bit stream is performed by using a sliding window; if the consistency check fails, strong decision compensation is triggered, the preliminary fusion decision is backtracked and corrected by using the strong decision compensation, and high-speed logging communication is performed by using the backtracking correction result.

[0119] Further, the trigger analysis module 60 in the embodiment of the present application is further configured to perform the following steps:

[0120] During the backtracking check process, multiple data nodes are backtracked to establish a check chain, and the check chain includes a timing check chain and a signal check chain; starting from the node where strong decision is triggered, item-by-item check and backtracking are performed based on the check chain to establish the backtracking correction result.

[0121] Further, the trigger analysis module 60 in the embodiment of the present application is further configured to perform the following steps:

[0122] Activate the timing check sub-channel, and use the timing check sub-channel to perform time consistency check based on the timing check chain to generate the first concerned check result; activate the spectrum check sub-channel, and use the spectrum check sub-channel to perform spectrum feature consistency check based on the signal check chain to generate the second concerned check result; perform double-check verification on the first concerned check result and the second concerned check result to complete item-by-item check backtracking.

[0123] Further, the result sending module 30 in the embodiment of the present application is further configured to execute the following steps:

[0124] After extracting the frequency domain information of the preprocessing result by using the first demodulation module, perform fast Fourier transform; perform energy detection on the output of the fast Fourier transform to determine the main frequency point with the maximum energy; compare the main frequency point with a preset frequency threshold to complete preliminary bit decision; perform confidence level identification of the preliminary bit decision based on the energy value corresponding to the main frequency point; perform smoothing of the preliminary bit decision based on the confidence level identification based on the energy-timing window, and output the output result of the first demodulation module.

[0125] Further, the result sending module 30 in the embodiment of the present application is further configured to execute the following steps:

[0126] Perform adaptive time window segmentation on the preprocessing result to establish a time window set; perform local spectrum extraction on each time window set; calculate the spectrum change amount features of adjacent time windows, where the spectrum change amount features include main frequency offset amount, energy distribution diffusion feature, energy distribution contraction feature, and local noise rise rate feature; combine the spectrum change amount features to construct a local feature vector, and perform bit decision to establish the output result of the second demodulation module.

[0127] Further, the identification mode update module 20 in the embodiment of the present application is further configured to execute the following steps:

[0128] When the signal granularity discrimination result is the first granularity discrimination result, activate the long-time window smoothing mode, and update the long-time window smoothing mode to the internal identification mode of dual-mode FSK, where the first granularity discrimination result is a coarse granularity discrimination result; when the signal granularity discrimination result is the second granularity discrimination result, activate the high-confidence tracking mode, and update the high-confidence tracking mode to the internal identification mode of dual-mode FSK, where the second granularity discrimination result is a fine granularity discrimination result.

[0129] Further, the signal granularity discrimination module 10 in the embodiment of the present application is further configured to execute the following steps:

[0130] Obtain the working condition characteristics of the front-end working condition, adaptively match the noise suppression parameters based on the working condition characteristics; use the adaptively matched noise suppression parameters to perform noise suppression preprocessing on the simulated FSK modulation signal.

[0131] Furthermore, the communication execution module 70 in the embodiment of the present application is further configured to execute the following steps:

[0132] Establish an early warning monitoring index set, where the early warning monitoring index set includes a signal strength change index, a main frequency drift rate index, a bit error rate index, and a confidence distribution anomaly index; use the early warning monitoring index set to perform early warning trigger analysis on the simulated FSK modulation signal and report an early warning signal.

[0133] Through the foregoing detailed description of a high-speed logging communication method for dual-mode FSK demodulation in this specification, those skilled in the art can clearly know a high-speed logging communication system for dual-mode FSK demodulation in this embodiment. For the system disclosed in Embodiment 2, since it corresponds to the method disclosed in Embodiment 1, it has corresponding functional modules and beneficial effects. For the relevant parts, refer to the description in the method part.

[0134] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A high-speed logging communication method for dual-mode FSK demodulation, characterized in that, The method includes: Receiving an analog FSK modulation signal of an underground communication link, and after preprocessing the analog FSK modulation signal, performing signal granularity discrimination on the preprocessing result; Dynamically updating the internal recognition mode of dual-mode FSK by using the signal granularity discrimination result, including: When the signal granularity discrimination result is the first granularity discrimination result, activating the long-time window smoothing mode and updating the long-time window smoothing mode as the internal recognition mode of dual-mode FSK, where the first granularity discrimination result is a coarse granularity discrimination result; When the signal granularity discrimination result is the second granularity discrimination result, activating the high-confidence tracking mode and updating the high-confidence tracking mode as the internal recognition mode of dual-mode FSK, where the second granularity discrimination result is a fine granularity discrimination result; Sending the preprocessing result to the first demodulation module and the second demodulation module of dual-mode FSK respectively, where the first demodulation module is a main frequency-energy demodulation module and the second demodulation module is a local spectrum dynamic feature extraction module; Respectively obtaining the output results corresponding to the first demodulation module and the second demodulation module, and fusing the output results by using the internal recognition mode to establish a preliminary fusion decision; Performing adjacent frequency point search on the preliminary fusion decision to establish an adjacent frequency point set; Performing soft decision trigger analysis based on the preliminary fusion decision and the adjacent frequency point set. If soft decision compensation is triggered, updating the preliminary fusion decision by using the adjacent frequency point set; Performing high-speed logging communication by using the updated preliminary fusion decision.

2. The high-speed logging communication method for dual-mode FSK demodulation according to claim 1, wherein The performing soft decision trigger analysis based on the preliminary fusion decision and the adjacent frequency point set includes: If soft decision compensation is not triggered, establishing a preliminary candidate bit stream based on the preliminary fusion decision; Performing consistency check on the preliminary candidate bit stream by using a sliding window; If the consistency check fails, triggering strong decision compensation, performing backtracking correction on the preliminary fusion decision by using the strong decision compensation, and performing high-speed logging communication by using the backtracking correction result.

3. The high-speed logging communication method for dual-mode FSK demodulation according to claim 2, wherein The performing backtracking correction on the preliminary fusion decision by using the strong decision compensation includes: During the backtracking check process, backtracking multiple data nodes to establish a check chain, where the check chain includes a timing check chain and a signal check chain; Starting from the node where strong decision is triggered, performing item-by-item check backtracking based on the check chain to establish the backtracking correction result.

4. A high-speed logging communication method for dual-mode FSK demodulation according to claim 3, characterized in that, The performing item-by-item check backtracking based on the check chain includes: Activating a timing check sub-channel and performing time consistency check based on the timing check chain by using the timing check sub-channel to generate a first attention check result; Activating a spectrum check sub-channel and performing spectrum feature consistency check based on the signal check chain by using the spectrum check sub-channel to generate a second attention check result; Performing double-check verification on the first attention check result and the second attention check result to complete item-by-item check backtracking.

5. A high-speed logging communication method for dual-mode FSK demodulation according to claim 1, characterized in that, The sending the preprocessing result to the first demodulation module and the second demodulation module of dual-mode FSK respectively includes: After extracting the frequency domain information of the preprocessing result by using the first demodulation module, performing fast Fourier transform; Perform energy detection on the output of the fast Fourier transform to determine the main frequency point with the maximum energy; Compare the main frequency point with a preset frequency threshold to complete preliminary bit decision; Perform confidence identification for the preliminary bit decision based on the energy value corresponding to the main frequency point; Perform smoothing of the preliminary bit decision based on the confidence identification based on the energy-time window, and output the output result of the first demodulation module.

6. The high-speed logging communication method for dual-mode FSK demodulation according to claim 5, characterized in that The step of sending the preprocessing result to the first demodulation module and the second demodulation module of the dual-mode FSK further includes: Perform adaptive time window segmentation on the preprocessing result to establish a time window set; Extract local spectra for each time window set; Calculate the spectral change characteristics of adjacent time windows, and the spectral change characteristics include main frequency offset, energy distribution diffusion characteristics, energy distribution contraction characteristics, and local noise rise rate characteristics; Combine the spectral change characteristics to construct a local feature vector, and perform bit decision to establish the output result of the second demodulation module.

7. The high-speed logging communication method for dual-mode FSK demodulation according to claim 1, wherein The preprocessing of the analog FSK modulation signal includes: Obtain the working condition characteristics of the front-end working condition, and adaptively match the noise suppression parameters based on the working condition characteristics; Perform noise suppression preprocessing on the analog FSK modulation signal using the adaptively matched noise suppression parameters.

8. A high-speed logging communication method for dual-mode FSK demodulation according to claim 1, characterized in that The high-speed logging communication using the updated fusion preliminary decision includes: Establish a warning monitoring index set, and the warning monitoring index set includes signal strength change index, main frequency drift rate index, bit error rate index, and confidence distribution anomaly index; Perform warning trigger analysis on the analog FSK modulation signal using the warning monitoring index set, and report a warning signal.

9. A high-speed logging communication system with dual-mode FSK demodulation, characterized in that, The system is used to execute a high-speed logging communication method for dual-mode FSK demodulation according to any one of claims 1-8, including: A signal granularity discrimination module, configured to receive the analog FSK modulation signal of the downhole communication link, and perform signal granularity discrimination on the preprocessing result after preprocessing the analog FSK modulation signal; An identification mode update module, configured to dynamically update the internal identification mode of the dual-mode FSK using the signal granularity discrimination result; A result sending module, configured to send the preprocessing result to the first demodulation module and the second demodulation module of the dual-mode FSK respectively, wherein the first demodulation module is a main frequency-energy demodulation module, and the second demodulation module is a local spectrum dynamic feature extraction module; A result fusion module, configured to respectively obtain the output results corresponding to the first demodulation module and the second demodulation module, and fuse the output results using the internal identification mode to establish a fusion preliminary decision; An adjacent frequency point search module, configured to search for adjacent frequency points for the fusion preliminary decision to establish an adjacent frequency point set; A trigger analysis module, configured to perform soft decision trigger analysis based on the fusion preliminary decision and the adjacent frequency point set. If soft decision compensation is triggered, update the fusion preliminary decision using the adjacent frequency point set; A communication execution module, configured to perform high-speed logging communication using the updated fusion preliminary decision.

Images