Deep well geological exploration tunneling parameter control system and method

Abstract

The invention relates to the technical field of petroleum and natural gas drilling engineering, and discloses a deep well geological exploration tunneling parameter control system and method, and the system comprises a signal collection and synchronization module which collects drilling parameters and aligns timestamps; the initialization and modeling module is used for calculating fluid transmission time lag and blade skim-over frequency; the active disturbance generation module generates a target bit pressure instruction superposed with sine disturbance to drive load application; the two-channel feature extraction and verification module aligns the historical torque and the current pressure according to time delay and calculates the confidence coefficient in combination with the frequency; the adaptive decision and closed-loop control module uses the confidence coefficient to correct the gain and update the reference bit pressure to form a closed loop. A dual-channel verification framework based on fluid-structure interaction transmission time lag and blade characteristic frequency is constructed, the covering effect of long-distance drill column friction torque on real cutting torque in deep well operation is effectively stripped, and it is ensured that a control system only responds to effective rock breaking action.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oil and gas drilling engineering, in particular to a deep well geological exploration tunneling parameter control system and method. BACKGROUND

[0002] With the extension of energy exploration to deep and ultra-deep geological structures, the working conditions of drilling operations are becoming increasingly complex. In order to protect downhole tools while improving rock breaking efficiency, it is necessary to accurately control the drilling pressure, a core parameter.

[0003] The existing automatic drilling technology mainly relies on the data of hook load, standpipe pressure and top drive torque collected by ground sensors for feedback control. The general practice is to use a PID control strategy to maintain the drilling pressure or the mechanical drilling speed near the preset target value to achieve constant drilling pressure or constant drilling speed drilling. Before the operation starts, the engineers usually conduct a drilling test to determine an ideal drilling pressure setting value by analyzing the drilling speed performance under different drilling pressures. This method can effectively reduce the labor intensity of the driller and provide basic automated operation capability in shallow or medium deep well operations.

[0004] However, in deep well operations of several kilometers, this mode based on direct feedback from the ground has obvious physical limitations. First, there is a time-space dislocation problem in signal transmission. The speed of mechanical wave transmission along the drill string steel body is much faster than the propagation speed of pressure wave in the drilling fluid, resulting in a misalignment of the torque signal and the standpipe pressure signal on the time axis. Direct coupling calculation will cause a phase difference between the control command and the actual state at the bottom of the well. Second, the friction effect of the long open hole section in deep wells is obvious. The huge pipe column friction torque often overwhelms the small torque change produced by the drill bit cutting rock, making it difficult for the system to distinguish whether the torque rise is due to the drill bit eating into the formation or the pipe column and the well wall friction, thereby causing distorted control decisions. In addition, the lithology of the formation is changing in real time. The pre-set fixed parameters cannot adapt to the sudden geological environment, and the passive control strategy lacks the ability to actively explore the drillability boundary of the formation, making it difficult to find the optimal working point without interrupting the drilling. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a deep well geological exploration tunneling parameter control system and method to solve the technical problems that, in the existing deep well automated drilling technology, due to the significant difference in transmission speed between the drilling fluid pressure wave and the drill string mechanical wave, the ground monitoring signals are not synchronized in time and space, and the high friction interference of the long open hole section in deep wells masks the real cutting characteristics of the drill bit, making it impossible for the control system to accurately identify the downhole working conditions and adaptively lock the optimal drilling pressure parameter.

[0006] To achieve the above object, the present application is implemented by the following technical solutions: a deep well geological exploration tunneling parameter control system, comprising: A signal acquisition and synchronization module is configured to acquire the hook load, top drive torque and standpipe pressure and perform timestamp alignment, and output a real-time drilling parameter sequence; An initialization and modeling module is configured to calculate a fluid transmission time lag based on the drilling fluid properties and the wellbore depth, and calculate a blade sweep frequency based on the real-time rotational speed and the bit structure; An active disturbance generation module is configured to generate a target drilling pressure instruction superimposed with a sinusoidal waveform disturbance based on a reference drilling pressure, and drive an actuator to apply a dynamic axial load at the bottom of the well; A double-channel feature extraction and verification module is configured to receive the real-time drilling parameter sequence, extract ground torque data at a historical time based on the fluid transmission time lag to align the standpipe pressure data at the current time, and calculate a confidence value representing the effectiveness of the torque response in combination with the blade sweep frequency; An adaptive decision and closed-loop control module is configured to correct the differential response gain of the ground torque with respect to the drilling pressure using the confidence value, update the reference drilling pressure based on the corrected effective gain value, and provide the updated reference drilling pressure to the active disturbance generation module for generating a new target drilling pressure instruction.

[0007] Preferably, the signal acquisition and synchronization module comprises: A sensor array is configured to acquire analog signals or digital signals of the hook load, top drive torque and standpipe pressure; Low-pass filtering is applied to the hook load, top drive torque and standpipe pressure respectively, and a unified timestamp is marked for each channel data based on a system unified clock; The filtered data is reorganized according to the same timestamp index to construct the real-time drilling parameter sequence containing time-aligned ground drilling pressure, ground torque and standpipe pressure; The real-time drilling parameter sequence is stored in the system's circular buffer for subsequent module calling.

[0008] Preferably, the initialization and modeling module performs the following: The mixed fluid density and equivalent bulk modulus of the drilling fluid are obtained, and the propagation speed of pressure waves in the wellbore is calculated; The fluid transmission time lag required for fluid pressure waves to transmit from the bottom of the well to the ground surface is calculated based on the measured depth along the wellbore axis trajectory and the propagation speed; The number of main blades of the drill bit and the real-time rotational speed of the top drive or rotary table are obtained; The theoretical reference characteristic frequency when the drill bit cuts rock is calculated based on the kinematic relationship, which is used as the blade sweep frequency.

[0009] Preferably, the active disturbance generation module comprises: reading the reference WOB; calculating a sinusoidal signal varying with time according to a preset disturbance amplitude and a preset disturbance frequency; superimposing the sinusoidal signal on the reference WOB to generate the target WOB instruction; sending the target WOB instruction to an automatic pipe running actuator to control the axial release speed of the drill string at the wellhead to establish axial load fluctuation; wherein the disturbance frequency is set at a low frequency range outside the mud pump stroke frequency and the harmonic frequency of the mud pump stroke frequency.

[0010] Preferably, the adaptive decision and closed-loop control module comprises: selecting an integer multiple of the disturbance period as a data window to obtain the surface WOB sequence and the surface torque sequence in the data window; performing linear regression analysis on the surface WOB sequence and the surface torque sequence by least squares method to calculate the original differential response gain; weighting and correcting the original differential response gain by using the confidence value to obtain the effective cutting gain representing the true cutting effect; combining the effective cutting gain with a preset optimization step coefficient to calculate the reference WOB correction amount and update the reference WOB.

[0011] Preferably, the dual-channel feature extraction and verification module performs: running fluid-structure coupling time-domain verification logic and mechanical frequency spectrum frequency-domain verification logic in parallel; in the fluid-structure coupling time-domain verification logic, establishing the space-time correspondence between the surface torque and the riser pressure by using the fluid transmission time lag to calculate the fluid-structure coupling correlation coefficient; in the mechanical frequency spectrum frequency-domain verification logic, extracting frequency domain energy features by using the blade passing frequency to calculate the mechanical energy sensitivity; according to the drilling assembly configuration, assigning a weight factor to the fluid-structure coupling correlation coefficient and the mechanical energy sensitivity, and weighting and fusing the two to output the confidence value.

[0012] Preferably, the fluid-structure coupling time-domain verification logic comprises: reading the fluid transmission time lag and obtaining the torque sampling point that is earlier than the fluid transmission time lag by the number of seconds at the current time; reorganizing the torque sampling point to perform delay alignment processing to generate an aligned torque signal that is synchronized in time sequence with the riser pressure data; establishing a sliding time window, calculating a covariance of the aligned torque signal and real-time standpipe pressure data within the sliding time window; calculating a Pearson cross-correlation coefficient based on the covariance as the fluid-structure coupling correlation coefficient to quantify the synchronism of the lagged mechanical torque response and the fluid pressure response in waveform variation trend.

[0013] Preferably, the mechanical frequency spectrum frequency domain verification logic comprises: performing a fast Fourier transform on the ground torque data within the sliding window to convert into a frequency domain power spectral density function; taking a frequency band with a preset bandwidth centered on the blade sweep frequency, calculating an integral energy within the frequency band; calculating a differential ratio of the integral energy with respect to the ground weight on bit to obtain the mechanical energy sensitivity; mapping the mechanical energy sensitivity to a normalized interval using a piecewise normalization function with dead zone and saturation characteristics for subsequent weighted fusion calculation.

[0014] Preferably, the adaptive decision and closed-loop control module further comprises, when updating the reference weight on bit: calculating a difference between the current ground torque and a preset target threshold value; if the difference exceeds zero, calculating a rollback response amount according to the difference and a preset overload protection coefficient; subtracting the rollback response amount from the reference weight on bit correction amount to obtain a reference weight on bit set value at the next time; judging whether the reference weight on bit set value is within a preset safety interval, and if it exceeds, forcing to clamp to the boundary value of the safety interval.

[0015] Preferably, a deep well geological exploration excavation parameter control method comprises the following steps: collecting hook load, top drive torque and standpipe pressure and performing timestamp alignment to output real-time drilling parameter sequences; calculating fluid transmission time lag according to drilling fluid properties and wellbore depth, and calculating blade sweep frequency according to real-time rotational speed and bit structure; generating a target weight on bit instruction superimposed with a sinusoidal waveform disturbance based on the reference weight on bit to drive the actuator to apply dynamic axial load at the bottom of the well; receiving the real-time drilling parameter sequences, extracting ground torque data at a historical time according to the fluid transmission time lag to align the standpipe pressure data at the current time, and combining the blade sweep frequency to calculate a confidence value representing the effectiveness of the torque response; The differential response gain of the ground torque with respect to the WOB is corrected by using the confidence value, and the reference WOB is updated according to the corrected effective gain value, and the new target WOB instruction is generated.

[0016] The present application provides a deep well geological exploration excavation parameter control system and method. The present application has the following advantages: 1. The present application constructs a double-channel verification architecture based on fluid-structure coupling transmission time delay and blade characteristic frequency, and the control gain is weighted and corrected by calculating the confidence of multi-source signals, effectively eliminating the masking effect of long-distance drill string friction torque on the real cutting torque in deep well operation, ensuring that the control system only responds to effective rock breaking actions. Compared with the existing technology which directly relies on the feedback of the ground torque amplitude, the present application solves the engineering problem that the system cannot accurately identify the bit working state due to the interference of friction in the ultra-deep well section, thereby causing misjudgment or blind pressure increase leading to drill string buckling.

[0017] 2. According to the physical characteristics of the obvious difference between the transmission speed of the drilling fluid pressure wave and the drill string mechanical wave, the present application implements a history data reorganization and space-time alignment strategy based on fluid transmission time delay, eliminates the time phase difference of multi-physical field signals arriving at the ground sensor, and establishes accurate time sequence mapping between ground measurement data and instantaneous working conditions at the bottom of the well. Unlike traditional PID control or simple threshold feedback technology which often ignores signal transmission delay, leading to serious lag of control instructions or oscillation when facing sudden changes in the formation, the present application significantly improves the phase margin and stability of the closed-loop control system in the transient response process.

[0018] 3. The present application uses an active detection method of superimposing a micro-amplitude low-frequency sinusoidal disturbance, continuously excites the formation to produce a dynamic response during drilling, and excites the adaptive gradient search algorithm. Under the premise of not interrupting normal drilling operation, the present application can real-time capture the change trend of formation drillability and automatically lock the optimal drilling pressure working point. Compared with the existing static drilling test or stepwise pressure test method, the present application avoids the problem of increased non-production time and high data sample dispersion caused by stopping drilling for testing, and realizes the automatic optimization of drilling efficiency maximization and mechanical specific energy minimization. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The present application is a system architecture schematic diagram of the embodiment; Figure 2 The present application is a method flowchart schematic diagram of the embodiment.

[0020] Among them: 100, signal acquisition and synchronization module; 200, initialization and modeling module; 300, active disturbance generation module; 400, double-channel feature extraction and verification module; 500, adaptive decision and closed-loop control module. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the specification of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0022] Referring to the drawings Figure 1 , Figure 1 is a schematic diagram of the overall architecture of a deep well geological exploration and excavation parameter control system according to an embodiment of the present application. The present application provides a deep well geological exploration and excavation parameter control system applied to a deep well drilling engineering site. The deep well geological exploration and excavation parameter control system is constructed based on an existing industrial drilling rig hardware environment, which includes a surface rotary steering device, a drill string lifting device, a mud circulation device, and a downhole drilling assembly. The surface rotary steering device is specifically embodied as a top drive system or a rotary table system, which is used to drive the drill string to rotate and provide cutting torque. The drill string lifting device is embodied as a drawworks system or an automatic pipe handler, which is used to control the axial displacement of the drill string and the weight on bit applied to the drill bit. The downhole drilling assembly is connected to the bottom end of the drill string, including a drill bit and an optional downhole power drill.

[0023] The deep well geological exploration and excavation parameter control system is configured with a sensor array, which includes a hook load sensor, a top drive torque sensor, and a standpipe pressure sensor. The hook load sensor is installed at the hook or dead rope fixing end of the drilling rig, which is used to detect the suspended load data to calculate the surface weight on bit. The top drive torque sensor is installed at the output shaft of the top drive motor or reads data through frequency converter current feedback, which is used to detect the surface mechanical torque. The standpipe pressure sensor is installed at the standpipe manifold, which is used to detect the fluid pressure when the mud pump is pumped into the wellhead.

[0024] The core processing unit of the deep well geological exploration and excavation parameter control system is a programmable automation controller, which is communicatively connected with the sensor array, the top drive system, and the drawworks system through an industrial field bus. The programmable automation controller is internally configured with a logic operation module, which specifically includes: a signal acquisition and synchronization module 100, an initialization and modeling module 200, an active disturbance generation module 300, a double-channel feature extraction and verification module 400, and an adaptive decision and closed-loop control module 500.

[0025] The signal acquisition and synchronization module 100 is configured to receive analog signals or digital signals from the sensor array. The signal acquisition and synchronization module 100 performs analog-to-digital conversion and low-pass filtering on the data transmitted by the hook load sensor, the top drive torque sensor and the standpipe pressure sensor, and adds time stamps to the data of each channel based on a unified system clock to output a time-aligned real-time drilling parameter sequence.

[0026] The initialization and modeling module 200 is configured to establish a physical reference model of the system at the start of the drilling operation or at the change of the drilling assembly. The initialization and modeling module 200 calculates the propagation speed and transmission delay time of the fluid pressure wave in the wellbore according to the density and bulk modulus parameters of the drilling fluid. The initialization and modeling module 200 also calculates the theoretical blade passing frequency when the drill bit cuts the rock according to the number of blades of the drill bit and the current mechanical rotation speed.

[0027] The active disturbance generation module 300 is configured to generate a non-steady-state detection excitation signal. The active disturbance generation module 300 superimposes a low-frequency disturbance signal in the form of a sine wave on a preset reference WOB value to generate a target WOB instruction. The active disturbance generation module 300 sends the target WOB instruction to the drawworks system or the automatic pipe running device to drive the actuator to apply a dynamically changing axial load at the bottom of the well.

[0028] The dual-channel feature extraction and verification module 400 is configured to verify the validity of the physical source of the torque change monitored on the ground. The dual-channel feature extraction and verification module 400 receives the ground torque data and the standpipe pressure data processed by the signal acquisition and synchronization module 100. The dual-channel feature extraction and verification module 400 internally runs the fluid-structure coupling time-domain verification logic and the mechanical frequency spectrum frequency-domain verification logic in parallel to calculate the correlation coefficient based on the fluid pressure wave lag response and the sensitivity index based on the mechanical frequency spectrum characteristic energy, respectively. The dual-channel feature extraction and verification module 400 outputs a confidence value representing the validity of the current torque response based on the calculation results.

[0029] The adaptive decision and closed-loop control module 500 is configured to execute the final parameter optimization control strategy. The adaptive decision and closed-loop control module 500 calculates the differential response gain of the ground torque with respect to the WOB, and weights and corrects the differential response gain using the confidence value output by the dual-channel feature extraction and verification module 400. The adaptive decision and closed-loop control module 500 determines the current drilling state according to the corrected effective gain value, including the efficient cutting state, the friction blocking state and the transition saturation state. The adaptive decision and closed-loop control module 500 adjusts the reference WOB at the next time according to the determination result, and feeds back the updated control instruction to the active disturbance generation module 300, thereby forming a closed-loop control loop.

[0030] Referring to the drawings Figure 2 ,Figure 2 is a flowchart of a deep well geological exploration excavation parameter control method according to an embodiment of the present application. The present application provides a deep well geological exploration excavation parameter control method, which is executed in reliance on the aforementioned deep well geological exploration excavation parameter control system, and includes the following steps: S100, the signal acquisition and synchronization module 100 continuously acquires drilling parameters, and performs low-pass filtering and timestamp alignment processing on the hook load, top drive torque and standpipe pressure data; S200, the initialization and modeling module 200 calculates fluid transmission delay parameters and mechanical spectral characteristic parameters at the start of the operation, and constructs a physical reference model; S300, the active disturbance generation module 300 generates target drilling pressure instructions with micro-amplitude sinusoidal disturbances based on the reference drilling pressure, and drives the drawworks system to apply dynamic axial load to the formation; S400, the dual-channel feature extraction and verification module 400 synchronously analyzes the space-time correlation features and frequency domain energy features of the ground torque and standpipe pressure in the disturbance period, calculates the confidence of the torque response effectiveness, and eliminates invalid torque components caused by the lateral friction of the drill string; S500, the adaptive decision and closed-loop control module 500 determines the current drilling state and dynamically adjusts the reference drilling pressure at the next time according to the effective torque gain corrected by the confidence, and searches for the optimal operating point.

[0031] The signal acquisition and synchronization module 100 serves as the input front end of the system, and is responsible for providing time-aligned and denoised drilling state data. The running process of the signal acquisition and synchronization module 100 includes steps S101 to S103.

[0032] In step S101, the signal acquisition and synchronization module 100 accesses the multi-channel analog signals of the sensor array. The signal acquisition and synchronization module 100 is connected in parallel to the sensor array of the drilling platform through an industrial field bus or an analog input interface. The sensor array specifically includes a hook load sensor, a top drive torque sensor, a standpipe pressure sensor, and a rotational speed sensor.

[0033] The hook load sensor is usually installed on the dead line fixer or the traveling block, and is used to output a voltage signal or a current signal representing the weight on the drill string. For the specific hardware selection and installation method of the hook load sensor, those skilled in the art can make a routine selection according to the drilling tonnage and the site conditions, which belongs to the known technology in the art and will not be described here. The signal acquisition and synchronization module 100 receives the weight signal and obtains the hook load reference value when the drill string is lifted off the bottom and maintained in rotation. The real-time ground drilling pressure data is calculated by subtracting the current real-time weight signal from the reference value . This method based on the measured reference difference can automatically offset the influence of the drill string dead weight and buoyancy on the drilling pressure measurement.

[0034] The top drive torque sensor acquires raw data representing the ground mechanical torque. The raw data can be obtained directly by a strain gauge torque meter installed on the top drive output shaft, or by collecting the output current and voltage parameters of the top drive frequency converter, and calculating by using the pre-calibrated motor torque coefficient. The signal acquisition and synchronization module 100 converts the collected signals into ground torque data in standard units (such as kN·m) .

[0035] The standpipe pressure sensor is arranged at the standpipe manifold or high-pressure manifold of the drilling fluid circulation system, for detecting the fluid pressure when the mud pump pumps into the wellhead. The signal acquisition and synchronization module 100 collects the pressure signal as standpipe pressure data . The rotational speed sensor is used to collect the real-time rotational speed of the top drive or rotary table .

[0036] The signal acquisition and synchronization module 100 is configured with a high-precision analog-to-digital conversion unit, and the sampling frequency of the analog-to-digital conversion unit should meet the Nyquist sampling theorem to cover the high-frequency characteristics of the drill bit cutting and its harmonics, and is preferably set to a range of 50Hz to 100Hz to ensure effective analysis of the up to 20Hz blade passing frequency.

[0037] In step S102, the signal acquisition and synchronization module 100 performs hierarchical filtering on the raw data. Since the subsequent modules have different requirements for the frequency band of the signal, the signal acquisition and synchronization module 100 constructs two independent data processing paths: a control signal path and an analysis signal path.

[0038] For the control signal path, the signal acquisition and synchronization module 100 applies a strong low-pass filter with a low cutoff frequency (e.g. 1Hz to 2Hz) to the hook load data , ground torque data and standpipe pressure data . The smoothed data output by this path is used for gradient calculation and reference WOB update in S500 to eliminate the influence of high-frequency noise on control stability.

[0039] For the analysis signal path, the signal acquisition and synchronization module 100 retains the original signal with a wide frequency band (e.g. DC to 25Hz) or only applies an anti-aliasing filter. The high-frequency reserved data output by this path is specially provided to the dual-channel feature extraction and verification module 400 to ensure that the spectral energy at the blade passing frequency can be accurately extracted in S402.

[0040] In step S103, the signal acquisition and synchronization module 100 establishes a global clock synchronization and timestamp alignment mechanism. Since different types of sensors can access the control system via different transmission links (such as 4-20 mA analog transmission, Profibus bus transmission, or Ethernet transmission), there is a slight transmission time lag difference in the time when each channel data arrives at the processing unit. The signal acquisition and synchronization module 100 based on the system unified high-precision clock of the programmable automatic controller, marks a unified timestamp for each discrete data point collected .

[0041] The signal acquisition and synchronization module 100 reorganizes the filtered large hook load data, ground torque data, standpipe pressure data, and rotating speed data according to the same timestamp index, and constructs a time-aligned real-time drilling parameter state vector . Through this step, the signal acquisition and synchronization module 100 eliminates the time sequence deviation caused by the difference in data transmission link, and ensures that each physical quantity called by the subsequent processing module in the time dimension strictly corresponds when performing multi-physical field coupling analysis. The signal acquisition and synchronization module 100 stores the state vector into the system's circular buffer in real time, for the initialization and modeling module 200 and the double-channel feature extraction and verification module 400 to call.

[0042] The initialization and modeling module 200 is used to establish a physical reference model required for system operation, and provides a space-time alignment basis and a frequency domain feature reference for subsequent multi-source signal verification. The running process of the initialization and modeling module 200 mainly includes steps S201 and S202.

[0043] In step S201, the initialization and modeling module 200 calculates the fluid transmission lag parameter. In the process of deep well drilling, the pressure pulse signal generated by the downhole motor is transmitted to the surface standpipe pressure sensor through the drilling fluid column in the drill string, and there is a physical delay determined by the medium property and path length in this transmission process. The initialization and modeling module 200 obtains the physical property parameters of the drilling fluid in the current wellbore through the system parameter configuration interface or the real-time logging data interface, including the mixed fluid density and the equivalent bulk modulus . Among them, the equivalent bulk modulus preferably uses a corrected modulus value corrected by the average temperature and average pressure of the wellbore to compensate for the influence of the deep well environment on the fluid compressibility.

[0044] The initialization and modeling module 200 calculates the propagation speed of the pressure wave in the wellbore filled with drilling fluid according to the fluid acoustics propagation principle . The calculation process uses the following formula: ; Among them, denotes the pressure wave propagation speed in meters per second (m / s); denotes the equivalent bulk modulus of the drilling fluid in Pascals (Pa); denotes the drilling fluid density in kilograms per cubic meter (kg / m 3 ).

[0045] Subsequently, the initialization and modeling module 200 reads the current measured depth data . The measured depth data is defined as the measured depth (MD) along the wellbore axis trajectory, rather than the vertical depth, to accurately represent the actual transmission path length of the fluid in the wellbore. The initialization and modeling module 200 calculates the fluid transmission time lag required for the fluid pressure wave to transmit from the bottom of the well to the ground, calculated as follows: ; wherein, denotes the lag time in seconds (s). The initialization and modeling module 200 stores the calculated in the system memory. In this embodiment, the system considers the propagation of mechanical waves in the metal drill string as instantaneous transmission, and uses the above calculated fluid transmission time lag to shift the time sequence of the standpipe pressure data backward, aligning it with the torque signal transmitted through the drill string on the time axis, thereby establishing the spatiotemporal correspondence between the two.

[0046] In step S202, the initialization and modeling module 200 calculates the mechanical spectral feature parameters. When the drill bit breaks rocks at the bottom of the well, due to its discrete cutter distribution structure, it will generate a periodic excitation signal strictly related to the rotational speed and the number of blades, i.e., the blade passing frequency.

[0047] The initialization and modeling module 200 obtains the current drilling bit structure parameters, specifically the number of main blades of the drilling bit . This parameter is entered through the configuration interface before the drilling assembly is lowered into the well. At the same time, the initialization and modeling module 200 reads the real-time rotational speed of the top drive or rotary table provided by the signal acquisition and synchronization module 100 in real time .

[0048] Based on the kinematic relationship, the initialization and modeling module 200 calculates the theoretical reference feature frequency , calculated as follows: ; wherein, denotes the blade passing frequency in Hertz (Hz); denotes the drill string rotational speed in revolutions per minute (rpm); is an integer.

[0049] The initialization and modeling module 200 will calculate the dynamic update to the dual-channel feature extraction and verification module 400. The feature frequency constitutes the frequency domain benchmark for identifying the effective cutting state of the drill bit. When the drill bit is in the effective rock breaking state, there should be a clear energy peak at and its harmonics in the power spectral density of the surface torque signal; otherwise, if the drill string is subject to lateral friction or buckling, the torque signal will exhibit low-frequency random fluctuations, lacking the energy concentration of the characteristic frequency component. The initialization and modeling module 200 quantifies this physical feature to provide a speed-dependent dynamic adjustment criterion for subsequent frequency domain verification.

[0050] The active disturbance generation module 300 is used to superimpose a controlled dynamic excitation signal during the steady-state drilling process, providing the necessary input source for subsequent system identification. The operation process of the active disturbance generation module 300 mainly includes steps S301 and S302.

[0051] In step S301, the active disturbance generation module 300 calculates and generates a sinusoidal micro-disturbance target instruction. The active disturbance generation module 300 reads the current reference WOB set value . In the initial running period of system startup, the active disturbance generation module 300 reads the initial WOB value preset by the operator as ; in the subsequent running period, the active disturbance generation module 300 reads the updated feedback from the adaptive decision and closed-loop control module 500. And superimpose a periodic low-frequency sinusoidal wave signal on the basis of the reference value, generate the target WOB instruction that changes with time. The calculation formula of the target WOB instruction is as follows: ; Where, represents the target WOB instruction value at time , unit: kilo Newton (kN); represents the current reference WOB set by the adaptive decision and closed-loop control module 500, unit: kilo Newton (kN); represents the disturbance amplitude, unit: kilo Newton (kN); represents the disturbance frequency, unit: Hertz (Hz); represents the system time.

[0052] When the active disturbance generation module 300 determines the disturbance amplitude , it calls the pre-stored drill string mechanical parameters in the initialization and modeling module 200 to obtain the sinusoidal buckling critical load value of the drill string at the current depth. The value of is limited to a preset safety ratio of the critical load (e.g., 5% to 10%) to ensure that the applied axial dynamic force can overcome the static friction between the drill string and the well wall, while avoiding inducing plastic deformation of the drill string.

[0053] The active disturbance generation module 300 determines the disturbance frequency. At the same time, the frequency data of the mud pump stroke counter is read in real time through the signal acquisition interface. The active disturbance generation module 300 calculates the mud pump stroke frequency and its low-order harmonic frequencies, and generates the disturbance frequency. The frequency is set in a low-frequency band (preferably 0.05Hz to 0.2Hz) outside the aforementioned frequency range. This frequency selection logic ensures that the artificially applied drilling pressure excitation is completely separated from the fluid pulsations generated by the mud pump in the frequency domain, preventing signal aliasing from interfering with subsequent feature extraction.

[0054] In step S302, the active disturbance generation module 300 executes control command transmission and drive. The active disturbance generation module 300 transmits the calculated target drilling pressure command via an industrial fieldbus (such as Profibus-DP or EtherNet / IP). The automatic drilling actuator sends data to the drilling rig in real time.

[0055] The automatic drilling actuator is specifically embodied in a winch frequency converter drive or a hydraulic disc brake controller. This actuator responds to... The command adjusts the motor's output torque or brake pressure to control the axial release speed of the drill string at the wellhead, thereby establishing an amplitude of [value missing] at the bottom of the well. , frequency is The axial load fluctuation. This actively applied load fluctuation serves as the system's detection input signal, transmitted via the drill string to the contact surface between the drill bit and the formation, and elicits corresponding torque and pressure responses, which are then used by subsequent modules for correlation analysis.

[0056] The dual-channel feature extraction and verification module 400 is used to process multi-source sensor signals in parallel during the application of dynamic excitation by the active disturbance generation module 300, and to calculate and output a confidence index characterizing the validity of the torque response source. The operation flow of the dual-channel feature extraction and verification module 400 mainly includes steps S401 to S403.

[0057] In step S401, the dual-channel feature extraction and verification module 400 executes fluid-structure interaction time-domain verification logic. This logic utilizes the linear correlation between the screw motor output torque and the cross-motor pressure difference to identify bottom hole cutting behavior by detecting the synchronous fluctuation characteristics of surface torque and riser pressure.

[0058] The dual-channel feature extraction and verification module 400 reads the fluid transport time delay calculated by the initialization and modeling module 200. . Since the fluid pressure wave propagates slower than the drill string mechanical wave, the riser pressure signal received at the surface lags behind the surface torque signal in time. The dual-channel feature extraction and verification module 400 performs a time delay alignment process on the surface torque data input from the signal acquisition and synchronization module 100 to generate an aligned torque signal corresponding to the riser pressure in space and time . Specifically, the dual-channel feature extraction and verification module 400 reads the torque sample point that is earlier than the current time second in the data buffer, and the alignment calculation formula is as follows: ; wherein, denotes the surface torque data after time delay compensation; denotes the original surface torque data; denotes the current system time; denotes the fluid wave propagation lag time.

[0059] The dual-channel feature extraction and verification module 400 establishes a sliding time window with a length of . In the current window, the Pearson cross-correlation coefficient between the aligned torque and the real-time riser pressure is calculated, and the calculation formula is as follows: ; wherein, denotes the fluid-structure coupling correlation coefficient, and the value range is [-1, 1]; denotes the th torque sample point data after time delay compensation in the sliding window; denotes the th original riser pressure sample point data in the sliding window; and are the arithmetic mean values of the aligned torque data sequence and the riser pressure data sequence in the current window, respectively. The formula calculates the synchronization of the lagged mechanical torque response and the fluid pressure response in the waveform change trend through the standardized covariance calculation.

[0060] In step S402, the dual-channel feature extraction and verification module 400 performs a mechanical spectrum frequency domain verification logic. The dual-channel feature extraction and verification module 400 performs a fast Fourier transform on the surface torque signal in the sliding window to convert the time domain signal into a frequency domain power spectral density function .

[0061] The dual-channel feature extraction and verification module 400 reads the reference feature frequency provided by the initialization and modeling module 200 The dual-channel feature extraction and verification module 400 extracts the feature band energy with the center frequency and the bandwidth , and calculates the integral energy in the frequency band. The calculation formula is as follows: ; wherein represents the feature band energy; represents the power spectrum density of the torque signal; is a preset bandwidth parameter.

[0062] The dual-channel feature extraction and verification module 400 further calculates the differential sensitivity of the feature energy with respect to the change in the WOB . In order to avoid numerical instability, the dual-channel feature extraction and verification module 400 only performs the following differential calculation when the WOB change is greater than a preset noise threshold value, otherwise the sensitivity value of the last time is maintained: ; wherein represents the mechanical energy sensitivity; is the corresponding surface WOB data; is the differential time step.

[0063] In step S403, the dual-channel feature extraction and verification module 400 performs effectiveness confidence fusion calculation. The dual-channel feature extraction and verification module 400 assigns a weight factor to the two channels according to the current drilling assembly configuration. If the drilling assembly contains a downhole power drilling tool, a non-zero weight is given to the fluid-structure coupling channel; if it is pure rotary drilling, the weight of the fluid-structure coupling channel is set to zero.

[0064] The dual-channel feature extraction and verification module 400 maps the mechanical energy sensitivity to the interval [0, 1] using a nonlinear mapping function to obtain the normalized mechanical index . In a preferred embodiment, the mapping function uses a piecewise normalization function with a dead zone and saturation characteristic, and the specific calculation formula is as follows: ; wherein is a preset saturation threshold parameter, which is set according to the average torque response level of the field drilling tool, and is used to linearly scale the effective sensitivity value to the interval [0, 1]. The mapping process truncates the negative to 0, and the exceeding the preset saturation threshold is truncated.is mapped to 1 to eliminate the dimension effect. The dual-channel feature extraction and verification module 400 calculates the final torque validity confidence , which is calculated as follows: ; wherein, denotes the final confidence; and are weighting factors, and satisfy . In the formula, the normalized can ensure that the mechanical channel index and the fluid channel index are in the same order of magnitude, ensuring the mathematical validity of the weighted calculation. The dual-channel feature extraction and verification module 400 outputs the calculated to the adaptive decision and closed-loop control module 500, which is used to quantify the probability of the contribution of the real cutting action in the current ground torque increment.

[0065] The adaptive decision and closed-loop control module 500 is responsible for calculating and outputting the weight-on-bit control instruction according to the input physical feature recognition result. The adaptive decision and closed-loop control module 500 dynamically adjusts the reference weight-on-bit by real-time evaluation of the effective cutting gain, realizing the automatic optimization of drilling parameters. The running process of the adaptive decision and closed-loop control module 500 mainly includes steps S501 to S503.

[0066] In step S501, the adaptive decision and closed-loop control module 500 calculates the original differential response gain of the ground torque with respect to the weight-on-bit. Since the active disturbance generation module 300 applies a sinusoidal weight-on-bit disturbance with a frequency of to the system input end, the ground torque signal contains the corresponding response component.

[0067] The adaptive decision and closed-loop control module 500 selects a data window with a time length of to obtain the ground weight-on-bit sequence and the ground torque sequence within the window. The window length is set as an integer multiple of the disturbance period to ensure the periodic integrity of the sampled data. The adaptive decision and closed-loop control module 500 uses the least squares method to perform linear regression analysis on the data to calculate the original differential response gain . The calculation formula is as follows: ; wherein, denotes the original differential response gain, and the physical meaning is the ratio of the ground torque change amount to the ground weight-on-bit change amount; is the total number of sampling points in the window; is the sampling point index.

[0068] In step S502, the adaptive decision-making and closed-loop control module 500 performs a weighted correction of the effective cutting gain. This step uses signal verification results to eliminate invalid torque increments caused by friction or buckling.

[0069] The adaptive decision-making and closed-loop control module 500 reads the torque validity confidence level output by the dual-channel feature extraction and verification module 400. The adaptive decision and closed-loop control module 500 utilizes this confidence level to adjust the gain of the original differential response. Make corrections and calculate the effective cutting gain. The formula is as follows: ; This calculation determines the confidence level when the system is in an ineffective cutting state (such as drill string buckling or severe lateral friction). The decrease in the value of reduces the calculated effective cutting gain. The torque is reduced or zeroed accordingly to prevent the control algorithm from misinterpreting non-cutting torque increases as pressure increases.

[0070] In step S503, the adaptive decision-making and closed-loop control module 500 performs state determination and baseline parameter update. The adaptive decision-making and closed-loop control module 500 uses a gradient ascent algorithm to search for the optimal drilling pressure operating point.

[0071] The adaptive decision-making and closed-loop control module 500 will calculate the effective cutting gain. Compared with the preset optimization step size coefficient and system damping coefficient Combined, calculate the baseline drilling pressure correction for the next time step. The control law formula is as follows: ; in, This indicates the baseline drilling pressure setting value for the next control cycle; This represents the baseline drilling pressure for the current cycle; The positive optimization gain coefficient, whose dimensions correspond to force / length, is used to set the rate at which drilling pressure increases with the improvement of cutting efficiency; is the effective cutting gain currently calculated; is the overload protection coefficient, used to set the retraction response strength when the torque exceeds the threshold; This refers to the preset optimal operating torque or safety alarm threshold for the top drive system. (The formula contains...) The logic ensures that the decompression operation is triggered only when the actual torque exceeds the target threshold.

[0072] The adaptive decision-making and closed-loop control module 500 calculates... Saturation clipping is performed. The adaptive decision and closed loop control module 500 judges whether it is in a preset safe range of WOB . If the calculated value is out of the range, the reference WOB is forced to clamp to the boundary value of the range to protect the drilling equipment safety. Then, the adaptive decision and closed loop control module 500 sends the updated reference WOB to the active disturbance generation module 300 to complete the closed loop control.

[0073] While the embodiments of the application have been illustrated and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the spirit and scope of the application, which is defined by the appended claims and their equivalents.​

Claims

1. A deep well geological exploration and tunneling parameter control system, characterized in that, include: The signal acquisition and synchronization module is used to acquire the hook load, top drive torque and riser pressure, perform timestamp alignment, and output a real-time drilling parameter sequence. The initialization and modeling module is used to calculate the fluid transport time delay based on drilling fluid properties and wellbore depth, and to calculate the blade sweep frequency based on real-time rotation speed and drill bit structure. The active disturbance generation module is used to generate a target drilling pressure command superimposed with a sinusoidal waveform disturbance based on the reference drilling pressure, and drive the actuator to apply a dynamic axial load at the bottom of the well. The dual-channel feature extraction and verification module is used to receive the real-time drilling parameter sequence, extract the ground torque data of historical moments based on the fluid transmission time delay to align with the riser pressure data of the current moment, and calculate the confidence value characterizing the effectiveness of the torque response in combination with the blade sweep frequency. The adaptive decision-making and closed-loop control module is used to correct the differential response gain of the ground torque with respect to the drilling pressure using the confidence value, update the reference drilling pressure based on the corrected effective gain value, and provide it to the active disturbance generation module to generate a new target drilling pressure command.

2. The deep well geological exploration and tunneling parameter control system according to claim 1, characterized in that, The signal acquisition and synchronization module includes: The analog or digital signals of the hook load, top drive torque and riser pressure are acquired through a sensor array. Low-pass filtering is applied to the hook load, top drive torque and riser pressure respectively, and a unified timestamp is marked for the data of each channel according to the unified clock of the system. The filtered data is reassembled according to the same timestamp index to construct the real-time drilling parameter sequence containing time-aligned surface drill pressure, surface torque, and riser pressure. The real-time drilling parameter sequence is stored in the system's circular buffer for subsequent modules to call.

3. The deep well geological exploration and tunneling parameter control system according to claim 1, characterized in that, The initialization and modeling module performs the following: Obtain the mixed fluid density and equivalent bulk modulus of the drilling fluid, and calculate the propagation speed of pressure waves in the wellbore; Based on the measured depth along the wellbore axis trajectory and the propagation velocity, the fluid transmission time delay required for the fluid pressure wave to travel from the bottom of the well to the surface is calculated; Obtain the number of main blades of the drill bit and the real-time rotational speed of the top drive or rotary table; The theoretical reference characteristic frequency of the drill bit cutting rock is calculated based on kinematic relationships and used as the sweep frequency of the cutter blade.

4. The deep well geological exploration and tunneling parameter control system according to claim 1, characterized in that, The active perturbation generation module includes: Read the currently set reference drilling pressure; Based on the preset disturbance amplitude and disturbance frequency, calculate the sinusoidal signal that changes with time; The sinusoidal signal is superimposed on the reference drill pressure to generate the target drill pressure command; The target drilling pressure command is sent to the automatic drilling actuator to control the axial release speed of the drill string at the wellhead to establish axial load fluctuation. The disturbance frequency is set in the low-frequency range outside the range of the mud pump stroke frequency and the harmonic frequency range of the mud pump stroke frequency.

5. The deep well geological exploration and tunneling parameter control system according to claim 1, characterized in that, The adaptive decision-making and closed-loop control module includes: Select an integer multiple of the disturbance period as the data window, and obtain the ground drilling pressure sequence and ground torque sequence within the data window; The least squares method was used to perform linear regression analysis on the ground drilling pressure sequence and the ground torque sequence to calculate the original differential response gain. The original differential response gain is weighted and corrected using the confidence level value to obtain the effective cutting gain that characterizes the actual cutting action. The effective cutting gain is combined with a preset optimization step size coefficient to calculate the baseline drill pressure correction and update the baseline drill pressure.

6. The deep well geological exploration and tunneling parameter control system according to claim 1, characterized in that, The dual-channel feature extraction and verification module performs the following: Parallel operation of fluid-structure interaction time-domain verification logic and mechanical spectrum frequency-domain verification logic; In the fluid-structure interaction time-domain verification logic, the spatiotemporal correspondence between ground torque and riser pressure is established using the fluid transport time delay, and the fluid-structure interaction correlation coefficient is calculated. In the mechanical spectrum frequency domain verification logic, the frequency domain energy features are extracted using the blade sweep frequency to calculate the mechanical energy sensitivity. Based on the drilling tool assembly configuration, weighting factors are assigned to the fluid-structure interaction correlation coefficient and the mechanical energy sensitivity, and the confidence score is output.

7. A deep well geological exploration and tunneling parameter control system according to claim 6, characterized in that, The fluid-structure interaction time-domain verification logic includes: Read the fluid transmission time delay and obtain torque sampling points that are earlier than the current time by the number of seconds of the fluid transmission time delay; The torque sampling points are recombined to perform delayed alignment processing, generating an aligned torque signal that is synchronized with the riser pressure data in time. Establish a sliding time window and calculate the covariance between the aligned torque signal and the real-time riser pressure data within the sliding time window; The Pearson cross-correlation coefficient is calculated based on the covariance and used as the fluid-structure interaction correlation coefficient to quantify the synchronicity of the hysteretic mechanical torque response and the fluid pressure response in terms of waveform change trends.

8. A deep well geological exploration and tunneling parameter control system according to claim 6, characterized in that, The mechanical spectrum frequency domain verification logic includes: Perform a fast Fourier transform on the ground torque data within the sliding window to convert it into a frequency domain power spectral density function; A frequency band with a preset bandwidth is extracted using the blade sweep frequency as the center frequency, and the integrated energy within the frequency band is calculated. The mechanical energy sensitivity is obtained by calculating the differential ratio of the integrated energy with respect to the change in ground drilling pressure. The mechanical energy sensitivity is mapped to a normalized interval using a piecewise normalization function with dead zone and saturation characteristics for subsequent weighted fusion calculations.

9. A deep well geological exploration and tunneling parameter control system according to claim 5, characterized in that, The adaptive decision-making and closed-loop control module further includes the following when updating the baseline drilling pressure: Calculate the difference between the current ground torque and the preset target threshold; If the difference exceeds zero, the backoff response amount is calculated based on the difference and the preset overload protection coefficient. The reference drilling pressure setting value for the next moment is obtained by subtracting the back-off response amount from the reference drilling pressure correction amount. Determine whether the reference drilling pressure setting value is within the preset safe range. If it exceeds the limit, force clamping to the boundary value of the safe range.

10. A method for controlling deep well geological exploration tunneling parameters, applied to the deep well geological exploration tunneling parameter control system described in any one of claims 1-9, characterized in that, Includes the following steps: Collect the hook load, top drive torque and riser pressure, perform timestamp alignment, and output a real-time drilling parameter sequence; The fluid transport time delay is calculated based on the drilling fluid properties and wellbore depth, and the blade sweep frequency is calculated based on the real-time rotation speed and drill bit structure. Based on the benchmark drilling pressure, a target drilling pressure command superimposed with a sinusoidal waveform disturbance is generated, which drives the actuator to apply a dynamic axial load at the bottom of the well. The system receives the real-time drilling parameter sequence, extracts the ground torque data from historical moments based on the fluid transport time delay to align with the riser pressure data at the current moment, and calculates a confidence level value characterizing the effectiveness of the torque response by combining the blade sweep frequency. The confidence level value is used to correct the differential response gain of the surface torque with respect to the drilling pressure. The reference drilling pressure is updated based on the corrected effective gain value and is then used to generate the new target drilling pressure command.

Images