CN-121996881-A - Method, system, medium and equipment for optimizing drilling parameters of coal-rock gas horizontal well

CN121996881ACN 121996881 ACN121996881 ACN 121996881ACN-121996881-A

Abstract

The invention relates to the technical field of deep coal gas horizontal well drilling parameter optimization, in particular to a coal gas horizontal well drilling parameter optimization method, a system, a medium and equipment, wherein the method comprises the steps of acquiring logging while drilling parameters, constructing a rock breaking energy utilization rate calculation model by utilizing the logging while drilling parameters to form a sample data set, carrying out multi-stage pretreatment on the sample data set, constructing a long-term and short-term memory neural network model, the trained long-short-term memory neural network model is embedded into a logging system, the maximum mechanical drilling speed and the minimum real breaking energy utilization rate are used as optimization targets based on real-time prediction of real breaking energy utilization rate, and the drilling pressure, the rotating speed and the torque parameters are dynamically adjusted.

Inventors

SONG WEIQIANG
LI YUANHANG
YU HAIYANG
ZHANG WEI
SHAN QINGLIN
SUN XUE
LI SHUPENG
WANG XU

Assignees

山东科技大学

Dates

Publication Date: 20260508
Application Date: 20260408

Claims (10)

1. The method for optimizing the drilling parameters of the horizontal well of the coal and rock gas is characterized by comprising the following steps of: Acquiring logging while drilling parameters, wherein the logging while drilling parameters comprise drilling pressure, rotating speed, torque and mechanical drilling speed, and acquiring effective rock breaking torque of a drill bit based on torque separation, wherein the effective rock breaking torque of the drill bit is the difference between real-time torque and idle torque of normal drilling; Constructing a rock breaking energy utilization rate calculation model by using logging while drilling parameters and the effective rock breaking torque of the drill bit, determining characteristic parameters of the rock breaking energy utilization rate, and forming a sample data set; performing a multi-stage preprocessing on the sample dataset, the multi-stage preprocessing including based on Abnormal value detection of a criterion, input-output characteristic double smoothing processing and sliding window time sequence reconstruction; Constructing a long-term and short-term memory neural network model, capturing long time sequence dependency relationship among drilling parameters through a sliding window, and establishing nonlinear dynamic mapping between the drilling parameters and the real rock breaking energy utilization rate; The trained long-term and short-term memory neural network model is embedded into a logging system, real rock breaking energy utilization rate based on real-time prediction is combined with a deep coal gas reservoir working condition threshold, mechanical drilling speed maximization and real rock breaking energy utilization rate minimization are used as optimization targets, and drilling weight, rotating speed and torque parameters are dynamically adjusted.
2. The method for optimizing drilling parameters of a horizontal well of coal and rock gas according to claim 1, wherein the obtaining the logging while drilling parameters comprises separating rotational torque of a drill string from effective breaking torque of a drill bit based on torque, and obtaining real-time torque recorded by a logging system under normal contact downhole rotational drilling state of the drill bit And drill string rotation torque recorded by the logging system under the condition that the drill bit is lifted off the bottom of the well and the original rotation speed idle state is maintained Based on real-time torque Torque to drill string rotation And (3) carrying out differential operation on the drill bit to obtain the effective rock breaking torque of the drill bit 。
3. The method for optimizing drilling parameters of a horizontal well of coal and rock gas according to claim 2, wherein the constructing a calculation model of the rock breaking energy utilization rate by using the logging while drilling parameters and the effective rock breaking torque of the drill bit comprises the steps of Substituting the real rotation speed of the drill bit into the corrected mechanical specific energy formula, and calculating to obtain the real rock breaking energy utilization rate And will As a characteristic parameter of the rock breaking energy utilization rate, wherein the real rotating speed of the drill bit is the sum of the top drive rotating speed and the output rotating speed of the underground power drilling tool, and the calculation formula of the real rock breaking energy utilization rate is as follows: ; Wherein, the For the real rock breaking energy utilization rate, In order to achieve a weight-on-bit, Is the cross-sectional area of the borehole, Is the rotation speed of the top drive, The output rotational speed for the downhole motor, For real-time torque recorded by the logging system during normal drilling, Drill string rotational torque recorded by the logging system when the drill bit is lifted off the bottom hole to idle, Is the rate of penetration.
4. The method for optimizing drilling parameters of a horizontal well of coal and rock gas according to claim 1, wherein the multi-stage preprocessing is performed on a sample data set, the method comprises the steps of performing data cleaning based on physical feasible region constraint and statistical distribution characteristics of drilling parameters, performing input-output characteristic double smoothing processing on the cleaned data set, suppressing high-frequency random noise by adopting exponential moving average filtering for the input characteristic, suppressing local fluctuation by adopting a sliding window averaging method for the output characteristic, converting the smoothed data set into a sliding window format, taking drilling parameter vectors at a plurality of continuous historical moments as the input characteristic, and using a broken rock energy utilization rate at the next moment as an output label, and constructing a time sequence sample pair suitable for long-term and short-term memory network training, wherein the sliding window averaging expression of the output characteristic is as follows: ; Wherein, the The output characteristic value after the smoothing at the moment t, The input characteristic value after the smoothing processing at the moment i, To take the smaller value of the current time sequence number and the sliding window length.
5. The method for optimizing drilling parameters of a horizontal well of coal and rock according to claim 1, wherein the constructing a long-term and short-term memory neural network model comprises: A long-period memory network unit comprising a forgetting gate, an input gate and an output gate is constructed, a plurality of long-period memory network units are cascaded to form a long-period memory network layer, and forgetting of historical information and input of current information are controlled through a gating mechanism; Constructing a fully-connected output layer, introducing a time step attention mechanism between the long-period memory network layer and the output layer, and calculating attention weights of all time steps based on the correlation between the hidden state and the current task; adopting a batch gradient descent algorithm fused with L2 regularization to carry out iterative optimization on network parameters, and implementing early shutdown to save optimal model parameters; The weight calculation of the time-step attention mechanism introduces a physical a priori constraint: ; Wherein, the As the attention weight at time t, In order to be the hidden state at the instant t, For the state of the cell at time t, 、 And As a matrix of weights, the weight matrix, Is a physical prior function based on drilling working conditions and is used for quantifying the correlation between the rock breaking energy utilization rate at historical moment and the current prediction target, For the physical a priori influencing factors, In time steps.
6. A method for optimizing drilling parameters of a horizontal well of a coal and rock gas according to claim 3, wherein the threshold of the working condition of the deep coal and rock gas reservoir is combined, comprising: Obtaining the rock breaking energy utilization rate and the actually measured mechanical drilling rate predicted in real time by the long-short-term memory neural network model, and constructing a double-target working condition discrimination index; comparing the working condition discrimination index with a preset multi-stage working condition threshold interval, and identifying the grade to which the current drilling working condition belongs, wherein: when the rock breaking energy utilization rate is in the range of 80-120 MPa and the mechanical drilling speed is not lower than 36m/h, judging the rock breaking energy utilization rate as an optimal working condition; When the rock breaking energy utilization rate is in the range of 120-150 MPa and the mechanical drilling speed is in the range of 24-36 m/h, judging the rock breaking energy utilization rate as a suboptimal working condition; When the rock breaking energy utilization rate is greater than 150MPa or the mechanical drilling speed is lower than 24m/h, judging a risk working condition; When the rock breaking energy utilization rate is lower than 80MPa and the mechanical drilling speed is lower than 30m/h, judging that the working condition is low-efficiency; the working condition discrimination index is constructed based on cooperative normalization of rock breaking energy efficiency and drilling efficiency: ; Wherein, the Is the working condition discrimination index, As the weight coefficient of the light-emitting diode, In order to achieve the utilization rate of the rock breaking energy, And The lower limit and the upper limit of the rock breaking energy utilization rate threshold are respectively defined, In order to measure the rate of penetration of the machine, And The lower and upper thresholds of the rate of penetration are respectively.
7. The method for optimizing drilling parameters of a coal gas horizontal well according to claim 6, wherein the optimizing target is maximized at a mechanical drilling speed and minimized at a real breaking energy utilization rate, the method comprises constructing a multi-objective collaborative optimization function integrating a breaking energy efficiency inhibition item, a mechanical drilling speed lifting item and a parameter adjustment amplitude constraint item based on the identified working condition level, solving an optimal solution of the multi-objective collaborative optimization function, determining a drilling pressure adjustment amount, a rotation speed adjustment amount and a torque adjustment amount, and converging drilling parameters to the optimizing target, wherein the drilling pressure adjustment amount is 5-10kN, and the rotation speed adjustment amount is 5-10RPM.
8. A coal and rock gas horizontal well drilling parameter optimization system, implemented in the method of claim 1, comprising: The rock breaking energy efficiency calculation module is configured to acquire logging while drilling parameters, construct a rock breaking energy utilization rate calculation model by utilizing the logging while drilling parameters, determine characteristic parameters of the rock breaking energy utilization rate under different well depths and form a sample data set, wherein the logging while drilling parameters comprise weight on bit, rotating speed, torque and mechanical drilling rate; The time sequence prediction model construction module is configured to carry out multi-stage pretreatment on a sample data set, construct a long-term and short-term memory neural network model, capture long time sequence dependency relationship among drilling parameters in a sliding window mode, and establish nonlinear dynamic mapping between the drilling parameters and the real rock breaking energy utilization rate through parameter optimization and training; The real-time optimization control module is configured to embed the trained long-short-period memory neural network model into a logging system, dynamically adjust drilling pressure, rotating speed and torque parameters based on real-time prediction of real rock breaking energy utilization rate and combining deep coal gas reservoir working condition threshold values by taking maximum mechanical drilling speed and minimum real rock breaking energy utilization rate as optimization targets.
9. A computer readable storage medium having stored therein a plurality of instructions adapted to be loaded by a processor of a terminal device and to perform a method of optimizing a drilling parameter of a gas horizontal well as claimed in claim 1.
10. A terminal device comprising a processor for implementing instructions and a computer-readable storage medium for storing instructions, characterized in that the instructions are adapted to be loaded by the processor and to perform a method for optimizing drilling parameters of a gas horizontal well according to claim 1.

Description

Method, system, medium and equipment for optimizing drilling parameters of coal-rock gas horizontal well Technical Field The invention relates to the technical field of optimization of drilling parameters of deep coal-rock gas horizontal wells, in particular to a method, a system, a medium and equipment for optimizing the drilling parameters of the coal-rock gas horizontal wells. Background The deep coal-rock gas resource is used as an important unconventional natural gas resource, a reservoir is generally provided with geological features of deep burial, high ground stress, complex lithology and strong heterogeneity, reasonable configuration of drilling parameters such as weight on bit, rotating speed, torque and the like directly influences mechanical drilling speed, drill bit service life and well wall stability in the horizontal well drilling process, at present, a drilling parameter optimization method mainly comprises a manual adjustment method based on field engineer experience, a mathematical model method based on mechanical mechanism and an intelligent prediction method based on machine learning, wherein the manual adjustment method depends on individual experience and response lag, the mathematical model method is difficult to adapt to dynamic changes of complex stratum, while the machine learning method based on BP neural network can utilize historical data for modeling, but is limited by inherent gradient disappearance defect of a network structure and cannot effectively capture long time sequence dependency relationship among drilling parameters. The prior art scheme has the remarkable limitations in practical application, namely firstly, the traditional experience method and the mathematical model method lack of excavating capability on time sequence characteristics of logging while drilling data, cannot identify complex working conditions such as drill bit stick slip, torque abrupt change and the like in real time, cause serious delay of parameter adjustment to stratum change and are easy to induce underground accidents such as drill bit tooth collapse, borehole instability and the like, secondly, the traditional optimization method based on a shallow neural network cannot fully consider torque transmission loss between a drill string and the drill bit, the rock breaking energy utilization ratio is larger in calculation error, or an underground vibration sensor and well ground communication equipment are additionally required to be additionally arranged, so that the operation cost and the technical implementation difficulty are remarkably increased, thirdly, the traditional deep learning model mainly adopts a standard feedforward structure, long-period dynamic mapping between drilling parameters and rock breaking energy efficiency is difficult to be established, and engineering requirements of deep coal gas reservoirs on parameter real-time optimization and time sequence correlation analysis cannot be met, and a coal gas horizontal well drilling parameter optimization method, system, medium and equipment are needed at the present. Disclosure of Invention The invention provides a method, a system, a medium and equipment for optimizing drilling parameters of a coal-rock horizontal well, aiming at solving the problems of insufficient mining of time sequence dependency and insufficient real-time identification capability of complex working conditions in the existing method for optimizing drilling parameters of a deep coal-rock horizontal well. In a first aspect, the invention provides a method for optimizing drilling parameters of a horizontal well of coal and rock, which adopts the following technical scheme: a method for optimizing drilling parameters of a coal-rock gas horizontal well, comprising: Acquiring logging while drilling parameters, wherein the logging while drilling parameters comprise drilling pressure, rotating speed, torque and mechanical drilling speed, and acquiring effective rock breaking torque of a drill bit based on torque separation, wherein the effective rock breaking torque of the drill bit is the difference between real-time torque and idle torque of normal drilling; Constructing a rock breaking energy utilization rate calculation model by using logging while drilling parameters and the effective rock breaking torque of the drill bit, determining characteristic parameters of the rock breaking energy utilization rate, and forming a sample data set; performing a multi-stage preprocessing on the sample dataset, the multi-stage preprocessing including based on Abnormal value detection of a criterion, input-output characteristic double smoothing processing and sliding window time sequence reconstruction; Constructing a long-term and short-term memory neural network model, capturing long time sequence dependency relationship among drilling parameters through a sliding window, and establishing nonlinear dynamic mapping between the drilling parameters and the real rock breaking