Automatic bit feeding optimization control system

Abstract

The invention relates to the technical field of optimization control, in particular to an automatic bit feeding optimization control system which comprises a bit pressure frequency extraction module, an amplification trend recognition module, a linkage signal monitoring module, a slippage reaction judgment module and a bit feeding speed adjustment module. According to the method, a frequency set is constructed through extreme value extraction, time span and amplitude difference are combined, a bit pressure fluctuation sequence is formed, resolution of dynamic characteristics is enhanced, trend sections are screened through frequency change cumulants, quantitative recognition of fluctuation trends is achieved, well bottom tracks and torque fluctuation signals are introduced, and a period direction consistency comparison mechanism is established. The signal coupling judgment precision is improved, slip judgment logic is constructed according to the amplitude growth rate and the direction consistency, risk grade classification is completed through a strain energy density index, and multi-dimensional signal linkage recognition, trend quantitative grading and energy index evaluation are achieved; and the risk identification accuracy, the response sensitivity and the control instruction adaptability under the complex drilling working condition are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of optimization control, in particular to an automatic drill feeding optimization control system. Background Art

[0002] The field of optimization control technology includes automated decision-making control and parameter adjustment methods in industrial production processes. Its core content is to achieve the optimization of the control objective function through real-time detection and feedback of system state variables, based on certain control strategies, and on the premise of meeting system constraints. It is mainly used in process industries, manufacturing systems, energy equipment and other scenarios. It covers multivariable control theory, constraint modeling, optimal solution solving mechanism and its integrated implementation in control systems. Through the establishment of parameter models and the setting of control laws, it can achieve stable control and optimal scheduling of dynamic systems.

[0003] The automatic drill feed optimization control system is a system that collects data and dynamically adjusts drilling parameters such as drilling pressure, rotational speed, pump pressure, and mechanical penetration rate during oil or gas drilling. It aims to achieve precise control of drill string load and bottomhole operation status by selecting a reasonable combination of drilling parameters based on formation feedback signals during the drilling process through optimized control logic based on working condition classification. Using working condition identification rules, drilling pressure setting calculation formulas, and speed and pump pressure setting strategies, combined with real-time data from the drilling phase, parameter adjustment and drilling control closed-loop execution are completed by adjusting the actuator action through control commands.

[0004] In the existing drilling control process, parameter setting and working condition classification are the core mechanisms, and the drilling status is controlled by preset adjustment logic. During the drilling process, there is a lack of synchronous fusion analysis of multi-source data, and a dynamic identification mechanism for the coupling characteristics between downhole signals has not been established. The system feedback response mainly revolves around a single parameter or independent variable, and the trend consistency and energy changes between signals are not effectively extracted, resulting in a limited dimension of risk behavior identification. Especially in the context of complex disturbances such as slip, the existing technology lacks real-time trend judgment and behavior persistence analysis of the risk evolution process, making it difficult to achieve feedforward assessment of potential drill string deformation status. When the bottomhole stress state undergoes a periodic disturbance, the system cannot make an effective judgment based on the signal direction coupling and amplitude change rules, resulting in delayed risk identification and control decision deviations. It is easy to cause problems such as drill string force accumulation, increased drill tool wear, and increased system misjudgment rate, which have an adverse impact on drilling efficiency and equipment reliability. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an automatic drill feeding optimization control system.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions: an automatic drill feeding optimization control system comprises:

[0007] The WOB frequency extraction module obtains dynamic sampling data of WOB during deep well drilling, identifies pairs of local extreme value points, establishes a set of WOB fluctuation frequencies based on the time window span and amplitude difference between extreme values, sequentially constructs a time series of WOB fluctuations, and generates WOB spectrum sequence data.

[0008] The increasing trend identification module calculates the frequency change accumulation according to the WOB spectrum sequence data, compares it with the change rate threshold, selects the continuously increasing frequency segments and classifies them into categories, and generates a frequency increasing interval set;

[0009] The linkage signal monitoring module extracts the bottom hole trajectory coordinate sequence and the drill string torque fluctuation sequence within the corresponding period according to the frequency increment interval set, compares the fluctuation directions, selects the same-direction fluctuation feature points, and generates a same-direction fluctuation feature point set;

[0010] The slip response judgment module determines whether the directional consistency coefficient and the amplitude synchronization growth rate characteristics are present in continuous cycles based on the set of unidirectional fluctuation feature points, determines it as slip risk behavior, constructs a strain energy density index, defines the risk level classification of the drilling stage, and generates a slip risk quantification matrix.

[0011] As a further solution of the present invention, the drilling pressure spectrum sequence data includes frequency time series, fluctuation amplitude distribution, and periodic characteristic indicators; the frequency increment interval set includes frequency change trend category, duration classification label, and threshold matching record; the same-direction fluctuation feature point set includes signal correlation coefficient distribution, amplitude integral value, and trend persistence level; and the slip risk quantification matrix includes strain energy density index, directional consistency coefficient level, and risk level classification.

[0012] As a further solution of the present invention, the weight-on-bit frequency extraction module includes:

[0013] The WOB extreme value identification submodule detects the changing trend of the WOB curve within each cycle based on the continuous cycle WOB dynamic sampling data collected by sensors, identifies the pairing relationship between local maximum and local minimum points, determines the time series distribution and differences between extreme value points, and generates a sequence of WOB extreme value pairs;

[0014] The fluctuation interval construction submodule calculates the time difference and amplitude difference between the extreme value pairs according to the time span and the difference in the drilling pressure amplitude of adjacent extreme value pairs in the drilling pressure extreme value pair sequence, combines them to form a periodic fluctuation unit, and forms a periodic grouping according to the amplitude difference and time span between adjacent units to generate a drilling pressure fluctuation interval set;

[0015] The WOB spectrum generation submodule extracts the time span, amplitude difference, and extreme value density of each set of fluctuation intervals based on the WOB fluctuation interval set. It integrates the fluctuation characteristic values ​​of each set, calculates the fluctuation frequency within a continuous period, and constructs the spectrum trend change by combining the fluctuation amplitude and time parameters to generate WOB spectrum sequence data.

[0016] As a further solution of the present invention, the growth trend identification module includes:

[0017] The frequency change calculation submodule obtains the frequency value difference of adjacent cycles item by item based on the frequency values ​​in any consecutive adjacent cycles in the bit weight spectrum sequence data, traverses the entire spectrum sequence in chronological order, constructs a continuous frequency change sequence according to the sliding window, calculates the cumulative trend of frequency change, and generates a frequency change trend value sequence;

[0018] The growth segment screening submodule reads each segment of continuous change according to the frequency change trend value sequence, compares the change rate with the frequency change rate threshold, screens the continuous growth segments, establishes the start and end periods and duration, and obtains a set of continuous growth frequency segments;

[0019] The interval trend division submodule calculates the total frequency change and the rate of change per unit time within the paragraph based on the set of continuously increasing frequency paragraphs. Based on whether the duration exceeds the set time threshold and whether the frequency growth direction vector is consistent, it establishes the numbering and marking of each category of frequency segments and generates a set of frequency increasing intervals.

[0020] As a further solution of the present invention, the linkage signal monitoring module includes:

[0021] The periodic data extraction submodule extracts the bottom hole trajectory coordinate sequence and the drill string torque fluctuation sequence within the period based on the time period covered by each interval in the frequency increment interval set, and synchronizes the two signals within the same period to a unified time scale to generate a periodic linkage signal data pair set;

[0022] The same-direction feature recognition submodule extracts the numerical change trend of the trajectory signal and the torque signal at each time point within the same period based on the periodic linkage signal data pair set, counts the trajectory change value and the torque change value of two adjacent sampling points, determines the same-direction fluctuation, and calculates the correlation proportionally to generate a same-direction fluctuation feature index sequence;

[0023] The feature integral calculation submodule obtains the position of each feature point in the original signal sequence in turn based on the same-direction fluctuation feature index sequence, calculates the joint amplitude integral value, calculates the length of the continuous time period in combination with the time series of each feature point, and jointly establishes a same-direction fluctuation feature point set.

[0024] As a further solution of the present invention, the slip reaction judgment module includes:

[0025] The direction consistency judgment submodule judges the direction consistency based on the same-direction fluctuation feature point set and the angle difference of the change direction, calculates the amplitude growth rate of the amplitude change values ​​of adjacent cycles according to the period difference, marks the slip response, and obtains the set of sections with consistent slip directions;

[0026] The energy index calculation submodule extracts the bottom hole displacement amplitude and torque amplitude in each period based on the set of segments with consistent sliding directions, calculates the strain energy density index per unit time by combining the equivalent stiffness parameters and rotational inertia of the drill string, and generates a strain energy density value sequence;

[0027] The risk level assessment submodule extracts the strain energy density index corresponding to each period segment according to the strain energy density value sequence, classifies and judges the index, and establishes a slip risk quantification matrix.

[0028] As a further embodiment of the present invention, the system further comprises:

[0029] The drilling speed adjustment module determines the direction and amplitude of the current drilling speed based on the risk level classification identified in the slip risk quantification matrix, matches the preset drilling speed adjustment rules, and generates a drilling speed control instruction;

[0030] The drilling speed control instruction includes a direction correction value, an amplitude adjustment factor, and a control parameter index.

[0031] As a further solution of the present invention, the drill speed adjustment module includes:

[0032] The risk matching submodule extracts the level information corresponding to each drilling cycle based on the cycle risk level label in the slip risk quantification matrix, matches the risk level identified by each cycle with the PID control parameter table one by one, and generates a risk response parameter set;

[0033] The parameter correction submodule collects historical error records within a continuous period based on the risk response parameter set, performs control calculations on the proportional term, integral term, and differential term, calculates the corresponding drilling speed change range based on the control output after dynamic change of the error trend, and obtains the drill speed correction value;

[0034] The control output submodule forms a new cycle drilling speed instruction based on the drill speed correction value and the current drilling speed as a benchmark, converts the instruction into a standard control signal, synchronously records the instruction issuance timestamp, and establishes a drilling speed control instruction.

[0035] Compared with the prior art, the advantages and positive effects of the present invention are:

[0036] In the present invention, a frequency set is constructed by extreme value extraction, and a drilling pressure fluctuation sequence is formed by combining the time span and the amplitude difference, thereby enhancing the resolution of dynamic characteristics. The trend segment is screened by the cumulative amount of frequency change, and quantitative identification of the fluctuation trend is realized. The bottom hole trajectory and torque fluctuation signal are introduced to establish a periodic direction consistency comparison mechanism to improve the accuracy of signal coupling judgment. According to the amplitude growth rate and direction consistency, the slip judgment logic is constructed, and the risk level classification is completed through the strain energy density index. The overall multi-dimensional signal linkage identification, trend quantitative classification and energy index evaluation are realized, which significantly improves the risk identification accuracy, system response sensitivity and control instruction adaptability under complex drilling conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a system flow chart of the present invention;

[0038] Figure 2 This is a flow chart of the weight-on-bit frequency extraction module of the present invention;

[0039] Figure 3 This is a flow chart of the amplification trend identification module of the present invention;

[0040] Figure 4 This is a flow chart of the linkage signal monitoring module of the present invention;

[0041] Figure 5 This is a flow chart of the slip reaction judgment module of the present invention;

[0042] Figure 6 This is a flow chart of the drill speed adjustment module of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0044] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0045] See also Figure 1 , an automatic drill feeding optimization control system includes:

[0046] The WOB frequency extraction module uses sensors to acquire dynamic sampling data of WOB within continuous cycles of deep well drilling. It identifies local maximum / minimum pairs of the WOB curve in each cycle, establishes a WOB fluctuation frequency set based on the time window span and amplitude difference between the extreme values, and sequentially constructs a time series of WOB fluctuations to generate WOB spectrum sequence data.

[0047] The increasing trend identification module calculates the cumulative frequency change based on the frequency values ​​within any consecutive adjacent cycles of the WOB spectrum sequence data. This is then compared with the rate of change threshold (based on the criteria for abnormal WOB fluctuations defined in API RP59, "Recommended Practice for Dynamic Monitoring of Drilling Systems"). This module then filters out continuously increasing frequency segments and classifies them according to the duration threshold and trend vector direction, generating a set of frequency increasing intervals.

[0048] The linkage signal monitoring module extracts the bottomhole trajectory coordinate sequence and drill string torque fluctuation sequence within the corresponding period based on the time period covered by each interval in the frequency increment interval set. It then compares the fluctuation directions of the two signals within the same period, selects the characteristic points of the same-direction fluctuation (determined by calculating the Pearson correlation coefficient between the bottomhole trajectory and torque signals, in accordance with the signal correlation analysis method defined in the SPE154893 standard), and records the corresponding amplitude integral value and trend persistence to generate a set of characteristic points of the same-direction fluctuation.

[0049] The slip response judgment module, based on each set of signals from the same-direction fluctuation feature points, determines whether they exhibit directional consistency coefficients and synchronous amplitude growth rates over consecutive cycles. If these conditions are met, a slip risk behavior is determined to have occurred. Based on the displacement and torsional changes corresponding to each trend, a strain energy density index (a drill string deformation assessment parameter based on energy principles, the calculation method is described in Appendix C of ISO 10407-1, "Petroleum and Natural Gas Industry - Rotary Drilling Equipment") is constructed. This defines the risk level classification for the drilling phase and generates a slip risk quantification matrix.

[0050] The drill speed adjustment module, based on the risk level classification identified in the slip risk quantification matrix, matches the preset PID control parameter table (developed according to the closed-loop control system parameter configuration specifications provided by the drilling equipment manufacturer and in compliance with the ISA-75.25 industrial control valve standard), determines the direction and amplitude of the current drill speed, and generates a drilling speed control command.

[0051] The drilling pressure spectrum sequence data includes frequency time series, fluctuation amplitude distribution, and periodic characteristic indicators. The frequency increment interval set includes frequency change trend category, duration classification label, and threshold matching record. The same-direction fluctuation characteristic point set includes signal correlation coefficient distribution, amplitude integral value, and trend persistence level. The slip risk quantification matrix includes strain energy density index, directional consistency coefficient level, and risk level classification. The drilling speed control command includes direction correction value, amplitude adjustment factor, and control parameter index.

[0052] See also Figure 2 , the WOB frequency extraction module includes:

[0053] The WOB extreme value identification submodule detects the changing trend of the WOB curve within each cycle based on the continuous cycle WOB dynamic sampling data collected by sensors, identifies the pairing relationship between local maximum and local minimum points, determines the time series distribution and differences between extreme value points, and generates a sequence of WOB extreme value pairs;

[0054] Based on the continuous periodic drilling pressure dynamic sampling data collected by the sensor, the corresponding relationship between the unified time axis and the drilling pressure sampling value should be established first to ensure that the drilling pressure value is obtained once per second. A total of 30 sets of drilling pressure data are obtained within the 30-second sampling time range. After that, the drilling pressure value sequence is traversed and the local extreme value judgment is performed based on the data of three adjacent points. When the middle value is greater than the two sides, it is a local maximum value, otherwise it is a minimum value. For example, in the sampling sequence, if the drilling pressure values ​​from the 1st to the 3rd second are 160kN, 175kN, and 160kN respectively, then the data point 175kN at the 2nd second is identified as the local maximum value, and the time point and drilling pressure value are recorded. For example, the values ​​from the 3rd to the 5th second are If the maximum value is 160kN, 185kN, and 160kN, then the maximum value at the 4th second is 185kN. All extreme values ​​that meet the conditions in the continuous sequence are identified and numbered. After all extreme values ​​are identified, minimum-maximum or maximum-minimum pairing is performed. If the minimum value at the 1st second is 160kN and the maximum value at the 2nd second is 175kN, a group of extreme value point pairs is formed. The recording time span Δt is 10s and the amplitude difference Δp is 15kN. Multiple extreme value point pairs are paired in sequence to construct an extreme value point pair sequence. Table 1 lists the extreme value feature comparison information of some actual sampling data points, which is used to assist in distinguishing pairing rules and fluctuation structure trends.

[0055] Table 1 Extreme values ​​and fluctuation parameters of bit pressure

[0056] Weight on bit (kN) Sampling time (s) Extreme point type Amplitude difference (kN) Time span (s) 160 0 Minimum 175 10 Maximum 15.0 10.0 185 20 Minimum 10.0 10.0 200 30 Maximum 15.0 10.0

[0057] As shown in Table 1, multiple extreme value points can be identified according to the sampling time series order and paired according to the maximum and minimum rules. At the same time, the corresponding amplitude difference and time span are calculated to form a complete sequence of extreme bit pressure pairs.

[0058] The fluctuation interval construction submodule calculates the time difference and amplitude difference between extreme value pairs in the WOB extreme value pair sequence based on the time span and WOB amplitude difference of adjacent extreme value pairs, combines them to form periodic fluctuation units, and forms periodic groups based on the amplitude difference and time span between adjacent units to generate a set of WOB fluctuation intervals.

[0059] According to the extreme value pair sequence of bit pressure, the time span Δt and amplitude difference Δp in the extreme value pair are used as basic characteristic parameters to construct multiple fluctuation intervals. For each extreme value pair, the timestamp difference between the minimum point and the subsequent maximum point is taken to obtain Δt. For example, if the minimum value appears at the 0th second and the maximum value appears at the 10th second, then Δt is 10s, and the amplitude difference Δp is 175kN minus 160kN, which is 15kN. Then the pair of parameters is judged and classified according to the set amplitude difference threshold and time span benchmark. The Δt interval threshold is set to 1. 0s, the Δp classification threshold is 10kN. If Δp ≥ 10kN and Δt ≤ 10s, it is classified as a high-frequency fluctuation interval, otherwise it is classified as a low-frequency slow-changing interval. In this example, the pairing of the minimum value 160kN and the maximum value 175kN is classified as a high-frequency fluctuation interval. After traversing all data pairs in the extreme value sequence, all units that meet the above intervals are counted and merged into a set with the same fluctuation structure characteristics. The time period and bit weight difference parameters of each group are added, and the complete sequence is arranged in chronological order to finally generate the bit weight fluctuation interval set.

[0060] The WOB spectrum generation submodule extracts the time span, amplitude difference, and extreme value density of each set of fluctuation intervals based on the WOB fluctuation interval set. It then integrates the fluctuation characteristic values ​​of each set, calculates the fluctuation frequency within a continuous period, and constructs the spectrum trend change based on the fluctuation amplitude and time parameters to generate WOB spectrum sequence data.

[0061] Based on the set of WOB fluctuation intervals, core parameter values ​​are extracted for each group of fluctuation units, including the amplitude difference Δp, duration Δt, number of extreme value points per unit time ρ, and periodic standard deviation σ. Assuming the system records continuously for 30 seconds during the sampling period, if three extreme value pairs are detected during this period, the extreme value density per unit time ρ is 3 / 30 = 0.1 pairs / s. Cyclic stability σ is defined as the standard deviation of Δt in the fluctuation interval. If the three Δt values ​​are 10s, 10s, and 10s, respectively, σ = 0, indicating uniform and stable periodic fluctuations. If Δt is 10s, 9s, and 12s, σ = 1.247. By comparing the changes in σ values, the changing trend of the fluctuation structure can be determined. Combined with basic calculations such as frequency f = 1 / Δt, the frequency type of each fluctuation group can be further determined and sorted in time series order. Ultimately, a set of associations between multiple fluctuation frequency values ​​and structural features is obtained, and a standardized WOB spectrum structure is constructed for outputting sorted WOB spectrum sequence data.

[0062] See also Figure 3, the growth trend identification module includes:

[0063] The frequency change calculation submodule obtains the frequency value difference of adjacent cycles one by one based on the frequency value in any consecutive adjacent cycles in the bit weight spectrum sequence data, traverses the entire spectrum sequence in chronological order, and constructs a continuous frequency change sequence according to the sliding window. The formula is:

[0064]

[0065] Calculate the frequency change cumulative trend V corresponding to the kth period when the window starting point is k , generate a frequency change trend value sequence, where f i Represents the frequency value of the i-th cycle, f i+1 Represents the frequency value of the i+1th cycle, is the average frequency in the current sliding window, and m is the window length;

[0066] Based on the frequency values ​​within any consecutive adjacent cycles in the WOB spectrum sequence data, a correspondence table between cycle numbers and corresponding frequency values ​​is first constructed in chronological order. For example, the frequency of the first cycle is 2.4 Hz, the second cycle is 2.7 Hz, the third cycle is 3.0 Hz, the fourth cycle is 3.4 Hz, and the fifth cycle is 3.8 Hz. The frequency changes between adjacent cycles are then calculated. The frequency change is equal to the frequency value of the next cycle minus the frequency value of the previous cycle. Therefore, the change of the second cycle relative to the first cycle is 0.3 Hz, and the change of the third cycle relative to the second cycle is also 0.3 Hz. Similarly, the change of the fourth cycle is 0.4 Hz, and the change of the fifth cycle is 0.4 Hz. Subsequently, a cumulative frequency change sequence is constructed. The starting cumulative value of the second cycle is 0.3 Hz, the third cycle is 0.6 Hz, the fourth cycle is 1.0 Hz, and the fifth cycle is 1.4 Hz, as shown in the table below:

[0067] Table 2 Frequency trend calculation example table

[0068] Cycle Number Frequency value (Hz) Frequency change (Hz) Cumulative change (Hz) 1 2.4 2 2.7 0.3 0.3 3 3.0 0.3 0.6 4 3.4 0.4 1.0 5 3.8 0.4 1.4

[0069] Assume that the window length is 4, that is, m=3, and the sliding starts from period 1. The frequency values ​​are as follows (from Table 2):

[0070] Cycle Number <![CDATA[Frequency value f i (Hz)]]> 1 2.4 2 2.7 3 3.0 4 3.4

[0071] Calculate the sum of frequency changes

[0072] |f2-f1|=|2.7-2.4|=0.3;

[0073] |f3-f2|=|3.0-2.7|=0.3;

[0074] |f4-f3|=|3.4-3.0|=0.4;

[0075] The sum of the frequency changes is:

[0076] 0.3+0.3+0.4=1.0;

[0077] Calculate the frequency average

[0078]

[0079] Calculate the frequency standard deviation part

[0080]

[0081] The sum of squares is:

[0082] 0.2256+0.0306+0.0156+0.2756=0.5474;

[0083] Divide by m+1=4, and take the square root:

[0084]

[0085] Adding the two parts together gives:

[0086] V1=1.0+0.370=1.370;

[0087] The cumulative trend value of the frequency change in the 4-period window starting from the first period is 1.370. This value is used to measure the strength of the frequency fluctuation trend and can be used as a basis for screening subsequent continuous growth trends to form a frequency change trend value sequence.

[0088] The cumulative trend of frequency change is a comprehensive indicator used to measure the intensity of the frequency change trend between adjacent cycles in a WOB spectrum sequence. In essence, it reflects the comprehensive characteristics of the total amplitude of frequency change and the consistency of fluctuation within a certain time window. This indicator consists of two parts: one is the sum of the absolute values ​​of the frequency changes between each cycle, which is used to reflect the overall increase or decrease in frequency within the time period; the other is the standard deviation of the frequency values ​​within the window, which is used to measure the consistency and stability of frequency fluctuations. The smaller the standard deviation, the more concentrated the frequency change and the more obvious the trend. The cumulative trend formed by adding the two can reflect both whether there is a trend of rapid frequency growth and whether the growth is continuous and stable. Therefore, this quantity can be used as the core criterion for screening continuous rising segments in the process of identifying frequency increase trends, and plays a key role in identifying the stage-by-stage increasing characteristics of the WOB fluctuation process.

[0089] The calculation logic of the formula is based on the fact that the frequency change trend is jointly determined by the "total amplitude" and "fluctuation stability". First, the absolute value of the frequency difference of adjacent periods in the window is taken and summed up in order to quantify the overall frequency change intensity within the time period. The absolute value avoids the possibility of adjacent changes offsetting each other, thereby truly reflecting the cumulative change of frequency; then, the deviations of all frequency values ​​in the window and their average frequency value are squared, and the square root is taken after averaging, that is, the standard deviation is calculated. This operation is used to measure the stability of frequency changes. The smaller the standard deviation, the more concentrated the changes and the more consistent the trend; finally, the sum of the frequency changes is added to the standard deviation in order to comprehensively reflect the "growth scale" and "fluctuation consistency" of the frequency, so as to determine whether the frequency shows a continuous and obvious increasing trend, thereby improving the accuracy and completeness of trend segment identification.

[0090] The growth segment screening submodule reads each segment of continuous change according to the frequency change trend value sequence, compares the change rate with the frequency change rate threshold, screens the continuous growth segments, establishes the start and end periods and duration, and obtains the continuous growth frequency segment set;

[0091] According to the frequency change trend value sequence, the frequency change trend value of each continuous segment is extracted, and the abnormal judgment standard of drilling pressure fluctuation defined in the APIRP59 standard is set as the threshold benchmark. The standard defines that when the drilling pressure fluctuation frequency continuously increases and the average change rate exceeds 0.3Hz / s, it is considered an abnormal trend. Based on this, the frequency change rate threshold is set to 0.3Hz / s, and each trend sequence is judged. If the frequency change values ​​of three consecutive cycles are all positive and the average change amount of each step is greater than or equal to 0.3Hz, the segment is judged as a frequency continuous growth segment. Taking the second to fifth cycles in Table 2 as an example, their frequency values ​​are 2.7 and 3.7, respectively. Hz, 3.0Hz, 3.4Hz, and 3.8Hz, with continuous changes of 0.3Hz, 0.4Hz, and 0.4Hz respectively, and an average of 0.367Hz, which exceeds the threshold and meets the judgment condition. Therefore, this sequence segment is extracted as a growth segment for identification, and its starting cycle number and ending cycle number are recorded as cycles 2-5, and its duration is 4 cycles. The corresponding duration is further accumulated (if the single cycle time is 1s, the duration is 4s). Finally, the segment information that meets the criterion is screened out from all frequency trend sequences and recorded in the form of segment number and start and end cycle to obtain a set of continuously growing frequency segments.

[0092] The interval trend division submodule is based on a set of continuously increasing frequency segments, calculates the total frequency change and the rate of change per unit time within the segment, and determines whether the duration exceeds the set time threshold and whether the frequency growth direction vector is consistent. It establishes the number and mark of each category of frequency segment and generates a set of frequency increasing intervals.

[0093] For the set of continuously growing frequency paragraphs, the start and end period numbers and duration values ​​of each paragraph are read in turn, and the frequency growth vector direction and growth rate in each paragraph are further calculated. The starting frequency value and the ending frequency value are directly called to find the difference and divided by the number of continuous periods to obtain the average growth rate. If the rate is stable and the duration is greater than the set time threshold of 5s, it is classified as a stable growth trend segment, otherwise it is classified as a short-term fluctuation trend segment. Taking the growth segment 2-5 period in paragraph 2 as an example, its duration is 4s, which does not reach the 5s threshold and is divided into a short-term growth interval. Then, based on whether the growth direction is consistent, that is, whether all frequency change values ​​are positive, it is judged whether the trend direction is unified. If it is unified, it is marked as a single vector segment. If there is fluctuation, it is marked as a mixed trend segment. In the above example, the frequency rises continuously and the direction is unified, which is classified as a single vector short-term growth interval, and is assigned number and label attributes for subsequent analysis, and finally a set of frequency increasing intervals is established.

[0094] See also Figure 4 , the linkage signal monitoring module includes:

[0095] The periodic data extraction submodule obtains the corresponding period number based on the time period covered by each interval in the frequency increment interval set, and extracts the bottom hole trajectory coordinate sequence and drill string torque fluctuation sequence within the period. For the trajectory coordinate sequence, the bottom hole position coordinate value is recorded every second. For the torque fluctuation data, the torque value change curve per unit time is extracted from the same time starting point. The two signals within the same period are synchronized to a unified time scale to generate a period linkage signal data pair set.

[0096] Based on the time period covered by each interval in the set of frequency-increasing intervals, the cycle number information corresponding to the time period is first obtained. The start and end time points of each cycle are further located. Based on this time period range, a three-dimensional coordinate sequence is extracted from the original bottomhole trajectory database. This sequence records the bottomhole position changes at a second-level sampling frequency, corresponding to one coordinate point per second, and consists of x, y, and z three-dimensional positions. At the same time, the drill string torque sensor data recorded in the corresponding time period is called. The torque signal is recorded at a 5 Hz sampling rate, that is, 5 torque values ​​are recorded per second. Based on this, the bottomhole trajectory sequence and the torque sequence are unified to the same time axis. The bottomhole trajectory is increased from the original 1 Hz to 5 Hz through linear interpolation to ensure that both signals contain an equal number of sampling points within the same cycle, forming a time-synchronized signal pair structure. For example, cycle 1 covers the time period from 0 to 6 seconds, corresponding to 6 groups of trajectory sampling points and 30 groups of torque sampling points. After interpolation, the trajectory also becomes 30 groups, forming 30 one-to-one corresponding trajectory-torque linkage sample groups. The synchronized signal pairs of each cycle are then packaged and stored to obtain a cycle-linked signal data pair set.

[0097] The same-direction feature recognition submodule extracts the numerical change trend of the trajectory signal and torque signal at each time point within the same period based on the periodic linkage signal data pair set, and calculates the change direction value respectively. The trajectory change value and torque change value of two adjacent sampling points are counted, and the trend direction is determined by the positive and negative signs. If the two have the same sign, it is judged as a same-direction fluctuation. The number of points with the same trend direction is counted, and the correlation is calculated proportionally to generate a same-direction fluctuation feature index sequence.

[0098] According to the periodic linkage signal data pair set, the trajectory sequence and torque sequence in each cycle are compared and processed. First, the two signals are differentiated point by point in the time series, and the change between each pair of adjacent sampling points is calculated respectively. For the j-th sampling point, its change direction is determined by f(j)-f(j-1). If the corresponding change value signs of the trajectory sequence and the torque sequence at the same time point are consistent, it is determined to be a point of change in the same direction. The index information of the point is recorded, the total number of points in the same direction in each cycle is counted and the proportion of all points is calculated. Further comparison of the trajectory and torque is performed. The Pearson correlation coefficient is calculated for the full-cycle sequence. Assume that the sequences are {x1, x2, ..., xn} and {y1, y2, ..., yn}, with means μx and μγ. The coefficient is calculated as the covariance divided by the product of the standard deviation. The result is judged to be greater than the critical value of 0.7. If it is satisfied, it is considered that the trajectory and torque fluctuation in this cycle have a high correlation. The set of same-direction points obtained by screening is confirmed again as significant same-direction fluctuation feature points. All the point numbers, corresponding time indexes and sequence features that pass the judgment in each cycle are integrated to establish a same-direction fluctuation feature index sequence.

[0099] The feature integral calculation submodule obtains the position of each feature point in the original signal sequence based on the same-direction fluctuation feature index sequence, extracts the corresponding amplitude value according to the bottom hole trajectory and torque signal of the position, accumulates the joint amplitude value of all feature points, and records the maximum duration of continuous appearance of each group of feature points, using the formula:

[0100]

[0101] Calculate the joint amplitude integral value A s , calculate the length of the continuous time period T by combining the time series of each feature point s , together establish a set of characteristic points of the same-direction fluctuation, where Represents the bottom hole trajectory amplitude value at point j, μ x is the average amplitude of the trajectory signal, is the torque amplitude value at point j, μ y is the average amplitude of the torque signal, d j is the duration of the corresponding point in the feature segment, T s is the trend duration, n is the number of sampling points;

[0102] The same-direction fluctuation feature index sequence is called, and the amplitude value of each feature point is extracted in the corresponding sampling window in the trajectory and torque signal. The amplitude of the trajectory and torque signal are calculated as the difference between the maximum and minimum values ​​within 1 second. The sampling frequency is 5Hz, that is, the window length is 10 data points, and the trajectory mean μ is 10. x and the mean torque μ y Set them to 2.3mm and 4.0N·m respectively. Three characteristic points are selected, and their amplitude values ​​and durations are shown in Table 3:

[0103] Table 3 Example of calculation of characteristic points of same-direction fluctuations

[0104]

[0105] Substituting the above parameters into the formula, the calculation is as follows:

[0106] Point 1:

[0107] Point 2:

[0108] Point 3:

[0109] A s =2.132+1.814+2.000=5.946;

[0110] The results show that the total joint amplitude change of the three feature points in their respective corresponding periods is 5.946, and the maximum duration is 4 seconds, which are respectively formed into the joint amplitude integral value and trend duration, and together constitute the same-direction fluctuation feature point set.

[0111] The joint amplitude integral value represents the cumulative effect of the joint fluctuation intensity of the bottomhole trajectory signal and the drill string torque signal at each sampling point in the same-direction fluctuation feature point. By measuring the degree of deviation of the trajectory amplitude and torque amplitude of each feature point from their average value, and combining it with the duration of the trend at that point, it reflects a comprehensive indicator of the fluctuation synchronization and amplitude superposition of the two types of signals on the time scale. The larger the value, the more significant the same-direction response of the trajectory and torque in that period and the more sufficient the energy superposition. It is an important measure of the linkage strength of the two signals within the period.

[0112] The operational logic of the formula is to comprehensively evaluate the impact of the amplitude deviation of the two signals at the characteristic points of the same-direction fluctuation and the duration of the trend on the overall fluctuation intensity. The absolute value of the difference between the trajectory amplitude and the torque amplitude and their average values ​​is used to quantify the degree of deviation of the fluctuation amplitude at this point relative to the global level, reflecting the local amplitude significance of this point. The sum of the two items represents the joint amplitude contribution of the trajectory and torque at this point, and then the duration d of the trend segment at this point is introduced. j, and taking its square root makes the duration weigh the integral value, but avoids too strong amplification effect of its numerical value. The square root form can gently reflect the importance of trend continuity, while maintaining the overall formula's ability to suppress abnormal short-term amplitude peaks. Therefore, the formula forms a balanced calculation of the intensity of the same-direction linkage fluctuations with both transient and time characteristics through the "sum of the offset intensity terms plus the square root of the duration".

[0113] See also Figure 5 , the slip reaction judgment module includes:

[0114] The direction consistency judgment submodule is based on each group of signals in the same-direction fluctuation feature point set. It records the trajectory displacement change vector sequence and torque fluctuation direction in consecutive cycles, and judges the direction consistency based on the angular difference of the change direction. The amplitude change value of adjacent cycles is calculated according to the period difference to determine whether the amplitude of consecutive cycles is a positive increasing sequence. If both conditions are met, it is marked as a slip response. All marked period segments are uniformly sorted to obtain a set of segments with consistent slip directions.

[0115] Based on each group of signals in the same-direction fluctuation feature point set, the change direction characteristics in adjacent periods are first detected, and the change vector corresponding to each period in the trajectory signal and torque signal is obtained. The angular difference between the trajectory displacement vector change and the torque fluctuation direction in the unit period is calculated. When the angular difference is less than 15 degrees, it is judged to be in the same direction and marked as a candidate period segment. Then, the amplitude change sequence of the corresponding period segment is obtained, and the difference between the trajectory amplitude and torque amplitude of adjacent periods is calculated item by item. If the difference is positive, the amplitude of the period segment is judged to be in a synchronous growth state. Finally, only when both the direction consistency and the amplitude synchronous growth are met, it is judged to be a period with slip response characteristics, and these period sets that meet the conditions are integrated into a set of segments with consistent slip directions.

[0116] The energy index calculation submodule extracts the bottom hole displacement amplitude and torque amplitude in each period based on the set of segments with consistent sliding directions, and combines the equivalent stiffness parameters and rotational inertia of the drill string to use the formula:

[0117]

[0118] Calculate the strain energy density index W per unit time q , all periodic index values ​​are compared with the defined threshold and normalized to the percentage level to generate a strain energy density value sequence, where U q is the equivalent axial stiffness of the drill string in the qth cycle, Δx q is the displacement change, Δt q is the torque change in this cycle, I q is the torsional inertia of the corresponding period;

[0119] Based on the set of segments with consistent slip directions, the displacement change, torque change, axial stiffness, and torsional inertia of each cycle are extracted to construct the strain energy density calculation process. Taking cycles 1 to 3 as an example, the parameters are shown in Table 4:

[0120] Table 4 Strain energy density calculation example

[0121] Cycle Number <![CDATA[Displacement change Δx q (mm)]]> <![CDATA[Torque change Δt q (N·m)]]> <![CDATA[Axial stiffness U q (kN / mm)]]> <![CDATA[Moment of inertia I q (kg·m 2 )]]> 1 3.2 80 1.2 0.015 2 2.5 65 1.2 0.015 3 3.0 75 1.2 0.015

[0122] According to the displacement and torque fluctuation characteristics, the strain energy density formula is used for calculation. The calculation of each cycle is as follows:

[0123] Cycle 1:

[0124] Cycle 2:

[0125] Cycle 3:

[0126] Combining the above calculations, the strain energy density index sequence corresponding to each cycle is obtained, and the strain energy density value sequence is constructed.

[0127] The strain energy density index reflects the energy per unit volume accumulated by the drill string due to the combined effects of axial displacement and torsional deformation during a specific drilling cycle. This index comprehensively measures the deformation energy consumption state of the drill string under complex downhole loading conditions. It reflects the axial tensile and compressive deformation strength through displacement changes, and the torsional shear strength through torque changes. These two parts of deformation energy are converted into a unified numerical form, thereby measuring the deformation risk of the drill string at different drilling stages. A larger value of this index indicates more severe deformation of the drill string during the cycle and more energy accumulation, which usually indicates an increased risk of slippage, sticking, or instability in the well. Therefore, it can be used as a quantitative assessment benchmark for slip response during drilling.

[0128] The operational logic of this formula is based on the theory of energy conservation and elastic mechanics, and the deformation behavior of the drill string is modeled using the angle of energy accumulation, where the displacement term (Δx q ) 2 It represents the square of the deformation of the drill string in the axial direction, multiplied by the axial stiffness U q And multiplied by 1 / 2 because the axial strain energy calculation formula in the elastic stage is It is the elastic potential energy stored per unit axial deformation; and the torque term (Δt q ) 2 / I qIt reflects the torsional energy density under unit moment of inertia, and is also multiplied by 1 / 2 to express the cumulative form of torsional energy. The entire formula is formed by the sum of these two parts to form the total strain energy density, which reflects the total deformation energy storage generated by the drill string due to force during this cycle. The two are equivalent transformations of forces in different directions, and the superposition represents the corresponding relationship between the overall energy input and the structural response.

[0129] The risk level assessment submodule extracts the strain energy density index corresponding to each cycle segment based on the strain energy density value sequence and classifies and judges it. Based on the strain energy density value, a two-dimensional matrix diagram is constructed with the drilling cycle on the horizontal axis and the energy level on the vertical axis. The risk level of each cycle segment is filled into the matrix, and the cycle segment marks with consistent directions and amplitude growth index marks are superimposed. Matrix mapping is performed on the intersection of multidimensional factors to establish a slip risk quantification matrix.

[0130] According to the strain energy density value sequence, the energy index W corresponding to each cycle is q The value is mapped to the risk level distribution, and the risk classification standard is: W q ≤100,000 is low risk, 100,000 is <W q ≤180,000 is medium risk, W q A value greater than 180,000 indicates high risk. For the three periods in the table above, period 1 is 213,339.47, indicating high risk; period 2 is 140,837.08, indicating medium risk; and period 3 is 187,505.4, also indicating high risk. By entering the number, risk level, direction consistency indicator, and amplitude synchronization indicator into a two-dimensional matrix, we construct a slippage risk quantification matrix with period number on the horizontal axis and risk level on the vertical axis. This yields the slippage risk quantification matrix.

[0131] See also Figure 6 , the drill speed adjustment module includes:

[0132] The risk matching submodule extracts the level information corresponding to each drilling cycle based on the cycle risk level label in the slip risk quantification matrix. Combined with the preset PID control parameter table, the risk level identified by each cycle is matched one by one with the control parameter table, and the P, I, and D control parameter sets corresponding to each cycle are retrieved to generate a risk response parameter set.

[0133] Based on the risk level classification identified in the slip risk quantification matrix, the corresponding risk level field is extracted for each drilling cycle, and this field is used as the key matching key to perform field matching with the preset PID control parameter table. The PID parameter table is established according to the ISA-75.25 industrial control valve standard. The table records the proportional coefficient, integral coefficient, and differential coefficient corresponding to the three risk levels. The system automatically calls the corresponding parameters according to different levels for switching control strategies. During the operation, the cycle number is used as the index to traverse the risk level label corresponding to each cycle one by one, match and extract the corresponding PID control parameter triples to form an independent control parameter configuration for each cycle. For example, P = 0.5, I = 0.8, and D = 0.2 are extracted for the high-risk cycle, P = 1.0, I = 0.5, and D = 0.3 for the medium-risk cycle, and P = 1.5, I = 0.2, and D = 0.5 for the low-risk cycle, forming a cycle-risk-control parameter three-way relationship. The data in the table are as follows:

[0134] Table 5 Comparison table of risk levels and PID parameters

[0135] Risk Level Proportional coefficient P Integration coefficient I Differential coefficient D Low risk 1.5 0.2 0.5 Medium risk 1.0 0.5 0.3 High risk 0.5 0.8 0.2

[0136] As shown in Table 5, the relationship between different risk levels and control parameters has been solidified through the configuration file, and dynamic matching is completed by the risk level field to ultimately generate a set of risk response parameters.

[0137] The parameter correction submodule calculates the current control error based on the risk response parameter set and the difference between the actual drilling speed in each cycle and the preset target drilling speed. It collects historical error records within consecutive cycles, performs control calculations on the proportional term, integral term, and differential term respectively, reads the current drilling pressure disturbance level and friction change trend, and uses the historical drill string stiffness estimation value to adjust the control amount. Based on the control output amount after the dynamic change of the error trend, it calculates the corresponding drilling speed change amplitude and obtains the drill speed correction value.

[0138] The PID parameter combination matched in Table 5 is read. The control error is constructed by combining the difference between the current drilling speed and the target drilling speed in each cycle. The current error value is obtained and the error data series of the previous three cycles are collected for integration processing. At the same time, the difference between the current error and the error of the previous cycle is obtained for differential estimation. The drilling pressure disturbance amplitude (set to 1.2 MPa), drill string stiffness (set to 130 kN / mm), and friction coefficient (set to 0.65) are then combined as adjustment factors and act on the proportional, integral, and differential output channels respectively to adjust the sensitivity of the control output. The speed correction value is then obtained by summing up the various adjustment values. If the current error is 0.3 m / min, the integral accumulates to 0.4 m / min, the differential change is -0.2 m / min, and the calculated correction value is +0.15 m / min. Based on the current drilling speed of 1.70 m / min, after adding the control correction value, the target speed is changed to 1.85 m / min. This correction value is the drilling speed correction value.

[0139] The control output submodule uses the drill feed speed correction value and the current drilling speed as a benchmark, adds the correction value as a superposition item, and forms a new cycle drilling speed instruction. The instruction is converted into a standard control signal, encapsulated into a PLC-recognizable format, written into the underlying actuator drive channel, and synchronously records the instruction issuance timestamp to establish the drilling speed control instruction.

[0140] Based on the drill speed correction value of 1.85 m / min obtained above, the system converts the target speed into the binary control quantity required by the control interface standard protocol. The actuator adopts a 10-bit bit width control mode. The system encodes the speed value as "0000111011" with a step size of 0.01 m / min, and then encapsulates it into a hexadecimal command "0106000100EB" according to the Modbus protocol. The system then writes the command into the control buffer of the underlying actuator and adds a timestamp mark "14:23:45.132" to the data packet as the queue index of the instruction in the multi-cycle control. At the same time, the record is registered in the control log and marked as "ACK" status, indicating that the instruction has been successfully issued and confirmed by the system, and finally the drilling speed control instruction is established.

[0141] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. An automatic drill feeding optimization control system, characterized in that: The system comprises: The WOB frequency extraction module obtains dynamic sampling data of WOB during deep well drilling, identifies pairs of local extreme value points, establishes a set of WOB fluctuation frequencies based on the time window span and amplitude difference between extreme values, sequentially constructs a time series of WOB fluctuations, and generates WOB spectrum sequence data. The increasing trend identification module calculates the frequency change accumulation according to the WOB spectrum sequence data, compares it with the change rate threshold, selects the continuously increasing frequency segments and classifies them into categories, and generates a frequency increasing interval set; The linkage signal monitoring module extracts the bottom hole trajectory coordinate sequence and the drill string torque fluctuation sequence within the corresponding period according to the frequency increment interval set, compares the fluctuation directions, selects the same-direction fluctuation feature points, and generates a same-direction fluctuation feature point set; The slip response judgment module determines whether the directional consistency coefficient and the amplitude synchronization growth rate characteristics are present in continuous cycles based on the set of unidirectional fluctuation feature points, determines it as slip risk behavior, constructs a strain energy density index, defines the risk level classification of the drilling stage, and generates a slip risk quantification matrix.

2. The automatic drill feeding optimization control system according to claim 1, characterized in that: The drilling pressure spectrum sequence data includes frequency time series, fluctuation amplitude distribution, and periodic characteristic indicators. The frequency increment interval set includes frequency change trend category, duration classification label, and threshold matching record. The same-direction fluctuation feature point set includes signal correlation coefficient distribution, amplitude integral value, and trend persistence level. The slip risk quantification matrix includes strain energy density index, directional consistency coefficient level, and risk level classification.

3. The automatic drill feeding optimization control system according to claim 1, characterized in that: The weight-on-bit frequency extraction module includes: The WOB extreme value identification submodule detects the changing trend of the WOB curve within each cycle based on the continuous cycle WOB dynamic sampling data collected by sensors, identifies the pairing relationship between local maximum and local minimum points, determines the time series distribution and differences between extreme value points, and generates a sequence of WOB extreme value pairs; The fluctuation interval construction submodule calculates the time difference and amplitude difference between the extreme value pairs according to the time span and the difference in the drilling pressure amplitude of adjacent extreme value pairs in the drilling pressure extreme value pair sequence, combines them to form a periodic fluctuation unit, and forms a periodic grouping according to the amplitude difference and time span between adjacent units to generate a drilling pressure fluctuation interval set; The WOB spectrum generation submodule extracts the time span, amplitude difference, and extreme value density of each set of fluctuation intervals based on the WOB fluctuation interval set. It integrates the fluctuation characteristic values ​​of each set, calculates the fluctuation frequency within a continuous period, and constructs the spectrum trend change by combining the fluctuation amplitude and time parameters to generate WOB spectrum sequence data.

4. The automatic drill feed optimization control system according to claim 1, characterized in that: The increasing trend identification module includes: The frequency change calculation submodule obtains the frequency value difference of adjacent cycles item by item based on the frequency values ​​in any consecutive adjacent cycles in the bit weight spectrum sequence data, traverses the entire spectrum sequence in chronological order, constructs a continuous frequency change sequence according to the sliding window, calculates the cumulative trend of frequency change, and generates a frequency change trend value sequence; The growth segment screening submodule reads each segment of continuous change according to the frequency change trend value sequence, compares the change rate with the frequency change rate threshold, screens the continuous growth segments, establishes the start and end periods and duration, and obtains a set of continuous growth frequency segments; The interval trend division submodule calculates the total frequency change and the rate of change per unit time within the paragraph based on the set of continuously increasing frequency paragraphs. Based on whether the duration exceeds the set time threshold and whether the frequency growth direction vector is consistent, it establishes the numbering and marking of each category of frequency segments and generates a set of frequency increasing intervals.

5. The automatic drill feeding optimization control system according to claim 1, characterized in that: The linkage signal monitoring module includes: The periodic data extraction submodule extracts the bottom hole trajectory coordinate sequence and the drill string torque fluctuation sequence within the period based on the time period covered by each interval in the frequency increment interval set, and synchronizes the two signals within the same period to a unified time scale to generate a periodic linkage signal data pair set; The same-direction feature recognition submodule extracts the numerical change trend of the trajectory signal and the torque signal at each time point within the same period based on the periodic linkage signal data pair set, counts the trajectory change value and the torque change value of two adjacent sampling points, determines the same-direction fluctuation, and calculates the correlation proportionally to generate a same-direction fluctuation feature index sequence; The feature integral calculation submodule obtains the position of each feature point in the original signal sequence in turn based on the same-direction fluctuation feature index sequence, calculates the joint amplitude integral value, calculates the length of the continuous time period in combination with the time series of each feature point, and jointly establishes a same-direction fluctuation feature point set.

6. The automatic drill feeding optimization control system according to claim 1, characterized in that: The slip reaction judgment module includes: The direction consistency judgment submodule judges the direction consistency based on the same-direction fluctuation feature point set and the angle difference of the change direction, calculates the amplitude growth rate of the amplitude change values ​​of adjacent cycles according to the period difference, marks the slip response, and obtains the set of sections with consistent slip directions; The energy index calculation submodule extracts the bottom hole displacement amplitude and torque amplitude in each period based on the set of segments with consistent sliding directions, calculates the strain energy density index per unit time by combining the equivalent stiffness parameters and rotational inertia of the drill string, and generates a strain energy density value sequence; The risk level assessment submodule extracts the strain energy density index corresponding to each period segment according to the strain energy density value sequence, classifies and judges the index, and establishes a slip risk quantification matrix.

7. The automatic drill feeding optimization control system according to claim 1, characterized in that: The system further comprises: The drilling speed adjustment module determines the direction and amplitude of the current drilling speed based on the risk level classification identified in the slip risk quantification matrix, matches the preset drilling speed adjustment rules, and generates a drilling speed control instruction; The drilling speed control instruction includes a direction correction value, an amplitude adjustment factor, and a control parameter index.

8. The automatic drill feeding optimization control system according to claim 7, characterized in that: The drilling speed adjustment module includes: The risk matching submodule extracts the level information corresponding to each drilling cycle based on the cycle risk level label in the slip risk quantification matrix, matches the risk level identified by each cycle with the PID control parameter table one by one, and generates a risk response parameter set; The parameter correction submodule collects historical error records within a continuous period based on the risk response parameter set, performs control calculations on the proportional term, integral term, and differential term, calculates the corresponding drilling speed change range based on the control output after dynamic change of the error trend, and obtains the drill speed correction value; The control output submodule forms a new cycle drilling speed instruction based on the drill speed correction value and the current drilling speed as a benchmark, converts the instruction into a standard control signal, synchronously records the instruction issuance timestamp, and establishes a drilling speed control instruction.

Images