Adaptive Control Method and System for Drilling Rig Electrical Control Combining Artificial Intelligence

Abstract

This invention provides an adaptive control method and system for drilling rig electrical control that incorporates artificial intelligence, relating to the field of artificial intelligence technology. First, drilling geological data and drilling operating condition data are collected. Then, coupled feature extraction processing is performed on the drilling geological data and drilling operating condition data to generate a set of coupled geological and operating condition features. This set of coupled geological and operating condition features is input into a pre-trained electrical control parameter evolution model to generate a set of control parameters. The control parameter set is transmitted to the execution unit, and feedback data is collected. Finally, the feedback data is correlated and compared with the feature set, and the internal parameters of the model are adjusted. This invention achieves adaptive control of the drilling rig electrical control system, which can dynamically adjust control parameters according to real-time geological conditions, possesses self-learning and optimization capabilities, improves drilling efficiency, ensures safety, and reduces costs.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and more specifically, to an adaptive control method and system for drilling rig electrical control that incorporates artificial intelligence. Background Technology

[0002] In drilling operations, precise control of the drilling rig's electrical control system is crucial for improving drilling efficiency, ensuring drilling safety, and reducing operating costs. Traditional drilling rig electrical control methods mainly rely on human experience and preset fixed parameter modes. During drilling, geological conditions are complex and variable. Different rock formations, compositions, structures, and distributions can significantly affect drilling conditions. For example, hard rock formations may increase drill string load, while soft rock formations may cause excessively fast drilling speeds, affecting drilling quality.

[0003] Meanwhile, drilling data such as motor operating status, drill string load, and drilling fluid flow are also changing in real time. However, traditional methods struggle to perceive these complex geological and operational changes in real time and cannot dynamically adjust the control parameters of the electrical control system according to actual conditions. This often leads to problems such as unreasonable motor speed, uneven drill string load, and mismatched drilling fluid flow during drilling, which in turn affects drilling efficiency, increases equipment wear, and may even cause safety accidents. Furthermore, traditional methods lack self-learning and optimization capabilities, making it impossible to continuously improve control strategies over long-term operations to adapt to drilling needs under different geological conditions. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide an adaptive control method for drilling rig electrical control incorporating artificial intelligence, the method comprising: Collect drilling geological data and drilling condition data during the drilling operation. The drilling geological data includes rock formation composition data, rock formation structure data, and rock formation distribution data. The drilling condition data includes motor operation data, drill string load data, and drilling fluid flow data. The drilling geological data and the drilling operating condition data are subjected to coupling feature extraction processing to identify the correlation between the drilling geological data and the drilling operating condition data, and a geological operating condition coupling feature set is generated, which contains feature combinations corresponding to the correlation. The set of coupled geological conditions is input into a pre-trained electrical control parameter evolution model to generate a set of control parameters for the drilling rig electrical control system. The set of control parameters includes motor speed parameters, drill string load adjustment parameters, and drilling fluid flow rate parameters. The set of control parameters is transmitted to the execution component of the drilling rig's electrical control system, driving the execution component to perform control operations. At the same time, drilling condition feedback data after the execution of control operations by the execution component is collected. The characteristic type of the drilling condition feedback data is consistent with that of the drilling condition data. The drilling condition feedback data is correlated and compared with the geological condition coupling feature set to calculate the feature deviation value between the two, and the internal operating parameters of the electrical control parameter evolution model are adjusted according to the feature deviation value.

[0005] Furthermore, embodiments of the present invention also provide an adaptive control system for drilling rig electrical control that incorporates artificial intelligence, characterized in that it includes: A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to execute the aforementioned adaptive control method for drilling rig electrical control incorporating artificial intelligence by executing the machine-executable instructions.

[0006] In another aspect, embodiments of the present invention also provide a computer program product, the computer program product including machine-executable instructions, the machine-executable instructions being stored in a computer-readable storage medium, a processor of a computer device reading the machine-executable instructions from the computer-readable storage medium, the processor executing the machine-executable instructions, causing the computer device to execute the above-described adaptive control method for drilling rig electrical control combined with artificial intelligence.

[0007] Based on the above, by comprehensively collecting drilling geological data and drilling condition data during the drilling operation, and performing coupling feature extraction processing on the collected data, the inherent correlation between the two data can be deeply identified. The set of coupled geological and drilling condition features is then input into a pre-trained electrical control parameter evolution model to generate a set of control parameters. This achieves intelligent generation of control parameters, enabling dynamic adjustment of motor speed, drill string load, and drilling fluid flow parameters according to different geological conditions, thus improving the flexibility and adaptability of drilling operations. The control parameter set is transmitted to the actuators, and feedback data is collected. By correlating and comparing the feedback data with the set of coupled geological and drilling condition features and adjusting the model's internal operating parameters, the electrical control parameter evolution model acquires self-learning and optimization capabilities. This allows for continuous improvement of control strategies during long-term drilling operations, further enhancing drilling efficiency, ensuring drilling safety, and reducing operating costs. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the execution flow of the adaptive control method for drilling rig electrical control that combines artificial intelligence, provided in an embodiment of the present invention.

[0009] Figure 2This is a schematic diagram of exemplary hardware and software components of an adaptive control system for drilling rig electrical control that incorporates artificial intelligence, provided in an embodiment of the present invention. Detailed Implementation

[0010] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating an embodiment of the adaptive control method for drilling rig electrical control that incorporates artificial intelligence. The following is a detailed description of this adaptive control method for drilling rig electrical control that incorporates artificial intelligence.

[0011] Step S110: Collect drilling geological data and drilling condition data during the drilling operation. The drilling geological data includes rock formation composition data, rock formation structure data, and rock formation distribution data. The drilling condition data includes motor operation data, drill string load data, and drilling fluid flow data.

[0012] In this embodiment, the drilling operation of a shale gas horizontal well on a land-based oil drilling platform is taken as the application scenario. In this scenario, it is necessary to acquire various data in real time during the drilling process to achieve adaptive control of the electrical control system. For the acquisition of drilling geological data, it is acquired in real time during the drilling process through logging-while-drilling instruments. Among them, rock composition data includes the composition information of various rocks such as shale, sandstone, and limestone at different depths within the shale gas horizontal well section, such as the content of clay minerals and quartz in shale; rock structure data covers the bedding structure and fracture development of rocks, such as the spacing of shale bedding and the porosity of sandstone; rock distribution data involves the distribution of different lithologies along the wellbore trajectory direction, such as the alternation pattern of shale and sandstone within a certain well section. When collecting this geological data, sensitive information such as the well site's geographical location and specific reservoir information is involved. Therefore, data encryption transmission technology is employed. AES encryption is used to ensure the data is not leaked during transmission. Sensitive fields are anonymized during data storage; for example, specific wellhead coordinates are converted to relative coordinates. Drilling condition data is collected by sensors installed on various key parts of the drilling equipment. Motor operation data includes parameters such as the drilling motor's real-time speed, output power, current, and voltage. Drill string load data includes the torque, axial pressure, and lateral force borne by the drill pipe. Drill fluid flow data includes the circulating flow rate, inlet and outlet pressure, density, and viscosity of the drilling fluid. All of this condition data is transmitted in real-time to the data processing center via an industrial bus. Timestamp synchronization technology is used during transmission to ensure the time consistency of all types of data.

[0013] Step S120: Perform coupling feature extraction processing on the drilling geological data and the drilling operating condition data to identify the correlation between the drilling geological data and the drilling operating condition data, and generate a geological operating condition coupling feature set, which contains feature combinations corresponding to the correlation.

[0014] The above step S120 specifically includes the following sub-steps: Step S121: Perform time-series slicing on the drilling condition data according to the order of acquisition time to obtain multiple condition data slices. Add a time identifier to each condition data slice. Each condition data slice contains motor operation data, drill string load data and drilling fluid flow data within the corresponding time period.

[0015] In this shale gas horizontal well drilling scenario, when executing step S121, the time interval for time-series slices must first be determined. By reviewing historical drilling operation records for the block where the well is located, the average duration of a complete drilling operation (such as drilling, connecting single sections, tripping, etc.) in the shale gas horizontal well drilling operation of that block is calculated. A time interval matching the duration of this operation is selected to ensure that each time interval contains a complete drilling operation. For example, if the average drilling operation duration is calculated to be 30 minutes, then the time interval for time-series slices is set to 30 minutes. Next, the acquisition time information in the drilling condition data is extracted. This time information is usually accurate to the second. All condition data are arranged in chronological order of acquisition time to form an ordered drilling condition data sequence. Each data point in this drilling condition data sequence is accompanied by corresponding acquisition time information. Then, using a defined 30-minute time interval, the ordered drilling condition data sequence is segmented. Starting from the first data point, all data points within the first 30 minutes are extracted as the first condition data slice, and so on, extracting data points within subsequent 30-minute intervals to form multiple condition data slices. Next, the amount of missing data in the motor operation data, drill string load data, and drilling fluid flow data within each slice is counted. If the amount of missing data in a slice exceeds a preset missing threshold (this threshold is set according to data integrity requirements, for example, 5% of the total slice data), the time interval is readjusted, for example, to 25 minutes or 35 minutes, and the ordered drilling condition data sequence is re-sliced ​​into time-series slices until the amount of missing data in all slices is below the preset missing threshold. A time identifier is added to each condition data slice that passes the integrity check. This time identifier consists of the slice's start and end acquisition times. For example, if the first slice's start acquisition time is 8:00:00 on a certain day and its end acquisition time is 8:30:00, then its time identifier is for that time period. Each data slice is bound to its corresponding time identifier, forming a data slice unit containing both data content and time identifier. The drilling procedure name corresponding to each slice unit is also recorded, such as the drilling procedure. Finally, all data slice units are sorted according to the chronological order of their time identifiers, forming an ordered data slice sequence.

[0016] Step S1211: Determine the time interval of the time sequence slice. The time interval is determined based on the process duration of the drilling operation. The process duration is obtained by statistically analyzing the process records of historical drilling operations. Select the time interval that matches the process duration so that each time interval contains a complete drilling process.

[0017] In this shale gas horizontal well drilling scenario, to determine the time interval for time-series slices, drilling operation records for shale gas horizontal wells in the area over the past three years were first collected. These records detailed the start and end times of each operation (such as drilling, connecting single sections, circulating drilling fluid, tripping, etc.) during different well sections of each well. Statistical analysis was performed on these records to calculate the average duration of each operation. For example, the average duration of the drilling operation was found to be 30 minutes, the average duration of the connecting single section operation was 15 minutes, and the average duration of the circulating drilling fluid operation was 20 minutes. Since drilling is the main operation in well drilling, and the operating data changes frequently during drilling, the average duration of the drilling process is selected as the primary reference. At the same time, the duration of other processes is also taken into account, and a time interval that can include the complete process of most processes is selected. For example, after comprehensive consideration, 30 minutes is selected as the time interval for time slices. This time interval can include the complete drilling process. For processes with shorter durations, such as connecting a single joint or circulating drilling fluid, multiple such processes may be included in one time interval, but they can be distinguished by the subsequent process name records.

[0018] Step S1212: Extract the acquisition time information from the drilling condition data, arrange all data according to the order of acquisition time information to form an ordered drilling condition data sequence, wherein each data point in the drilling condition data sequence is accompanied by acquisition time information.

[0019] The acquisition time information for each data point is read from the storage file of the drilling condition data. This acquisition time information is usually stored in the form of timestamps, accurate to the millisecond level. All data points are sorted according to the order of their timestamps, for example, arranged in chronological order from morning to evening, thus forming an ordered drilling condition data sequence. During the sorting process, if data points with the same timestamp are found, they are further sorted according to the priority of their data types. For example, the priority of motor operation data is higher than that of drill string load data, and the priority of drill string load data is higher than that of drilling fluid flow data, to ensure the uniqueness and orderliness of the data sequence. Each data point in this ordered drilling condition data sequence is accompanied by its original acquisition time information for subsequent slice processing and time stamp addition.

[0020] Step S1213: Divide the ordered drilling condition data sequence into segments based on defined time intervals. Starting from the first data point, extract all data points within the first time interval as the first condition data slice, and then extract data points within subsequent time intervals to form multiple condition data slices.

[0021] Using the 30-minute time interval determined in step S1211 as a unit, the ordered drilling condition data sequence formed in step S1212 is segmented. Starting from the first data point in the sequence, its acquisition time is checked. Taking this time as the starting point, all data points within a 30-minute interval are extracted and combined to form the first condition data slice. Then, taking the end time of the first condition data slice as the starting point, data points within a 30-minute interval are extracted to form the second condition data slice, and so on, until the entire ordered drilling condition data sequence is segmented, resulting in multiple condition data slices. During the extraction process, if the number of data points in a certain time interval is too small, it may be due to drilling operations being suspended or sensor malfunctions during that time period. In this case, it is necessary to record the situation and consider it in the subsequent missing data statistics.

[0022] Step S1214: Calculate the amount of missing data in motor operation data, drill string load data and drilling fluid flow data in each slice. If the amount of missing data exceeds the preset missing threshold, readjust the time interval and re-process the ordered drilling condition data sequence into time-series slices.

[0023] For each data slice of the drilling conditions, the missing data volume for motor operation data, drill string load data, and drilling fluid flow data is calculated separately. The missing data volume is calculated as follows: for each type of data, the difference between the total number of data points that should have been collected within the slice interval (calculated based on the sensor sampling frequency and time interval) and the actual number of data points collected is calculated. This difference is the missing data volume for that type of data. The missing data volumes of the three types of data are added together to obtain the total missing data volume for that slice. A preset missing data threshold is set according to the importance of the data and the needs of subsequent processing, for example, set to 5% of the total number of data points that should be collected within the slice interval. If the total missing data volume of a slice exceeds the preset missing data threshold, the time interval needs to be readjusted, for example, shortened to 25 minutes or extended to 35 minutes. Then, the ordered drilling condition data sequence is re-sliced ​​according to the new time interval, and the missing data volume of each slice is counted again until the missing data volume of all slices is below the preset missing data threshold.

[0024] Step S1215: Add a time stamp to each working condition data slice that passes the integrity check. The time stamp consists of the start and end acquisition times of the slice. The start acquisition time is the acquisition time of the first data point in the slice, and the end acquisition time is the acquisition time of the last data point in the slice.

[0025] For each data slice that passes the integrity check (i.e., the amount of missing data is less than a preset missing threshold), the acquisition time of the first data point within the slice is extracted as the start acquisition time, and the acquisition time of the last data point is extracted as the end acquisition time. The start and end acquisition times are combined into a timestamp. For example, if the start acquisition time is 8:00:00 AM on May 10, 2024, and the end acquisition time is 8:30:00 AM on May 10, 2024, then the timestamp for this data slice is "2024-05-10 08:00:00-2024-05-10 08:30:00". This timestamp is then associated with the data slice and stored for subsequent association processing.

[0026] Step S1216: Bind each working condition data slice to its corresponding time identifier to form a working condition data slice unit containing data content and time identifier, and record the drilling process name corresponding to each slice unit.

[0027] The data content of each working condition data slice (i.e., motor operation data, drill string load data, and drilling fluid flow data within that slice) is bound to its corresponding time identifier to form a complete working condition data slice unit. Simultaneously, based on real-time drilling operation records, the drilling procedure name corresponding to each working condition data slice unit is determined. For example, by viewing the operation log corresponding to the time identifier, it can be determined whether the procedure in progress during that time period is a drilling procedure or a single-joint connection procedure, and the procedure name is recorded in the attribute information of the working condition data slice unit.

[0028] Step S1217: Sort all operating condition data slice units according to the order of their time identifiers to form an ordered operating condition data slice sequence.

[0029] All drilling data slices are sorted according to their start acquisition times in their timestamps, starting from the earliest start acquisition time, to form an ordered drilling data slice sequence. This drilling data slice sequence can clearly reflect the changes in drilling data over time.

[0030] Step S122: Spatial partitioning of the drilling geological data is performed according to the depth range of the drilling operation to obtain multiple geological data partitions. Each geological data partition contains rock layer composition data, rock layer structure data and rock layer distribution data within the corresponding depth range. A depth identifier is added to each geological data partition.

[0031] In this shale gas horizontal well drilling scenario, according to the drilling design, the target formation of the horizontal well is located at a certain depth range underground. Therefore, the drilling geological data is spatially partitioned according to the depth interval of the drilling operation. First, the criteria for dividing the depth intervals are determined. Referring to the geological design profile of this shale gas horizontal well, the entire well section is divided into multiple continuous depth intervals, for example, each interval is 50 meters. For the geological data obtained from logging while drilling, based on the measurement depth, the rock composition data, rock structure data, and rock distribution data at each depth are assigned to the corresponding depth interval, thus obtaining multiple geological data partitions. Each geological data partition contains rock composition data for all measurement points within the 50-meter depth interval, such as the shale content and sandstone content at each measurement point within the interval; rock structure data, such as parameters reflecting the density of the structure, such as rock density and sonic transit time at each measurement point; and rock distribution data, such as the vertical distribution range of different lithologies within the interval. Add a depth identifier to each geological data partition. The depth identifier consists of the starting depth and ending depth of the partition. For example, if the starting depth of a partition is 2000 meters and the ending depth is 2050 meters, then its depth identifier is the depth range.

[0032] Step S123: Establish the correspondence between time markers and depth markers. Based on the progress record of drilling operations, associate the working condition data slices and geological data partitions of the same operation stage to form multiple geological working condition association units. Each geological working condition association unit contains a set of working condition data slices and a set of geological data partitions.

[0033] In this shale gas horizontal well drilling scenario, the drilling progress log details the drilling depths at different time points. By analyzing this progress log, a correspondence can be established between the time markers of the drilling data slices and the depth markers of the geological data partitions. For example, according to the progress log, if the drilling depth increases from 2000 meters to 2050 meters within the time period corresponding to the time marker of a certain drilling data slice, then the time marker of this drilling data slice corresponds to the geological data partition with a depth marker of 2000 meters to 2050 meters. Associating drilling data slices with the same operational phase (i.e., the phase corresponding to time and depth) with geological data partitions—for example, associating a drilling data slice with a time marker of 8:00:00 to 8:30:00 on a certain day with a geological data partition with a depth marker of 2000 meters to 2050 meters—creates a geological operational data association unit. This geological operational data association unit includes the drilling data slices within that time period and the geological data partitions within that depth range, thus achieving an organic combination of operational data in the time dimension and geological data in the spatial dimension.

[0034] Step S124: Extract features from the working condition data slices in each geological working condition associated unit, extracting the speed change features in the motor operation data, the load fluctuation features in the drill string load data, and the flow stability features in the drilling fluid flow data to form a subset of working condition features.

[0035] In this shale gas horizontal well drilling scenario, feature extraction was performed on the data slices of the working conditions within each geological condition associated unit. For motor operation data, the speed variation over time was analyzed to extract speed variation features, such as average, maximum, minimum, variance, and rate of change. These features reflect the motor's operational stability and trend over that time period. For drill string load data, the fluctuations in parameters such as torque and axial pressure were analyzed to extract load fluctuation features, such as load fluctuation amplitude, fluctuation frequency, and the number of peak occurrences. These features reflect the changes in load experienced by the drill string during drilling. For drilling fluid flow data, the stability of the circulating flow rate was studied to extract flow stability features, such as the standard deviation of the flow rate, the duration of deviation from the set value, and the time to recover stability. These features reflect the working state of the drilling fluid circulation system. The extracted speed variation features, load fluctuation features, and flow stability features were combined to form a subset of working condition features for this geological condition associated unit.

[0036] Step S125: Extract features from the geological data partitions in each geological condition association unit, extracting the component proportion features from the rock stratum composition data, the structural compactness features from the rock stratum structure data, and the distribution continuity features from the rock stratum distribution data, forming a subset of geological features.

[0037] In this shale gas horizontal well drilling scenario, feature extraction is performed on the geological data partitions within each geological condition associated unit. For rock composition data, the proportion of various rock components within the depth range is statistically analyzed, and compositional proportion features are extracted, such as the average proportion of shale, the maximum proportion of sandstone, and the minimum proportion of limestone within this range. These features reflect the lithological composition of this depth range. For rock structure data, parameters such as rock density, porosity, and sonic transit time are analyzed to extract structural compactness features, such as the average density, average porosity, and maximum sonic transit time within this range. These parameters characterize the density of the rock structure. For rock layer distribution data, the continuity of the distribution of different lithologies within the depth range is observed, and distribution continuity features are extracted, such as the continuous distribution length of shale layers, the number and thickness of sandstone interlayers, etc. These features reflect the spatial distribution pattern of the rock layers. The extracted compositional proportion features, structural compactness features, and distribution continuity features are combined to form a subset of geological features for this geological condition associated unit.

[0038] Step S126: Calculate the coupling degree between the working condition feature subset and the geological feature subset in each geological working condition association unit. The coupling degree is determined by the association frequency between the working condition feature subset and the geological feature subset. The association frequency is the number of times the two features change simultaneously within the same time period.

[0039] Step S126 above specifically includes the following sub-steps: Step S1261: Perform change point detection on the subset of working condition features in each geological working condition associated unit. For each feature in the subset of working condition features, identify the time point when the feature value changes. The change point is the time point when the difference between the feature value and the value at the previous moment exceeds the preset change threshold.

[0040] In this shale gas horizontal well drilling scenario, for each subset of working condition features within a geological working condition associated unit, change point detection is performed on each working condition feature based on the time series of working condition data slices. Taking the average rotational speed in the rotational speed variation feature as an example, the preset change threshold is twice the standard deviation of the historical data for this feature. The difference between the average rotational speed at two adjacent time points is calculated sequentially. When the absolute value of this difference exceeds the preset change threshold, that time point is determined to be a change point in the average rotational speed. Following the same method, change point detection is performed on each feature in the working condition feature subset, such as load fluctuation features (e.g., load fluctuation amplitude) and flow stability features (e.g., flow standard deviation), identifying time points where all feature values ​​change significantly.

[0041] Step S1262: Detect change points in the subset of geological features in each geological condition associated unit and identify the time points when the values ​​of each geological feature change.

[0042] For a subset of geological features, since geological data varies with depth, and depth and time are correlated through drilling progress, the depth variations of geological features are converted into time variations. For example, based on drilling operation progress records, the drilling time corresponding to a geological feature at a certain depth is known. For each geological feature, such as the average shale proportion in the compositional features, a preset change threshold is set at three times the average amplitude of the feature's variation across adjacent depth intervals. The difference in the average shale proportion between adjacent depth intervals is calculated. When this difference exceeds the preset change threshold, the time point at which the geological feature value changes is determined based on the drilling time corresponding to that depth interval. Similarly, change point detection is performed for each geological feature, such as structurally dense features (e.g., average density) and continuous distribution features (e.g., continuous shale distribution length), to obtain the time information of its change points.

[0043] Step S1263: Collect all change points of working condition characteristics to form a set of working condition change points, and collect all change points of geological characteristics to form a set of geological change points. The change points in the set of working condition change points and the set of geological change points are accompanied by corresponding time information.

[0044] All the changes in working condition characteristics detected in step S1261 are summarized to form a set of working condition change points. Each change point in this set is accompanied by its specific time information, such as 8:15:20 on a certain day. Similarly, all the changes in geological features detected in step S1262 are summarized to form a set of geological change points. Each change point is also accompanied by corresponding time information, which is obtained by converting the relationship between depth and time.

[0045] Step S1264: Set a time matching window. The duration of the time matching window is determined based on the delay in the impact of geological features on working condition features during drilling operations. The impact delay is obtained by statistically analyzing the time difference of feature changes in historical data.

[0046] In this shale gas horizontal well drilling scenario, changes in geological features affect operating conditions, but these effects are delayed. For example, when encountering tight sandstone formations, the drill string load does not change immediately but only becomes apparent after drilling to a certain depth. By analyzing the time difference between the geological feature change points and the corresponding operating condition change points in the historical drilling data of this block, the average delay of these effects was calculated, for example, an average delay of 5 minutes. The duration of the time matching window was set to twice this average delay, i.e., 10 minutes, to ensure that the possible time range of the impact of geological feature changes on operating conditions is covered.

[0047] Step S1265: Traverse each change point in the set of working condition change points. Using the time information of the change point as a reference, search for change points in the set of geological change points within the time matching window. If a geological change point exists, it is determined that the working condition change point and the geological change point are related.

[0048] Iterate through each change point in the set of working condition change points. For each change point, using its time information as a baseline, extend forward and backward by 5 minutes (i.e., the time matching window is the time of the change point ± 5 minutes). Search the set of geological change points for any geological change points within this time window. If such a geological change point exists, it is determined that the working condition change point and the geological change point are related, forming a change point pair. For example, if the time of a working condition change point is 8:20, and within its time matching window (8:15 to 8:25), there is a geological change point in the set with a time of 8:18, then these two change points are determined to be related.

[0049] Step S1266: Count the number of associated change point pairs in each geological working condition association unit, and use the number of change point pairs as the association frequency. Each change point pair contains one working condition change point and one associated geological change point.

[0050] For each geological condition associated unit, the total number of change point pairs that were determined to be associated in step S1265 is counted. This total number is the association frequency of that associated unit. For example, if 10 associated change point pairs are found in a certain geological condition associated unit through the above search process, then the association frequency of that associated unit is 10.

[0051] Step S1267: Calculate the coupling degree, which is the ratio of the correlation frequency to the total number of change points in the set of operating condition change points. If the total number of change points in the set of operating condition change points is zero, then the coupling degree is zero.

[0052] For each geological condition associated unit, its coupling degree is calculated. The formula for calculating the coupling degree is the association frequency divided by the total number of change points in the set of condition change points. For example, if the association frequency of a certain associated unit is 10 and the total number of change points in the set of condition change points is 20, then its coupling degree is 10 divided by 20, resulting in 0.5. If the total number of change points in the set of condition change points is zero, that is, the condition characteristics of the associated unit have not changed during this operation phase, then the coupling degree is directly set to zero.

[0053] Step S1268: Record the coupling degree of each geological condition associated unit in the associated unit information table. The associated unit information table also includes the associated unit identifier, the condition feature subset identifier, and the geological feature subset identifier.

[0054] Create an association unit information table. The fields of this table include association unit identifier, working condition feature subset identifier, geological feature subset identifier, and coupling degree. Assign a unique association unit identifier to each geological working condition association unit. Fill the corresponding fields in the table with the identifiers of its working condition feature subset, geological feature subset, and the calculated coupling degree, thus completing the recording of the coupling degree for each association unit.

[0055] Step S127: Filter the combination of working condition features and geological features with a coupling degree exceeding the preset coupling threshold, and mark them as coupled feature pairs. Each coupled feature pair contains a working condition feature and a corresponding geological feature.

[0056] In this shale gas horizontal well drilling scenario, the preset coupling threshold is determined based on empirical data of the correlation between geological features and drilling conditions in historical drilling operations, for example, set to 0.3. The association unit information table of all geological and drilling condition related units is traversed. For each related unit with a coupling degree exceeding 0.3, its included subsets of drilling conditions and geological features are examined. Combinations of drilling conditions and geological features that result in high coupling degrees are identified and marked as coupling feature pairs. For example, in a certain related unit, the correlation frequency between changes in shale content and changes in drill string torque is high, causing the coupling degree to exceed the threshold; therefore, the shale content feature and the drill string torque feature are marked as a coupling feature pair.

[0057] Step S128: According to the type of working condition characteristics, the coupling feature pairs are divided into motor-related coupling feature group, load-related coupling feature group and drilling fluid-related coupling feature group, and the motor-related coupling feature group, load-related coupling feature group and drilling fluid-related coupling feature group are integrated to form a geological working condition coupling feature set.

[0058] The identified coupling feature pairs are categorized according to the type of operating condition characteristics. Coupled feature pairs related to motor operation data are classified into a motor-related coupling feature group, such as the coupling pair between motor speed variation characteristics and rock formation density characteristics; coupled feature pairs related to drill string load data are classified into a load-related coupling feature group, such as the coupling pair between drill string torque fluctuation characteristics and rock formation composition ratio characteristics; coupled feature pairs related to drilling fluid flow data are classified into a drilling fluid-related coupling feature group, such as the coupling pair between drilling fluid flow stability characteristics and rock formation distribution continuity characteristics. These three coupling feature groups are integrated to form a geological operating condition coupling feature set, which comprehensively reflects the correlation characteristics between drilling geological data and drilling operating condition data.

[0059] Step S130: Input the set of coupled geological conditions features into the pre-trained electrical control parameter evolution model to generate a set of control parameters for the drilling rig electrical control system. The set of control parameters includes motor speed parameters, drill string load adjustment parameters, and drilling fluid flow rate parameters.

[0060] The above step S130 specifically includes the following sub-steps: Step S131: Perform format conversion on the geological condition coupling feature set, convert the coupling feature pairs in the geological condition coupling feature set into a feature vector format that can be accepted by the electronic control parameter evolution model, retain the correlation and feature attributes in the coupling feature pairs during the conversion process, and obtain the converted feature vector.

[0061] In this shale gas horizontal well drilling scenario, the electrical control parameter evolution model requires the input feature vector to have a specific format. First, a feature dictionary is constructed, containing all possible feature names and corresponding feature codes from the geological condition coupled feature set. The feature codes are obtained by hashing the feature names, with each feature name corresponding to a unique numerical code. For example, hashing "motor speed change feature" yields code 1001, and "shale content feature" yields code 2001, etc. Each coupled feature pair in the geological condition coupled feature set is traversed, and the condition feature name and geological feature name are extracted from each coupled feature pair. The corresponding feature code is then searched in the feature dictionary. Finally, the condition feature code and geological feature code of each coupled feature pair are combined to form a two-dimensional code array. For example, if the condition feature code is 1001 and the geological feature code is 2001, then the two-dimensional code array is [1001, 2001]. Extract the feature attribute values ​​from each coupled feature pair, such as the rate of change of motor speed and the content value of shale content, and normalize these feature attribute values ​​to the range of 0-1. Combine the normalized feature attribute values ​​with a two-dimensional encoding array to form feature tuples. Each feature tuple contains the working condition feature code, the geological feature code, and the corresponding normalized attribute value. Determine the dimension of the feature vector to be the total number of feature codes in the feature dictionary, with each dimension corresponding to one feature code. Construct an initial feature vector with all elements set to zero. Based on the feature code in each feature tuple, assign the corresponding normalized attribute value to the element of the corresponding dimension in the initial feature vector to obtain the assigned feature vector. Each feature vector corresponds to a coupled feature pair, and all feature vectors together constitute the feature vector set input to the electronic control parameter evolution model.

[0062] Step S1311: Construct a feature dictionary, which contains all possible feature names and corresponding feature codes in the geological condition coupling feature set. The feature codes are unique numerical codes obtained by performing a hash operation on the feature names.

[0063] In this shale gas horizontal well drilling scenario, the first step is to collect all possible feature names from the coupled geological condition feature set. This includes all operational feature names (such as motor speed variation characteristics, drill string torque fluctuation characteristics, drilling fluid flow stability characteristics, etc.) and geological feature names (such as shale content characteristics, sandstone content characteristics, rock density characteristics, etc.). A hash operation is performed on each feature name, for example using the MD5 hash algorithm, to convert the feature name into a fixed-length hash value. This hash value is then mapped to a unique numerical code, ensuring that each feature name corresponds to a unique feature code. The feature names and their corresponding feature codes are stored in a dictionary structure, forming a feature dictionary. This feature dictionary allows for quick lookup of the corresponding feature code using the feature name, and also allows for reverse lookup of the feature name using the feature code.

[0064] Step S1312: Traverse each coupled feature pair in the geological condition coupled feature set, extract the condition feature name and geological feature name in each coupled feature pair, and look up the corresponding feature code in the feature dictionary.

[0065] Iterate through each coupled feature pair in the set of geological working condition coupled features. For each coupled feature pair, identify the working condition feature and geological feature it contains, and extract their feature names. For example, if a coupled feature pair is "motor speed change feature - shale content feature", then the extracted working condition feature name is "motor speed change feature", and the geological feature name is "shale content feature". Then, search for the feature codes corresponding to these two feature names in the feature dictionary constructed in step S1311. Assume that the code for "motor speed change feature" is found to be 1001, and the code for "shale content feature" is found to be 2001.

[0066] Step S1313: Combine the working condition feature code and geological feature code of each coupled feature pair to form a two-dimensional code array. The first element of the two-dimensional code array is the working condition feature code, and the second element is the geological feature code.

[0067] For each coupled feature pair, the found working condition feature codes and geological feature codes are combined into an array in a fixed order. The first element of the array is the working condition feature code, and the second element is the geological feature code. For example, the two-dimensional code array of the above coupled feature pair is [1001, 2001].

[0068] Step S1314: Extract the feature attribute values ​​from each coupled feature pair, and normalize the feature attribute values. Combine the normalized feature attribute values ​​with the two-dimensional encoding array to form feature tuples. Each feature tuple contains working condition feature codes, geological feature codes, and corresponding normalized attribute values. The feature attribute values ​​include the variation range of the working condition features and the attribute parameters of the geological features.

[0069] For each coupled feature pair, the variation range of the operating condition feature and the attribute parameters of the geological feature are extracted as feature attribute values. For example, in the coupled pair of "motor speed variation feature - shale content feature", the variation range of the operating condition feature is the maximum change in motor speed within a certain time period, and the attribute parameter of the geological feature is the average content of shale within a certain depth range. The above feature attribute values ​​are normalized using the min-max normalization method, which transforms the feature attribute values ​​to the range of 0-1. The formula is that the normalized value equals (original value minus minimum value) divided by (maximum value minus minimum value), where the minimum and maximum values ​​are the value range of the feature attribute in historical data. The normalized feature attribute values ​​are combined with a two-dimensional encoding array to form a feature tuple, such as [1001, 2001, 0.65], where 0.65 is the normalized feature attribute value.

[0070] Step S1315: Determine the dimension of the feature vector, where the dimension of the feature vector is the total number of feature codes in the feature dictionary, and each dimension corresponds to one feature code.

[0071] The total number of feature codes in the feature dictionary is the dimension of the feature vector. For example, if there are 200 different feature codes in the feature dictionary, then the dimension of the feature vector is 200, and each dimension corresponds to a unique feature code.

[0072] Step S1316: Construct an initial feature vector, assign the corresponding normalized attribute values ​​to the elements of the corresponding dimension in the initial feature vector to obtain the assigned feature vector. All element values ​​of the initial feature vector are zero, based on the feature encoding in each feature tuple.

[0073] Construct an initial feature vector whose length is equal to the dimension of the feature vector, and whose initial values ​​for all elements are zero. For each feature tuple, based on the working condition feature code and geological feature code within it, find the corresponding dimension position in the initial feature vector, and assign the normalized feature attribute values ​​to the elements of these dimensions. For example, in the feature tuple [1001, 2001, 0.65], the working condition feature code 1001 corresponds to the m-th dimension of the feature vector, and the geological feature code 2001 corresponds to the n-th dimension. Therefore, set the values ​​of the m-th and n-th elements of the initial feature vector to 0.65, while keeping the values ​​of the elements in other dimensions zero, thus obtaining the assigned feature vector.

[0074] Step S1317: The output after converting the format of the assigned feature vectors, each feature vector corresponds to a coupled feature pair, and all feature vectors together constitute the feature vector set input to the electronic control parameter evolution model.

[0075] All the assigned feature vectors obtained in step S1316 are used as the output after format conversion, and each feature vector corresponds to a coupled feature pair. The above feature vectors are arranged in order of the type of coupled feature pairs to form the feature vector set input to the electronic control parameter evolution model.

[0076] Step S132: Input the transformed feature vector into the electronic control parameter evolution model for dimension alignment, perform feature enhancement processing on the dimension-aligned feature vector, establish a correspondence table between coupling degree and signal strength amplification factor through historical data statistics, divide the coupling degree into multiple continuous intervals, assign a fixed signal strength amplification factor to each interval, and implement the higher the interval where the coupling degree value is located, the larger the amplification factor value is assigned. After calling the correspondence table to determine the amplification factor of each feature vector, amplify the signal strength of the corresponding feature vector to obtain the enhanced feature vector.

[0077] In this shale gas horizontal well drilling scenario, the electrical control parameter evolution model has specific requirements for the dimensionality of the input feature vectors. Therefore, it is necessary to perform dimensional alignment processing on the transformed feature vectors to ensure that the dimensionality of all feature vectors is consistent with the model requirements. If the dimensionality of the transformed feature vectors is lower than the model requirements, it is expanded by padding with zeros; if the dimensionality is higher than the model requirements, it is processed by dimensionality reduction methods such as principal component analysis to retain the main feature information. After dimensional alignment, feature enhancement processing is performed. By analyzing the signal strength of the influence of features on electrical control parameters under different coupling degrees in the historical drilling data of this block, a table of correspondence between coupling degree and signal strength amplification factor is statistically established. The coupling degree is divided into multiple continuous intervals, such as 0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, and 0.8-1.0. A fixed signal strength amplification factor is assigned to each interval; for example, the amplification factor is 1.0 for the coupling degree in the 0-0.2 interval, 1.2 for the 0.2-0.4 interval, 1.5 for the 0.4-0.6 interval, 1.8 for the 0.6-0.8 interval, and 2.0 for the 0.8-1.0 interval. For each feature vector, the corresponding amplification factor is looked up in the correspondence table based on the coupling degree of its corresponding coupled feature pair. Then, each element value of the feature vector is multiplied by this amplification factor to amplify the signal strength, resulting in the enhanced feature vector.

[0078] Step S133: The enhanced feature vector is mapped to an initial parameter vector through a feature mapping layer. The dimension of the initial parameter vector is consistent with the number of control parameters of the drilling rig's electrical control system.

[0079] The electrical control parameter evolution model includes a feature mapping layer, which employs a fully connected neural network structure. The input is a strengthened feature vector, and the output is an initial parameter vector. In this shale gas horizontal well drilling scenario, the control parameters of the drilling rig's electrical control system include motor speed parameters, drill string load adjustment parameters, and drilling fluid flow rate parameters. Assuming there are 10 control parameters, the initial parameter vector has a dimension of 10. The number of neurons in the feature mapping layer is determined by the dimensions of the input feature vector and the output initial parameter vector. For example, if the input feature vector has a dimension of 200 and the output dimension is 10, then the weight matrix of this layer is 200×10, and the bias vector is 10×1. After the strengthened feature vector is input into the feature mapping layer, it undergoes multiplication with the weight matrix and addition with the bias vector to obtain the initial parameter vector.

[0080] Step S134: Referring to the current geological condition association unit information of the drilling operation, the initial parameter vector is processed for parameter adjustment. The adjusted parameter vector is split into multiple sub-parameter vectors. Each sub-parameter vector corresponds to a type of control parameter, including motor speed sub-parameter vector, drill string load adjustment sub-parameter vector, and drilling fluid flow rate sub-parameter vector.

[0081] Step S134 above specifically includes the following sub-steps: Step S1341: Extract the current geological condition associated unit information of the drilling operation, and extract the coupling feature pairs and corresponding coupling degrees in the current geological condition associated unit information. The current geological condition associated unit information includes the condition feature subset and the geological feature subset in the current associated unit.

[0082] In this shale gas horizontal well drilling scenario, the current progress of the drilling operation is acquired in real time, and the current geological condition associated unit is determined. All coupling feature pairs and their corresponding coupling degrees are extracted from the information table of this associated unit. For example, the current associated unit contains 5 coupling feature pairs, each with its corresponding coupling degree value.

[0083] Step S1342: Establish a correspondence table between coupling degree and parameter adjustment coefficient. The correspondence table is obtained by statistically analyzing parameter adjustment data from historical drilling operations. The actual range of parameter adjustment coefficient values ​​under different coupling degrees in the historical data is statistically analyzed. The coupling degree is divided into multiple continuous intervals, and a fixed parameter adjustment coefficient value is assigned to each coupling degree interval. The higher the value of the coupling degree interval, the larger the assigned parameter adjustment coefficient value.

[0084] Adjustment data of electrical control parameters under different geological conditions in historical drilling operations within this block were collected. This data includes coupling degree values ​​and the actual values ​​of the corresponding parameter adjustment coefficients. Statistical analysis was performed on this data, dividing the coupling degree into multiple continuous intervals, such as 0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, and 0.8-1.0. The average value of the parameter adjustment coefficient within each coupling degree interval was calculated and used as the fixed parameter adjustment coefficient value for that interval. For example, the coupling degree interval 0-0.2 corresponds to an adjustment coefficient of 1.0, 0.2-0.4 corresponds to 1.1, 0.4-0.6 corresponds to 1.2, 0.6-0.8 corresponds to 1.3, and 0.8-1.0 corresponds to 1.4, forming a correspondence table between coupling degree and parameter adjustment coefficient.

[0085] Step S1343: Based on the coupling degree of each coupled feature pair in the current associated unit, find the corresponding parameter adjustment coefficient in the corresponding relationship table, and assign an adjustment coefficient to each coupled feature pair.

[0086] For each coupled feature pair in the current associated unit, the corresponding parameter adjustment coefficient is looked up in the correspondence table established in step S1342 according to its coupling degree value. For example, if the coupling degree of a certain coupled feature pair is 0.5, which is in the range of 0.4-0.6, then a parameter adjustment coefficient of 1.2 is assigned to it.

[0087] Step S1344: Split the initial parameter vector into sub-parameter segments corresponding to the coupling feature pairs. Each sub-parameter segment corresponds to a coupling feature pair. The length of the sub-parameter segment is determined according to the number of control parameters corresponding to the coupling feature pair.

[0088] Based on the correspondence between coupling feature pairs and control parameters, the initial parameter vector is divided into multiple sub-parameter segments, each corresponding to a coupling feature pair. The length of a sub-parameter segment depends on the number of control parameters affected by that coupling feature pair. For example, if a coupling feature pair only affects the motor speed parameter, its sub-parameter segment length is 1; if another coupling feature pair affects two specific parameters in the drill string load adjustment parameters, its sub-parameter segment length is 2.

[0089] Step S1345: Adjust the parameter values ​​in each sub-parameter segment by multiplying the parameter values ​​by the corresponding parameter adjustment coefficient to obtain the adjusted sub-parameter segment.

[0090] Multiply each parameter value in each sub-parameter segment by the parameter adjustment coefficient of the corresponding coupling feature pair. For example, if the parameter value of a certain sub-parameter segment is [200, 220] and the corresponding adjustment coefficient is 1.2, then the adjusted sub-parameter segment value is [200×1.2, 220×1.2]=[240, 264], thus obtaining the adjusted sub-parameter segment.

[0091] Step S1346: Collect historical parameter adjustment records of the current drilling operation, find historical associated units similar to the current geological conditions, and determine the similarity by the overlap of coupling feature pairs.

[0092] Collect past parameter adjustment records for the current drilling operation. These records contain coupling feature pair information for each historical associated unit. Calculate the overlap between the coupling feature pairs of the current geological condition associated unit and the coupling feature pairs of each historical associated unit. The overlap is calculated as the number of common coupling feature pairs divided by the total number of coupling feature pairs in the current associated unit. Set an overlap threshold, such as 0.7, and classify historical associated units with an overlap exceeding this threshold as similar historical associated units.

[0093] Step S1347: Extract the adjusted parameter values ​​corresponding to similar historical associated units, and compare the adjusted parameter values ​​as reference parameter values ​​with the currently adjusted sub-parameter segment values.

[0094] Extract the adjusted parameter values ​​from the records of similar historically related units and use them as reference parameter values. Compare the currently adjusted values ​​of each sub-parameter segment with the corresponding reference parameter values ​​and calculate the difference between them.

[0095] Step S1348: If the difference between the current adjusted sub-parameter segment value and the reference parameter value exceeds the preset reference threshold, calculate the difference between the reference parameter value and the current adjusted sub-parameter segment value, determine the secondary adjustment amount based on the difference, and perform calculations between the current adjusted sub-parameter segment value and the secondary adjustment amount to obtain the sub-parameter segment value after secondary adjustment, so that the difference between the sub-parameter segment value after secondary adjustment and the reference parameter value is less than the preset reference threshold.

[0096] The preset reference threshold is set according to the accuracy requirements of the control parameters. For example, for the motor speed parameter, the reference threshold is set to 5 revolutions per minute. If the difference between the currently adjusted sub-parameter value and the reference parameter value exceeds this threshold, the difference between the reference parameter value and the currently adjusted sub-parameter value is calculated. This difference is used as a secondary adjustment amount. The currently adjusted sub-parameter value is then added to the secondary adjustment amount to obtain the sub-parameter value after secondary adjustment. For example, if the currently adjusted sub-parameter value is 240 and the reference parameter value is 230, the difference is -10, which exceeds the preset reference threshold of 5. Therefore, the secondary adjustment amount is -10, and the sub-parameter value after secondary adjustment is 240 + (-10) = 230. At this point, the difference between the sub-parameter value and the reference parameter value is 0, which is less than the threshold.

[0097] Step S1349: Recombine all the sub-parameter segments after secondary adjustment to form the adjusted parameter vector, and record the adjustment process and adjustment coefficient of each sub-parameter segment.

[0098] All the sub-parameter segments obtained in step S1348 after secondary adjustment are recombined in their original order to form the adjusted parameter vector. At the same time, detailed information such as the parameter adjustment coefficients and secondary adjustment amounts used in the adjustment process for each sub-parameter segment is recorded for subsequent analysis and traceability.

[0099] Step S135: Compare the values ​​in the sub-parameter vector with the hardware parameter range of the drilling rig's electrical control system. If the values ​​in the sub-parameter vector exceed the hardware parameter range, adjust the values ​​that exceed the hardware parameter range to the boundary values ​​of the hardware parameter range to obtain the adjusted sub-parameter vector. The hardware parameter range includes the maximum motor speed, the maximum load of the drilling tool, and the maximum flow rate of the drilling fluid.

[0100] In this shale gas horizontal well drilling scenario, the hardware parameter range of the drilling rig's electrical control system is preset based on the equipment's performance indicators. For example, the maximum motor speed is 3000 rpm, and the minimum speed is 500 rpm; the maximum load of the drill string is 500 kN, and the minimum load is 100 kN; the maximum flow rate of the drilling fluid is 50 L / s, and the minimum flow rate is 10 L / s. Each value in the motor speed sub-parameter vector, drill string load adjustment sub-parameter vector, and drilling fluid flow rate sub-parameter vector obtained in step S134 is compared with its corresponding hardware parameter range. If a value in a sub-parameter vector exceeds its corresponding hardware parameter range—for example, if a value in the motor speed sub-parameter vector is 3200 rpm, exceeding the maximum speed of 3000 rpm, then that value is adjusted to 3000 rpm; if a value is 400 rpm, lower than the minimum speed of 500 rpm, then it is adjusted to 500 rpm. After making the above adjustments to all sub-parameter vectors, the adjusted sub-parameter vectors are obtained.

[0101] Step S136: Integrate all the adjusted sub-parameter vectors to form a set of control parameters for the drilling rig electrical control system. Each control parameter is accompanied by a corresponding coupling feature pair identifier.

[0102] The adjusted motor speed sub-parameter vector, drill string load adjustment sub-parameter vector, and drilling fluid flow rate sub-parameter vector are integrated together according to the type of control parameter to form the control parameter set of the drilling rig electrical control system. A corresponding coupling feature pair identifier is added to each control parameter. This identifier indicates which coupling feature pair was used to adjust the control parameter. For example, the coupling feature pair identifier corresponding to a certain value in the motor speed parameter is "motor speed change feature - shale content feature".

[0103] Step S140: Transmit the set of control parameters to the execution component of the drilling rig's electrical control system, drive the execution component to perform control operations, and simultaneously collect drilling condition feedback data after the execution of control operations by the execution component. The drilling condition feedback data and the drilling condition data have the same characteristic type.

[0104] The above step S140 specifically includes the following sub-steps: Step S141: The control parameter set is encoded using the communication encoding protocol supported by the drilling rig electrical control system. The encoded control parameter set is then transmitted to the controller of the execution component, so that the controller receives the encoded stream and performs decoding processing. Based on the parameter type identifier, the decoded control parameters are allocated to the corresponding execution modules. The execution modules include a motor control module, a drill string load control module, and a drilling fluid control module.

[0105] In this shale gas horizontal well drilling scenario, the drilling rig's electrical control system supports the MODBUS protocol for communication. Each control parameter in the control parameter set is encoded according to the MODBUS protocol format, including parameter address, parameter type, and parameter value. The encoded control parameter set is transmitted to the controller of the actuator via an industrial Ethernet network. Upon receiving the encoded stream, the controller decodes it according to the MODBUS protocol, extracting the value of each control parameter and its corresponding parameter type identifier. Based on the parameter type identifier, the decoded motor speed parameter is assigned to the motor control module, the drill string load adjustment parameter to the drill string load control module, and the drilling fluid flow rate parameter to the drilling fluid control module.

[0106] Step S142: After receiving the motor speed parameters, the motor control module adjusts the output of the motor drive circuit. By changing the voltage and frequency of the drive circuit, the motor speed is adjusted so that the motor speed meets the parameter requirements.

[0107] Step S142 above specifically includes the following sub-steps: Step S1421: After receiving the motor speed parameters, the motor control module parses the target speed value and speed change rate requirement in the motor speed parameters. The target speed value is the final speed that the motor needs to reach, and the speed change rate requirement is the rate at which the motor increases or decreases from the current speed to the target speed.

[0108] After receiving the motor speed parameters, the motor control module parses the parameters and extracts the target speed value and the required rate of change of speed. For example, the motor speed parameters may be expressed as "target speed: 2000 rpm, rate of change of speed: 50 rpm²", then the target speed value is 2000 rpm, and the required rate of change of speed is 50 rpm², meaning that the change in motor speed per minute does not exceed 50 rpm.

[0109] Step S1422: The current speed value of the motor is obtained in real time by the speed sensing device installed on the motor shaft, and the current speed value is compared with the target speed value to determine the speed adjustment direction and adjustment range.

[0110] A speed sensor (such as an encoder) mounted on the motor shaft collects the current motor speed in real time and transmits this value to the motor control module. The motor control module compares the current speed value with the target speed value obtained from the analysis. If the current speed value is lower than the target speed value, the adjustment direction is to increase the speed, and the adjustment amount is the target speed value minus the current speed value; if the current speed value is higher than the target speed value, the adjustment direction is to decrease the speed, and the adjustment amount is the current speed value minus the target speed value.

[0111] Step S1423: Based on the required speed change rate and adjustment range, and combined with the correlation between motor speed and drive circuit voltage and frequency, calculate the adjustment curves of drive circuit voltage and frequency. Implement the adjustment curve as the trajectory of voltage and frequency change over time. Verify through simulation that the trajectory can make the motor speed change from the current value to the target value meet the speed change rate requirements.

[0112] Based on the required rate of change of motor speed and the adjustment range, calculate the time required for the motor to reach the target speed from the current speed. Combining the motor's speed characteristic curve—the functional relationship between motor speed and the voltage and frequency of the drive circuit (determined through factory parameters and experimental data)—calculate the trajectory of the voltage and frequency of the drive circuit over time to ensure the motor speed changes at the required rate; this is the adjustment curve. The adjustment curve is then verified using motor simulation software. The simulated rate of change of motor speed is compared with the required rate of change. If they match, the adjustment curve is confirmed; otherwise, the voltage and frequency trajectories are readjusted until the simulation results meet the requirements.

[0113] Step S1424: Convert the adjustment curve into a control signal for the drive circuit. The control signal is a pulse width modulation signal, and the output voltage and frequency are adjusted by changing the pulse width.

[0114] The voltage and frequency changes over time in the adjustment curve are converted into duty cycle changes of a pulse width modulation (PWM) signal. The duty cycle of the PWM signal is proportional to the output voltage; changing the duty cycle adjusts the output voltage. The frequency of the PWM signal directly determines the output frequency of the drive circuit. Based on the adjustment curve, the duty cycle and frequency of the PWM signal at each moment are calculated to form the control signal.

[0115] Step S1425: Send the control signal to the power amplifier unit of the drive circuit so that the power amplifier unit amplifies the control signal to enhance the driving capability of the signal and drive the stator winding of the motor.

[0116] The converted control signal is sent to the power amplifier unit of the drive circuit via an internal bus. The power amplifier unit, composed of power transistors and other devices, amplifies the power of the control signal to a level sufficient to drive the stator windings of the motor. After power amplification, the control signal is applied to the stator windings of the motor, generating a rotating magnetic field that drives the motor rotor to rotate.

[0117] Step S1426: During the adjustment process, the motor speed data is collected in real time, and the collected speed data is compared with the theoretical speed corresponding to the adjustment curve to calculate the speed deviation.

[0118] During motor speed regulation, the speed sensor continuously collects the actual motor speed data in real time and transmits it to the motor control module. The motor control module compares the collected actual speed data with the theoretical speed data corresponding to that moment on the adjustment curve, and calculates the difference between the two, i.e., the speed deviation. For example, if the theoretical speed corresponding to the adjustment curve at a certain moment is 1800 rpm, and the actual collected speed is 1780 rpm, then the speed deviation is -20 rpm.

[0119] Step S1427: Adjust the pulse width of the control signal according to the speed deviation. If the actual speed is lower than the theoretical speed, increase the pulse width to increase the output voltage and frequency; if the actual speed is higher than the theoretical speed, decrease the pulse width to decrease the output voltage and frequency.

[0120] Based on the calculated speed deviation, the pulse width of the control signal is adjusted. If the actual speed is lower than the theoretical speed, it indicates that the current output voltage and frequency are too low, requiring an increase in the pulse width to improve the output voltage and frequency of the drive circuit, thus increasing the motor speed. Conversely, if the actual speed is higher than the theoretical speed, the pulse width needs to be decreased to reduce the output voltage and frequency, causing the motor speed to decrease. The adjustment range is determined by the magnitude of the speed deviation; the larger the deviation, the greater the adjustment range.

[0121] Step S1428: Continuously adjust the control signal until the motor speed reaches the target speed value and the speed fluctuation amplitude is less than the preset fluctuation threshold. At this time, maintain the current control signal output to make the motor run stably at the target speed.

[0122] Repeat steps S1426 and S1427, continuously adjusting the control signal until the actual motor speed reaches the target speed value, and the fluctuation range near the target speed is less than the preset fluctuation threshold. The preset fluctuation threshold is set according to the stability requirements of motor operation, for example, 5 revolutions per minute. When the motor speed stabilizes at the target speed and the fluctuation range is less than the threshold, stop adjusting the control signal, maintain the current control signal output, and ensure stable motor operation.

[0123] Step S1429: Record the control signal change data, speed change data, and speed deviation data throughout the entire speed adjustment process, and store the control signal change data, speed change data, and speed deviation data in the adjustment process database.

[0124] Throughout the speed regulation process, the motor control module records real-time changes in control signal pulse width, frequency, and other data, as well as actual motor speed changes and speed deviations. This data is then organized chronologically and stored in the regulation process database for subsequent data analysis and model optimization.

[0125] Step S143: After receiving the drill string load adjustment parameters, the drill string load control module adjusts the power distribution mechanism of the drill string, and adjusts the load distribution by changing the power input ratio of each drill string, so that the drill string load meets the parameter requirements.

[0126] After receiving the drill string load adjustment parameters, the drill string load control module analyzes the requirements for load distribution to each drill string (such as drill bit, drill collar, drill pipe, etc.). The drill string power distribution mechanism consists of a hydraulic system or a mechanical transmission system, which changes the power input ratio by controlling the opening degree of the hydraulic valves or the engagement degree of the clutches corresponding to each drill string. For example, if the parameters require increasing the load ratio of the drill bit, the opening degree of the hydraulic valve in the drill bit power input channel is increased, allowing more power to be transmitted to the drill bit, thereby increasing its load; at the same time, the power input ratio of other drill strings is reduced, reducing their load. During the adjustment process, the actual load of each drill string is monitored in real time by load sensors and compared with the parameter requirements, continuously adjusting the power distribution mechanism until the load of each drill string meets the parameter requirements.

[0127] Step S144: After receiving the drilling fluid flow parameters, the drilling fluid control module adjusts the operating status of the drilling fluid pump and the opening of the outlet valve. By changing the pump speed and valve opening, the flow rate of the drilling fluid is adjusted to meet the parameter requirements.

[0128] After receiving the drilling fluid flow rate parameters, the drilling fluid control module determines the target flow rate value. The operating status of the drilling fluid pump is controlled by the speed of its motor, and the opening of the outlet valve is adjusted by the electric actuator. Based on the target flow rate value, combined with the characteristic curves of the drilling fluid pump (relationship between pump speed and flow rate) and the relationship between valve opening and flow rate, the required pump speed and valve opening are calculated. The drilling fluid pump motor speed is controlled to reach the calculated value, and the electric actuator is controlled to adjust the valve opening to the calculated value. The actual flow rate of the drilling fluid is monitored in real time by a flow sensor and compared with the target flow rate. If a deviation exists, the pump speed and valve opening are further adjusted until the actual flow rate meets the parameter requirements.

[0129] Step S145: While the execution module is performing control operations, the sensing devices distributed in key parts of the drilling equipment are activated. The sensing devices include speed sensing devices, load sensing devices, flow sensing devices, and pressure sensing devices.

[0130] While the motor control module, drill string load control module, and drilling fluid control module begin executing control operations, a start signal is sent through the control system to activate various sensing devices distributed in key parts of the drilling equipment. Speed ​​sensors are installed on the motor shaft, drill string turntable, etc., to monitor the speed of the motor and drill string; load sensors are installed at drill pipe connections, drill bits, etc., to monitor load parameters such as drill string torque and axial force; flow sensors are installed on the inlet and outlet pipes of the drilling fluid circulation system to measure the flow rate of the drilling fluid; pressure sensors are installed at the drilling fluid pump outlet, wellhead, etc., to monitor the pressure of the drilling fluid. After these sensors are activated, they begin to collect corresponding operating condition data in real time.

[0131] Step S146: For each type of control parameter, determine its sensitivity to change; for each type of control parameter, statistically analyze the sensitivity range of that type of parameter using historical data, and set a fixed sampling frequency value for each sensitivity range. The higher the sensitivity range value of that type of parameter, the larger the set sampling frequency value, forming a sensitivity-frequency correspondence table for each type of parameter; call the corresponding correspondence table to determine the sampling frequency for each control parameter.

[0132] For the three types of control parameters—motor speed, drill string load adjustment, and drilling fluid flow rate—the sensitivity of each parameter is determined. Taking motor speed as an example, by analyzing historical data on the changes in motor speed under different operating conditions, the sensitivity range is statistically analyzed. Sensitivity is defined as the amount of change in the parameter per unit time. The sensitivity is divided into three ranges: low, medium, and high. For example, the low sensitivity range is a change of less than 10 revolutions per minute, the medium sensitivity range is 10-30 revolutions per minute, and the high sensitivity range is greater than 30 revolutions per minute. A fixed sampling frequency is set for each range: a low sensitivity range corresponds to a lower sampling frequency, such as 1 time / second; a medium sensitivity range corresponds to a medium sampling frequency, such as 5 times / second; and a high sensitivity range corresponds to a higher sampling frequency, such as 10 times / second, forming a sensitivity-frequency correspondence table for motor speed parameters. Similarly, similar statistical analysis and settings are performed for drill string load adjustment and drilling fluid flow rate parameters, forming their respective sensitivity-frequency correspondence tables. During the data acquisition process, the sensitivity of each control parameter is evaluated in real time, and the corresponding correspondence table is used to determine its sampling frequency.

[0133] Step S147: Arrange the collected drilling condition data in chronological order to form drilling condition feedback data, and format the drilling condition feedback data to ensure that the data's feature type and format are consistent with the drilling condition data.

[0134] The drilling condition data (including motor speed, drill string load, drilling fluid flow rate, pressure, etc.) collected by various sensors are arranged in chronological order of collection time to form time series data. The data is then formatted, including standardizing units, adjusting data precision, and removing outliers. For example, the unit for speed data is standardized to revolutions per minute (rpm), retaining one decimal place; data significantly outside the normal range (such as outliers caused by sensor malfunctions) is removed. The format of the formatted data (such as motor speed, torque, flow rate, etc.) is kept consistent with the previously collected drilling condition data to facilitate subsequent correlation and comparison processing.

[0135] Step S150: The drilling condition feedback data and the geological condition coupling feature set are correlated and compared, the feature deviation value between the two is calculated, and the internal operating parameters of the electrical control parameter evolution model are adjusted according to the feature deviation value.

[0136] The above step S150 specifically includes the following sub-steps: For example, step S151: extract the rotation speed change characteristics, load fluctuation characteristics and flow rate stability characteristics from the drilling condition feedback data to form a set of feedback characteristics.

[0137] Feature extraction is performed on the drilling condition feedback data, using the same method as the feature extraction from the data slices in step S124. Features of motor speed variation, such as average speed, variance, and rate of change, are extracted from the feedback data; features of drill string load fluctuation, such as load fluctuation amplitude and frequency, are extracted; and features of drilling fluid flow stability, such as the standard deviation of flow rate and the duration of deviation from the set value, are extracted. These extracted features are then combined to form a feedback feature set.

[0138] Step S152: Extract the coupling feature pairs corresponding to the feedback feature set from the geological working condition coupling feature set, match the corresponding working condition features according to the type of feedback features, and then determine the corresponding coupling feature pairs to form a comparison coupling feature set.

[0139] Based on the type of each feature in the feedback feature set, the corresponding working condition feature is searched in the geological working condition coupling feature set. For example, the speed change feature in the feedback feature set corresponds to the motor speed change working condition feature in the geological working condition coupling feature set. All coupling feature pairs containing this working condition feature are found, and these coupling feature pairs are the coupling feature pairs corresponding to the feedback feature. These coupling feature pairs are then combined to form the comparison coupling feature set.

[0140] Step S153: Compare each feature in the feedback feature set with the corresponding working condition feature in the comparison coupling feature set one by one, and calculate the deviation value of a single feature. The deviation value of a single feature is the difference between the value of the feedback feature and the value of the working condition feature in the coupling feature pair.

[0141] For each feature in the feedback feature set, a numerical comparison is performed with the operating condition feature of the corresponding coupled feature pair in the comparison coupled feature set. For example, if the value of the speed change feature in the feedback feature set is A, and the value of the operating condition feature of the corresponding coupled feature pair in the comparison coupled feature set is B, then the deviation value of a single feature is AB.

[0142] Step S154: Assign a feature weight to each individual feature deviation value. The feature weight is determined based on the coupling degree of the coupled feature pair corresponding to that feature. By statistically analyzing the actual matching data between coupling degree and feature weight through historical data, the coupling degree is divided into multiple continuous intervals. A fixed feature weight value is assigned to each coupling degree interval. The higher the value of the coupling degree interval, the larger the assigned feature weight value. A coupling degree and weight correspondence table is formed. The correspondence table is called to determine the weight of each individual feature deviation value.

[0143] By analyzing the impact of feature bias on model output under different coupling degrees in historical data, the actual matching data between coupling degree and feature weights are statistically analyzed. The coupling degree is divided into multiple continuous intervals, such as 0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, and 0.8-1.0. A fixed feature weight value is assigned to each interval; for example, the weight is 0.1 for the coupling degree interval of 0-0.2, 0.2 for 0.2-0.4, 0.3 for 0.4-0.6, 0.4 for 0.6-0.8, and 0.5 for 0.8-1.0, forming a table corresponding to coupling degree and weight. For each individual feature bias value, the corresponding feature weight is looked up in the table based on the coupling degree of its corresponding coupled feature pair.

[0144] Step S155: Normalize each individual feature deviation value to convert it into a proportional deviation value relative to the normal range of that feature; then calculate the overall feature deviation value, which is the sum of the products of each proportional deviation value and the corresponding feature weight, and then divide by the sum of the feature weights to obtain the weighted average proportional deviation value.

[0145] Each individual feature deviation value is normalized by dividing it by the feature's normal range (the difference between its maximum and minimum values ​​in historical data). For example, if a single feature has a deviation value of 10 and its normal range is 50, the proportional deviation value is 10 / 50 = 0.2. To calculate the overall feature deviation value, each proportional deviation value is multiplied by its corresponding feature weight. All these products are then summed and divided by the sum of all feature weights to obtain the weighted average proportional deviation value, which is the overall feature deviation value.

[0146] Step S156: Set the deviation threshold range. The deviation threshold range is determined according to the control accuracy requirements of the drilling rig's electrical control system. Clarify the allowable deviation range corresponding to different control accuracy levels of the drilling rig's electrical control system. The higher the control accuracy level, the smaller the corresponding allowable deviation range value. Select the corresponding allowable deviation range as the deviation threshold range according to the control accuracy level required for the current operation.

[0147] Based on the design specifications of the drilling rig's electrical control system, the allowable deviation ranges corresponding to different control accuracy levels are defined. For example, control accuracy levels are divided into three categories: ordinary, higher, and high precision. The allowable deviation range for the ordinary level is ±0.2, for the higher level it is ±0.1, and for the high precision level it is ±0.05. If the current drilling operation is in the reservoir drilling stage, where higher control accuracy is required, the allowable deviation range of ±0.1 for the higher level is selected as the deviation threshold range.

[0148] Step S157: Compare the overall characteristic deviation value with the deviation threshold range. If the overall characteristic deviation value is within the deviation threshold range, it is determined that the current internal operating parameters of the electronic control parameter evolution model do not need to be adjusted.

[0149] The calculated overall characteristic deviation value is compared with the set deviation threshold range. If the overall characteristic deviation value is within the deviation threshold range (e.g., an overall characteristic deviation value of 0.08 is within ±0.1), it indicates that the control parameters generated by the current electronic control parameter evolution model can meet the control requirements, and the internal operating parameters of the model do not need to be adjusted.

[0150] Step S158: If the overall feature deviation value exceeds the deviation threshold range, determine the internal operating parameters of the model that need to be adjusted. By statistically analyzing the historical data, match the difference between the overall feature deviation value exceeding the threshold and the parameter adjustment range. Divide the difference into multiple continuous intervals and assign a fixed parameter adjustment range value to each difference interval. The larger the difference interval value, the larger the assigned adjustment range value. Form a difference-range correspondence table and call the difference-range correspondence table to determine the parameter adjustment range.

[0151] If the overall feature deviation value exceeds the deviation threshold range, calculate the difference of the excess portion (overall feature deviation value minus the upper or lower limit of the deviation threshold, taking the absolute value). By statistically analyzing historical data, the relationship between this difference and the adjustment range of the model's internal operating parameters is established. The difference is divided into multiple continuous intervals, such as 0-0.1, 0.1-0.2, 0.2-0.3, etc., and a fixed parameter adjustment range is assigned to each interval; the larger the difference, the larger the adjustment range. For example, the 0-0.1 interval corresponds to an adjustment range of 0.05, the 0.1-0.2 interval corresponds to 0.1, etc., forming a table corresponding to the difference and the adjustment range. The corresponding parameter adjustment range is then looked up in the table based on the current difference.

[0152] Step S159: Adjust the internal operating parameters of the electronic control parameter evolution model according to the determined adjustment range. The adjusted internal operating parameters include the dimension alignment coefficient in the feature adaptation process and the mapping coefficient in the parameter generation process.

[0153] The internal operating parameters of the electronic control parameter evolution model include the weight matrix and bias vector of the feature mapping layer, as well as the amplification factor correspondence in feature enhancement processing. These internal operating parameters are adjusted according to the determined parameter adjustment range. For example, if the weight matrix of the feature mapping layer needs adjustment, each element value in the weight matrix is ​​multiplied by (1 + adjustment range) or (1 - adjustment range), with the specific direction determined by the sign of the deviation; if the deviation is positive, the weight value is decreased; if the deviation is negative, the weight value is increased.

[0154] Step S1510: Substitute the adjusted internal operating parameters into the electrical control parameter evolution model, re-input the geological condition coupling feature set to generate control parameters, collect the corresponding condition feedback data and calculate the feature deviation value.

[0155] The adjusted internal operating parameters are updated in the electrical control parameter evolution model. Then, the geological condition coupling feature set is re-inputted into the model to generate a new set of control parameters. Following step S140, the new control parameters are transmitted to the execution unit to perform control operations and collect corresponding drilling condition feedback data. Steps S151 to S155 are repeated to calculate the new overall characteristic deviation value.

[0156] Step S1511: Repeat the above adjustment and verification process until the overall feature deviation value is within the deviation threshold range, and record the magnitude of each parameter adjustment and the corresponding change in deviation value.

[0157] Repeat steps S157 to S1510, continuously adjusting and verifying the model's internal operating parameters until the overall feature deviation value is within the set deviation threshold range. During each adjustment, record the magnitude of parameter adjustment, the adjusted internal operating parameter values, and the corresponding changes in the overall feature deviation value, and store these records in the model optimization record database.

[0158] Based on the same inventive concept, please refer to Figure 2 The diagram shows a schematic block diagram of an AI-integrated drilling rig electrical control adaptive control system 100 for executing the above-described AI-integrated drilling rig electrical control adaptive control method, provided in an embodiment of this application. The AI-integrated drilling rig electrical control adaptive control system 100 may include a communication unit 110, a machine-readable storage medium 120, and a processor 130.

[0159] In this embodiment, both the machine-readable storage medium 120 and the processor 130 are located within the AI-integrated drilling rig electrical control adaptive control system 100 and are separately configured. However, it should be understood that the machine-readable storage medium 120 may also be independent of the AI-integrated drilling rig electrical control adaptive control system 100 and may be accessed by the processor 130 via a bus interface. Alternatively, the machine-readable storage medium 120 may be integrated into the processor 130 and may communicate with external systems via the communication unit 110.

[0160] The processor 130 is the control center of the AI-integrated drilling rig electrical adaptive control system 100. It connects various parts of the AI-integrated drilling rig electrical adaptive control system 100 via various interfaces and lines. By running or executing software programs and / or modules stored in the machine-readable storage medium 120, and by calling data stored in the machine-readable storage medium 120, it performs various functions and processes data of the AI-integrated drilling rig electrical adaptive control system 100, thereby providing overall monitoring of the AI-integrated drilling rig electrical adaptive control system 100. Optionally, the processor 130 may include one or more processing cores; for example, the processor 130 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor. The machine-readable storage medium 120 is used to store machine-executable instructions for executing the scheme of this application, and the processor 130 is used to execute the machine-executable instructions stored in the machine-readable storage medium 120 to implement the AI-integrated drilling rig electrical adaptive control method provided in the foregoing method embodiments.

[0161] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A drilling rig electrically controlled adaptive control method combined with artificial intelligence, characterized in that, The method includes: Collect drilling geological data and drilling condition data during the drilling operation. The drilling geological data includes rock formation composition data, rock formation structure data, and rock formation distribution data. The drilling condition data includes motor operation data, drill string load data, and drilling fluid flow data. The drilling geological data and the drilling operating condition data are subjected to coupling feature extraction processing to identify the correlation between the drilling geological data and the drilling operating condition data, and a geological operating condition coupling feature set is generated, which contains feature combinations corresponding to the correlation. The set of coupled geological conditions is input into a pre-trained electrical control parameter evolution model to generate a set of control parameters for the drilling rig electrical control system. The set of control parameters includes motor speed parameters, drill string load adjustment parameters, and drilling fluid flow rate parameters. The set of control parameters is transmitted to the execution component of the drilling rig's electrical control system, driving the execution component to perform control operations. At the same time, drilling condition feedback data after the execution of control operations by the execution component is collected. The characteristic type of the drilling condition feedback data is consistent with that of the drilling condition data. The drilling condition feedback data and the geological condition coupling feature set are correlated and compared to calculate the feature deviation value between the two, and the internal operating parameters of the electrical control parameter evolution model are adjusted according to the feature deviation value. The process of performing coupled feature extraction on the drilling geological data and the drilling operating condition data identifies the correlation between the drilling geological data and the drilling operating condition data, and generates a set of geological and operating condition coupled features, including: The drilling data is processed into time-series slices according to the order of collection time to obtain multiple data slices. A time identifier is added to each data slice. Each data slice contains motor operation data, drill string load data and drilling fluid flow data within the corresponding time period. The drilling geological data is spatially partitioned according to the depth range of the drilling operation to obtain multiple geological data partitions. Each geological data partition contains rock layer composition data, rock layer structure data and rock layer distribution data within the corresponding depth range. A depth identifier is added to each geological data partition. Establish the correspondence between time markers and depth markers. Based on the progress record of drilling operations, associate the working condition data slices and geological data partitions of the same operation stage to form multiple geological working condition association units. Each geological working condition association unit contains a set of working condition data slices and a set of geological data partitions. Feature extraction is performed on the working condition data slices in each geological working condition associated unit. The features of speed change in motor operation data, load fluctuation in drill string load data, and flow stability in drilling fluid flow data are extracted to form a subset of working condition features. Feature extraction is performed on the geological data partitions in each geological condition association unit. The feature extraction includes the component proportion feature in the rock stratum composition data, the structural compactness feature in the rock stratum structure data, and the distribution continuity feature in the rock stratum distribution data, forming a subset of geological features. Calculate the coupling degree between the working condition feature subset and the geological feature subset in each geological working condition association unit. The coupling degree is determined by the association frequency between the working condition feature subset and the geological feature subset. The association frequency is the number of times the two features change simultaneously within the same time period. The combination of working condition features and geological features with a coupling degree exceeding a preset coupling threshold is selected and marked as a coupled feature pair. Each coupled feature pair contains a working condition feature and a corresponding geological feature. According to the type of working condition characteristics, the coupling feature pairs are divided into motor-related coupling feature group, load-related coupling feature group and drilling fluid-related coupling feature group. The motor-related coupling feature group, load-related coupling feature group and drilling fluid-related coupling feature group are integrated to form a geological working condition coupling feature set. The drilling data is processed into time-series slices according to the chronological order of acquisition time, resulting in multiple data slices. A time identifier is added to each data slice, including: The time interval for the time slice is determined based on the process duration of the drilling operation. The process duration is obtained by statistically analyzing the process records of historical drilling operations. Time intervals that match the process duration are selected so that each time interval contains a complete drilling process. Extract the acquisition time information from the drilling condition data, arrange all data according to the chronological order of the acquisition time information to form an ordered drilling condition data sequence, wherein each data point in the drilling condition data sequence is accompanied by acquisition time information; Using a defined time interval as a unit, the ordered drilling condition data sequence is divided. Starting from the first data point, all data points within the first time interval are extracted as the first condition data slice, and data points within subsequent time intervals are extracted in turn to form multiple condition data slices. The amount of missing data in motor operation data, drill string load data, and drilling fluid flow data within each slice is counted. If the amount of missing data exceeds the preset missing threshold, the time interval is readjusted and the ordered drilling condition data sequence is re-processed into time-series slices. Add a time stamp to each data slice that passes the integrity check. The time stamp consists of the start and end times of the data acquisition. The start time is the acquisition time of the first data point in the slice, and the end time is the acquisition time of the last data point in the slice. Each working condition data slice is bound to its corresponding time identifier to form a working condition data slice unit containing data content and time identifier, and the drilling process name corresponding to each slice unit is recorded at the same time. All operating condition data slice units are sorted according to the chronological order of their time identifiers to form an ordered sequence of operating condition data slices. The calculation of the coupling degree between the subset of working condition features and the subset of geological features in each geological working condition association unit includes: Change point detection is performed on the subset of working condition features in each geological working condition associated unit. For each feature in the subset of working condition features, the time point when the feature value changes is identified. The change point is the time point when the difference between the feature value and the value at the previous moment exceeds the preset change threshold. Change point detection is performed on the subset of geological features in each geological condition association unit to identify the time point when the value of each geological feature changes. All change points of working condition characteristics are collected to form a set of working condition change points, and all change points of geological characteristics are collected to form a set of geological change points. The change points in the set of working condition change points and the set of geological change points are accompanied by corresponding time information. A time matching window is set, the duration of which is determined based on the delay in the impact of geological features on working condition features during drilling operations. The impact delay is obtained by statistically analyzing the time difference of feature changes in historical data. Iterate through each change point in the set of working condition change points, and use the time information of the change point as a reference to search for change points in the set of geological change points within the time matching window. If a geological change point exists, it is determined that the working condition change point and the geological change point are related. The number of associated change point pairs in each geological working condition association unit is counted, and this number of change point pairs is used as the association frequency. Each change point pair contains one working condition change point and one associated geological change point. Calculate the coupling degree, which is the ratio of the correlation frequency to the total number of change points in the set of operating condition change points. If the total number of change points in the set of operating condition change points is zero, then the coupling degree is zero. The coupling degree of each geological condition associated unit is recorded in the associated unit information table, which also includes the associated unit identifier, the condition feature subset identifier, and the geological feature subset identifier.

2. The electrically controlled self-adaptive control method of a drilling rig combined with artificial intelligence according to claim 1, characterized in that, The step of inputting the set of coupled geological conditions features into a pre-trained electrical control parameter evolution model to generate a set of control parameters for the drilling rig electrical control system includes: The geological condition coupling feature set is converted into a format that can be accepted by the electronic control parameter evolution model. The correlation and feature attributes in the coupling feature pair are preserved during the conversion process to obtain the converted feature vector. The transformed feature vectors are input into the electronic control parameter evolution model for dimension alignment. The dimension-aligned feature vectors are then subjected to feature enhancement processing. A correspondence table between coupling degree and signal strength amplification factor is established through historical data statistics. The coupling degree is divided into multiple continuous intervals, and a fixed signal strength amplification factor is assigned to each interval. The higher the interval in which the coupling degree value is located, the larger the amplification factor value is assigned. The amplification factor of each feature vector is determined by calling the correspondence table, and the signal strength of the corresponding feature vector is amplified to obtain the enhanced feature vector. The enhanced feature vector is mapped to an initial parameter vector through a feature mapping layer. The dimension of the initial parameter vector is consistent with the number of control parameters of the drilling rig's electrical control system. Referring to the current geological conditions associated with the drilling operation, the initial parameter vector is adjusted, and the adjusted parameter vector is split into multiple sub-parameter vectors. Each sub-parameter vector corresponds to a type of control parameter, including motor speed sub-parameter vector, drill string load adjustment sub-parameter vector, and drilling fluid flow rate sub-parameter vector. The values ​​in the sub-parameter vector are compared with the range of hardware parameters of the drilling rig's electrical control system. If the values ​​in the sub-parameter vector exceed the range of hardware parameters, the values ​​exceeding the range are adjusted to the boundary values ​​of the range to obtain the adjusted sub-parameter vector. The range of hardware parameters includes the maximum speed of the motor, the maximum load of the drilling tool, and the maximum flow rate of the drilling fluid. All the adjusted sub-parameter vectors are integrated to form a set of control parameters for the drilling rig electrical control system. Each control parameter is accompanied by a corresponding coupling feature pair identifier.

3. The adaptive control method for drilling rig electrical control combined with artificial intelligence according to claim 2, characterized in that, The process involves format conversion of the geological condition coupled feature set, transforming the coupled feature pairs in the set into a feature vector format acceptable to the electronic control parameter evolution model. During the conversion, the correlation and feature attributes within the coupled feature pairs are preserved, resulting in the converted feature vector, including: Construct a feature dictionary, which contains all possible feature names and corresponding feature codes in the geological condition coupling feature set. The feature codes are unique numerical codes obtained by performing a hash operation on the feature names. Traverse each coupled feature pair in the set of geological working conditions coupled features, extract the working condition feature name and geological feature name in each coupled feature pair, and look up the corresponding feature code in the feature dictionary; The working condition feature code and geological feature code of each coupled feature pair are combined to form a two-dimensional code array. The first element of the two-dimensional code array is the working condition feature code, and the second element is the geological feature code. Extract the feature attribute values ​​from each coupled feature pair and normalize the feature attribute values. Combine the normalized feature attribute values ​​with a two-dimensional encoding array to form a feature tuple. Each feature tuple contains the working condition feature code, the geological feature code, and the corresponding normalized attribute value. The feature attribute value includes the variation range of the working condition feature and the attribute parameters of the geological feature. The dimension of the feature vector is determined, which is the total number of feature codes in the feature dictionary, with each dimension corresponding to one feature code; Construct an initial feature vector, and assign the corresponding normalized attribute values ​​to the elements of the corresponding dimension in the initial feature vector to obtain the assigned feature vector. All element values ​​of the initial feature vector are zero, based on the feature encoding in each feature tuple. The output after the format conversion of the assigned feature vectors is such that each feature vector corresponds to a coupled feature pair, and all feature vectors together constitute the feature vector set input to the electronic control parameter evolution model.

4. The adaptive control method for drilling rig electrical control combined with artificial intelligence according to claim 2, characterized in that, The reference drilling operation's current geological condition association unit information is used to adjust the initial parameter vector, splitting the adjusted parameter vector into multiple sub-parameter vectors, including: Extract the current geological condition associated unit information of the drilling operation, and extract the coupling feature pairs and corresponding coupling degrees in the current geological condition associated unit information. The current geological condition associated unit information includes the condition feature subset and the geological feature subset in the current associated unit. A correspondence table between coupling degree and parameter adjustment coefficient is established. The correspondence table is obtained by statistically analyzing parameter adjustment data in historical drilling operations. The actual value range of parameter adjustment coefficient under different coupling degrees in the historical data is statistically analyzed. The coupling degree is divided into multiple continuous intervals, and a fixed parameter adjustment coefficient value is assigned to each coupling degree interval. The higher the value of the coupling degree interval, the larger the assigned parameter adjustment coefficient value. Based on the coupling degree of each coupling feature pair in the current associated unit, the corresponding parameter adjustment coefficient is found in the corresponding relationship table, and an adjustment coefficient is assigned to each coupling feature pair; The initial parameter vector is split into sub-parameter segments corresponding to the coupling feature pairs. Each sub-parameter segment corresponds to a coupling feature pair. The length of the sub-parameter segment is determined according to the number of control parameters corresponding to the coupling feature pair. The parameter values ​​in each sub-parameter segment are adjusted by multiplying the parameter value by the corresponding parameter adjustment coefficient to obtain the adjusted sub-parameter segment. Collect historical parameter adjustment records of the current drilling operation, find historical related units similar to the current geological conditions, and determine the similarity by the overlap of coupled feature pairs; Extract the adjusted parameter values ​​corresponding to similar historical associated units, and use the adjusted parameter values ​​as reference parameter values ​​to compare with the currently adjusted sub-parameter segment values; If the difference between the current adjusted sub-parameter segment value and the reference parameter value exceeds the preset reference threshold, calculate the difference between the reference parameter value and the current adjusted sub-parameter segment value, determine the secondary adjustment amount based on the difference, and perform calculations between the current adjusted sub-parameter segment value and the secondary adjustment amount to obtain the sub-parameter segment value after secondary adjustment, so that the difference between the sub-parameter segment value after secondary adjustment and the reference parameter value is less than the preset reference threshold. All the sub-parameter segments after secondary adjustment are recombined to form the adjusted parameter vector, and the adjustment process and adjustment coefficient of each sub-parameter segment are recorded.

5. The drilling rig electrical control adaptive control method combining artificial intelligence according to claim 1, characterized in that, The process of transmitting the set of control parameters to the actuators of the drilling rig's electrical control system, driving the actuators to perform control operations, and simultaneously collecting drilling condition feedback data after the actuators perform control operations includes: The control parameter set is encoded using a communication encoding protocol supported by the drilling rig electrical control system. The encoded control parameter set is then transmitted to the controller of the execution component, so that the controller receives the encoded stream, performs decoding processing, and allocates the decoded control parameters to the corresponding execution modules according to the parameter type identifier. The execution modules include a motor control module, a drill string load control module, and a drilling fluid control module. The motor control module receives the motor speed parameters and adjusts the output of the motor drive circuit. By changing the voltage and frequency of the drive circuit, the motor speed is adjusted so that the motor speed meets the parameter requirements. After receiving the drill load adjustment parameters, the drill load control module adjusts the power distribution mechanism of the drill bit, and adjusts the load distribution by changing the power input ratio of each drill bit, so that the drill load meets the parameter requirements. After receiving the drilling fluid flow parameters, the drilling fluid control module adjusts the operating status of the drilling fluid pump and the opening of the outlet valve. By changing the pump speed and valve opening, the drilling fluid flow rate is adjusted to meet the parameter requirements. While the execution module performs control operations, it activates the sensing devices distributed in key parts of the drilling equipment, including speed sensing devices, load sensing devices, flow sensing devices, and pressure sensing devices. For each type of control parameter, its sensitivity to change is determined. For each type of control parameter, the sensitivity range of change for that type of parameter is statistically analyzed using historical data. A fixed sampling frequency value is set for each sensitivity range. The higher the sensitivity range value of that type of parameter, the larger the set sampling frequency value, forming a sensitivity-frequency correspondence table for each type of parameter. The corresponding correspondence table is then called to determine the sampling frequency for each control parameter. The collected drilling condition data are arranged in chronological order to form drilling condition feedback data. The drilling condition feedback data is then formatted to ensure that the data characteristics and format are consistent with the drilling condition data.

6. The drilling rig electrical control adaptive control method combining artificial intelligence according to claim 5, characterized in that, After receiving the motor speed parameters, the motor control module adjusts the output of the motor drive circuit. By changing the voltage and frequency of the drive circuit, the motor speed is adjusted to meet the parameter requirements, including: After receiving the motor speed parameters, the motor control module parses the target speed value and speed change rate requirement in the motor speed parameters. The target speed value is the final speed that the motor needs to reach, and the speed change rate requirement is the rate at which the motor increases or decreases from the current speed to the target speed. The current speed of the motor is obtained in real time by a speed sensor installed on the motor shaft. The current speed value is compared with the target speed value to determine the direction and magnitude of speed adjustment. Based on the required rate of change of speed and the adjustment range, and combined with the correlation between motor speed and the voltage and frequency of the drive circuit, the adjustment curves of the voltage and frequency of the drive circuit are calculated. The adjustment curve is the trajectory of voltage and frequency changing over time. Simulation verifies that this trajectory can make the rate of change of motor speed from the current value to the target value meet the requirements of the rate of change of speed. The adjustment curve is converted into a control signal for the drive circuit. The control signal is a pulse width modulation signal, and the output voltage and frequency are adjusted by changing the pulse width. The control signal is sent to the power amplifier unit of the drive circuit so that the power amplifier unit amplifies the control signal to enhance the driving capability of the signal and drive the stator winding of the motor. During the adjustment process, the motor speed data is collected in real time, and the collected speed data is compared with the theoretical speed corresponding to the adjustment curve to calculate the speed deviation. The pulse width of the control signal is adjusted according to the speed deviation. If the actual speed is lower than the theoretical speed, the pulse width is increased to increase the output voltage and frequency; if the actual speed is higher than the theoretical speed, the pulse width is decreased to decrease the output voltage and frequency. Continuously adjust the control signal until the motor speed reaches the target speed value and the speed fluctuation is less than the preset fluctuation threshold. At this time, maintain the current control signal output to make the motor run stably at the target speed. Record the control signal change data, speed change data, and speed deviation data throughout the entire speed regulation process, and store the control signal change data, speed change data, and speed deviation data in the regulation process database.

7. A drilling rig electrical control adaptive control system incorporating artificial intelligence, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the drilling rig electrical control adaptive control method incorporating artificial intelligence as described in any one of claims 1 to 6 by executing the machine-executable instructions.

Images