A method and system for multi-parameter regulation of a drilling fluid equipment
Patent Information
- Application Number
- CN202610922759.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-25
- Publication Date
- 2026-10-30
AI Technical Summary
[0003]针对现有技术中所存在的不足，本发明提供了一种钻井液设备多参数调控方法，其解决了现有技术中存在的调控滞后以及调控周期长的问题
本发明钻井液机理数据中多项参数转换成矩阵形式的性能参数向量，通过性能标定矩阵来量化各项参数之间的影响程度，根据当前性能参数向量和性能标定矩阵来生成对故障进行调控的最优预测参数矩阵，在控制钻井液设备以最优预测参数矩阵进行执行，多参数的共同调控，避免各类参数之间出现“循环震荡”的问题，同时根据获取执行参数对性能标定矩阵进行实时优化，使钻井液能够实时进行故障调控。
Smart Images
Figure CN123013623A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of oil drilling technology, and in particular to a method and system for multi-parameter control of drilling fluid equipment. Background Technology
[0002] In oil and gas drilling operations, drilling fluid is the "blood" that ensures drilling safety. Precise control of its performance is crucial for preventing downhole accidents, improving drilling efficiency, and reducing operating costs. However, current drilling fluid performance control often employs an open-loop, manual, single-parameter control model. Because downhole accidents develop extremely rapidly, by the time the early warning system detects that a monitored parameter is about to exceed its limit and issues a warning, and the on-site engineer manually formulates a control plan, the accident has often already occurred. This control model is severely lagging, only allowing for post-accident remediation, often missing the optimal window for intervention. Furthermore, when controlling an accident, adjustments are made only to a single parameter exceeding its limit, completely ignoring the coupling effects between parameters. This frequently leads to "cyclical oscillations" between various parameters, resulting in long control cycles and severely delaying the drilling cycle. Summary of the Invention
[0003] To address the shortcomings of existing technologies, this invention provides a multi-parameter control method for drilling fluid equipment, which solves the problems of control lag and long control cycles in existing technologies.
[0004] According to an embodiment of the present invention, a method for multi-parameter control of drilling fluid equipment includes: S1: Obtain drilling fluid mechanism data, generate performance parameter vectors and performance calibration matrices based on drilling fluid mechanism data, and then use performance parameter vectors and performance coupling matrices to generate prediction parameter matrices; S2: Construct a parameter optimization function based on drilling fluid mechanism data, and use the NSGA-III algorithm to optimize the predicted parameter matrix with the goal of minimizing the parameter optimization function, so as to obtain the optimal predicted parameter matrix; S3: Control the drilling fluid equipment to execute with the optimal prediction parameter matrix and obtain the execution parameters. Based on the execution parameters, use the LMS algorithm to optimize the performance parameter vector. Then repeat steps S2-S3 until the execution parameters meet the preset requirements.
[0005] Preferably, the performance parameter vector includes a control performance vector and a control action vector, and the performance calibration matrix includes a performance coupling matrix and an effect coupling matrix; The dot product of the regulation performance vector and the performance coupling matrix is used to obtain the prediction performance matrix. The dot product of the regulation action vector and the effect coupling matrix is used to obtain the prediction effect matrix. The prediction performance matrix and the prediction effect matrix are collectively referred to as the prediction parameter matrix.
[0006] Preferably, the parameter optimization function is as follows:
[0007]
[0008]
[0009]
[0010] in, The weight coefficient for the i-th parameter; The actual value of the parameter. For the target value of the parameter, Let j be the unit cost of the control action. For the amount of movement, The rate of change of the parameter.
[0011] Preferably, the method for optimizing the prediction parameter matrix using the NSGA-III algorithm to obtain the optimal prediction parameter matrix includes: A1: Initialize the population and the number of iterations. The population includes multiple parent individuals. Each parent individual is assigned random state parameters according to the prediction parameter matrix. A2: Calculate the individual fitness of each parent individual based on the parameter optimization function and the state parameters of each parent individual; A3: Sort all parent individuals in a non-dominated order based on their fitness, and select the top N parent individuals for crossbreeding to generate N offspring individuals; A4: Sort all parent and child individuals in a non-dominated order, and use the reference point screening method to select the top K individuals as elite individuals; A5: Repeat steps A2-A5 with the managed individual as the new parent individual until the maximum number of iterations is reached, and take all elite individuals obtained in the last iteration as the optimal solution set; A6: Based on the fuzzy membership function, select the elite individual with the lowest parameter optimization function from all elite individuals obtained from the last iteration, and use the parameter state parameters corresponding to this elite individual as the optimal prediction parameter matrix.
[0012] Preferably, the method for optimizing the prediction parameter matrix using the NSGA-III algorithm to obtain the optimal prediction parameter matrix includes: A1: Initialize the population and the number of iterations. The population includes multiple parent individuals. Each parent individual is assigned random state parameters according to the prediction parameter matrix. A2: Calculate the individual fitness of each parent individual based on the parameter optimization function and the state parameters of each parent individual; A3: Sort all parent individuals in a non-dominated order based on their fitness, and select the top N parent individuals for crossbreeding to generate N offspring individuals; A4: Sort all parent and child individuals in a non-dominated order, and use the reference point screening method to select the top K individuals as elite individuals; A5: Repeat steps A2-A5 with the managed individual as the new parent individual until the maximum number of iterations is reached, and take all elite individuals obtained in the last iteration as the optimal solution set; A6: Based on the fuzzy membership function, select the elite individual with the lowest parameter optimization function from all elite individuals obtained from the last iteration, and use the parameter state parameters corresponding to this elite individual as the optimal prediction parameter matrix.
[0013] Preferably, the formula for calculating the regulation rate is as follows:
[0014]
[0015] in, The optimal control activation time (time before the accident occurs); The timeframe for risk prediction output by the advanced early warning system; For the control and response time of on-site equipment; This is a safety buffer period.
[0016] Preferably, the method for optimizing the performance parameter vector using the LMS algorithm based on the execution parameters includes: Calculate the control effect deviation vector based on the execution parameters and the prediction performance matrix; Based on the control deviation vector, incremental optimization is performed on the control performance vector and the control action vector.
[0017] On the other hand, according to embodiments of the present invention, a multi-parameter control system for drilling fluid equipment is also provided. This system uses the above-described multi-parameter control method for drilling fluid equipment, including: The data acquisition module is used to acquire drilling fluid mechanism data; The parameter coupling module is used to generate a performance parameter vector and a performance calibration matrix based on drilling fluid mechanism data, and then use the performance parameter vector and the performance coupling matrix to generate a prediction parameter matrix. The performance prediction module, the parameter optimization module is used to construct a parameter optimization function based on drilling fluid mechanism data, and with the goal of minimizing the parameter optimization function, use the NSGA-III algorithm to optimize the prediction parameter matrix to obtain the optimal prediction parameter matrix; The parameter optimization module is used to control the drilling fluid equipment to execute with the optimal prediction parameter matrix, obtain the execution parameters, optimize the performance parameter vector using the LMS algorithm based on the execution parameters, and then repeat steps S2-S3 until the execution parameters meet the preset requirements.
[0018] Compared with the prior art, the present invention has the following beneficial effects: This invention converts multiple parameters in drilling fluid mechanism data into a matrix-form performance parameter vector. The influence between various parameters is quantified through a performance calibration matrix. Based on the current performance parameter vector and the performance calibration matrix, an optimal prediction parameter matrix for fault control is generated. The drilling fluid equipment is controlled to execute according to the optimal prediction parameter matrix. The joint control of multiple parameters avoids the problem of "cyclic oscillation" between various parameters. At the same time, the performance calibration matrix is optimized in real time based on the acquired execution parameters, enabling the drilling fluid to perform fault control in real time. Attached Figure Description
[0019] Figure 1 This is a diagram illustrating a multi-parameter control method according to an embodiment of the present invention. Detailed Implementation
[0020] The technical solutions of the present invention will be further described below with reference to the accompanying drawings and embodiments.
[0021] like Figure 1 As shown in the figure, this invention proposes a method for multi-parameter control of drilling fluid equipment, including: S1: Obtain drilling fluid mechanism data, generate performance parameter vectors and performance calibration matrices based on drilling fluid mechanism data, and then use performance parameter vectors and performance coupling matrices to generate prediction parameter matrices; Two types of basic data were extracted from the obtained drilling fluid mechanism data: (1) Regulation of performance vector The regulation performance vector is as follows:
[0022] Where η is the apparent viscosity, ρ is the density, pH is the acidity / alkalinity, and G is the gel shear force; this vector is a 4-dimensional column vector, all of which are core performance parameters that can be directly controlled by reagents and circulation processes, directly determining the wall protection, rock carrying, and lubrication performance of the drilling fluid.
[0023] (2) Adjusting the action vector The control action vector is as follows:
[0024] in To increase the amount of material added, This refers to the dosage of viscosity reducer. This refers to the amount of pH adjuster added. To reduce the amount of filtrate added, This represents the amount of flocculant added; this vector is a 5-dimensional column vector, covering all commonly used additives for on-site drilling fluid performance control.
[0025] Based on drilling fluid mechanism data, a 4×4 performance coupling matrix M is constructed to accurately quantify the degree of coupling influence between any two performance parameters.
[0026] The standard calibration values for the performance coupling matrix M of conventional water-based drilling fluids are as follows:
[0027] Matrix elements Represents the synchronous change rate of the i-th parameter when the j-th parameter changes by 1%. The diagonal elements are 1 (the parameter itself). Positive values represent positive correlation, negative values represent negative correlation, and the larger the absolute value, the stronger the coupling effect.
[0028] For example, in matrix M 12 =0.28, which means that for every 1% increase in density, the viscosity will increase by 0.28% simultaneously. When adjusting the density, the coupling effect on viscosity can be predicted in advance, and the viscosity-reducing agent can be used to offset it in advance, avoiding the problem of "adjusting the density but exceeding the viscosity limit", thus solving the core pain point of the existing technology from the root.
[0029] Next, the performance coupling matrix and the regulation performance vector are multiplied by a dot product to obtain the prediction performance matrix:
[0030] in This is the initial performance parameter vector (4×1 column vector) before adjustment. The prediction vector (4×1 column vector).
[0031] For actual on-site control scenarios, a 4×5 dimension effect coupling matrix B is constructed to accurately quantify the impact of different agent injection volumes on drilling fluid performance parameters. This allows for early prediction of control effects and avoids overshoot and parameter conflicts, forming the core foundation of multi-objective optimization algorithms.
[0032] The standard calibration values for the effect coupling matrix B of conventional water-based drilling fluids are as follows:
[0033] Matrix elements Representative: When 1 unit of the j-th agent is added, the change in the i-th performance parameter is determined by indoor formulation experiments and field drilling data, and can be dynamically adjusted according to different drilling fluid systems.
[0034] Next, the dot product of the control action vector and the effect coupling matrix is performed to obtain the prediction effect matrix:
[0035] in To control the action vector (5×1 column vector).
[0036] S2: Construct a parameter optimization function based on drilling fluid mechanism data, and use the NSGA-III algorithm to optimize the predicted parameter matrix with the goal of minimizing the parameter optimization function, so as to obtain the optimal predicted parameter matrix; To address the core requirements of drilling fluid control, three core optimization objectives are constructed, combined into a parametric optimization function, and all objectives must be minimized simultaneously: Parameter optimization function:
[0037] The three core optimization objectives are as follows: (1) Minimize control deviation
[0038] in, The weighting coefficient for the i-th parameter is determined by the control priority. For the actual value of the parameter, The target value for the parameter.
[0039] (2) Minimize regulation costs
[0040] in, Let j be the unit cost of the control action. This refers to the amount of motion.
[0041] (3) Minimize fluctuations in regulation
[0042] in, The rate of change of the parameter is used to achieve stable control and avoid secondary risks caused by large fluctuations.
[0043] Hard constraints based on drilling engineering mechanisms are incorporated into the optimization algorithm to ensure that the solved solution meets on-site safety requirements and avoids results that do not conform to engineering logic. The specific constraints are as follows: (1) Safety range constraint: Xi_min≤Xi≤Xi_max (All parameters must be within the safety range specified in the industry standard SY / T 5374-2018) (2) Coupling logic constraints: Based on the coupling matrix M, ensure that parameter changes conform to the physical and chemical mechanism of drilling fluid. (3) Execution capacity constraint: Uj_min≤Uj≤Uj_max (the amount of drug added and the range of equipment adjustment must be within the execution capacity of the on-site equipment) (4) Timing constraints: Based on the advance warning time, ensure that the control action is completed before the risk occurs. The advance warning time is obtained by the advance warning device configured in the drilling fluid equipment. The advance warning device can determine whether the monitoring parameters are about to exceed the limit. When they are about to exceed the limit, it will issue an early warning and predict the time when the limit will be exceeded.
[0044] Then, with the goal of minimizing the parameter optimization function, the NSGA-III algorithm is used to optimize the parameters of the prediction parameter matrix. The specific process is as follows: A1: Initialize the population and the number of iterations. The population includes multiple parent individuals. Each parent individual is assigned random state parameters according to the prediction parameter matrix. A2: Calculate the individual fitness of each parent individual based on the parameter optimization function and the state parameters of each parent individual; A3: Sort all parent individuals in a non-dominated order based on their fitness, and select the top N parent individuals for crossbreeding to generate N offspring individuals; A4: Sort all parent and child individuals in a non-dominated order, and use the reference point screening method to select the top K individuals as elite individuals; A5: Repeat steps A2-A5 with the managed individual as the new parent individual until the maximum number of iterations is reached, and take all elite individuals obtained in the last iteration as the optimal solution set; A6: Based on the fuzzy membership function, select the elite individual with the lowest parameter optimization function from all elite individuals obtained from the last iteration, and use the parameter state parameters corresponding to this elite individual as the optimal prediction parameter matrix.
[0045] If the obtained optimal prediction parameter matrix satisfies all constraints, output the optimal prediction parameter matrix; otherwise, repeat steps A1-A6.
[0046] S3: Control the drilling fluid equipment to execute with the optimal prediction parameter matrix and obtain the execution parameters. Based on the execution parameters, use the LMS algorithm to optimize the performance parameter vector. Then repeat steps S2-S3 until the execution parameters meet the preset requirements.
[0047] To control the drilling fluid equipment to adjust from its current state to the parameters in the optimal prediction parameter matrix and execute the adjustment, in order to avoid large fluctuations in parameters due to excessively fast adjustment rate, which could cause short-term overload of the drilling fluid equipment, this invention sets a stable adjustment rate.
[0048] (1) Regulation time window
[0049] in The optimal time to initiate control measures; The estimated time when the parameters output by the early warning system for drilling fluid exceed the limit; For the control and response time of on-site equipment; For safety buffer time, a fixed 30 minutes is set.
[0050] Utilize the control time window to ensure that all control actions are completed before the risk occurs, while reserving a safety buffer time to completely eliminate the risk.
[0051] (2) Stable regulation rate To avoid large parameter fluctuations caused by excessively fast control rates, the optimal stable control rate is calculated using the following formula:
[0052] in, The optimal control rate for the parameter; The target value for the parameter; This is the current value of the parameter.
[0053] Then, based on the control rate, the drilling fluid equipment is gradually adjusted from its current state to the optimal predicted parameter matrix in an incremental manner.
[0054] The drilling fluid equipment is controlled to execute with the optimal predicted parameter matrix, and the execution parameters at this time are obtained. Due to model errors and the influence of various environmental factors, the standard calibrated performance coupling matrix M and effect coupling matrix B are not applicable to all situations. Therefore, the drilling fluid equipment can not perfectly handle the fault when it executes with the optimal predicted parameters (and cannot fully meet the preset requirements). Therefore, it is necessary to further adjust the performance coupling matrix and effect coupling matrix according to the working environment of the drilling fluid equipment, so that the drilling fluid equipment can execute with the optimal predicted parameters.
[0055] (1) Quantitative calculation of the deviation of the regulation effect
[0056] in It is a 4×1 dimensional column vector of the control effect deviation; The execution parameters are measured by a self-developed monitoring device after the control is implemented; These are the parameters in the optimal prediction parameter matrix. The smaller the deviation, the higher the model's prediction accuracy.
[0057] (2) Incremental update of performance coupling matrix M
[0058] in, This is the updated 4×4 performance coupling matrix; The original 4×4 performance coupling matrix; The learning rate is fixed at 0.01 to 0.05 to avoid model oscillation. This is a transpose (1×4-dimensional row vector) of the control performance vector of the drilling fluid equipment before execution using the optimal prediction parameter matrix.
[0059] (3) Incremental update of effect coupling matrix B
[0060] in The updated 4×5 effect coupling matrix; The original 4×5 effect coupling matrix; The learning rate is fixed at 0.005 to 0.02. This is a transpose (1×5-dimensional row vector) of the drilling fluid equipment control action vector before execution using the optimal prediction parameter matrix.
[0061] Then repeat steps S2-S3 until the execution parameters meet the preset requirements.
[0062] On the other hand, embodiments of the present invention also provide a multi-parameter control system for drilling fluid equipment, which uses the above-mentioned multi-parameter control method for drilling fluid equipment, including: The data acquisition module is used to acquire drilling fluid mechanism data; The parameter coupling module is used to generate a performance parameter vector and a performance calibration matrix based on drilling fluid mechanism data, and then use the performance parameter vector and the performance coupling matrix to generate a prediction parameter matrix. The performance prediction module, the parameter optimization module is used to construct a parameter optimization function based on drilling fluid mechanism data, and with the goal of minimizing the parameter optimization function, use the NSGA-III algorithm to optimize the prediction parameter matrix to obtain the optimal prediction parameter matrix; The parameter optimization module is used to control the drilling fluid equipment to execute with the optimal prediction parameter matrix, obtain the execution parameters, optimize the performance parameter vector using the LMS algorithm based on the execution parameters, and then repeat steps S2-S3 until the execution parameters meet the preset requirements.
[0063] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A method for multi-parameter control of drilling fluid equipment, characterized in that: include: S1: Obtain drilling fluid mechanism data, generate performance parameter vectors and performance calibration matrices based on drilling fluid mechanism data, and then use performance parameter vectors and performance coupling matrices to generate prediction parameter matrices; S2: Construct a parameter optimization function based on drilling fluid mechanism data, and use the NSGA-III algorithm to optimize the predicted parameter matrix with the goal of minimizing the parameter optimization function, so as to obtain the optimal predicted parameter matrix; S3: Control the drilling fluid equipment to execute with the optimal prediction parameter matrix and obtain the execution parameters. Based on the execution parameters, use the LMS algorithm to optimize the performance parameter vector. Then repeat steps S2-S3 until the execution parameters meet the preset requirements.
2. The method for multi-parameter control of drilling fluid equipment as described in claim 1, characterized in that: The performance parameter vector includes a control performance vector and a control action vector, and the performance calibration matrix includes a performance coupling matrix and an effect coupling matrix. The dot product of the regulation performance vector and the performance coupling matrix is used to obtain the prediction performance matrix, and the dot product of the regulation action vector and the effect coupling matrix is used to obtain the prediction effect matrix. The prediction performance matrix and the prediction effect matrix are collectively referred to as the prediction parameter matrix.
3. The method for multi-parameter control of drilling fluid equipment as described in claim 2, characterized in that: The parameter optimization function is as follows: in, The weight coefficient for the i-th parameter; For the actual value of the parameter, For the target value of the parameter, Let j be the unit cost of the control action. For the amount of movement, The rate of change of the parameter.
4. The method for multi-parameter control of drilling fluid equipment as described in claim 1, characterized in that: Methods for optimizing the prediction parameter matrix using the NSGA-III algorithm to obtain the optimal prediction parameter matrix include: A1: Initialize the population and the number of iterations. The population includes multiple parent individuals. Each parent individual is assigned random state parameters according to the prediction parameter matrix. A2: Calculate the individual fitness of each parent individual based on the parameter optimization function and the state parameters of each parent individual; A3: Sort all parent individuals in a non-dominated order based on their fitness, and select the top N parent individuals for crossbreeding to generate N offspring individuals; A4: Sort all parent and child individuals in a non-dominated order, and use the reference point screening method to select the top K individuals as elite individuals; A5: Repeat steps A2-A5 with the managed individual as the new parent individual until the maximum number of iterations is reached, and take all elite individuals obtained in the last iteration as the optimal solution set; A6: Based on the fuzzy membership function, select the elite individual with the lowest parameter optimization function from all elite individuals obtained from the last iteration, and use the parameter state parameters corresponding to this elite individual as the optimal prediction parameter matrix.
5. The method for multi-parameter control of drilling fluid equipment as described in claim 1, characterized in that: The control rate is calculated based on drilling fluid mechanism data, and the drilling fluid equipment is adjusted from its current state to the optimal predicted parameter matrix in an incremental manner based on the control rate.
6. The method for multi-parameter control of drilling fluid equipment as described in claim 5, characterized in that: The formula for calculating the regulation rate is as follows: in, The optimal control activation time (time before the accident occurs); The timeframe for risk prediction output by the advanced early warning system; For the control and response time of on-site equipment; This is a safety buffer period.
7. The method for multi-parameter control of drilling fluid equipment as described in claim 1, characterized in that: Based on the execution parameters, methods for optimizing the performance parameter vector using the LMS algorithm include: Calculate the control effect deviation vector based on the execution parameters and the prediction performance matrix; Based on the control deviation vector, incremental optimization is performed on the control performance vector and the control action vector.
8. A multi-parameter control system for drilling fluid equipment, characterized in that, The system uses a multi-parameter control method for drilling fluid equipment as described in any one of claims 1-7, comprising: The data acquisition module is used to acquire drilling fluid mechanism data; The parameter coupling module is used to generate a performance parameter vector and a performance calibration matrix based on drilling fluid mechanism data, and then use the performance parameter vector and the performance coupling matrix to generate a prediction parameter matrix. The performance prediction module, the parameter optimization module is used to construct a parameter optimization function based on drilling fluid mechanism data, and with the goal of minimizing the parameter optimization function, use the NSGA-III algorithm to optimize the prediction parameter matrix to obtain the optimal prediction parameter matrix; The parameter optimization module is used to control the drilling fluid equipment to execute with the optimal prediction parameter matrix, obtain the execution parameters, optimize the performance parameter vector using the LMS algorithm based on the execution parameters, and then repeat steps S2-S3 until the execution parameters meet the preset requirements.