A flexible load-based pumping unit well group control method

Abstract

A flexible load-based pumping unit group control method belongs to the technical field of pumping unit group control. The method comprises the following steps: step 1, modeling the pumping system; step 2, establishing a flexible load model suitable for the pumping unit group; step 3, taking the minimum difference between the well group power and the input power as the objective function, considering the power constraint and the motor frequency constraint of each pumping unit, and solving the constraints to obtain the pumping unit group scheduling scheme; and step 4, forming a pumping unit group predictive control algorithm. In the flexible load-based pumping unit group control method, the flexible load characteristics of the pumping unit group are fully utilized, and green power resources can be efficiently absorbed and utilized. By constructing an accurate flexible load model of the pumping unit group and optimizing the control of the system, dynamic matching of the load and energy supply can be realized.

Description

Technical Field

[0001] A method for controlling a group of pumping unit wells based on flexible loads, belonging to the field of pumping unit well group control technology. Background Technology

[0002] The importance of clean energy in daily life and industrial production is becoming increasingly prominent. Clean energy power generation, represented by wind and solar power, is subject to certain instability due to weather and seasonal variations. Flexible regulation methods have become a key strategy to address this instability. The core advantage of flexible load technology lies in its high flexibility in electricity demand within a certain time frame. This flexibility allows flexible loads to dynamically adjust according to the power grid's supply status, increasing electricity consumption when power supply is sufficient and decreasing it when power supply is tight. In the oilfield sector, pumping unit well clusters are systems with significant flexible load characteristics, and their energy consumption accounts for more than 40% of the total energy consumption of the oilfield. Significant progress has been made in utilizing flexible regulation methods to absorb green electricity; therefore, applying flexible regulation methods to pumping units has broad development prospects.

[0003] Traditional control methods for pumping unit well clusters rely on manual operation, which has limited efficiency and struggles to cope with the variability of reservoir conditions. To improve efficiency, automated and intelligent control technologies have emerged. In current mainstream system design, model predictive control (MPC) is typically used for controlling nonlinear processes like pumping unit well clusters, which involve disturbances and constraints. Essentially, MPC achieves its control objective by solving an open-loop optimal control problem. While the concept of MPC is independent of the specific model, its implementation is highly dependent on the accuracy of the model.

[0004] In existing technologies, the traditional control solutions for pumping unit well groups include:

[0005] Reference: Yang Xuefeng. Research on Co-optimization of Peak-Shaving and Intermittent Pumping Operation System of Pumping Unit Well Group [D]. China University of Petroleum (Beijing), 2023. This paper discloses a technical solution. In this solution, for cluster well groups with insufficient fluid supply, the optimal single-well intermittent pumping system is taken as the premise, and the minimum energy consumption cost of the pumping unit is taken as the objective function. Combining the characteristics of pumping unit well power data and the variation law of pump cavitation, the single-well production capacity of single-well intermittent pumping is guaranteed. Considering the constraints of peak and off-peak electricity prices for industrial electricity, the characteristic value change rate constraint, and the minimum and maximum number of wells to be opened, a collaborative scheduling model for peak-shaving well opening and intermittent pumping of cluster well groups is established to achieve the coordination of single-well intermittent pumping and peak-shaving well opening of the well group. However, this technical solution obtains the relationship between pump cavitation and peak power by analyzing the characteristics of the pumping unit's working state and electrical parameters. A collaborative optimization model for intermittent pumping and peak-shaving of the well group is established and solved by combining particle swarm optimization algorithm and genetic algorithm to obtain the optimal intermittent pumping system for oil wells. However, the issue of green energy consumption is not considered.

[0006] Chinese invention patent application number 202410523927.7, filed on April 29, 2024, entitled "A Method for Intermittent Pumping Control of Oil Pumping Units Using Wireless Measurement," discloses a technical solution. This solution includes establishing an oil well production simulation model based on geological, engineering, and production data. First, the control parameter relationships of the pumping unit are determined. Then, the current oil well status information and the current pumping unit status information are clarified to determine the response complexity level and formulate a coarse intermittent pumping control strategy. Further, the response detection time is determined based on the response complexity level, and within this response detection time, the start-up and shutdown of the pumping unit and control parameters are controlled using the coarse intermittent pumping control strategy. Finally, a fine intermittent pumping control strategy is formulated based on the control effect to further refine the control of the pumping unit's start-up and shutdown and controllable parameters. However, the constructed oil pumping unit production simulation model and the determination of the intermittent pumping system do not consider the issue of flexible regulation.

[0007] Chinese invention patent application number 202111423156.7, filed on November 26, 2021, entitled "A Method and System for Non-Stop Pumping Control of Oil Pumping Units," discloses a technical solution that can effectively improve the efficiency of oil pumping units and enable oil wells to achieve optimal supply and drainage coordination. However, this technical solution only considers a single oil pumping unit and does not address the coordinated control of a group of wells.

[0008] Chinese invention patent application number 202110588435.2, filed on May 28, 2021, entitled "A Self-Learning-Based Method for Adjusting the Stroke Frequency of Rod Pumping Unit Wells," discloses a technical solution. This solution, combined with oilfield production processes, provides a self-learning-based parameter optimization method for oilfield production, achieving a "one well, one policy" approach. It enables automatic evaluation and verification of the adjusted well parameters during the optimization process, resulting in higher reliability. Furthermore, by combining historical data analysis, it selects the optimal stroke frequency value for optimization. This multi-objective optimization strategy is applicable to different production needs in oilfields and can reduce costs and increase efficiency in oil wells, providing a technical foundation for green oilfield construction. However, this technical solution only considers a single pumping unit, neglecting the coordinated control of well groups and the issue of system constraints. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a flexible load-based pumping unit well group control method that makes full use of the flexible load characteristics of the pumping unit well group, can efficiently absorb and utilize green power resources, and achieves dynamic matching of load and energy supply by constructing an accurate flexible load model of the pumping unit well group and performing scheduling optimization control on the system.

[0010] The technical solution adopted by this invention to solve its technical problem is: a method for controlling a group of pumping wells based on flexible load, characterized by including the following steps:

[0011] Step 1: Model the oil pumping system;

[0012] Step 2: Establish a flexible load model suitable for pumping unit well groups;

[0013] Step 3: Taking the minimum difference between the well group power and the input power as the objective function, consider the power constraints of each pumping unit and the motor frequency constraints, and solve them to obtain the pumping scheduling scheme between the pumping unit well group.

[0014] Step 4: Develop a predictive control algorithm for the pumping unit well group.

[0015] Preferably, in step 2, the flexible load model defines three operating conditions for the pumping unit: Condition 1, power is not adjustable, under which the power of the pumping unit cannot be adjusted due to factors such as the production volume and the dynamic fluid level, and always remains constant; Condition 2, power is adjustable, the power consumed per day is a fixed value, under which the power consumed per day is a fixed value, but is adjusted every time period; Condition 3, power is adjustable, but the power is limited, under which the power of the pumping unit is adjusted, but does not exceed its limit value.

[0016] Preferably, in step 3, the inter-well pumping scheduling scheme includes the following steps:

[0017] Step 3-1: Establish an optimization scheduling objective function to minimize the difference between the power of the pumping well group system and the input power;

[0018]

[0019] Where P0(t) represents the real-time output power on the power supply side, which is known; V = [v 1 ,…,v i ,…,v N ] T v i Let P be a 0-1 variable representing the on / off state of the i-th pumping unit, to be determined; P = [P 1 ,…,P i ,…,P N ], P i This represents the predicted power output after the i-th pumping unit is turned on;

[0020] Step 3-2: Set constraints to ensure the safe operation of the well group system;

[0021] Step 3-3: The scheduling optimization problem is described as a mixed integer nonlinear programming model, and the Gurobi solver is applied to solve the model to obtain the pumping scheduling scheme among the pumping well groups.

[0022] Preferably, in step 3-2, the constraints specifically include:

[0023] (1) Power constraint:

[0024]

[0025] in, Let be the instantaneous input power of the i-th pumping unit;

[0026] (2) Single well operating time constraints:

[0027] t min ≤t i ≤t max

[0028] Among them, t min The minimum required well opening time; t max The maximum allowable well opening time; t i This refers to the operating time of a single well.

[0029] (3) Well group operation status constraints:

[0030] O min ≤N L ≤O max

[0031] Among them, O min Minimum number of wells to be opened; O max N represents the maximum number of wells to be opened. L The number of wells opened in the pumping unit well group;

[0032] (4) Single-well production capacity constraints:

[0033]

[0034] in, The production volume of the i-th oil well is represented by T; the sampling period is T; and Δt is the time interval between two adjacent samplings. This represents the minimum production rate of the i-th oil well;

[0035] (5) Dynamic liquid surface constraint:

[0036]

[0037] in, This represents the dynamic fluid level height of the i-th oil well; These are the maximum and minimum dynamic fluid level heights allowed for the i-th oil well, respectively.

[0038] Preferably, in step 2, the flexible load model applicable to pumping unit well groups is:

[0039]

[0040] in, This indicates that the power of the i-th pumping unit is not adjustable and is always P1; This indicates that the power of the i-th pumping unit is adjustable, but the total power consumed in a single day is a constant value P2; This indicates that the power of the i-th pumping unit is adjustable (flexible load), but there are upper and lower limits to the power during the adjustment process.

[0041] The following overall power load of the pumping well group was obtained:

[0042]

[0043] Where N1, N2, and N3 represent the number of pumping units that meet the three operating conditions, and N is the total number of pumping units in the well group, N = N1 + N2 + N3. Let i = 1, 2, ..., N1, ..., N1 + N2, ..., N, that is, the first N1 pumping units are the first operating condition, the N1+1 to N1+N2 pumping units are the second operating condition, and the N1+N2+1 to N pumping units are the third operating condition.

[0044] Preferably, in step 4, the pumping well group prediction and control algorithm includes the following steps:

[0045] Step 4-1: Determine the optimal control objective function to ensure that the difference between the power consumed by the well group and the input power is minimized.

[0046]

[0047] Where P0(t) represents the real-time output power on the power supply side, which is known; V = [v 1 ,…,v i ,…,v N ] T v i The variable is 0-1, representing the on / off state of the i-th pumping unit, which has been obtained in step three; This represents the real-time power of the i-th pumping unit obtained from the power model in step one, from which the control law of the i-th pumping unit, i.e., the motor frequency, is obtained.

[0048] Step 4-2: Set constraints to ensure the safe and stable operation of the well group system;

[0049] Step 4-3: Design the predictive control algorithm.

[0050] Preferably, the constraints in step 4-2 specifically include:

[0051] (1) Total production capacity constraint of well group:

[0052]

[0053] Where N is the number of pumping units in the well group; The production rate of the i-th oil well is represented by T; the sampling period is T; Δt is the time interval between two adjacent samplings; Q represents the production rate of the i-th oil well. min This indicates the minimum required production rate for this well group;

[0054] (2) Single-well production capacity constraint:

[0055]

[0056] in, The production volume of the i-th oil well is represented by T; the sampling period is T; and Δt is the time interval between two adjacent samplings. This represents the minimum production rate of the i-th oil well;

[0057] (2) Dynamic liquid surface constraint:

[0058]

[0059] in, This represents the dynamic fluid level height of the i-th oil well; These are the maximum and minimum dynamic fluid level heights allowed for the i-th oil well, respectively;

[0060] (3) Indicator diagram constraints:

[0061] ΔS min <ΔS k <ΔS max

[0062] ΔS k =S k -S k-1

[0063] Among them, S k S represents the area of ​​the dynamometer diagram at the current moment. k-1 This represents the area of ​​the indicator diagram at the previous moment.

[0064] Preferably, step 4-3 includes the following steps:

[0065] Step 4-3-1, the mechanism model of the pumping unit well group model (including the four-bar linkage model, pump load model, dynamic fluid level model, production model, and electric power model of the pumping unit well group) established in Step 1 is as follows:

[0066]

[0067] Step 4-3-2: Based on the mechanism model established in Step 1, set the control time domain and prediction time domain, and predict N at time k. p Output within the range:

[0068]

[0069] Step 4-3-3: Under the given constraints, the designed prediction algorithm is used to solve the objective function to obtain the optimal real-time control strategy for each pumping unit.

[0070]

[0071] ΔS min <ΔS k <ΔS max

[0072] Where x1…x7 represent suspension point displacement, pump load of oil pump, pump displacement, suspension point displacement, liquid production, dynamic liquid level height, and instantaneous motor input power, respectively; U is the system input; and Y is the system output.

[0073] Compared with the prior art, the beneficial effects of this invention are:

[0074] This method for controlling pumping unit well groups based on flexible loads fully utilizes the flexible load characteristics of pumping unit well groups, enabling efficient absorption and utilization of green power resources. Furthermore, by constructing an accurate flexible load model for the pumping unit well groups and optimizing the scheduling control of the system, dynamic matching between load and energy supply can be achieved.

[0075] In this flexible load-based control method for pumping unit well groups, a flexible load model suitable for pumping unit well groups is designed. Based on the actual situation of oilfield pumping units, three operating conditions of pumping units are defined, thus solving the problem of flexible expression of pumping units.

[0076] In this flexible load-based pumping unit well group control method, a pumping unit well group model under flexible load is first established. Then, based on the model, an optimized scheduling method is designed to obtain the start-up and shutdown times of each pumping unit in the well group. Finally, based on the above content and the constraints of the well group system, an optimized control strategy is developed to further ensure the stability of the well group system and the maximum absorption of green electricity.

[0077] In this flexible load-based pumping unit well cluster control method, the start-up and shutdown times of pumping units under different operating conditions are determined by tracking the input power curve of green electricity, while ensuring system stability and safety, thereby realizing the absorption of green electricity power. This enables the pumping unit well cluster system to utilize green electricity more efficiently, reducing the pressure on the power grid, while improving system stability and operating efficiency. Attached Figure Description

[0078] Figure 1 Flowchart of a control method for pumping well groups based on flexible load. Detailed Implementation

[0079] Figure 1 This is the preferred embodiment of the present invention, which is described below in conjunction with the accompanying drawings. Figure 1 The present invention will be further described below.

[0080] like Figure 1 As shown, a control method for a pumping unit well group based on flexible load (hereinafter referred to as the control method) includes the following steps:

[0081] Step 1: Model the oil pumping system;

[0082] In this step, the following model is established for the oil pumping system:

[0083] 1. By analyzing the motion laws and structural parameters of the pumping unit, the kinematic equations of the pumping unit are established, and a four-bar linkage model is built:

[0084] θ = 2π·u·t

[0085]

[0086] θ2=2π-θ+α

[0087]

[0088] ψ=χ+β

[0089]

[0090] x1(t)=A(ψ-ψ min )

[0091] Where u is the control cabinet frequency, which is also the control input; x1(t) represents the suspension point displacement; R is the torque factor; P is the crank radius; A is the connecting rod length; C is the walking beam front arm length; K is the walking beam rear arm length; I is the horizontal distance from the center of the gearbox output shaft to the center of the support shaft; L is the torque factor; R is the crank radius; P is the connecting rod length; A is the walking beam front arm length; C is the walking beam rear arm length; K is the base rod length; I is the horizontal distance from the center of the gearbox output shaft to the center of the support shaft; L is kθ1 is the length from the crank end to the walking beam axis; θ2 is the angle between the base rod and the crank, with counterclockwise as the positive direction; ψ is the angle between the rear arm of the walking beam and the base rod; α is the angle between the base rod and the vertical direction; β is the angle between the base rod and the line connecting the crank and the walking beam axis; ψ2 is the angle between the base rod and the line connecting the crank and the walking beam axis; min The minimum angle between the walking beam rear arm and the base rod is given by t, where t represents the crank running time.

[0092] 2. By analyzing the structure and forces acting on the sucker rod string, the fluctuation equation of the sucker rod string and the pump load model of the oil pump are established:

[0093]

[0094] p d =p o +[w0ρ w +(1-w0)ρ0]gL p

[0095] p s =p c +[w0ρ w +(1-w0)ρ0]g(L p -H)

[0096] W0 = A p (p d -p s )

[0097]

[0098] Where x2 represents the pump load of the oil pump; p d It is the pump's discharge pressure, p s It is sunk pressure; p o It is the wellhead oil pressure; w0 is the water cut; ρ w ρ0 is the density of water; ρ0 is the density of crude oil; L p It is the depth of the pump; p c It is the wellhead casing pressure; H is the dynamic fluid level height; A p It is the cross-sectional area of ​​the plunger; L is the total length of the sucker rod string; E r It is the elastic modulus; A r It is the cross-sectional area of ​​the sucker rod; S p For pump stroke; t m W0 represents the time it takes for the plunger to reach top dead center, W0 represents the difference between the discharge pressure and the sinking pressure acting on the plunger, and g represents the acceleration due to gravity.

[0099] Sucker rod undulation equation:

[0100]

[0101] Where s(l,t) represents the displacement of the sucker rod section of length l at time t; x1 is the suspension point displacement; x2 represents the pump load of the oil pump; A r It is the cross-sectional area of ​​the sucker rod string; E r It is the elastic modulus; L is the total length of the sucker rod string; g is the acceleration due to gravity; It is the speed of sound in the sucker rod; ρ r υ is the density of the sucker rod; υ is the rod fluid damping coefficient.

[0102] The pump displacement and suspension load are obtained by solving the fluctuation equation of the sucker rod string using the finite difference method.

[0103]

[0104] Where x3 represents pump displacement, x4 represents suspension load, and W r This is the weight of the sucker rod string in air.

[0105] 3. Establish dynamic liquid level model and liquid production model.

[0106] The dynamic liquid surface model is as follows:

[0107]

[0108] In the formula, x5 represents the production volume; x6 represents the dynamic liquid level height; u represents the control cabinet frequency; D is the pump diameter of the oil pump; S is the suspension stroke; n is the number of strokes; α p It is pump efficiency; s1 is slip rate; k is reduction ratio; n p Q is the extreme logarithm; in It is the amount of fluid flowing from the reservoir into the oil well; D ci D te These are the inner diameter of the casing and the outer diameter of the tubing, respectively. The derivative representing the height of the moving liquid surface.

[0109] The liquid production model is as follows:

[0110] α p =η S η F η L η V

[0111]

[0112] Where, η S It is the plunger stroke coefficient; η F It is the pump fill factor; η L It is the leakage coefficient; η V It is the volume coefficient of the mixture; A r It is the cross-sectional area of ​​the sucker rod string; ρ lL is the well fluid density; n is the pump mounting depth; E is the pump stroke rate; A is the elastic modulus of the steel. p It is the cross-sectional area of ​​the plunger; w0 is the water content; ρ w It is the density of water; ρ o ΔP is the density of crude oil; ΔP is the pressure difference between the liquid columns at both ends of the plunger; S is the stroke; n represents the number of strokes; A t It is the weighted average cross-sectional area of ​​the oil pipe; R x It is the solubility coefficient; p s It is sunk pressure; p h Z is the saturation pressure; Z is the gas compressibility coefficient; B l δ is the volume coefficient of the liquid inside the pump; δ is the clearance between the plunger and the pump barrel; L c It is the effective plunger length; δ l It is the relative density of the liquid; B o B w These are the volume coefficients of the crude oil and water in the pump, respectively.

[0113] 4. Establish an electrical power model for the pumping unit well group;

[0114] The electrical power model for the pumping unit well group is as follows:

[0115]

[0116] Where x7(t) represents the instantaneous motor input power; x4 is the suspension load; P0 is the motor no-load power; P N η is the rated power of the motor. N The rated efficiency of the motor; β = N MO / P N N represents the instantaneous power utilization rate of the motor. MO ω is the instantaneous output power of the electric motor; ω is the angular velocity of the pumping unit crank. The transmission efficiency between the pumping unit belt and the gearbox is given by k1 = ±1 (when v A When v > 0, k1 = -1; when v A <0, k1=1, v A (Velocity at the suspension point). M N It is the net torque of the crankshaft; It is the pumping unit torque factor; B W The problem lies in the unbalanced structure of the oil pump. It is the mechanical transmission efficiency from the crank to the suspension point, k2=±1(when N MO When N > 0, k2 = -1; when N > 0, k2 = -1; MO <0, k2=1); M C θ is the maximum balancing torque of the crank counterweight; θ is the crank angle; τ is the offset angle of the crank counterweight.

[0117] Step 2: Establish a flexible load model suitable for pumping unit well groups;

[0118] A flexible load model suitable for pumping unit well groups is established. In this control method, the flexible load model defines three operating conditions for the pumping unit based on actual conditions: First, power is not adjustable. Under this condition, the power of the pumping unit cannot be adjusted due to factors such as production volume and dynamic fluid level, and remains constant. Second, power is adjustable, but the daily power consumption is constant. In this case, although the daily power consumption of the pumping unit is constant, it can be adjusted according to production needs in each time period. Third, power is adjustable, but power is limited. Under this condition, the power of the pumping unit can be appropriately adjusted according to production needs, but it cannot exceed its limit. The model is as follows:

[0119]

[0120] in, This indicates that the power of the i-th pumping unit is not adjustable and is always P1; This indicates that the power of the i-th pumping unit is adjustable, but the total power consumed in a single day is a constant value P2; This indicates that the power of the i-th pumping unit is adjustable (flexible load), but there are upper and lower limits to the power during the adjustment process.

[0121] Based on the specific operating conditions of the pumping unit well group, the following overall power load of the pumping unit well group is obtained:

[0122]

[0123] Where N1, N2, and N3 represent the number of pumping units that meet the above three operating conditions, and N is the total number of pumping units in the well group. For ease of subsequent expression, N = N1 + N2 + N3.

[0124] For ease of explanation later, the pumping units in the well group are ordered according to three operating conditions, let:

[0125] i = 1, 2, ..., N1, ..., N1+N2, ..., N, meaning the first N1 pumping units are the first type of operating condition, the N1+1 to N1+N2 pumping units are the second type of operating condition, and the N1+N2+1 to N pumping units are the third type of operating condition.

[0126] Step 3: Obtain the pumping schedule plan among the pumping unit well groups;

[0127] To meet the requirements of flexible load and green energy absorption, staggered well operation and intermittent pumping of the pumping unit well group are necessary. However, scheduling schemes based on manual experience have significant uncertainties, which can easily lead to increased energy consumption and costs. Therefore, in this control method, the objective function is to minimize the difference between the well group power and the input power. Considering the power constraints of each pumping unit and the motor frequency constraints, the algorithm is solved to obtain the intermittent pumping scheduling scheme for the pumping unit well group. The algorithm flow is as follows:

[0128] Step 3-1: Establish an optimization scheduling objective function to minimize the difference between the power of the pumping well group system and the input power;

[0129]

[0130] Where P0(t) represents the real-time output power on the power supply side, which is known; V = [v 1 ,…,v i ,…,v N ] T v i P is a 0-1 variable representing the on / off state of the i-th pumping unit; P = [P 1 ,…,P i ,…,P N ], P i Let J represent the predicted power after the i-th pumping unit is turned on, where J is the objective function and T is the prediction time domain.

[0131] Step 3-2: Set constraints to ensure the safe operation of the well group system;

[0132] The constraints specifically include:

[0133] (1) Power constraint:

[0134]

[0135] in, Let be the instantaneous input power of the i-th pumping unit.

[0136] (2) Single well operating time constraints:

[0137] t min ≤t i ≤t max

[0138] Among them, t min The minimum required well opening time; t max The maximum allowable well opening time; t i The time is the operating time of a single well; the total scheduling time is 24 hours, and the minimum scheduling unit is assumed to be 2 hours.

[0139] (3) Well group operation status constraints:

[0140] O min ≤N L ≤O max

[0141] Among them, O min Minimum number of wells to be opened; O max N represents the maximum number of wells to be opened. L This refers to the number of wells opened in the pumping unit well group.

[0142] (4) Single-well production capacity constraints:

[0143]

[0144] in, The production volume of the i-th oil well is represented by T; the sampling period is T; and Δt is the time interval between two adjacent samplings. This represents the minimum production rate of the i-th oil well.

[0145] (5) Dynamic liquid surface constraint:

[0146]

[0147] in, This represents the dynamic fluid level height of the i-th oil well; These are the maximum and minimum dynamic fluid level heights allowed for the i-th oil well, respectively.

[0148] Step 3-3: The scheduling optimization problem is described as a mixed integer nonlinear programming model, and the Gurobi solver is applied to solve the model to obtain the pumping scheduling scheme among the pumping well groups.

[0149] Step 4: Develop a predictive control algorithm for the pumping unit well group;

[0150] To better utilize green energy and achieve refined control of pumping unit well clusters, this control method develops a predictive control algorithm for pumping unit well clusters. The specific steps are as follows:

[0151] Step 4-1: Determine the optimal control objective function to ensure that the difference between the power consumed by the well group and the input power is minimized.

[0152]

[0153] Where P0(t) represents the real-time output power on the power supply side, which is known; V = [v 1 ,…,v i ,…,v N ] T v i The variable is 0-1, representing the on / off state of the i-th pumping unit, which has been obtained in step three; The real-time power of the i-th pumping unit is obtained from the power model in step 1, from which the control law of the i-th pumping unit, i.e., the motor frequency, is obtained; T is the prediction time domain.

[0154] Step 4-2: Set constraints to ensure the safe and stable operation of the well group system;

[0155] The constraints specifically include:

[0156] (1) Total production capacity constraint of well group:

[0157]

[0158] Where N is the number of pumping units in the well group; The production rate of the i-th oil well is represented by T; the sampling period is T; Δt is the time interval between two adjacent samplings; Q represents the production rate of the i-th oil well. min This indicates the minimum required production volume for this well group.

[0159] (2) Single-well production capacity constraint:

[0160]

[0161] in, The production volume of the i-th oil well is represented by T; the sampling period is T; and Δt is the time interval between two adjacent samplings. This represents the minimum production rate of the i-th oil well.

[0162] (2) Dynamic liquid surface constraint:

[0163]

[0164] in, This represents the dynamic fluid level height of the i-th oil well; These are the maximum and minimum dynamic fluid level heights allowed for the i-th oil well, respectively.

[0165] (3) Indicator diagram constraints:

[0166] ΔS min <ΔS k <ΔS max

[0167] ΔS k =S k -S k-1

[0168] Among them, S k S represents the area of ​​the dynamometer diagram at the current moment. k-1 This represents the area of ​​the indicator diagram at the previous moment. ΔS min ΔS max These represent the minimum and maximum values ​​of the difference between the area of ​​the indicator diagram at the current moment and the area of ​​the indicator diagram at the previous moment.

[0169] Step 4-3: Design the predictive control algorithm;

[0170] This algorithm can predict the future output of the system based on the historical information and future inputs of the controlled object. The predictive models can take the form of convolutional models, mechanistic models, neural network models, etc.

[0171] The mechanistic model of the pumping unit well group model (including the four-bar linkage model, pump load model, dynamic fluid level model, production model, and electric power model of the pumping unit well group) established in step 1 is as follows:

[0172]

[0173] Based on the mechanism model established in step 1, the control time domain and prediction time domain are set up to predict N at time k. p Output within the range:

[0174]

[0175] Under the given constraints, the objective function is solved using the designed prediction algorithm to obtain the optimal real-time control strategy for each pumping unit:

[0176]

[0177] ΔS min <ΔS k <ΔS max

[0178] Where x1…x7 represent suspension point displacement, pump load of oil pump, pump displacement, suspension point displacement, liquid production, dynamic liquid level height, and instantaneous motor input power, respectively; U is the system input; and Y is the system output.

[0179] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for controlling a group of pumping wells based on flexible load, characterized in that: Includes the following steps: Step 1: Model the oil pumping system; Step 2: Establish a flexible load model suitable for pumping unit well groups; Step 3: Taking the minimum difference between the well group power and the input power as the objective function, consider the power constraints of each pumping unit and the motor frequency constraints, and solve them to obtain the pumping scheduling scheme between the pumping unit well group. Step 4: Develop a predictive control algorithm for the pumping unit well group; In step 2, the flexible load model defines three operating conditions for the pumping unit: Condition 1, power is not adjustable, under which the power of the pumping unit cannot be adjusted due to factors such as the production volume and the dynamic fluid level, and always remains constant; Condition 2, power is adjustable, the power consumed per day is a fixed value, under which the power consumed per day is a fixed value, but is adjusted every time period; Condition 3, power is adjustable, but the power is limited, under which the power of the pumping unit is adjusted, but does not exceed its limit value. Step 4, the predictive control algorithm for the pumping unit well group, includes the following steps: Step 4-1: Determine the optimal control objective function to ensure that the difference between the power consumed by the well group and the input power is minimized. Where P0(t) represents the real-time output power on the power supply side, which is known; V = [v 1 ,…,v i ,…,v N ] T v i The variable is 0-1, representing the on / off state of the i-th pumping unit, which has been obtained in step three; This represents the real-time power of the i-th pumping unit obtained from the power model in step one, from which the control law of the i-th pumping unit, i.e., the motor frequency, is obtained. Step 4-2: Set constraints to ensure the safe and stable operation of the well group system; Step 4-3: Design the predictive control algorithm; The constraints in step 4-2 specifically include: (1) Total production capacity constraint of well group: Where N is the number of pumping units in the well group; The production rate of the i-th oil well is represented by T; the sampling period is T; Δt is the time interval between two adjacent samplings; Q represents the production rate of the i-th oil well. min This indicates the minimum required production rate for this well group; (2) Single-well production capacity constraint: in, The production volume of the i-th oil well is represented by T; the sampling period is T; and Δt is the time interval between two adjacent samplings. This represents the minimum production rate of the i-th oil well; (2) Dynamic liquid surface constraint: in, This represents the dynamic fluid level height of the i-th oil well; These are the maximum and minimum dynamic fluid level heights allowed for the i-th oil well, respectively; (3) Indicator diagram constraints: ΔS min <ΔS k <ΔS max ΔS k =S k -S k-1 Among them, S k S represents the area of ​​the dynamometer diagram at the current moment. k-1 The area of ​​the indicator diagram at the previous moment; Step 4-3 includes the following steps: Step 4-3-1, the mechanism model of the pumping unit well group model (including the four-bar linkage model, pump load model, dynamic fluid level model, production rate model, and electric power model of the pumping unit well group) established in Step 1 is as follows: Step 4-3-2: Based on the mechanism model established in Step 1, set the control time domain and prediction time domain, and predict N at time k. p Output within the range: Step 4-3-3: Under the given constraints, the designed prediction algorithm is used to solve the objective function to obtain the optimal real-time control strategy for each pumping unit. ΔS min <ΔS k <ΔS max Where x1…x7 represent suspension point displacement, pump load of oil pump, pump displacement, suspension point displacement, liquid production, dynamic liquid level height, and instantaneous motor input power, respectively; U is the system input; and Y is the system output.

2. The method for controlling a group of pumping wells based on flexible load according to claim 1, characterized in that: In step 3, the pumping schedule among the pumping unit wells includes the following steps: Step 3-1: Establish an optimization scheduling objective function to minimize the difference between the power of the pumping well group system and the input power; Where P0(t) represents the real-time output power on the power supply side, which is known; V = [v 1 ,…,v i ,…,v N ] T v i Let P be a 0-1 variable representing the on / off state of the i-th pumping unit, to be determined; P = [P 1 ,…,P i ,…,P N ], P i This represents the predicted power output after the i-th pumping unit is turned on; Step 3-2: Set constraints to ensure the safe operation of the well group system; Step 3-3: The scheduling optimization problem is described as a mixed integer nonlinear programming model, and the Gurobi solver is applied to solve the model to obtain the pumping scheduling scheme among the pumping well groups.

3. The method for controlling a group of pumping wells based on flexible load according to claim 2, characterized in that: In step 3-2, the constraints specifically include: (1) Power constraint: in, Let be the instantaneous input power of the i-th pumping unit; (2) Single well operating time constraints: t min ≤t i ≤t max Among them, t min The minimum required well opening time; t max The maximum allowable well opening time; t i This refers to the operating time of a single well. (3) Well group operation status constraints: THE min ≤N L ≤O max Among them, O min Minimum number of wells to be opened; O max N represents the maximum number of wells to be opened. L The number of wells opened in the pumping unit well group; (4) Single-well production capacity constraints: in, The production volume of the i-th oil well is represented by T; the sampling period is T; and Δt is the time interval between two adjacent samplings. This represents the minimum production rate of the i-th oil well; (5) Dynamic liquid surface constraint: in, This represents the dynamic fluid level height of the i-th oil well; These are the maximum and minimum dynamic fluid level heights allowed for the i-th oil well, respectively.

4. The method for controlling a group of pumping wells based on flexible load according to claim 1, characterized in that: In step 2, the flexible load model applicable to pumping unit well groups is as follows: in, This indicates that the power of the i-th pumping unit is not adjustable and is always P1; This indicates that the power of the i-th pumping unit is adjustable, but the total power consumed in a single day is a constant value P2; This indicates that the power of the i-th pumping unit is adjustable (flexible load), but there are upper and lower limits to the power during the adjustment process. The following overall power load of the pumping well group was obtained: Where N1, N2, and N3 represent the number of pumping units that meet the three operating conditions, and N is the total number of pumping units in the well group, N = N1 + N2 + N3. Let i = 1, 2, ..., N1, ..., N1 + N2, ..., N, that is, the first N1 pumping units are the first operating condition, the N1+1 to N1+N2 pumping units are the second operating condition, and the N1+N2+1 to N pumping units are the third operating condition.

Images