Multi-model predictive control method for rescue drilling rigs drilling in complex formations

Abstract

The present invention discloses a multi-model predictive control method for a rescue drilling rig for drilling in complex strata, comprising the following steps: establishing a drill pressure control system model for the rescue drilling rig using voltage as input and drilling pressure as output for each stratum based on the finite element method; performing rigid-flexible mode separation on the drill pressure control system model for each stratum based on the frequency domain analysis method, and establishing a reduced-order model of the drill pressure control system model for each stratum; determining a corresponding feed damping coefficient for each stratum, and assigning a one-to-one correspondence between multiple strata and multiple reduced-order models, wherein each stratum is pre-configured with a controller; and when the stratum conditions change, switching the corresponding controller to predictively control the drill pressure according to the stratum conditions. The method of the present invention can quickly respond to stratum changes and adjust the drill pressure to maintain the stability and efficiency of the drilling process, thereby ensuring that the drill pressure can always be stable within a safe range of the set value in different strata.

Description

Technical Field

[0001] The present invention relates to the technical field of rescue drilling rigs, and in particular to a multi-model predictive control method for rescue drilling rigs for drilling in complex formations. Background Art

[0002] To ensure rapid rescue after coal mine accidents, China urgently needs fast, mobile rescue equipment with efficient drilling capabilities as the primary rescue force. Rescue drills, widely used in coal mine accidents, must possess stable drilling pressure control capabilities in various ground conditions.

[0003] In the related art, PID control and fuzzy control methods are usually used to control the drilling pressure, which has problems such as the need to adjust parameters in a timely manner and steady-state errors. Summary of the Invention

[0004] The present invention aims to address, at least to some extent, one of the technical problems in the related art. To this end, a first object of the present invention is to propose a multi-model predictive control method for a rescue drilling rig for drilling in complex formations. This method can rapidly respond to formation changes and adjust the drilling pressure to maintain drilling stability and efficiency, thereby ensuring that the drilling pressure remains stable within a safe range of a set value in various formations.

[0005] To achieve the above-mentioned purpose, an embodiment of the first aspect of the present invention proposes a multi-model predictive control method for a rescue drilling rig for drilling in complex formations, the method comprising the following steps: based on the finite element method, a drilling pressure control system model of the rescue drilling rig with voltage as input and drilling pressure as output is established for each formation, the drilling pressure control system model comprising a proportional relief valve model, a hydraulic cylinder system model, a drill string axial model and a drill bit rock action model; based on the frequency domain analysis method, the drilling pressure control system model of each formation is subjected to rigid-flexible mode separation, and a reduced-order model of the drilling pressure control system model of each formation is established; the corresponding feed damping coefficient is determined under each formation, and multiple formations are matched one-to-one with multiple reduced-order models, wherein each formation is pre-configured with a controller; when the formation conditions change, the corresponding controller is switched to perform predictive control of the drilling pressure according to the formation conditions.

[0006] The multi-model predictive control method for a rescue drilling rig for drilling in complex formations, according to an embodiment of the present invention, combines the characteristics of the drill string system and uses the finite element method to establish a WOB control system model. Based on the frequency domain analysis method, the rigid and flexible modes of the WOB control system model are first separated. Then, a corresponding reduced-order model is approximately designed. Multiple reduced-order models are established for different formations, and predictive controllers are designed for each. Based on the formation conditions, a corresponding model predictive control strategy is selected to track WOB when the formation changes. This method can thus quickly respond to formation changes and adjust WOB to maintain the stability and efficiency of the drilling process, ensuring that WOB remains stable within a safe range of set values ​​in different formations.

[0007] Compared with the prior art, the present invention has the following beneficial effects:

[0008] The present invention considers a rescue drilling rig that uses a proportional relief valve-hydraulic cylinder system to provide drilling pressure, and designs an adaptive control method for complex formations for the first time. The drilling pressure control system model is established by the finite element method. In order to reduce the complexity, the rigid mode and the flexible mode in the drilling pressure control system model are separated based on the frequency domain analysis method, and the two parts are approximated to obtain a reduced-order model; different reduced-order models are established for different formations, and predictive control methods for different reduced-order models are designed. The optimal model predictive control strategy for the formation is selected according to the difference between the estimated output and the actual output of each model, thereby completing the accurate tracking of the drilling pressure. The existing technology usually uses PID control and fuzzy control methods, which have problems such as the need to adjust parameters in a timely manner and steady-state errors. The method mentioned in the present invention breaks through the difficult problem of adaptive control of drilling pressure in variable formations, which is of great significance for a safe and efficient drilling process.

[0009] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 Flowchart of a multi-model predictive control method for a rescue drilling rig drilling in complex formations according to an embodiment of the present invention;

[0011] Figure 2 is a schematic diagram of a weight-on-bit control system model according to one embodiment of the present invention;

[0012] Figure 3 is a structural block diagram of a reduced-order model according to an embodiment of the present invention;

[0013] Figure 4 This is a diagram showing output tracking effects for different formations corresponding to different weight-on-bit setting values ​​according to one embodiment of the present invention. DETAILED DESCRIPTION

[0014] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0015] The following describes a multi-model predictive control method for a rescue drilling rig for drilling in complex formations proposed in an embodiment of the present invention with reference to the accompanying drawings.

[0016] Figure 1 The figure is a flow chart of a multi-model predictive control method for a rescue drilling rig for drilling in complex formations according to an embodiment of the present invention.

[0017] like Figure 1 As shown, the multi-model predictive control method for a rescue drilling rig for drilling in complex formations according to an embodiment of the present invention includes the following steps:

[0018] S1. Based on the finite element method, a drilling pressure control system model of the rescue drilling rig is established for each formation, with voltage as input and drilling pressure as output. The drilling pressure control system model includes a proportional relief valve model, a hydraulic cylinder system model, a drill string axial model, and a drill bit rock action model.

[0019] Specifically, the rescue drilling rig provides bit pressure with a proportional relief valve-hydraulic cylinder system, such as Figure 2 As shown in Figure 1, a WOB control system model is established based on the process mechanism of applying WOB by connecting the hydraulic cylinder to the drill string. The WOB control system model can be divided into four main parts: proportional relief valve model, hydraulic cylinder system model, drill string axial model and drill bit rock action model. The actual output pressure of the proportional relief valve is p r Affected by its command voltage v and the flow rate q through the relief valve r The oil pressure actually output by the proportional relief valve will be transmitted to the rodless chamber of the hydraulic cylinder, thereby controlling the output pressure of the hydraulic cylinder. The hydraulic cylinder is connected to the power head through a pulley, thereby controlling the drilling pressure of the drill string system. The drill string system consists of multiple drill tools. According to the finite element method, each drill rod can be regarded as a discrete unit, and their connection points can be regarded as generalized nodes. Assume that the drill string system consists of N drill rods, each drill rod can be numbered 1 to N from top to bottom, and there are N+1 generalized nodes in total. The top of the drill string is rigidly connected to the power head, and the speeds of the two are equal.

[0020] It should be noted that the proportional relief valve model is responsible for regulating the pressure in the hydraulic system to control the drilling pressure of the drilling rig; the hydraulic cylinder system model simulates the behavior of the hydraulic cylinder, converting hydraulic energy into mechanical energy to generate drilling pressure; the drill string axial model describes the mechanical behavior of the drill string in the axial direction, including its force and deformation; the drill bit rock interaction model simulates the mechanical process of interaction between the drill bit and the rock, which affects the drilling pressure and drilling efficiency.

[0021] S2, based on the frequency domain analysis method, the drilling pressure control system model of each formation is separated into rigid and flexible modes, and a reduced-order model of the drilling pressure control system model of each formation is established.

[0022] S3, determining a corresponding feed damping coefficient under each stratum, and making a one-to-one correspondence between the multiple strata and the multiple reduced-order models, wherein each stratum is pre-configured with a controller.

[0023] S4: When the formation conditions change, the corresponding controller is switched to perform predictive control of the drilling pressure according to the formation conditions.

[0024] Specifically, during drilling, the feed damping coefficient changes with strata changes, leading to changes in the model of the controlled object. Therefore, the fixed-parameter reduced-order model has significant limitations. For each stratum, the corresponding feed damping coefficient is determined, a well-matched reduced-order model is established, and a controller is pre-designed for that stratum. When the system operating conditions change from one stratum to another, the previous controller will have difficulty maintaining the original control quality. Therefore, selecting the controller that best matches the stratum based on the stratum conditions will achieve better control quality.

[0025] Specifically, first, the finite element method is used to establish a drilling pressure control system model for each formation. This model takes voltage as input and drilling pressure as output. Then, the frequency domain analysis method is used to separate the rigid and flexible modes of the drilling pressure control system model of each formation to establish a reduced-order model. This can simplify the model and reduce the computational complexity while retaining the key dynamic characteristics of the model to facilitate real-time control.

[0026] Furthermore, a corresponding feed damping coefficient is determined for each formation. This can be determined experimentally or empirically based on the formation's physical properties. Multiple formations are then mapped to multiple reduced-order models, and a controller is preconfigured for each formation. When formation conditions change, the system switches to the corresponding controller for predictive control based on real-time monitoring. This allows for rapid response to formation changes and adjustment of the drilling pressure to maintain drilling stability and efficiency.

[0027] According to one embodiment of the present invention, a drilling pressure control system model of a rescue drilling rig with voltage as input and drilling pressure as output is established for each formation based on the finite element method, including: determining the control input, output and state variables of the drilling pressure control system; and establishing a state space model of the drilling pressure control system according to the input, output and state variables of the drilling pressure control system.

[0028] Furthermore, according to one embodiment of the present invention, the control input of the weight-on-bit control system is expressed by the following expression:

[0029]

[0030] Among them, u is the control input of the bit pressure control system, v is the command voltage of the proportional relief valve, q r is the flow through the proportional relief valve;

[0031] The output of the bit weight control system is the bit weight at the top of the drill string, which is expressed by the following expression:

[0032] y s =F s

[0033] Among them, y s is the output of the bit pressure control system, F s The weight on bit at the top of the drill string;

[0034] The state variables of the bit pressure control system are expressed by the following expressions:

[0035]

[0036] Among them, x s is the state variable, p r is the output pressure of the proportional relief valve, R n represents n-dimensional Euclidean space, x n Expressed as the displacement of the nth generalized node, Expressed as the velocity of the nth generalized node;

[0037] The state space model of the bit pressure control system is established according to the following formula:

[0038]

[0039] in, is the first-order derivative of the state variable, which indicates the change of the system state over time; u(t) is the control input of the system, A s is the system matrix, B s is the input matrix, y s (t) is the output of the system, C s is the output matrix;

[0040] Determine the parameter A in the state space model of the bit pressure control system according to the following formula s 、B s 、C s :

[0041] A s =A sr +Hk r T

[0042] C s =C1-n1mQA s -n1bQ

[0043]

[0044] Among them, k r represents the feed damping coefficient, which indicates the magnitude of the damping experienced by the drill bit at a unit drilling speed and is an uncertain parameter related to the formation strength; the constant n1 is the coefficient of the feed mechanism, which indicates the relationship between the drilling pressure on the well and the output pressure of the hydraulic cylinder and is determined by the pulley structure; b is the viscous damping coefficient, which is obtained from the hydraulic cylinder motion test; m is the total mass of the drill string, O i×j is an i×j dimensional matrix whose elements are all 0;

[0045] The above parameters A, B, C1, D a 、E1、H、H1、K t 、M a , T, and Q are expressed by the following formulas:

[0046]

[0047] C1=[n1A a O 1×(2N+4) ]

[0048] D a =αM+βK+n1bE0Q

[0049]

[0050]

[0051] K t =KE3

[0052] M a =M+n1mE0Q

[0053] T=[00...1]∈R 1×(2N+5)

[0054] Q=[O 1×(N+4) 1O 1×N ]

[0055] Among them, Aa is the effective area of ​​the rodless cavity, K is the global stiffness matrix, M is the global mass matrix, and constants α and β are the model coefficients of the global damping model obtained using the Rayleigh damping model;

[0056] The above parameters A1, A2, A3, B1, B2, B3, E0, E2, E3, W, K, and M are expressed by the following formulas:

[0057]

[0058] E0=[10...0] T ∈R N+1

[0059] E2=[00...1] T ∈R N+1

[0060]

[0061]

[0062] Among them, K cr Indicates the equivalent flow pressure coefficient of the proportional relief valve, ω rv Indicates the dominant turning frequency of the main valve movement, ζ ry Indicates the damping ratio of the pilot valve, ω ry Indicates the natural frequency of the pilot valve, K qr represents flow pressure gain, ω rm Indicates the main valve equivalent natural frequency, ζ rm Represents the equivalent damping ratio of the main valve, P∈R N×N is a matrix whose main diagonal and elements above the main diagonal are all 1;

[0063] The global stiffness matrix K and the global mass matrix M are composed of the element mass matrix M of each discrete element i and the element stiffness matrix K i Combined, M i and K i They are:

[0064]

[0065] i=1,2,…,N

[0066] Where ρ is the density of each drill rod, S is the cross-sectional area of ​​each drill rod, E is the Young's modulus of each drill rod, and l is the length of each drill rod.

[0067] According to one embodiment of the present invention, a rigid-flexible mode separation is performed on the weight-on-bit control system model of each formation based on a frequency domain analysis method, and a reduced-order model of the weight-on-bit control system model of each formation is established, including: separating the weight-on-bit control system model to obtain low-order rigid modes and high-order flexible modes; using a first-order inertia link to describe the rigid modal transfer function; using a fixed multiplicative uncertainty weighting function to approximately describe the high-order flexible modes, and using a trial-and-error method to optimize the amplitude of the uncertainty weighting function.

[0068] Furthermore, according to one embodiment of the present invention, the state space equation of the rigid mode transfer function is expressed according to the following formula:

[0069]

[0070] Among them, x l is the system status, y l Output drilling pressure value for the system, B l is the input matrix, and the state transfer matrix ε is equal to the real part of the dominant pole;

[0071] The uncertainty weighting function is expressed according to the following formula:

[0072]

[0073] Among them, ω1, ω2, ζ1, ζ2, and K1 are constants.

[0074] Specifically, the transfer function matrix corresponding to the above-mentioned weight-on-bit control system model is defined as G h (s). Use the first-order inertial link to describe the low-order rigid mode, and its transfer function matrix is ​​G l (s), the corresponding state space equation is:

[0075]

[0076] For different strata, you can choose B l is a matrix whose elements are all constants. In order to facilitate the parameter selection of the flexible mode, the input matrix B l It is necessary to make the relative error amplitude response of the entire bit weight control system model and the rigid mode under different formations fluctuate within the range of 0 to 60 dB.

[0077] A fixed multiplicative uncertainty weighting function W(s) is used to approximate high-order flexible modes. To reduce the conservatism of the model, it is necessary to ensure that W(s) is close enough to the frequency band corresponding to the flexible mode. It is designed as follows:

[0078]

[0079] In addition, the amplitude of W(s) must be larger than the amplitude of the relative error, and the amplitude of the relative error should be as close as possible in the frequency band where the high-order flexible mode is located. That is, W(s) is obtained by solving the following optimization problem through trial and error:

[0080]

[0081] The structural diagram of the reduced-order model is as follows Figure 3 shown.

[0082] According to one embodiment of the present invention, when the formation conditions change, the corresponding controller is switched to perform predictive control of the drilling pressure according to the formation conditions, including: determining the model parameters of the drilling pressure control system model of each formation according to the transfer function of the reduced-order model of the drilling pressure control system model of each formation; determining an indicator switching function according to the model parameters of the drilling pressure control system model of each formation and the current error and historical error of the current formation output, and judging the matching degree between the drilling pressure control system model of each formation and the current formation through the indicator switching function; when the matching degree between the drilling pressure control system model and the current formation meets the preset conditions, determining the drilling pressure control system model corresponding to the current moment as the target model; and using the controller corresponding to the target model to perform predictive control of the drilling pressure.

[0083] Specifically, the rigid modal transfer function is combined with the uncertainty weighting function under each formation to obtain the reduced-order model transfer function, which is converted into a state-space equation as follows:

[0084]

[0085] Where A i 、B i 、C i are the model parameters after order reduction, and the model parameters are different in different formations i.

[0086] For the above model, the first-order Euler method is used to convert the continuous-time model into a discrete-time model:

[0087]

[0088] Where, T is the sampling time.

[0089] The model parameters of the bit pressure control system model of each formation can be obtained through the above steps.

[0090] Furthermore, according to one embodiment of the present invention, the indicator switching function is determined according to the following formula:

[0091]

[0092] e i (k) = yi (k)-y(k)

[0093] Among them, y i (k) is the output value of the bit pressure control system model of the i-th formation, y(k) is the real output of the bit pressure control system, e i (k) is the difference between the output value of the drilling pressure control system model and the actual system output value of the i-th formation; λ1 is the weight at the current moment, λ2 is the weight of the error at the historical moment, λ1>0,λ2>0; θ is the forgetting factor, which represents the memory effect of historical error information, 0<θ<1; h is the length of the historical error.

[0094] At any time, when (δ represents the number of formations), appropriate model parameters are determined to design a predictive control strategy.

[0095] According to one embodiment of the present invention, a controller corresponding to a target model is used to predict and control the drilling pressure, including: obtaining the model parameters of the target model, and setting the prediction time domain N of the drilling pressure control system. p , control time domain N c , where N c ≤N p ; and define the predicted output vector Y of the bit weight control system k and the predictive control vector U k ; Determine the predicted output vector based on the model parameters of the target model, the prediction time domain, the control time domain, the predicted control vector and the prediction equation; Based on the predicted output vector, the predicted control vector and the quadratic objective function, the optimal control law is obtained; Use the convex optimization toolbox to obtain the future N in the control time domain c The first result in the control sequence is used as the actual control input to act on the WOB control system to control the WOB control system, and then enters the next cycle and repeats the above process to complete the rolling optimization.

[0096] Furthermore, the model parameters of the target model are:

[0097] The predicted output vector is:

[0098] Y k =[y(k+1 / k)y(k+2 / k)…y(k+N c / k)…y(k+N p / k)] T

[0099] Where u(k+η / k) represents the input value at time k+η determined at time k; y(k+η / k) represents the prediction of the output value at time k+η at time k;

[0100] The predictive control vector is:

[0101] U k =[u(k / k)u(k+1 / k)…u(k+N c -1 / k)] T

[0102] The prediction equation is: k =Φ1x k +Γ1U k

[0103] in,

[0104]

[0105] According to one embodiment of the present invention, the quadratic objective function is:

[0106]

[0107] e(k+η / k)=y(k+η / k)-y d

[0108] Among them, Q is the error weight coefficient, R is the input signal weight coefficient, S is the prediction terminal error weight coefficient, y d is the desired weight on bit value;

[0109] The quadratic objective function simultaneous prediction equation is calculated as follows:

[0110]

[0111] in,

[0112]

[0113] Y d By N p y d The column vector composed of is the Kronecker product.

[0114] In order to simplify the control strategy, the input flow rate q is set to a constant value, and the range of the command voltage v is considered. min ~v max , construct the optimization problem of the model predictive control algorithm:

[0115]

[0116] The convex optimization toolbox is used to obtain the future N in the control domain c The first result in the control sequence is used as the actual control input to act on the system, and then enters the next cycle and repeats the above process to complete the rolling optimization.

[0117] The corresponding control algorithm is as follows:

[0118]

[0119]

[0120] A specific embodiment of the present invention is as follows:

[0121] Under three different strata, the following feed damping coefficients and uncertain parameters in the rigid mode are taken, from "soft" to "medium" to "hard": ε1=-8.385×10 -2 , ε2=-4.741×10 -2 , ε3=-3.364×10 -2 .

[0122] The parameters used in the model can be taken as A1 = 0.0154m 2 , b=1000N / (m·s), ω rm =50rad / s,ζ rm =0.9,ω ry =500rad / s,ζ ry =0.8,ω rv =20rad / s, K qr =2.5Mpa / V,K cr =0.016Mpa / L / min, T=0.1s, n1=2, λ1=0.5, λ2=0.5, h=10, θ=0.75. The parameters used in the controller design are Q=1, S=1, B l =[1100 0.1 0.005 0.00002 0.00000002].

[0123] According to the frequency response characteristics of the model and the aforementioned design requirements, the relevant values ​​of the described weighting function W(s) are ω1=10.49, ω2=8.94, ζ1=625, ζ2=50, and K1=1.375.

[0124] In the present invention, the reference WOB value in "soft" formation is set to 20,000 N, the reference WOB value in "medium" formation is set to 25,000 N, and the reference WOB value in "hard" formation is set to 30,000 N.

[0125] like Figure 4 As shown, the weight on bit tracks drilling from the "soft" formation to the "hard" formation and finally to the "medium" formation.

[0126] From the simulation results, the multi-model predictive control method of the rescue drilling rig for drilling in complex formations of the present invention can ensure that the bit pressure can always be stable within the safe range of the set value in different formations.

[0127] The multi-model predictive control method for a rescue drilling rig for drilling in complex formations, according to an embodiment of the present invention, combines the characteristics of the drill string system and uses the finite element method to establish a WOB control system model. Based on the frequency domain analysis method, the rigid and flexible modes of the WOB control system model are first separated. Then, a corresponding reduced-order model is approximately designed. Multiple reduced-order models are established for different formations, and predictive controllers are designed for each. Based on the formation conditions, a corresponding model predictive control strategy is selected to track WOB when the formation changes. This method can thus quickly respond to formation changes and adjust WOB to maintain the stability and efficiency of the drilling process, ensuring that WOB remains stable within a safe range of set values ​​in different formations.

[0128] It should be noted that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0129] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0130] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0131] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0132] In the present invention, unless otherwise specified or limited, the terms "installed," "connected," "connect," "fixed," etc. should be understood in a broad sense. For example, they can refer to fixed connection, detachable connection, or integration; mechanical connection, electrical connection; direct connection, or indirect connection through an intermediate medium; internal communication between two components, or interaction between two components, unless otherwise specified. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0133] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A multi-model predictive control method for a rescue drilling rig drilling in complex formations, characterized in that: The method comprises the following steps: Based on the finite element method, a drilling pressure control system model of the rescue drilling rig is established for each formation, with voltage as input and drilling pressure as output. The drilling pressure control system model includes a proportional relief valve model, a hydraulic cylinder system model, a drill string axial model, and a drill bit rock action model. Based on the frequency domain analysis method, the rigid and flexible modes of the drilling pressure control system model of each formation are separated to establish a reduced-order model of the drilling pressure control system model of each formation; Determining a corresponding feed damping coefficient under each stratum, and assigning a one-to-one correspondence between the plurality of strata and the plurality of reduced-order models, wherein each stratum is pre-configured with a controller; When the formation conditions change, the corresponding controller is switched according to the formation conditions to perform predictive control of the drilling pressure. A drilling pressure control system model of the rescue drilling rig with voltage as input and drilling pressure as output is established for each formation based on the finite element method, including: determining control inputs, outputs, and state variables of the weight-on-bit control system; Establishing a state space model of the weight-on-bit control system according to the input, output and state variables of the weight-on-bit control system; The control input of the weight-on-bit control system is expressed by the following expression: Wherein, u is the control input of the bit pressure control system, v is the command voltage of the proportional relief valve, and q r is the flow rate through the proportional relief valve; The output of the weight-on-bit control system is the weight-on-bit at the top of the drill string, which is expressed by the following expression: y s =F s Among them, y s is the output of the bit pressure control system, F s The weight on bit of the top of the drill string; The state variables of the weight-on-bit control system are expressed by the following expressions: Among them, x s is the state variable, p r is the output pressure of the proportional relief valve, R n represents n-dimensional Euclidean space, x n Expressed as the displacement of the nth generalized node, Expressed as the velocity of the nth generalized node; The state space model of the weight-on-bit control system is established according to the following formula: in, is the first-order derivative of the state variable, which indicates the change of the system state over time; u(t) is the control input of the system, A s is the system matrix, B s is the input matrix, y s (t) is the output of the system, C s is the output matrix; The parameter A in the state space model of the bit pressure control system is determined according to the following formula s 、B s 、C s : A s =A sr +Hk r T C s =C1-n1mQA s -n1bQ Among them, k r represents the feed damping coefficient, which indicates the magnitude of the damping experienced by the drill bit at a unit drilling speed and is an uncertain parameter related to the formation strength; the constant n1 is the coefficient of the feed mechanism, which indicates the relationship between the drilling pressure on the well and the output pressure of the hydraulic cylinder and is determined by the pulley structure; b is the viscous damping coefficient, which is obtained from the hydraulic cylinder motion test; m is the total mass of the drill string, O i×j is an i×j dimensional matrix whose elements are all 0; Parameters A, B, C1, D a 、E1、H、H1、K t 、M a , T, and Q are expressed by the following formulas: C1=[n1A a O 1×(2N+4) ] D a =αM+βK+n1bE0Q K t =KE3 M a =M+n1mE0Q T=[00...1]∈R 1×(2N+5) Q=[O 1×(N+4) 1O 1×N ] Among them, A a is the effective area of ​​the rodless cavity, K is the global stiffness matrix, M is the global mass matrix, and constants α and β are the model coefficients of the global damping model obtained using the Rayleigh damping model; Parameters A1, A2, A3, B1, B2, B3, E0, E2, E3, W, K, and M are expressed using the following formulas: E0=[1 0 ... 0] T ∈R N+1 <h2 style=";text-align:left;direction:ltr">E2=[00...1]<h2 style=";text-align:left;direction:ltr"> T <h2 style=";text-align:left;direction:ltr"> ∈R<h2 style=";text-align:left;direction:ltr"> N+1 Among them, K cr Indicates the equivalent flow pressure coefficient of the proportional relief valve, ω rv Indicates the dominant turning frequency of the main valve movement, ζ ry Indicates the damping ratio of the pilot valve, ω ry Indicates the natural frequency of the pilot valve, K qr represents flow pressure gain, ω rm Indicates the main valve equivalent natural frequency, ζ rm Represents the equivalent damping ratio of the main valve, P∈R N×N is a matrix whose main diagonal and elements above the main diagonal are all 1; The global stiffness matrix K and the global mass matrix M are composed of the element mass matrix M of each discrete element i and the element stiffness matrix K i Combined, M i and K i They are: Where ρ is the density of each drill rod, S is the cross-sectional area of ​​each drill rod, E is the Young's modulus of each drill rod, and l is the length of each drill rod; When the formation conditions change, the corresponding controller is switched to perform predictive control of the drilling pressure according to the formation conditions, including: determining model parameters of the weight-on-bit control system model of each formation according to a transfer function of a reduced-order model of the weight-on-bit control system model of each formation; Determining an indicator switching function based on the model parameters of the weight-on-bit control system model of each formation and the current error and historical error output by the current formation, and evaluating the degree of matching between the weight-on-bit control system model of each formation and the current formation using the indicator switching function; When the degree of matching between the weight-on-bit control system model and the current formation meets a preset condition, determining the weight-on-bit control system model corresponding to the current moment as the target model; Using a controller corresponding to the target model to predict and control the bit weight; The indicator switching function is determined according to the following formula: and i (k)=y i (k)-y(k) Among them, y i (k) is the output value of the bit weight control system model of the i-th formation, y(k) is the real output of the bit weight control system, e i (k) is the difference between the output value of the drilling pressure control system model and the actual system output value of the i-th formation; λ1 is the weight at the current moment, λ2 is the weight of the error at the historical moment, λ1>0, λ2>0; θ is the forgetting factor, which represents the memory effect of historical error information, 0<θ<1; h is the length of the historical error.

2. The multi-model predictive control method for rescue drilling rigs for drilling in complex formations according to claim 1 is characterized in that: The rigid-flexible mode separation of the weight-on-bit control system model of each formation is performed based on the frequency domain analysis method, and a reduced-order model of the weight-on-bit control system model of each formation is established, including: Separating the weight-on-bit control system model to obtain low-order rigid modes and high-order flexible modes; The first-order inertia link is used to describe the rigid modal transfer function; A fixed multiplicative uncertainty weighting function is used to approximately describe the high-order flexible modes, and the amplitude of the uncertainty weighting function is optimized using a trial-and-error method.

3. The multi-model predictive control method for rescue drilling rigs for drilling in complex formations according to claim 2 is characterized in that: The state space equation for the rigid mode transfer function is expressed as follows: Among them, x l is the system status, y l Output drilling pressure value for the system, B l is the input matrix, and the state transfer matrix ε is equal to the real part of the dominant pole; The uncertainty weighting function is expressed according to the following formula: Among them, ω1, ω2, ζ1, ζ2, and K1 are constants.

4. The multi-model predictive control method for rescue drilling rigs for drilling in complex formations according to claim 2 is characterized in that: Using a controller corresponding to the target model to predict and control the weight on bit includes: Obtain the model parameters of the target model and set the prediction time domain N of the bit pressure control system p , control time domain N c , where N c ≤N p ; and define the predicted output vector Y of the bit weight control system k and the predictive control vector U k ; Determining the predicted output vector according to the model parameters of the target model, the prediction time domain, the control time domain, the predicted control vector, and a prediction equation; An optimal control law based on the predicted output vector, the predicted control vector and a quadratic objective function; The convex optimization toolbox is used to obtain the future N in the control domain c The first result in the control sequence is used as the actual control input to act on the weight-on-bit control system to control the weight-on-bit control system, and then enters the next cycle and repeats the above process to complete the rolling optimization.

5. The multi-model predictive control method for rescue drilling rigs for drilling in complex formations according to claim 4 is characterized in that: The model parameters of the target model are: The predicted output vector is: AND k =[y(k+1 / k) y(k+2 / k) … y(k+N c / k) … and (k+N p / k)] T Where u(k+η / k) represents the input value at time k+η determined at time k; y(k+η / k) represents the prediction of the output value at time k+η at time k; The predictive control vector is: YOU k =[u(k / k) u(k+1 / k) … u(k+N c (-1 / k)] T The prediction equation is: k =Φ1x k +Γ1U k in, 6. The multi-model predictive control method for rescue drilling rigs for drilling in complex formations according to claim 5 is characterized in that: The quadratic objective function is: e(k+η / k)=y(k+η / k)-y d Among them, Q is the error weight coefficient, R is the input signal weight coefficient, S is the prediction terminal error weight coefficient, y d is the desired weight on bit value; The quadratic objective function is calculated by combining the prediction equations: in, Y d By N p y d The column vector composed of is the Kronecker product.

Images