A method for coordinated control of flow and pressure under high-pressure fluid operation conditions in deep wells
By using a virtual cross-impedance decoupling model and phase alignment technology, the problems of signal delay and parameter coupling in deep well high-pressure fluid operations are solved, enabling coordinated control of flow and pressure, and ensuring the stability and high-precision regulation of the system under extreme conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUNAN PUTAI FILLING MINING EQUIP CO LTD
- Filing Date
- 2026-03-23
- Publication Date
- 2026-05-26
AI Technical Summary
In deep well high-pressure fluid operation, traditional fixed-parameter control strategies are difficult to achieve dynamic decoupling, resulting in signal delay and parameter coupling, system oscillation, and inconsistent actuator response, making it difficult to achieve high-precision flow and pressure regulation.
A virtual cross-impedance decoupling model is adopted. By establishing a state feature vector, a logical control channel between flow and pressure is constructed. Using cross-correlation algorithm and phase alignment technology, the gain matrix is adjusted in real time, and a nonlinear damping component is injected to achieve coordinated regulation of flow and pressure.
It achieves high-precision flow and pressure control in deep well high-pressure environments, eliminates energy coupling and oscillation, ensures the stability and smoothness of the system under extreme conditions, and avoids the risk of pressure runaway.
Smart Images
Figure CN121900539B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method for coordinated control of flow and pressure under deep well high-pressure fluid operation conditions, belonging to the field of automation control technology. Background Technology
[0002] In the process of precision control of industrial fluids, especially in current deep well fluid operations involving the control of surface pump displacement and wellhead pressure, the fluid column of several kilometers has compressibility and tubing elasticity, causing a lag in the transmission of control signals. This signal delay and parameter coupling caused by the characteristics of the transmission medium is a typical common technical problem faced in the field of non-electric variable control. Since the time constant of the controlled object will drift in real time with environmental parameters, traditional fixed parameter control strategies are difficult to achieve dynamic decoupling while ensuring system stability. Conventional solutions adopt an independent loop architecture of pump flow regulation and back pressure valve pressure regulation. Under the high pressure conditions in deep earth, the fluid is not a rigid medium. Adjusting the pump speed generates a pressure slope that triggers the back pressure valve to act. The back pressure valve regulation inversely interferes with the pump end displacement output. The mutual disturbance logic defect causes the system to fall into compensating oscillation and triggers pressure relief protection.
[0003] Simply relying on the physical form of mechanical actuators, such as pump and valve components, to improve response inertia and dead zone characteristics is insufficient to overcome the bottleneck of control accuracy. The industry is trying to shift to software control strategy optimization, attempting to avoid hardware performance shortcomings through algorithm compensation. For example, Chinese invention patent application CN115826397A discloses a pressure control method for long-distance natural gas pipelines. The control center writes the pressure setpoint once, and dynamically updates the pressure process setpoint SPT based on the gradient of the deviation between the real-time pressure value and the setpoint, guiding the PID controller to adjust smoothly. The solution introduces threshold limitation and gradient tracking mechanism in the setpoint generation stage to solve the pressure overshoot problem caused by sudden load changes in long-distance pipelines. The industry has also tried to reduce the regulation bandwidth or establish a static time delay compensation model. However, changes in temperature and pressure gradients cause fluid density and elastic modulus to drift in real time. Static models predict the compensation time, but the phase of the physical fluctuation peak is out of sync. Increasing the sampling frequency or physical damping cannot eliminate the energy coupling between flow and pressure.
[0004] Therefore, the technical problem to be solved by this invention is how to eliminate energy coupling between regulating loops to suppress system oscillations, solve the compensation phase deviation caused by the evolution of fluid properties with temperature and pressure environment, and overcome the problem of dynamic pressure fluctuation suppression caused by inconsistent response modes of actuators in deep well high-pressure fluid operation conditions. Summary of the Invention
[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: A method for coordinated control of flow rate and pressure under deep well high-pressure fluid operation conditions, comprising the following steps:
[0006] Step 101: Obtain the instantaneous output flow rate of the power actuator, the real-time pressure value of the controlled system output, the operating displacement parameters of the power actuator, and the transient pressure change rate of the controlled system output; combine the operating displacement parameters and the transient pressure change rate into a vector to establish a state feature vector characterizing the energy state of the system.
[0007] Step 102: Input the flow regulation command and pressure limit command into the virtual cross impedance decoupling model to calculate the coordinated control sequence consisting of the flow control component and the pressure compensation component.
[0008] Step 103: When issuing the coordinated control sequence, superimpose the encoded signal with a preset frequency and preset amplitude into the flow control component to obtain a control sequence containing logic probes;
[0009] Step 104: Obtain pressure feedback data at the output of the controlled system driven by the controlled sequence; calculate the difference between the pressure feedback data and the preset pressure reference value to obtain the pressure feedback residual; use the cross-correlation algorithm to calculate the dynamic phase angle difference between the encoded signal and the pressure feedback residual.
[0010] Step 105: Update the time constant of the long-distance transmission channel according to the dynamic phase angle difference; use the time constant to perform time-domain translation processing on the pressure compensation component so that the anti-phase pressure feedforward signal and the pressure fluctuation caused by the long-distance transmission channel are phase aligned at the output of the controlled system.
[0011] Step 106: While adjusting the output intensity of the power actuator, control the output of the adjustment actuator to output a counter-current signal after phase alignment processing;
[0012] Step 107: Calculate the product of instantaneous output flow rate and real-time pressure value in real time to obtain energy flow rate; calculate the change of the second derivative of energy flow rate with respect to time; when the change of the second derivative exceeds the preset singularity threshold, inject a nonlinear damping component into the coordinated control sequence and switch to single-variable pressure holding mode.
[0013] Preferably, the virtual cross-impedance decoupling model has a dynamic gain matrix; step 102 further includes: extracting the pressure slope of the real-time pressure value within a preset time window and the flow gradient of the instantaneous output flow within the same time window; calculating the nonlinear proportional relationship between the pressure slope and the flow gradient; and updating the dynamic gain matrix using the nonlinear proportional relationship to adjust the logical isolation weight between the flow control component and the pressure compensation component.
[0014] Preferably, the method further includes the following steps: Step 301, obtaining the dynamic response characteristics of the power actuator and the regulation actuator, and calculating the response bandwidth deviation value between them; Step 302, constructing a virtual phase matching operator based on the response bandwidth deviation value; Step 303, using the virtual phase matching operator to perform filtering and delay processing on the pressure compensation component, so that the action of the regulation actuator and the physical disturbance generated by the power actuator remain synchronized on the time axis.
[0015] Preferably, in step 107, after calculating the change in the second derivative of the energy flow rate, the method further includes: calculating the rate of change of the energy entropy value of the controlled system; and when the rate of change of the energy entropy value exceeds a preset entropy increase threshold, forcibly setting the weight coefficient of the virtual cross impedance decoupling model to zero.
[0016] Preferably, before step 103, the method further includes: extracting periodic flow pulsations present in the operation of the power actuator as an endogenous detection signal; calculating the frequency domain phase difference between the endogenous detection signal and the corresponding pressure response signal at the output end of the controlled system; and determining the evolution trend of the equivalent elastic modulus of the channel medium based on the frequency domain phase difference.
[0017] Preferably, the method further includes: using the evolution trend of the equivalent elastic modulus to correct the gain coefficient of the virtual cross impedance decoupling model online, so as to compensate for the property drift of the channel medium.
[0018] Preferably, the method further includes: extracting the power consumption data of the drive motor at the power execution end, calculating the conversion efficiency residual of the drive motor power consumption data relative to the instantaneous output flow rate; converting the conversion efficiency residual into the execution stiffness attenuation factor of the power execution end; and increasing the gain of the flow control loop according to the execution stiffness attenuation factor.
[0019] Preferably, the frequency of the encoded signal is lower than the stroke frequency of the power actuator; in step 104, the dynamic phase angle difference is calculated using a cross-correlation algorithm, specifically: calculating the correlation coefficient between the encoded signal and the pressure feedback residual within the sliding window; locking the time delay when the correlation coefficient reaches its maximum value, and converting the time delay into a dynamic phase angle difference.
[0020] Preferably, in step 107, the convergence of the system is determined by calculating the instantaneous energy entropy value H, and the instantaneous energy entropy value H follows the following rules: Where P is the real-time pressure value at the output of the controlled system, and Q is the instantaneous output flow rate at the power actuator. The preset system baseline power value is used; when the rate of change of H is continuously positive and exceeds the preset threshold, the system is determined to have entered an uncontrolled growth mode.
[0021] Preferably, in step 105, the advance amount of the time-domain translation processing is controlled by the real-time compensation of the system boundary temperature field parameters; the method further includes: obtaining the real-time temperature value of the channel medium at the system output end; querying a preset density mapping table based on the real-time temperature value to obtain the medium sound velocity correction amount, and using the medium sound velocity correction amount to correct the time constant.
[0022] Compared with the prior art, the beneficial effects of the present invention are:
[0023] 1. In the coordinated regulation of flow and pressure, the virtual cross-impedance mapping mechanism transforms the energy coupling relationship in the regulation process of non-electrical variables, including flow and pressure in the closed wellbore, into a non-interfering logical control channel. By constructing an abstract impedance mathematical model based on the physical system, complex cross-dimensional disturbances are transformed into predictable logical compensation components, thereby achieving high-precision regulation of non-electrical variables. This avoids logical interference caused by the speed regulation action of the surface injection pump and the back pressure valve, eliminates the reciprocating compensation oscillation induced by mutual disturbances in the linear single-loop architecture, and keeps the system regulation trajectory within the preset phase plane boundary.
[0024] 2. The periodic flow pulses generated by the injection pump stroke are repeatedly used as detection signals. The phase difference characteristic value between the signal and the corresponding wellhead pressure response signal is extracted. The evolution trend of the equivalent elastic modulus of the wellbore fluid is inverted in real time and the control model gain is corrected. This solves the model mismatch problem caused by the nonlinear drift of fluid physical properties in ultra-deep and high-pressure environments, and realizes online perception of fluid physical state in sensorless environments. A virtual phase matching operator is constructed to compensate for the response bandwidth deviation between the large inertia injection pump and the high-precision backpressure valve. The feedforward compensation signal generated by flow regulation is phase-aligned with the physical disturbance on the time axis, eliminating the secondary pressure shock induced by the asymmetry of the actuator response speed, so that the wellhead pressure trajectory presents a monotonous and smooth transition characteristic at the moment of displacement step adjustment.
[0025] 3. Monitor the change in the second derivative of the energy flow rate evolution characteristics of the controlled system. At the moment when the formation load impedance experiences a step drop singularity, inject a nonlinear damping component into the coordinated control sequence to constrain the transient output power gradient of the injection pump. This enables the control logic to identify sudden changes in physical state and smoothly switch control modes, avoiding the risk of pressure runaway at the moment of fracturing. Extract the energy input data of the injection pump drive end and calculate the residual of the conversion efficiency relative to the flow output response. Convert this residual into the stiffness attenuation factor of the actuator and dynamically adjust the gain coefficient of the flow control loop to compensate for the execution deviation caused by the physical degradation of the injection pump. This achieves deep integration of the control logic and the physical loss state of the actuator. Attached Figure Description
[0026] Figure 1 This is a flowchart of the flow and pressure coordinated control method of virtual impedance decoupling and phase alignment according to the present invention;
[0027] Figure 2 This is the logic diagram for the stability monitoring and mode switching control of energy flow rate and entropy value in this invention. Detailed Implementation
[0028] The following detailed description, in conjunction with the accompanying drawings and embodiments, illustrates a method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions provided by the present invention. The embodiments described below are intended to explain the present invention and are not intended to limit the scope of protection of the present invention.
[0029] This invention provides a method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions. Through multi-variable decoupling and phase adaptive calibration at the logic level, it addresses the energy coupling relationship between flow and pressure within the wellbore. This method essentially belongs to a non-electrical variable control or regulation system, providing a general adaptive control framework for complex parameter fields. It is applicable not only to deep well operations but can also be extended to industrial fluid control scenarios with similar variable time-delay characteristics. In the initial stage of system operation, an initial state definition procedure is executed, and its objects include the power actuator and the controlled system. The power actuator is selected with a rated displacement of... to The ground injection pump unit has a displacement sensor sampling frequency of no less than 50Hz. A pressure sensor with a range of 0MPa to 140MPa and an accuracy class of 0.1%FS is installed at the output of the controlled system. The system extracts the real-time pressure value P from the output of the controlled system and calculates its pressure slope within a 2.0s sliding window, along with the flow gradient of the instantaneous output flow rate Q from the power actuator within the same window. The controller divides the pressure slope by the flow gradient to calculate a nonlinear proportional relationship and uses this relationship to adjust the dynamic gain matrix in real time. The weights of the mutual coupling terms; when an increase in pressure slope is detected due to the near-wellbore flow restriction effect, the system will dynamically adjust the gain matrix. The weighting coefficient representing the impact of flow disturbance on pressure control was reduced from 0.15 to 0.08. This operation achieves logical isolation between the flow control component and the pressure compensation component, eliminating self-excited oscillations induced by loop cross-intervention. In deep well fluid operation environments, the fluid column, several kilometers long, has compressibility and tubing elasticity, causing a delay in the transmission of pressure fluctuations generated by the flow rate adjustment at the surface pump end to the wellhead. Furthermore, the pressure compensation action of the back pressure valve can reversely interfere with the flow output at the pump end, generating oscillations induced by loop intervention. To address this technical problem, this method collects the instantaneous output flow rate Q of the power actuator, the real-time pressure value P at the output end of the controlled system, the operating displacement parameter s of the power actuator, and the transient pressure change rate dP / dt at the output end of the controlled system. The operating displacement parameter s and the transient pressure change rate dP / dt are vector-combined to establish a state feature vector representing the energy state of the system. .
[0030] Because of energy coupling between the control loops, direct control would cause the system's control trajectory to deviate from the preset phase plane boundary. Therefore, the system employs a virtual cross-impedance decoupling model to reconstruct the control commands. The flow control command and pressure limiting command are input into the virtual cross-impedance decoupling model to calculate the coordinated control sequence composed of the flow control component and the pressure compensation component. This model has a dynamic gain matrix. The update procedure includes: extracting the pressure slope of the real-time pressure value P within a preset time window and the flow gradient of the instantaneous output flow rate Q within the same time window; calculating the nonlinear proportional relationship between the pressure slope and the flow gradient; and using this nonlinear proportional relationship to update the dynamic gain matrix in real time. The elements in the matrix adjust the logical isolation weights between the flow control component and the pressure compensation component. When the ratio of the pressure slope to the flow gradient increases by 15%, the dynamic gain matrix... The interference coefficient of the flow control component on the pressure compensation loop is reduced, the cross-variance between loops is decreased, and logical isolation is achieved. To address the issue of phase deviation in fixed parameter compensation caused by fluid property drift with temperature and pressure changes, the system implements a calibration mechanism. When issuing the coordinated control sequence, a pre-set frequency and pre-set amplitude encoded signal are superimposed onto the flow control component. As a logic probe, a control sequence containing the logic probe is obtained. Through this active excitation and detection method, the control system has the ability to sense changes in the transfer function of the transmission channel online without relying on external sensors; encoded signal The frequency is set to 0.1Hz, which is 0.5Hz lower than the stroke frequency of the power actuator, and its amplitude is set to 2% to 5% of the average flow command value; the system acquires the pressure feedback data at the output of the controlled system driven by the controlled sequence, calculates the difference between this data and the preset pressure reference value to obtain the pressure feedback residual. The coded signal is calculated using a cross-correlation algorithm. With pressure feedback residual The correlation coefficient C(t) within the sliding window is used; the time delay Δt at which the correlation coefficient C(t) reaches its maximum value is locked, and this time delay is converted into a dynamic phase angle difference. According to the dynamic phase angle difference Update the time constant τ of the long-distance transmission channel; use the time constant τ to perform time-domain shift processing on the pressure compensation component, so that the anti-phase pressure feedforward signal and the pressure fluctuation caused by the long-distance transmission channel are phase aligned at the output of the controlled system.
[0031] While adjusting the output intensity of the power actuator, the system controls and adjusts the phase-aligned offset signal output of the actuator to reduce the secondary pressure shock induced by the asymmetry of the actuator's response speed. To ensure the stability of the system under extreme conditions, the system calculates the product of the instantaneous output flow rate Q and the real-time pressure value P in real time to obtain the energy flow rate. ; Calculate the change in the second derivative of the energy flux W with respect to time. In the change of the second derivative When the preset singularity threshold is exceeded, a nonlinear damping component is injected into the coordinated control sequence and the system switches to a single-variable pressure-holding mode. The system determines its convergence by calculating the instantaneous energy entropy value H, which follows the calculation rules below: ,in The system uses a preset baseline power value. When the rate of change of H remains positive and exceeds the preset entropy increase threshold, the system is determined to have entered an uncontrolled growth mode. The weight coefficients of the virtual cross-impedance decoupling model are set to zero. For execution deviations caused by physical degradation at the power actuator, the system extracts the power consumption data of the drive motor at the power actuator. Calculate the power consumption data of the drive motor Conversion efficiency residual relative to instantaneous output flow rate Q The system will convert the efficiency residual Converted to the actuation stiffness attenuation factor of the power actuator According to the stiffness attenuation factor To increase the gain of the flow control loop and compensate for internal leakage caused by pump and valve erosion, and to reduce the response bandwidth deviation between the ground injection pump and the back pressure valve, the system acquires the dynamic response characteristics of both and calculates the response bandwidth deviation value between them. Based on the response bandwidth deviation value A virtual phase matching operator is constructed and used to filter and delay the pressure compensation component, so that the action of the regulating actuator and the physical disturbance generated by the power actuator are synchronized on the time axis. This method enables the system to adaptively adjust when faced with fluid property drift by virtual impedance mapping at the logic level and phase alignment in the time domain, ensuring that the pressure of the controlled system converges smoothly.
[0032] Example 1: In a hydraulic fracturing operation scenario with a well depth exceeding 7000 meters, a fluid column of several kilometers generates a transmission hysteresis of more than 12 seconds, and the equivalent elastic modulus of the fluid drifts in real time with temperature and pressure fields. This causes the conventional static compensation model to induce reciprocating oscillations of the system due to phase tearing. The system collects the instantaneous output flow rate Q of the power actuator, the real-time pressure value P of the controlled system output, the operating displacement parameter s, and the transient pressure change rate dP / dt; and establishes a state feature vector. In the formula, X is the system energy state characteristic vector; s is the operating displacement parameter; dP / dt is the transient pressure change rate at the output of the controlled system. The flow regulation command and pressure limit command are input into a virtual cross-impedance decoupling model, utilizing a dynamic gain matrix. The algorithm architecture transforms flow demand and pressure constraints into two logically isolated channels. To ensure logical isolation between flow control and pressure limiting, a virtual cross-impedance decoupling model is constructed based on the diagonal dominance principle of multivariable control theory. The model maps the strongly coupled physical layer system into two independent single-input single-output channels. The generation of the coordinated control sequence follows the following deterministic state-space operation procedure: Define the input error vector. , The difference between the instantaneous output flow rate and the flow rate adjustment command. The difference between the real-time pressure value and the pressure limit command is calculated using a dynamic gain matrix. Perform a linear transformation on the input error vector, and the calculation formula is as follows: ,in, The output flow control component is set to the power actuator and has a value range of [0,1]. The output pressure compensation component is set to the regulating actuator and has a value range of [0,1]. The main channel gain coefficient represents the independent adjustment strength of the flow loop and the pressure loop, respectively. This is a pressure-flow cross-decoupling term used to counteract the reverse suppression effect of pressure fluctuations on pumping capacity; This is a flow-pressure cross-coupling term used to counteract the forward impact of a displacement step on wellhead pressure; determine the cross-coupling terms in the matrix. and For specific values, the system executes the following online parameter calibration procedure: The ratio λ of the pressure slope to the flow gradient is calculated in real time, and this ratio is substituted into a preset nonlinear mapping function f(λ). The mapping function is derived based on the fluid network dynamics equations and is expressed as follows: Where κ is the system's inherent impedance coefficient, determined through step response testing during the system initialization phase. When the pressure slope λ increases due to decreased formation permeability, the cross-term... It automatically increases, enhancing the ability to resist virtual impedance to pressure interference.
[0033] The control logic superimposes a 0.1Hz coded signal onto the flow control component. As a logic probe, the cross-correlation algorithm is used to solve the encoded signal. With pressure feedback residual The correlation coefficient between them is used to lock the dynamic phase angle difference. According to the dynamic phase angle difference The time constant τ of the transmission channel is updated in real time, and the pressure compensation component in the coordinated control sequence is time-domain shifted. At this time, the anti-phase pressure feedforward signal is phase-aligned with the fluid physical fluctuations on the time axis. The phase-calibrated counter-shock signal output by the adjustment terminal is used to offset the pressure shock caused by the displacement step. The real-time phase characteristics provided by the logic probe provide closed-loop feedback for the correction of the time constant τ, ensuring that the feedforward counter-shock signal can accurately cover the peak value of the fluid fluctuations. When the operation reaches the formation fracturing point, the load impedance drops sharply. The system calculates the energy flow rate W by real-time monitoring the product of the instantaneous output flow rate Q and the real-time pressure value P, and performs a second derivative operation on the energy flow rate W with respect to time to obtain the change. When the change was detected When the preset singularity threshold is exceeded, the control logic injects a nonlinear damping component into the coordinated control sequence and switches to a single-variable pressure-holding mode to limit the transient output power gradient at the power actuator. The system determines its convergence by calculating the instantaneous energy entropy value H, and the calculation rules are as follows: Where H is the instantaneous energy entropy value; P is the real-time pressure value; and Q is the instantaneous output flow rate. Assuming a preset system baseline power value, and with the instantaneous energy entropy value H changing within a preset threshold range, flow and pressure coordinated control is implemented, and the power consumption data of the drive motor at the power execution end is recorded. Used to calculate the conversion efficiency residual relative to the instantaneous output flow rate Q Conversion efficiency residual Converted into an execution stiffness attenuation factor The system is based on the stiffness attenuation factor. Increase the gain coefficient of the flow control loop to compensate for the adjustment deviation caused by the physical loss of the pump set; the driving action of the power actuator and the counteracting action of the adjustment actuator are kept synchronized under the coordination of the virtual phase matching operator, eliminating the secondary pressure shock caused by the actuator response bandwidth deviation. The system maintains the energy state of the controlled system within the preset phase plane boundary by impedance decoupling in the energy dimension and phase alignment in the time dimension.
[0034] Example 2: During the experimental phase, a fluid dynamics test platform with high-pressure operating condition simulation capabilities was used, and the rated displacement of the high-pressure pump set at the actuator end was set to 2.0. The displacement sensor sampling frequency was set to 50Hz, and the pressure acquisition accuracy at the output of the controlled system was set to 0.05%FS. The test platform simulated a wellbore environment at a depth of 7000 meters by adjusting the acoustic parameters of the long-distance pipeline string. The test data was acquired in real time by the sensor array of the physical test platform, with a sampling frequency of... The sampling frequency is set to 50Hz; this setting is based on the balance between the spectral width of the monitored signal and the processing load of the controller; as the cutoff frequency of the monitored signal increases, the sampling frequency... To avoid Nyquist signal aliasing, the flow regulation command is shifted to the upper limit of its value range accordingly. In the flow step response test, the flow regulation command is changed from 1.0. Step jump to 1.5 Additive white Gaussian noise with a signal-to-noise ratio of 18dB is superimposed on the pressure feedback signal, while simulating power frequency interference at a frequency of 50Hz. The encoded signal... The amplitude is set to 3% of the average flow rate; this setting considers the trade-off between extraction accuracy and the mechanical life of the actuator; when the background noise power spectral density of the detection environment increases, the encoded signal... The amplitude is increased accordingly to maintain the extraction of dynamic phase difference. The required signal-to-noise ratio.
[0035] After the test is started, the power actuator receives a superimposed coded signal. The system extracts the pressure feedback residual from the control sequence. The cross-correlation coefficient C(t) was calculated, and the time constant τ of the current transmission channel was determined to be 12.43s. The pressure compensation component was then shifted accordingly. See Table 1 for a comparison of the performance of flow and pressure coordinated control. Control group 1 used an independent proportional-integral-differential loop, control group 2 used virtual impedance decoupling without phase calibration, and the experimental group used the aforementioned specific implementation method. Control group 1 generated a pressure fluctuation with a peak value of 8.45MPa after a sudden change in flow, and the system was still in an oscillating state after 45.2s. Control group 2 reduced the peak value of the fluctuation to 4.12MPa, but due to a phase deviation of 1.86s between the compensation signal and the physical fluctuation, a secondary impact was generated. The experimental group suppressed the peak value of the pressure fluctuation to within 1.15MPa by updating the time constant τ, and the pressure trajectory converged smoothly within 18.3s.
[0036] Table 1: Comparison of Flow and Pressure Coordinated Regulation Performance
[0037]
[0038] In gradient verification at different depths, the simulated well depths were set to 5000 m, 6000 m, and 7000 m. As the well depth increased, the equivalent elastic modulus of the fluid column underwent a nonlinear shift, resulting in a dynamic phase angle difference. As the angle increases from 18.5° to 32.4°, the system automatically corrects the time constant τ and shifts the compensation sequence, ensuring that the control deviation at different depths is within ±0.25MPa, specifically for the encoded signal. Boundary tests of amplitude show that when the amplitude increases from 1% to 5%, the dynamic phase difference... The identification accuracy was improved by 24.6%; when the amplitude exceeded 6.2%, the displacement fluctuation at the power actuator affected the process stability, and the nonlinear harmonic components in the pressure feedback signal increased, confirming that the preset amplitude range is the optimal working window that balances identification accuracy and system steady state; the experimental results confirm that the rate of change of the instantaneous energy entropy value H can capture the critical point of the system entering the uncontrolled growth mode, when the simulated formation fracturing leads to During a sudden increase, the system injects a nonlinear damping component into the coordinated control sequence, reducing the transient power gradient by 31.8% and avoiding the risk of pressure runaway. This method enables the perception and recognition of transmission hysteresis through a cross-correlation algorithm, transforming the physical time delay challenge into an adjustable parameter that can be logically offset, and completing the full-link logical closed-loop verification from feature extraction to safety boundary arbitration.
[0039] Example 3: During the transmission of high-pressure fluid through a pipe column with a vertical depth of 8500 meters, the system faces the challenge of compensating for phase deviation caused by the fluctuation of sound velocity due to temperature and pressure gradients. The transmission delay from ground commands to the output of the controlled system exhibits nonlinear drift characteristics. The controller runs a calibration program, and the system synchronously acquires the coded signal from the power actuator. The pressure feedback residual at the output of the sequence and the controlled system The sequence has a sliding window length L set to 1024 sampling points. This parameter is used to balance the accuracy of correlation calculation with the real-time performance of the operation. The controller performs discrete shift-multiply-add operations on the sequence within the time domain window to determine the correlation coefficient C(m). The calculation rules are as follows: Where C(m) is the discrete correlation coefficient sequence; m is the discrete shift step size; These are the discrete sampled values of the encoded signal; L represents the discrete sampled value of the pressure feedback residual after the corresponding shift; L is the window length, and the controller locks the maximum value of the correlation coefficient C(m) within the preset time delay search interval, where the locked discrete shift step size is... The time constant τ of the corresponding transmission channel is 14.62s. The system uses this value to shift the pressure compensation component. At this time, the anti-phase pressure feedforward signal achieves phase alignment at the output of the controlled system, and the physical fluctuations caused by the flow regulation action are canceled by the counteracting signal.
[0040] Under the condition that the load impedance drops due to formation fracturing, the system monitors the energy flow rate W and its second derivative with respect to time. The singularity threshold was determined through a standardization procedure. During the system's stable operation period, 2000 data samples were collected, and the singularity threshold was set to the standard deviation of the second derivative of the sample sequence. 3.5 times that, the measured change in the second derivative From 4.2 It suddenly increased to 18.5 The system determines that it has entered a singularity mode. The control logic injects a nonlinear damping component into the coordinated control sequence and switches to a single-variable voltage holding mode. This action limits the power gradient at the power actuator. The system determines the instantaneous energy entropy value using the following formula. : Where H is the instantaneous energy entropy value; P is the real-time pressure value at the output end of the controlled system; and Q is the instantaneous output flow rate at the power actuator. The system's baseline power value is used as the reference value. The entropy increase threshold is set to 1.25 times the average instantaneous energy entropy value H during the steady period. When the rate of change of H remains positive and exceeds this threshold, the system disables the weighting coefficients of the virtual cross-impedance decoupling model, and the pressure trajectory remains within the safety boundary. The system uses the power consumption data of the drive motor at the power execution end... To perform compensation, the system calculates the instantaneous conversion efficiency residual using a three-dimensional efficiency benchmark model to obtain the conversion efficiency residual. Conversion efficiency residual For every 10% increase, the gain coefficient of the flow control loop increases by 8.5%, and the execution stiffness attenuation factor... It compensates for the displacement response deviation caused by physical damage to the pump set, reducing the flow regulation deviation from 12.4% to 1.8%. The wellhead pressure trajectory exhibits a monotonous and smooth characteristic. The compensation action of the regulating actuator and the physical disturbance of the power actuator remain synchronized under the action of the virtual phase matching operator. The pressure of the controlled system achieves stable convergence under ultra-deep and high-pressure conditions.
[0041] Example 4: In a monitoring scenario where the physical properties of the fluid medium drift, the system executes an online identification program for the equivalent elastic modulus of the wellbore medium; and extracts the stroke frequency generated during the operation of the power actuator in real time. The system generates periodic flow pulsations of 0.5 Hz and simultaneously acquires the pressure response signal from the output of the controlled system. It performs Fast Fourier Transform on the flow and pressure sequences with 2048 sampling points, extracting the complex modulus and phase angle of the two sequences at the 0.5 Hz frequency component. The system then calculates the frequency domain phase difference between the pressure response signal and the flow pulsation signal. The system uses this parameter to determine the trend of the change in the equivalent elastic modulus of the medium. When the temperature and pressure field fluctuations cause the fluid's equivalent elastic modulus to shift from 2.1 GPa to 2.35 GPa, the system corrects the gain coefficient of the virtual cross impedance decoupling model in real time by adjusting the change in phase difference, thus compensating for the model mismatch caused by the evolution of the medium's compressibility.
[0042] When the system performs time constant τ calibration based on the system boundary temperature field parameters, the density map table is organized as a two-dimensional discrete data matrix with temperature as the row index and pressure as the column index. The internal values represent the equivalent sound velocity of the medium under different temperature and pressure combinations. After obtaining the real-time temperature value, the density map table is retrieved through a bilinear interpolation algorithm, and the corresponding medium sound velocity correction is obtained. When the output temperature sensor reports that the temperature rise causes the retrieved medium sound velocity to drop from 1450 m / s to 1420 m / s, the controller linearly increases the time-domain shift advance of the pressure compensation component accordingly, ensuring that the anti-phase pressure feedforward signal can accurately anchor the arrival moment of physical fluctuations.
[0043] Example 5: In the pre-deployment calibration procedure, the system performs calibration on the response bandwidth deviation value. Measurement and calculation; the controller sends a displacement increment of 0.2 to the power actuator. The system receives a step command, collects the displacement feedback sequence from the power actuator, extracts the time interval corresponding to when the output response value reaches 63.2% of the steady-state value, and uses this as the time constant of the power actuator. The same step response test is then performed on the regulating actuator, and the corresponding time constant is extracted. The difference between the two is determined as the response bandwidth deviation. The virtual phase-matching operator uses this value to set the filtering parameters to process the pressure compensation component, and its calculation logic follows the formula below: ,in, Here, represents the filter coefficients of the virtual phase-matching operator, and ΔT is the sampling period of the discrete control system. This is the time constant difference between the power actuator and the regulation actuator.
[0044] When the system is applied to conditions where the sound velocity of the fluid medium fluctuates with the ambient temperature, the system runs a calibration procedure based on the time constant τ of the system boundary temperature field parameters. The controller acquires the real-time temperature value at the output of the controlled system and uses it as an index to query the density mapping table. Linear fitting is performed on the compressibility test data of the medium under different temperature gradients to construct the calibration. The system corrects the initial time constant τ based on the obtained medium sound velocity correction. When the real-time temperature increases from 80℃ to 115℃, the system retrieves the change in medium sound velocity from the density mapping table and adjusts the time-domain shift advance of the pressure compensation component accordingly, ensuring phase alignment of the anti-phase pressure feedforward signal at the output of the controlled system. During system startup, the controller performs a self-test procedure on the state feature vector X, monitoring the change in the second derivative of the energy flow rate W. The initial operating state of the system is determined; if the rate of change of the energy entropy value H is continuously positive and exceeds the preset entropy increase threshold, the controller sets the gain matrix coefficient of the virtual cross impedance decoupling model to zero; the system automatically switches to a single-variable regulation mode with the real-time pressure value P as the feedback variable; adaptive fault tolerance for fluctuations in the initial conditions of the work site is achieved, ensuring the smooth convergence of the pressure control sequence under various working conditions.
[0045] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0046] 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.
Claims
1. A method for coordinated control of flow rate and pressure under deep well high-pressure fluid operation conditions, characterized in that, Includes the following steps: Step 101: Obtain the instantaneous output flow rate of the power actuator, the real-time pressure value of the controlled system output, the operating displacement parameters of the power actuator, and the transient pressure change rate of the controlled system output. By combining the displacement parameters and the transient pressure change rate into vectors, a state feature vector characterizing the energy state of the system is established. Step 102: Input the flow regulation command and pressure limit command into the virtual cross impedance decoupling model to calculate the coordinated control sequence consisting of the flow control component and the pressure compensation component. Step 103: When issuing the coordinated control sequence, superimpose the encoded signal with a preset frequency and preset amplitude into the flow control component to obtain a control sequence containing logic probes; Step 104: Obtain the pressure feedback data at the output of the controlled system driven by the controlled sequence; The pressure feedback data is calculated to obtain the pressure feedback residual by comparing it with the preset pressure reference value. The dynamic phase angle difference between the encoded signal and the pressure feedback residual is calculated using a cross-correlation algorithm. Step 105: Update the time constant of the long-distance transmission channel according to the dynamic phase angle difference; use the time constant to perform time-domain translation processing on the pressure compensation component so that the anti-phase pressure feedforward signal and the pressure fluctuation caused by the long-distance transmission channel are phase aligned at the output of the controlled system. Step 106: While adjusting the output intensity of the power actuator, control the output of the adjustment actuator to output a counter-current signal after phase alignment processing; Step 107: Calculate the product of the instantaneous output flow rate and the real-time pressure value in real time to obtain the energy flow rate; Calculate the change in the second derivative of the energy flux with respect to time; When the change in the second derivative exceeds the preset singularity threshold, a nonlinear damping component is injected into the coordinated control sequence and the system switches to a single-variable pressure-holding mode.
2. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, The virtual cross-impedance decoupling model has a dynamic gain matrix; step 102 also includes: extracting the pressure slope of the real-time pressure value within a preset time window and the flow gradient of the instantaneous output flow within the same time window; calculating the nonlinear proportional relationship between the pressure slope and the flow gradient; and updating the dynamic gain matrix using the nonlinear proportional relationship to adjust the logical isolation weight between the flow control component and the pressure compensation component.
3. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, The method also includes the following steps: Step 301, obtaining the dynamic response characteristics of the power actuator and the regulation actuator, and calculating the response bandwidth deviation value between them; Step 302, constructing a virtual phase matching operator based on the response bandwidth deviation value; Step 303, using the virtual phase matching operator to perform filtering and delay processing on the pressure compensation component, so that the action of the regulation actuator and the physical disturbance generated by the power actuator are kept synchronized on the time axis.
4. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, In step 107, after calculating the change in the second derivative of the energy flow rate, the method further includes: calculating the rate of change of the energy entropy of the controlled system; when the rate of change of the energy entropy exceeds the preset entropy increase threshold, forcibly setting the weight coefficient of the virtual cross impedance decoupling model to zero.
5. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, Before step 103, the process also includes: extracting periodic flow pulsations present in the operation of the power actuator as an endogenous detection signal; calculating the frequency domain phase difference between the endogenous detection signal and the corresponding pressure response signal at the output of the controlled system; and determining the evolution trend of the equivalent elastic modulus of the channel medium based on the frequency domain phase difference.
6. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 5, characterized in that, The method also includes: using the evolution trend of the equivalent elastic modulus to correct the gain coefficient of the virtual cross impedance decoupling model online, so as to compensate for the property drift of the channel medium.
7. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, The method also includes: extracting the power consumption data of the drive motor at the power execution end, calculating the conversion efficiency residual of the drive motor power consumption data relative to the instantaneous output flow rate; converting the conversion efficiency residual into the execution stiffness attenuation factor of the power execution end; and increasing the gain of the flow control loop according to the execution stiffness attenuation factor.
8. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, The frequency of the encoded signal is lower than the stroke frequency of the power actuator. In step 104, the dynamic phase angle difference is calculated using a cross-correlation algorithm. Specifically, the correlation coefficient between the encoded signal and the pressure feedback residual within the sliding window is calculated. The time delay when the correlation coefficient reaches its maximum value is locked, and the time delay is converted into a dynamic phase angle difference.
9. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, In step 107, the convergence of the system is determined by calculating the instantaneous energy entropy value H, which follows the following rules: Where P is the real-time pressure value at the output of the controlled system, and Q is the instantaneous output flow rate at the power actuator. The preset system baseline power value is used; when the rate of change of H is continuously positive and exceeds the preset threshold, the system is determined to have entered an uncontrolled growth mode.
10. The method for coordinated flow and pressure control under deep well high-pressure fluid operation conditions according to claim 1, characterized in that, In step 105, the advance amount of the time-domain translation processing is controlled by the real-time compensation of the system boundary temperature field parameters; the method also includes: obtaining the real-time temperature value of the channel medium at the system output end; querying the preset density mapping table according to the real-time temperature value to obtain the medium sound velocity correction amount, and using the medium sound velocity correction amount to correct the time constant.