Adaptive load distribution control method and system for spring press treatment

Abstract

The application discloses a spring strong pressure treatment adaptive load distribution control method and system, and relates to the technical field of safety valve spring group cooperative control. The method comprises the following steps: setting a target system total load and a target load distribution ratio among springs; collecting real-time pressure values and compression displacement amounts of the springs; calculating predicted load values contributed by the springs in a design pre-tightening state under a current strong pressure state, and a predicted system total load of the entire spring group; generating independent compensation control instructions for each spring based on double comparison results; and differentially adjusting strong pressure loading end points of the springs during the strong pressure process according to the compensation control instructions. The technical problem that the load distribution among the springs is difficult to keep consistent during strong pressure treatment and assembly pre-tightening of the safety valve spring group in the prior art, leading to poor stability of safety valve opening and closing performance, is solved, and the technical effects of improving spring group load distribution uniformity and enhancing safety valve opening and closing stability are achieved.

Description

Technical Field

[0001] This invention relates to the field of safety valve spring assembly coordinated control technology, specifically to an adaptive load distribution control method and system for spring pressure processing. Background Technology

[0002] As a crucial safety protection component in pressure vessels, piping systems, and industrial equipment, the opening and closing performance of safety valves directly affects the safety and stability of system operation. In some large-diameter, high-pressure-rating, or multi-spring safety valves, multiple springs are typically used collaboratively to provide preload to meet the valve disc opening pressure and reseating performance requirements. To ensure the overall stress performance of the spring assembly after assembly, each spring usually undergoes a high-pressure treatment during production to reduce the impact of initial plastic deformation on subsequent performance. In existing technologies, spring high-pressure treatment mostly adopts a uniform loading method, applying the same high-pressure stroke or pressure value to multiple springs, without fully considering the inconsistencies in load response caused by differences in materials, processing errors, stiffness deviations, and residual stress among different springs. This leads to uneven load distribution among the springs after preload assembly, causing some springs to be under unbalanced load for extended periods. This affects the stability of the safety valve's opening pressure, reseating consistency, and overall opening and closing performance, and in severe cases, may even reduce the long-term operational reliability of the safety valve. Summary of the Invention

[0003] This application provides an adaptive load distribution control method and system for spring pressure treatment, which solves the technical problem in the prior art that the load distribution among the springs in the safety valve spring assembly is difficult to keep consistent during the pressure treatment and assembly pre-tightening process, resulting in poor stability of the safety valve opening and closing performance.

[0004] A first aspect of this application provides an adaptive load distribution control method for spring-loaded processing, the method comprising: The system sets the target total system load and the target load distribution ratio among the springs in the design preload state of the target safety valve spring assembly. Multiple springs belonging to the same safety valve are subjected to synchronous high-pressure treatment, and during the high-pressure loading process, the real-time pressure value and compression displacement of each spring are collected. Based on the real-time pressure value and compression displacement, the system calculates the predicted load value contributed by each spring in the design preload state, and the predicted total system load of the entire spring assembly, under the current high-pressure state, if the target safety valve spring assembly is directly installed into the target valve. The predicted load value is compared with the target load distribution ratio, and the predicted total system load is compared with the target total system load. Based on the comparison results, an independent compensation control command is generated for each spring. According to the compensation control command, the high-pressure loading endpoint of each spring is differentially adjusted during the high-pressure process, so that the predicted load value approaches the target load distribution ratio, and the predicted total system load approaches the target total system load.

[0005] A second aspect of this application provides an adaptive load distribution control system with spring-loaded processing, the system comprising: The system comprises the following modules: Target Setting Module: Sets the target total system load and target load distribution ratio among the springs in the design preload state of the safety valve spring assembly; Data Acquisition Module: Performs synchronous high-pressure processing on multiple springs belonging to the same safety valve, and collects the real-time pressure value and compression displacement of each spring during the high-pressure loading process; Calculation Module: Based on the real-time pressure value and compression displacement, calculates in real-time the predicted load value contributed by each spring in the design preload state, and the predicted total system load of the entire spring assembly, if the target safety valve spring assembly is directly installed into the target valve under the current high-pressure state; Comparison Module: Performs a dual comparison between the predicted load value and the target load distribution ratio, and between the predicted total system load and the target total system load, and generates an independent compensation control command for each spring based on the dual comparison results; Adjustment Module: According to the compensation control command, performs differentiated adjustment on the high-pressure loading endpoint of each spring during the high-pressure process, so that the predicted load value approaches the target load distribution ratio and the predicted total system load approaches the target total system load.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, the target total system load and the target load distribution ratio among the springs are set under the design preload state of the target safety valve spring assembly. Next, multiple springs belonging to the same safety valve are subjected to synchronous high-pressure treatment, and the real-time pressure value and compression displacement of each spring are collected during the high-pressure loading process. Further, based on the real-time pressure value and compression displacement, the predicted load value contributed by each spring under the design preload state, and the predicted total system load of the entire spring assembly, are calculated in real-time under the current high-pressure state if the target safety valve spring assembly were directly installed into the target valve. Then, the predicted load value is compared with the target load distribution ratio, and the predicted total system load is compared with the target total system load. Based on the results of the comparison, an independent compensation control command is generated for each spring. Finally, according to the compensation control command, the high-pressure loading endpoint of each spring is differentially adjusted during the high-pressure process, so that the predicted load value approaches the target load distribution ratio, and the predicted total system load approaches the target total system load. This invention solves the technical problem in the prior art where the load distribution among the springs in the safety valve spring assembly is difficult to maintain in a consistent manner during high-pressure treatment and assembly pre-tightening, resulting in poor stability of the safety valve's opening and closing performance. It achieves the technical effect of improving the load distribution balance of the spring assembly and enhancing the opening and closing stability of the safety valve. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic flowchart of the adaptive load distribution control method for spring-forced processing provided in the embodiments of this application; Figure 2 This is a schematic diagram of the adaptive load distribution control system with spring compression processing provided in an embodiment of this application.

[0009] Explanation of reference numerals in the attached diagram: Target setting module 11, data acquisition module 12, calculation module 13, comparison module 14, adjustment module 15. Detailed Implementation

[0010] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0011] Example 1, as Figure 1 As shown, this application provides an adaptive load distribution control method for spring-loaded processing, wherein the method includes: Set the target total system load and the target load distribution ratio among the springs in the safety valve spring assembly under the design preload state of the safety valve.

[0012] Specifically, the design parameters of the target safety valve are obtained, including at least the design opening pressure, effective pressure-bearing area of ​​the valve disc, valve stem guide friction coefficient, sealing surface preload compensation coefficient, number of springs, installation position of each spring, nominal stiffness of each spring, and maximum allowable load of each spring. The theoretical opening load required to balance the valve disc in the design opening state is calculated based on the design opening pressure and the effective pressure-bearing area of ​​the valve disc, where the theoretical opening load equals the product of the design opening pressure and the effective pressure-bearing area of ​​the valve disc. Subsequently, the theoretical opening load is corrected based on the valve stem guide friction coefficient and the sealing surface preload compensation coefficient to obtain the total load of the target system. The valve stem guide friction coefficient is used to compensate for mechanical friction losses during valve stem movement, and the sealing surface preload compensation coefficient is used to compensate for the additional clamping force generated when the valve disc sealing surface is in contact. Then, the nominal stiffness and position compensation coefficient corresponding to the installation position of each spring are obtained, and the load is calculated based on the nominal stiffness of the corresponding spring and the position compensation coefficient corresponding to the installation position. The position compensation coefficient is used to calculate the load-bearing weight of each spring, where the load-bearing weight is equal to the product of the nominal stiffness and the position compensation coefficient. The load-bearing weights of all springs are normalized to obtain the target load distribution ratio for each spring, where the target load distribution ratio for each spring is equal to the proportion of its load-bearing weight to the sum of the load-bearing weights of all springs. The total load of the target system is decomposed based on the target load distribution ratios for each spring to obtain the target load value for each spring. Then, it is determined whether the target load value of any spring exceeds the maximum allowable load of the corresponding spring. If it does, the load-bearing weight of the corresponding spring is reduced, and the normalization distribution calculation is re-executed until the target load values ​​of all springs are less than or equal to the maximum allowable load of the corresponding spring. Finally, the total load of the target system, the target load values ​​of each spring, and the target load distribution ratio of each spring are written into the control parameter table of the high-pressure control system as the benchmark constraint parameters for subsequent high-pressure prediction calculations, load balancing analysis, and compensation control.

[0013] Wherein, the valve stem guide friction compensation coefficient This is used to compensate for the additional driving force consumed by the mechanical friction between the valve stem and the guide sleeve, packing assembly, or sealing contact surface during the opening process of the safety valve. The valve stem friction compensation coefficient is obtained through friction testing and calibration, specifically including: after the safety valve is assembled, keeping the valve disc in a state without medium pressure, using an external loading device to drive the valve stem to move in a low-speed linear motion along the opening direction, and collecting the axial driving force during the valve stem movement in real time; the average driving force corresponding to the stable movement stage of the valve stem is recorded as the frictional additional load. Then, based on the safety valve's design opening pressure... Effective pressure-bearing area of ​​valve disc Calculation of theoretical starting load Then, calculate the valve stem friction compensation coefficient according to the following formula. Among them, when the valve stem guide clearance increases, the packing tightness increases, or the lubrication condition decreases, the additional frictional load... As the coefficient increases, the corresponding valve stem friction compensation coefficient also increases.

[0014] The sealing surface preload compensation coefficient This is used to compensate for the additional pre-tightening force that needs to be overcome between the valve disc and valve seat to ensure sealing performance of the safety valve; the sealing pre-tightening compensation coefficient is obtained through a sealing start-up test calibration, specifically including: real-time monitoring of the actual opening pressure corresponding to the initial micro-displacement of the valve disc during the gradual increase of medium pressure. The actual opening pressure is compared with the designed opening pressure. The additional load corresponding to the difference is defined as the sealing additional load. Then, calculate the sealing preload compensation coefficient according to the following formula: The additional sealing load The main factors are the sealing surface contact pressure between the valve disc and the valve seat, the influence of the sealing surface roughness, and the adhesion resistance formed by long-term pre-tightening; as the degree of sealing surface contact increases, the sealing pre-tightening compensation coefficient increases accordingly.

[0015] Wherein, the position compensation coefficient This is used to characterize the degree of influence of the i-th spring on the actual force contribution of the valve disc due to different installation positions. Since the installation positions of multiple springs in a safety valve are usually distributed with different radial distances and angles relative to the center of the valve disc, when there are machining errors, assembly eccentricity, or pressure eccentricity in the valve disc, spring seat, or guide structure, the transmission efficiency of the springs in different positions to the valve disc opening force will be different. Therefore, it is necessary to use the installation position compensation coefficient to correct the theoretical load distribution of each spring.

[0016] Specifically, after the safety valve is assembled, the positional offset of the valve disc center relative to the valve seat center is measured using a laser displacement sensor to obtain the valve disc eccentricity E; simultaneously, the off-center load direction angle is determined based on the angle between the valve disc center offset direction and the preset reference direction. When a database of assembly eccentricity for safety valves of the corresponding specifications has been established during the production process, the statistical average eccentricity and statistical average off-center load direction of the corresponding specifications can be directly used as the valve disc eccentricity E and the off-center load direction angle. Obtain the radial distance of the i-th spring center position relative to the geometric center of the valve disc. and installation azimuth angle Calculate the angle difference between the spring and the off-center loading direction: Subsequently, the installation position compensation coefficient is calculated based on the coupling relationship between the spring installation position and the off-center load direction: The position influence correction coefficient To pre-calibrate the parameters, multiple assembly tests were conducted on safety valves of the same specification. Specifically, the deviations between the actual load borne by each spring and the theoretical average load were recorded under different eccentricity conditions. The least squares method was then used for fitting and solving to obtain the coefficient that minimizes the predicted load error. Preferably, The value range is 0.1 to 0.5.

[0017] To avoid non-physical results caused by negative compensation coefficients, the installation position compensation coefficient is limited. At that time, take When calculated At that time, take This ensures that all springs bear the positive load and that there is no overload distribution.

[0018] When the i-th spring is located near the off-center load direction A positive value corresponds to an increase in the installation position compensation coefficient, indicating that the target load borne by the spring is appropriately increased; when the first... When the spring is in the opposite direction of the off-center load A negative value corresponds to a decrease in the installation position compensation coefficient, indicating that the target load borne by the spring is appropriately reduced. When the safety valve structure is an ideal symmetrical structure and the measured valve disc eccentricity is less than the preset eccentricity threshold, all installation position compensation coefficients are uniformly set to 1.

[0019] Multiple springs belonging to the same safety valve are subjected to synchronous high-pressure treatment, and the real-time pressure value and compression displacement of each spring are collected in real time during the high-pressure loading process.

[0020] First, multiple springs belonging to the same safety valve are installed in multiple independent loading stations of a multi-station synchronous high-pressure platform according to their actual assembly positions. Each loading station is equipped with an independent servo loading actuator, a pressure sensor, and a displacement acquisition unit. The servo loading actuator drives the corresponding spring to perform axial compression loading. The pressure sensor is installed in the loading force transmission path to detect the current axial pressure value of the corresponding spring in real time. The displacement acquisition unit detects the current compression displacement of the corresponding spring in real time. Then, a unified synchronization clock is established, and the central controller synchronously sends a high-pressure start command to each servo loading actuator based on the unified synchronization clock, causing multiple springs to enter the synchronous high-pressure state at the same initial loading time. During the high-pressure loading process, the central controller synchronously acquires the real-time pressure signals output by each pressure sensor and the real-time displacement signals output by each displacement acquisition unit according to a preset sampling period. The system collects signals and adds a unified timestamp to the collected data to construct a real-time force-displacement data stream for each spring. The preset sampling period is set according to the loading change rate during the spring compression process to ensure that the spring stiffness change and plastic deformation process can be continuously tracked. Then, the central controller performs synchronous verification on the real-time force-displacement data stream for each spring, and determines whether the time deviation of the data at each station exceeds the preset synchronization error threshold. If it does, data alignment compensation is triggered. At the same time, the system determines whether the current spring is experiencing loading abnormality, displacement stagnation, or local instability based on the real-time pressure value change rate and compression displacement change rate. If an abnormality is detected, an abnormality warning signal is generated for the corresponding station and the loading rate of the corresponding station is reduced. Finally, the real-time pressure value and compression displacement after synchronization verification are continuously written into the central controller's buffer area as real-time input data for the subsequent virtual assembly coupling model to perform instantaneous equivalent stiffness identification and predicted load calculation.

[0021] Based on the real-time pressure value and the compression displacement, the predicted load value contributed by each spring under the designed preload state is calculated in real time under the current high pressure state, if the target safety valve spring assembly is directly assembled into the target valve, as well as the predicted total system load of the entire spring assembly.

[0022] Furthermore, based on the real-time pressure value and the compression displacement, the predicted load value contributed by each spring under the designed preload state is calculated in real time under the current high pressure state, if the target safety valve spring assembly is directly assembled into the target valve, as well as the predicted total system load of the entire spring assembly, including: During the high-pressure loading process, the real-time force-displacement data stream, consisting of the real-time pressure value of each spring and the corresponding compression displacement, is input into a pre-constructed virtual assembly coupling model that includes the nonlinear stiffness of the valve disc-spring contact. Based on the input real-time force-displacement data stream, the virtual assembly coupling model dynamically identifies and outputs the instantaneous equivalent stiffness and residual deformation trend of each spring under the current high-pressure state. Based on the instantaneous equivalent stiffness and residual deformation trend of each spring, the virtual assembly coupling model continues to solve for the virtual assembly compression amount of each spring required to achieve the designed preload state under the geometric constraints of the target valve. Based on the virtual assembly compression amount and the instantaneous equivalent stiffness, the predicted load value contributed by each spring under the designed preload state is calculated, and the predicted load values ​​of all springs are summed to obtain the total predicted system load of the entire spring assembly.

[0023] During the high-pressure loading process, the central controller synchronously collects the real-time pressure value and compression displacement of each spring according to a preset sampling period, and constructs corresponding real-time force-displacement data streams in chronological order. Each set of real-time force-displacement data streams includes at least a sampling timestamp, real-time pressure value, real-time displacement, and corresponding loading rate. Subsequently, the real-time force-displacement data streams corresponding to each spring are input into a pre-constructed virtual assembly coupling model containing the nonlinear stiffness of the valve disc-spring contact. The virtual assembly coupling model establishes a nonlinear mechanical coupling relationship between the valve disc and spring groups based on the valve disc structure parameters, valve stem guiding parameters, spring installation position parameters, and design preload stroke parameters of the target safety valve. Then, the virtual assembly coupling model calculates the instantaneous equivalent stiffness of each spring based on the pressure increment and displacement increment between adjacent sampling points within the current sampling window. The instantaneous equivalent stiffness is equal to the ratio of the pressure increment to the displacement increment. The model also extrapolates the change in free length of the spring after unloading using the real-time force-displacement change trends in multiple consecutive sampling windows to obtain the residual deformation trend of the corresponding spring. Subsequently, the... The virtual assembly coupling model establishes displacement coordination constraints based on the target valve disc installation position under the designed pre-tightening state, the installation height constraints of each spring, and the current residual deformation trend of each spring. It then uses an iterative solution method to calculate the virtual assembly compression required for each spring to reach the designed pre-tightening state after actual assembly. In each iteration, the predicted load value of the corresponding spring is calculated based on its instantaneous equivalent stiffness and the current virtual assembly compression. The virtual assembly compression is corrected based on the error between the sum of all predicted load values ​​and the target valve disc balance load, until the load error between two consecutive iterations is less than a preset convergence threshold. Finally, the predicted load value contributed by each spring under the designed pre-tightening state is calculated based on the converged virtual assembly compression and the corresponding instantaneous equivalent stiffness. The predicted load values ​​of all springs are summed to obtain the total predicted system load for the entire spring group. Simultaneously, the predicted load values ​​of each spring and the total predicted system load are output to the load compensation control module in real time for subsequent load distribution deviation analysis and differentiated high-pressure compensation control.

[0024] Furthermore, the virtual assembly coupling model is an iterative solution model constructed based on the displacement coordination equation and the force balance equation; wherein, the displacement coordination equation is used to characterize that the sum of the compression displacements of each spring under the valve disc constraint is a constant value, and the force balance equation is used to characterize that the sum of the predicted load values ​​of each spring is equal to the resultant force on the valve disc.

[0025] Preferably, a valve disc constraint model is established based on the structural assembly parameters of the target safety valve. These structural assembly parameters include at least the target installation position of the valve disc, the installation height of the spring seat, the designed preload stroke, and the initial free length of each spring. Simultaneously, a virtual assembly coupling model incorporating the nonlinear stiffness of the valve disc-spring contact is constructed, where the contact stiffness of the i-th spring in the contact area with the valve disc is expressed as: ,in, Indicates contact stiffness, Indicates the initial contact stiffness. This indicates the amount of local contact deformation. and Represents the nonlinear correction coefficient; establishes the spring coupling matrix based on the structural relationship of multiple springs acting together on the valve disc: ,in, The coefficient representing the influence of the unit deformation of the j-th spring on the force state of the i-th spring is used to characterize the deformation transmission relationship generated by multiple springs acting together on the same valve disc. The virtual assembly compression of each spring is coupled and corrected according to the spring coupling matrix. The coupled correction compression corresponding to the i-th spring satisfies: ,in, This represents the coupling correction compression amount of the i-th spring. The original virtual assembly compression amount of the i-th spring is represented. After the coupling correction is completed, the coupling correction compression amount is used as the actual effective compression amount for subsequent displacement coordination constraints and force balance constraints calculations.

[0026] Then, the real-time force-displacement data stream is input into the virtual assembly coupling model, and the instantaneous equivalent stiffness of each spring is calculated using the least squares identification method within a sliding time window of length m: ,in, This represents the real-time pressure value corresponding to the t-th sampling point. This indicates the corresponding compression displacement. and These represent the average values ​​of pressure and displacement within the time window, respectively. Subsequently, a residual deformation prediction model is established based on the instantaneous equivalent stiffness sequence obtained from multiple consecutive time windows and the cumulative plastic deformation. ,in, This represents the predicted residual deformation of the i-th spring. This represents the cumulative amount of plastic deformation. and This represents the deformation correction coefficient. Subsequently, based on the coupling correction compression amount corresponding to each spring, the predicted residual deformation amount, and the assembly constraint relationship under the design preload state of the target safety valve, a displacement coordination equation is established. The effective compression displacement and residual deformation amount of each spring after coupling correction are jointly incorporated into the assembly displacement calculation process to characterize the overall displacement coordination relationship of the valve disc under the design preload state. ,in, This represents the total assembly compression displacement of the target safety valve under its design preload condition. Based on satisfying the displacement coordination relationship, and according to the instantaneous equivalent stiffness of each spring, the coupling-corrected compression, the predicted residual deformation, and the valve disc contact nonlinear stiffness, the predicted load value of each spring under its design preload condition is calculated. Subsequently, the predicted load values ​​corresponding to each spring are synthesized, and a force balance equation is established to characterize the balance relationship between the total load generated by the combined action of multiple springs and the target force under the designed preload state of the valve disc: ,in, This represents the target resultant force that the valve disc needs to withstand under the designed preload state; subsequently, under the joint constraints of the contact nonlinear stiffness relationship, the spring coupling matrix, the displacement compatibility equation, and the force balance equation, the original virtual assembly compression of each spring is used. The virtual assembly compression is used as an iterative variable for iterative solution; in each iteration, the corresponding coupling correction compression is calculated based on the virtual assembly compression obtained in the current iteration. Then, the predicted load values ​​for each spring and the total predicted load of the system are recalculated.

[0027] Subsequently, under the joint constraints of the contact nonlinear stiffness relationship, the spring coupling matrix, the displacement compatibility equation, and the force balance equation, the original virtual assembly compression of each spring is used. The solution is performed iteratively, using these variables as iteration variables. Before entering the iterative solution, the target load value for each spring is determined based on the total load of the target system and the target load distribution ratio of each spring. And combined with the instantaneous equivalent stiffness obtained from the current identification Contact stiffness And predicting residual deformation Calculate the initial virtual assembly compression for each spring: ,in, This represents the initial virtual assembly compression amount corresponding to the i-th spring; subsequently, using the initial virtual assembly compression amount as the starting point for iteration, in the i-th spring... During the round of iteration, the corresponding coupling correction compression amount is first calculated based on the current virtual assembly compression amount. And recalculate the predicted load values ​​for each spring. And predict the total system load; then, update and correct the virtual assembly compression amount based on the deviation between the current predicted load value and the target load value: ,in, Indicates the first The virtual assembly compression amount corresponding to each iteration. Indicates the first The virtual assembly compression amount corresponding to each iteration. This represents the iterative correction step size; when the predicted load value corresponding to the i-th spring is lower than the target load value, the corresponding virtual assembly compression amount is increased; when the predicted load value corresponding to the i-th spring is higher than the target load value, the corresponding virtual assembly compression amount is decreased; after completing the virtual assembly compression amount update, the coupling correction calculation, displacement coordination calculation, and force balance calculation are re-executed, and a new round of predicted load value and predicted total system load are generated; when the predicted total system load obtained from two consecutive iterative calculations satisfies: When determining whether the current iteration result satisfies the convergence condition, where, This indicates a preset convergence threshold, and the target resultant force is preferred. The load is 1% to 3% of the total load. Iteration stops when the convergence condition is met for three consecutive iterations, and the predicted load values ​​for each spring and the total predicted system load are output in the converged state. This serves as the input basis for subsequent load allocation comparison and compensation control.

[0028] Furthermore, the instantaneous equivalent stiffness is obtained by calculating the rate of change of the real-time pressure value near the current sampling point relative to the compression displacement, and the residual deformation trend is obtained by extrapolating and predicting the change in the free length of the spring after unloading by analyzing the real-time force-displacement data stream of the current and historical sampling points.

[0029] First, during the high-pressure loading process, the central controller continuously collects the real-time pressure value and compression displacement of each spring according to a preset sampling period, and forms a real-time force-displacement data stream in chronological order. Then, taking the current sampling point as the center, a local sampling window containing multiple consecutive sampling points is selected in the corresponding real-time force-displacement data stream. The pressure change and displacement change within the local sampling window are calculated differentially, where the pressure increment between adjacent sampling points is denoted as... The displacement increment is denoted as ; Calculate the instantaneous equivalent stiffness corresponding to the current sampling point based on the pressure increment and displacement increment: ,in, The instantaneous equivalent stiffness corresponding to the current sampling point is calculated. To reduce the impact of sampling noise on the stiffness calculation results, a moving average filter is applied to the calculation results of multiple instantaneous equivalent stiffnesses in the local sampling window to obtain a smoothed target instantaneous equivalent stiffness. Then, the central controller extracts the real-time force-displacement data stream corresponding to the current and historical sampling points, analyzes the relationship between the pressure increase rate, displacement increase rate, and unloading rebound trend during loading, and establishes a free length variation trend sequence based on the cumulative plastic deformation characteristics of multiple consecutive sampling windows. Specifically, the pressure and displacement values ​​at the end of each sampling window are used as window feature values. The elastic recovery of the spring in the current window is calculated based on the pressure and displacement values. The plastic deformation is obtained by subtracting the elastic recovery from the window compression displacement. The plastic deformation values ​​corresponding to multiple consecutive sampling windows are arranged in chronological order to form a cumulative plastic deformation sequence. Subsequently, a linear fitting model is established with the cumulative plastic deformation sequence as the independent variable and the corresponding sampling time as the dependent variable to obtain the plastic deformation... The plastic deformation growth trend equation is calculated, and the predicted plastic deformation amount is extrapolated by substituting the time point corresponding to a preset prediction cycle after the current sampling time into the plastic deformation growth trend equation. The predicted free length after unloading is calculated based on the predicted plastic deformation amount, where the predicted free length is equal to the initial free length of the spring minus the predicted plastic deformation amount. Then, the predicted free lengths corresponding to multiple consecutive prediction cycles are arranged in chronological order to form a free length change trend sequence. The residual deformation trend is determined based on the rate of change between adjacent predicted free lengths in the free length change trend sequence. When the free length corresponding to multiple consecutive prediction cycles continuously decreases, the residual deformation trend is determined to be an increasing trend; when the free length change corresponding to multiple consecutive prediction cycles tends to stabilize, the residual deformation trend is determined to be a stable trend. Finally, based on the difference between the predicted target free length and the initial free length of the spring, the predicted residual deformation amount corresponding to the current spring is obtained, and the predicted residual deformation amount and residual deformation trend are output to the virtual assembly coupling model in real time.

[0030] The predicted load value is compared with the target load allocation ratio, and the predicted total system load is compared with the target total system load. Based on the results of the comparison, an independent compensation control command is generated for each spring.

[0031] Furthermore, the predicted load value is compared with the target load allocation ratio, and the predicted total system load is compared with the target total system load. Based on the results of the comparison, an independent compensation control command is generated for each spring, including: The predicted load values ​​are converted into a predicted load distribution vector according to a proportional relationship, and compared with the target distribution vector determined by the target load allocation ratio. The vector angle and magnitude difference are calculated as the first set of indicators for evaluating the uniformity of load distribution. The relative error between the predicted total load of the system and the target total load is calculated as the second indicator for evaluating the consistency of system load. A comprehensive decision function is established, with the optimization objective being to minimize the first set of indicators and the second set of indicators. During the high-pressure process, the comprehensive decision function is dynamically solved based on the real-time values ​​of the first set of indicators and the second set of indicators to obtain a set of desired adjustment parameters for each spring that optimizes the current optimization objective. The desired adjustment parameters are mapped as independent compensation control commands to control the action of the actuators of the high-pressure equipment.

[0032] Preferably, the predicted load value corresponding to each spring is obtained, and the predicted load value of each spring is proportionally normalized according to the sum of the predicted load values ​​of all springs to construct a predicted load distribution vector: ,in, Let i be the predicted load value corresponding to the i-th spring; then, construct the target distribution vector according to the preset target load distribution ratio of each spring. ,in, The target load distribution ratio is assigned to the i-th spring. Then, the predicted load distribution vector is compared with the target distribution vector. The angle deviation index C and the modulus difference index D between the two vectors are calculated as the first set of indicators to evaluate the uniformity of load distribution. The angle deviation index... Calculate according to the following formula: ,in, This represents the dot product of the predicted load distribution vector and the target load distribution vector. This represents the magnitude of the predicted load distribution vector. The modulus of the target distribution vector is represented; the included angle deviation index is used to reflect the degree of directional deviation between the predicted load distribution and the target load distribution; the modulus difference index... Calculate according to the following formula: The modulus difference index is used to reflect the overall deviation between the predicted load distribution and the target load distribution; subsequently, the relative error between the predicted total system load and the target total system load is calculated: ,in, To predict the total system load, Let E be the total system load and E be the relative error of the total system load; the relative error serves as the second indicator for evaluating the consistency of system load; subsequently, the included angle deviation index... Length difference index and the relative error of total system load Perform normalization and calculate the total deviation: The corresponding optimization weights are dynamically determined based on the proportion of each evaluation indicator in the total deviation. , , And satisfy: Among them, when the included angle deviation index C increases, the corresponding weight The synchronous increase prioritizes reducing load distribution direction deviation during the optimization process; as the modulus difference index D increases, the corresponding weight... Synchronous increase allows the optimization process to prioritize reducing the dispersion of load distribution; when the relative error of the total system load... When it increases, the corresponding weight The system is increased synchronously to prioritize reducing the total system load deviation during the optimization process; subsequently, a comprehensive decision function is established: ,in, This is a comprehensive decision function; the optimization variables of the comprehensive decision function include the loading compensation increment corresponding to each spring. Unloading compensation increment Pressure holding duration and loading rate adjustment amount The value range of each optimization variable is jointly limited by the allowable adjustment range of the high-pressure equipment and the allowable deformation range of the spring material; then, during the high-pressure process, the comprehensive decision function is dynamically solved using an iterative optimization method, starting with the current adjustment parameter set corresponding to each spring. As the initial solution, multiple candidate solutions are generated by adding or subtracting perturbations to each optimization variable according to a preset perturbation step size. Subsequently, each candidate solution is input into the virtual assembly coupling model to recalculate the corresponding predicted load value, predicted load distribution vector, and predicted total system load, and the corresponding comprehensive decision function value is calculated. Select the candidate solution with the smallest comprehensive decision function value as the optimal solution for the current iteration; then, use the optimal solution for the current iteration as the initial solution for the next iteration and continue the above solution process, repeating prediction calculation, function evaluation, and parameter updates; when the change in the comprehensive decision function between two consecutive iterations satisfies: Or the number of iterations has reached the preset maximum number of iterations. When the iteration stops, the iteration is stopped. A preset convergence threshold is set; finally, the optimal solution of the current iteration is determined as the final desired adjustment parameter and mapped to the independent compensation control command of the corresponding high-pressure equipment actuator to drive each spring to perform differentiated high-pressure compensation control.

[0033] Furthermore, in the comprehensive decision function, the optimization weight set for the first set of indicators for evaluating the uniformity of load distribution is higher than the optimization weight set for the second indicator for evaluating the consistency of system load.

[0034] When establishing the comprehensive decision function, the central controller sets the vector angle deviation index and the modulus difference index, which characterize the uniformity of load distribution among the springs, as the primary optimization objectives, and sets the relative error of the total system load, which characterizes the consistency between the predicted total system load and the target total system load, as the secondary optimization objective. The sum of the optimization weights corresponding to the first set of indicators is greater than the optimization weight of the corresponding second indicator, ensuring priority is given to maintaining the load distribution balance among the springs during the high-pressure compensation control process. Subsequently, during the high-pressure loading process, the central controller dynamically adjusts the weight ratios in the comprehensive decision function based on the real-time calculated predicted load distribution state. When the predicted load distribution deviation among the springs is detected to exceed the preset equilibrium threshold, the optimization weights corresponding to the first set of indicators are increased to enhance load distribution. It has a strong ability to correct local load imbalances. When the predicted load distribution among the springs has become relatively balanced, but there is still a deviation between the predicted total system load and the target total system load, the optimization weight corresponding to the second indicator is appropriately increased to compensate and correct the overall system load. The weight adjustment process ensures that the total optimization weight corresponding to the first set of indicators is always higher than the optimization weight corresponding to the second indicator, thereby ensuring that the entire optimization process prioritizes reducing the degree of load distribution imbalance among the springs and avoids the situation where the total system load meets the target requirements but the local springs are in an off-center load state for a long time. Finally, the central controller re-solves the comprehensive decision function based on the adjusted optimization weights and outputs the corresponding optimal compensation adjustment parameters for subsequent differentiated compensation control of the high-pressure actuator.

[0035] According to the compensation control command, the end point of the high-pressure loading of each spring is adjusted differently during the high-pressure process, so that the predicted load value approaches the target load distribution ratio and the predicted total load of the system approaches the target total load.

[0036] First, the central controller parses the independent compensation control commands corresponding to each spring, obtaining the target loading increment, target unloading increment, loading rate adjustment parameters, pressure holding duration, and target endpoint position correction for each spring. Then, the central controller controls the servo loading actuators at each station to perform differentiated high-pressure adjustment according to the corresponding compensation control commands. For springs with predicted load values ​​higher than the target load distribution ratio, the endpoint loading depth of the corresponding station is reduced or the unloading rebound stage is extended to reduce the final equivalent stiffness and residual stress level of the corresponding spring. For springs with predicted load values ​​lower than the target load distribution ratio, the endpoint loading depth of the corresponding station is increased or the unloading rebound stage is shortened to increase the final equivalent stiffness and preload capacity of the corresponding spring. Afterward, during the differentiated high-pressure adjustment process for each spring, the central controller continuously collects the real-time pressure value and compression displacement of the corresponding spring and re-inputs them into the virtual assembly coupling model for predicted load calculation, updating the predicted load value for each spring and the total predicted load of the entire spring assembly in real time. Finally, based on the updated predicted load value and target load distribution ratio... The system dynamically adjusts the endpoint loading position and loading rate of the corresponding springs based on the degree of deviation between the predicted load and the target load, as well as the degree of deviation between the predicted total load of the system and the target total load. Specifically, when the predicted load value of the corresponding spring is consistently high, the subsequent loading increment of the corresponding station is gradually reduced; when the predicted load value of the corresponding spring is consistently low, the subsequent loading increment of the corresponding station is gradually increased. Simultaneously, when the overall predicted total load of the system is higher than the target total load, the endpoint loading correction magnitude of all stations is simultaneously reduced; when the overall predicted total load of the system is lower than the target total load, the endpoint loading correction magnitude of all stations is simultaneously increased. Subsequently, the central controller continuously executes the closed-loop adjustment process of "real-time acquisition - prediction calculation - deviation analysis - endpoint correction" until the deviation between the predicted load allocation ratio and the target load allocation ratio of each spring is less than the preset allocation error threshold, and the deviation between the predicted total load and the target total load is less than the preset total load error threshold. At this point, the current spring group is determined to have reached the target high-pressure state. Finally, the servo loading actuators of each station are controlled to stop loading, and the final high-pressure completion signal and the load consistency evaluation result of the corresponding spring group are output.

[0037] Furthermore, based on the aforementioned compensation control command, the endpoint of the high-pressure loading of each spring is differentially adjusted during the high-pressure process, including: The compensation control command is analyzed to obtain the desired residual plastic deformation and desired stiffness change direction for each spring. Based on the desired stiffness change direction, if stiffness needs to be reduced, after the spring completes the basic high-pressure stroke, the creep-induced rate is continued to be applied at a rate lower than the basic loading rate until the actual residual strain increment reaches the desired residual strain increment. If stiffness needs to be increased, after the spring completes the basic high-pressure stroke, the actuator is controlled to perform a stress relaxation cycle that includes partial unloading and a specific pressure holding stage.

[0038] First, the central controller parses the compensation control commands corresponding to each spring, extracting the desired residual plastic deformation, desired residual strain increment, target loading correction, and desired stiffness change direction for each spring. The desired stiffness change direction includes at least two directions: a decrease in stiffness and an increase in stiffness. Then, the servo loading actuator at the corresponding station first completes the basic high-pressure stroke of the corresponding spring according to the unified basic high-pressure parameters. Afterward, when the desired stiffness change direction of the corresponding spring is a decrease in stiffness, the central controller controls the corresponding actuator to continue executing additional high-pressure along the loading direction after completing the basic high-pressure stroke. During the additional high-pressure stage, a creep-induced rate lower than the basic loading rate is used for continuous loading, causing continuous plastic flow within the spring material. During the additional high-pressure process, the central controller continuously collects the real-time pressure value, real-time compression displacement, and real-time strain change of the corresponding spring, and calculates the current residual strain increment based on the real-time strain change. When the current residual strain increment reaches the corresponding desired residual strain increment, the additional high-pressure loading is immediately stopped to allow the corresponding spring to form the target plastic deformation state, thereby reducing the equivalent stiffness of the corresponding spring in the subsequent assembly state. The central controller, after the spring completes its basic high-pressure stroke, controls the corresponding actuator to perform a stress relaxation cycle including partial unloading and a specific pressure holding phase when the desired stiffness change direction of the corresponding spring is to increase stiffness. The partial unloading phase releases localized concentrated stress within the spring, while the specific pressure holding phase promotes the redistribution of residual stress within the spring. During the stress relaxation cycle, the central controller first controls the actuator to retract a preset unloading displacement along the unloading direction and maintains constant pressure at the corresponding unloading endpoint to enter the pressure holding phase. During the pressure holding phase, the pressure decay rate and displacement rebound rate of the corresponding spring are monitored in real time, and the pressure holding duration is dynamically adjusted based on the monitoring results. When both the pressure decay rate and displacement rebound rate are less than the corresponding stability threshold, the residual stress distribution within the spring is determined to have reached a stable state. Afterward, the actuator is controlled to reload to a set safe position to enable the corresponding spring to form a more stable elastic recovery capability and higher subsequent equivalent stiffness. Finally, the central controller re-collects real-time force-displacement data for each spring and re-executes predicted load calculation and deviation analysis for further differentiated compensation adjustments.

[0039] Furthermore, parsing the compensation control command to obtain the desired residual plastic deformation and desired stiffness change direction for each spring includes: The independent compensation control command is decoded to extract the compensation mode identifier and compensation amount parameter. If the compensation mode identifier corresponds to the first type of compensation mode, the compensation amount parameter is directly mapped to the expected residual plastic deformation amount, and the expected stiffness change direction is determined to require a reduction in stiffness. If the compensation mode identifier corresponds to the second type of compensation mode, the target residual stress distribution state is calculated based on the compensation amount parameter, and the expected residual plastic deformation amount and unloading control parameter are calculated based on the target residual stress distribution state. At the same time, the expected stiffness change direction is determined to require an increase in stiffness.

[0040] First, the central controller receives independent compensation control commands generated for each spring and decodes these commands. Each independent compensation control command includes at least a compensation mode identifier, a compensation amount parameter, a target load correction parameter, and a control priority parameter. Then, the central controller identifies the compensation control type corresponding to the current spring based on the compensation mode identifier. When the compensation mode identifier corresponds to the first type of compensation mode, it is determined that the current spring has a predicted load value that is too high or a local stiffness that is too large, requiring increased plastic deformation to reduce the equivalent stiffness of the corresponding spring in its subsequent assembly state. Afterward, the central controller... Directly mapped to the expected residual plastic deformation of the corresponding spring. The desired residual plastic deformation is calculated according to the following formula: ,in, The plastic compensation coefficient is pre-calibrated based on the target spring material type and high-pressure process test data. Simultaneously, the desired stiffness change direction of the current spring is determined as the direction requiring stiffness reduction, and corresponding creep-induced loading control parameters are generated. Subsequently, when the compensation mode identifier corresponds to the second type of compensation mode, it is determined that the current spring has a predicted load value that is too low or insufficient local stiffness, requiring optimization of the internal residual stress distribution to improve the equivalent stiffness of the corresponding spring in its subsequent assembly state. Afterwards, the central controller, based on the compensation parameters... Establish the target residual stress distribution state, wherein the target residual stress distribution state is represented by a target residual stress coefficient matrix as follows: Where m is the number of discrete elements of the effective cross-section of the spring. This represents the target residual stress level corresponding to the i-th discrete element; the target residual stress coefficient corresponding to each discrete element is calculated according to the following formula: ,in, The stress coefficient corresponding to the current residual stress state. Let be the stress adjustment coefficient corresponding to the i-th discrete element; then, calculate the desired residual plastic deformation and unloading control parameters required by the corresponding spring based on the target residual stress distribution state. Specifically, first calculate the average value of the target residual stress. : Then, calculate the expected residual plastic deformation using the following formula: ,in, This represents the current average residual stress. This is the residual stress conversion factor; simultaneously, the target unloading displacement is calculated based on the difference between the target mean residual stress and the current mean residual stress. : ,in, For unloading displacement conversion factor; holding time Calculate according to the following formula: ,in, Based on the holding time, This is the pressure holding time conversion factor; simultaneously, the target equivalent stiffness is calculated based on the difference between the target mean residual stress and the current mean residual stress. : ,in, For the current equivalent stiffness, The stiffness gain coefficient is used. When the target equivalent stiffness is higher than the current equivalent stiffness, it is determined that the corresponding spring needs to improve its load-bearing capacity in the subsequent assembly state. Therefore, the direction of the expected stiffness change corresponding to the current spring is determined as the direction of stiffness improvement. Finally, the central controller outputs the expected residual plastic deformation, the expected stiffness change direction, and the unloading control parameters of the corresponding spring to the subsequent high-pressure actuator to drive the corresponding spring to perform differentiated high-pressure endpoint adjustment.

[0041] Furthermore, if increased stiffness is required, after the spring completes its basic high-pressure stroke, the actuator is controlled to perform a stress relaxation cycle that includes partial unloading and a specific pressure-holding phase, including: The actuator of the high-pressure equipment is controlled to return to the preset unloading stroke from the end position of the basic high-pressure stroke in the unloading direction; the pressure is kept constant at the end position of the preset unloading stroke, and the specific pressure holding stage is entered and the pressure holding time is continued for a preset time; after the specific pressure holding stage ends, the actuator is controlled to reload to the set safe position.

[0042] First, when the central controller determines that the predicted load value corresponding to the current spring is lower than the target load distribution ratio, and the corresponding spring has a local stiffness deficiency, it generates a corresponding stress relaxation cycle control command. Then, the central controller controls the actuator of the corresponding high-pressure equipment to return to a preset unloading stroke from the current basic high-pressure endpoint position along the unloading direction after completing the basic high-pressure stroke. The preset unloading stroke is determined based on the instantaneous equivalent stiffness of the current spring, the residual stress distribution state, and the target stiffness increase. Specifically, the load deviation is first calculated based on the target load value and the predicted load value of the current spring. ,in, For the target load value, To predict the load value; then based on the current instantaneous equivalent stiffness of the spring. Calculate the theoretical displacement correction required to reach the target load: Simultaneously, the stress non-uniformity coefficient is calculated based on the current residual stress distribution of the spring: ,in, This represents the maximum residual stress inside the spring. This represents the minimum residual stress inside the spring. The average residual stress inside the spring is used as the reference value; the preset unloading stroke is finally determined based on the theoretical displacement correction and the stress non-uniformity coefficient. The preset unloading stroke is used to release localized concentrated stress inside the spring material and reduce the rate of plastic strain accumulation.

[0043] During the unloading process, the central controller continuously collects the real-time pressure and displacement values ​​of the corresponding springs, and monitors the pressure drop rate and displacement recovery rate during the unloading phase. Then, when the actuator reaches the preset unloading stroke endpoint, the central controller controls the actuator to stop displacement change and maintains the current loading pressure constant, entering a specific pressure holding phase. After entering the pressure holding phase, the central controller continuously collects pressure data according to a preset sampling period and calculates the pressure decay rate based on the pressure change between two adjacent sampling periods. ,in, This represents the pressure value corresponding to the current sampling period. This is the pressure value corresponding to the previous sampling period. The sampling time interval is defined as follows; subsequently, based on the current instantaneous equivalent stiffness... and the instantaneous equivalent stiffness of the target Calculate the target stiffness increase ratio: The average pressure decay rate is obtained by averaging the pressure decay rates across multiple consecutive sampling periods. Then, determine the preset holding time based on the target stiffness increase ratio and the average pressure decay rate: The specific pressure holding phase is used to induce the rearrangement of the dislocation structure inside the spring material and the redistribution of internal residual stress. During the pressure holding phase, the central controller continuously monitors the pressure fluctuation amplitude, displacement micro-change, and pressure decay trend of the corresponding spring, wherein the pressure decay rate is calculated according to the following formula: The displacement rebound rate is calculated according to the following formula: ,in, and These represent the displacement values ​​corresponding to the current sampling period and the previous sampling period, respectively. Based on historical qualified spring samples, a pressure decay rate sample set and a displacement rebound rate sample set are established, and the mean values ​​of the corresponding sample sets are calculated. , and standard deviation , The pressure decay stabilization threshold and displacement rebound stabilization threshold are determined according to the following formulas: , When the following conditions are met: and At that time, it is determined that the stress relaxation process inside the spring has reached a stable state; subsequently, after the specific pressure holding phase ends, the central controller controls the actuator to reload along the loading direction, and gradually loads to the set safe position at a reloading correction rate lower than the basic loading rate; where the basic loading rate is set to be... The reloading correction rate is then determined according to the following formula: The larger the target stiffness increase ratio, the smaller the corresponding reload correction rate. The set safety position is lower than the end position of the basic high pressure to avoid re-entering the excessive plastic deformation region and maintain a relatively stable elastic stress distribution state inside the spring. After that, the central controller re-acquires the real-time force-displacement data of the corresponding spring and recalculates the instantaneous equivalent stiffness and predicted load value of the current spring. When the recalculated predicted load value meets the corresponding target load distribution requirements, the stress relaxation cycle control of the corresponding spring is stopped. Otherwise, the preset unloading stroke, pressure holding duration and reload correction rate are readjusted, and the next round of stress relaxation cycle is executed.

[0044] For example, a three-spring assembly of a certain type of spring-loaded safety valve will be used as an example for illustration. This safety valve is designed to open at a pressure of 10 MPa, and the effective pressure-bearing area of ​​the valve disc is 500 mm². 2 The theoretical opening load was calculated to be 5000N; considering valve stem friction and sealing compensation, the total target system load was determined to be 5400N. The target load distribution ratio of the three springs was set to 1:1:1, therefore the target load value for each spring was 1800N.

[0045] Three springs were installed in a synchronous high-pressure device for high-pressure treatment. The basic high-pressure stroke was set to 35mm, the basic loading rate was set to 5mm / s, and the sampling period was set to 50ms. The pressure value and compression displacement of each spring were collected in real time during the high-pressure process. When the basic high-pressure treatment ended, the instantaneous equivalent stiffness of spring 1, spring 2, and spring 3 were 52N / mm, 49N / mm, and 46N / mm, respectively. The corresponding predicted load values ​​calculated using a virtual assembly coupling model were 1910N, 1805N, and 1655N, respectively, and the predicted total system load was 5370N.

[0046] Based on the predicted results, the current load distribution ratios are calculated to be 35.57%, 33.61%, and 30.82%, respectively. Compared with the target load distribution ratios of 33.33%, 33.33%, and 33.33%, spring 1 has a higher load and spring 3 has a lower load. The central controller generates differentiated compensation control commands accordingly. Spring 1 adopts the first type of compensation mode, reducing stiffness through additional creep loading; spring 3 adopts the second type of compensation mode, increasing stiffness through stress relaxation cycles; and spring 2 maintains a strong foundation pressure state.

[0047] For spring 3, the calculated target stiffness increase ratio is: The target stiffness is taken as 50 N / mm. Based on the theoretical displacement correction of 0.82 mm and the stress non-uniformity coefficient of 0.18 obtained from real-time calculation, the preset unloading stroke is determined as follows: The average pressure decay rate measured during the pressure holding phase was 0.0045 s. -1 Based on this, the holding time is calculated: s, take the pressure holding time as 20s. After the stress relaxation cycle is completed, perform the prediction calculation again, and adjust the predicted load values ​​of spring 1, spring 2 and spring 3 to 1825N, 1808N and 1762N respectively, and adjust the total predicted load of the system to 5395N.

[0048] At this point, the load distribution ratios corresponding to the three springs are 33.83%, 33.51%, and 32.66%, respectively. The maximum deviation from the target distribution ratio decreases from the initial 2.51% to 0.67%. The relative error between the predicted total system load and the target total system load decreases from 0.56% to 0.09%. The high-pressure processing ends when both the load distribution deviation and the total system load error meet the preset control thresholds.

[0049] In summary, the embodiments of this application have at least the following technical effects: First, the target total system load and the target load distribution ratio among the springs are set under the design preload state of the target safety valve spring assembly. Next, multiple springs belonging to the same safety valve are subjected to synchronous high-pressure treatment, and the real-time pressure value and compression displacement of each spring are collected during the high-pressure loading process. Further, based on the real-time pressure value and compression displacement, the predicted load value contributed by each spring under the design preload state, and the predicted total system load of the entire spring assembly, are calculated in real-time under the current high-pressure state if the target safety valve spring assembly were directly installed into the target valve. Then, the predicted load value is compared with the target load distribution ratio, and the predicted total system load is compared with the target total system load. Based on the results of the comparison, an independent compensation control command is generated for each spring. Finally, according to the compensation control command, the high-pressure loading endpoint of each spring is differentially adjusted during the high-pressure process, so that the predicted load value approaches the target load distribution ratio, and the predicted total system load approaches the target total system load. This invention solves the technical problem in the prior art where the load distribution among the springs in the safety valve spring assembly is difficult to maintain in a consistent manner during high-pressure treatment and assembly pre-tightening, resulting in poor stability of the safety valve's opening and closing performance. It achieves the technical effect of improving the load distribution balance of the spring assembly and enhancing the opening and closing stability of the safety valve.

[0050] Example 2, based on the same inventive concept as the adaptive load distribution control method with spring compression in the previous examples, such as... Figure 2 As shown, this application provides an adaptive load distribution control system with spring-loaded compression, wherein the system includes: Target setting module 11: Sets the target total system load of the target safety valve spring assembly under the design preload state of the safety valve and the target load distribution ratio among the springs; Data acquisition module 12: Performs synchronous high-pressure processing on multiple springs belonging to the same safety valve, and collects the real-time pressure value and compression displacement of each spring during the high-pressure loading process; Calculation module 13: Based on the real-time pressure value and the compression displacement, calculates in real-time the predicted load value contributed by each spring under the design preload state if the target safety valve spring assembly is directly assembled into the target valve under the current high-pressure state, as well as the predicted total system load of the entire spring assembly; Comparison module 14: Performs a double comparison between the predicted load value and the target load distribution ratio, and between the predicted total system load and the target total system load, and generates an independent compensation control command for each spring based on the double comparison results; Adjustment module 15: According to the compensation control command, performs differentiated adjustment on the high-pressure loading endpoint of each spring during the high-pressure process, so that the predicted load value approaches the target load distribution ratio and the predicted total system load approaches the target total system load.

[0051] Furthermore, the calculation module 13 is used to perform the following method: During the high-pressure loading process, the real-time force-displacement data stream, consisting of the real-time pressure value of each spring and the corresponding compression displacement, is input into a pre-constructed virtual assembly coupling model that includes the nonlinear stiffness of the valve disc-spring contact. Based on the input real-time force-displacement data stream, the virtual assembly coupling model dynamically identifies and outputs the instantaneous equivalent stiffness and residual deformation trend of each spring under the current high-pressure state. Based on the instantaneous equivalent stiffness and residual deformation trend of each spring, the virtual assembly coupling model continues to solve for the virtual assembly compression amount of each spring required to achieve the designed preload state under the geometric constraints of the target valve. Based on the virtual assembly compression amount and the instantaneous equivalent stiffness, the predicted load value contributed by each spring under the designed preload state is calculated, and the predicted load values ​​of all springs are summed to obtain the total predicted system load of the entire spring assembly.

[0052] Furthermore, the calculation module 13 is used to perform the following method: The virtual assembly coupling model is an iterative solution model built on the displacement coordination equation and the force balance equation; wherein, the displacement coordination equation is used to characterize that the sum of the compression displacements of each spring under the valve disc constraint is a constant value, and the force balance equation is used to characterize that the sum of the predicted load values ​​of each spring is equal to the resultant force on the valve disc.

[0053] Furthermore, the calculation module 13 is used to perform the following method: The instantaneous equivalent stiffness is obtained by calculating the rate of change of the real-time pressure value near the current sampling point relative to the compression displacement. The residual deformation trend is obtained by extrapolating and predicting the change in the free length of the spring after unloading by analyzing the real-time force-displacement data stream of the current and historical sampling points.

[0054] Furthermore, the comparison module 14 is used to perform the following method: The predicted load values ​​are converted into a predicted load distribution vector according to a proportional relationship, and compared with the target distribution vector determined by the target load allocation ratio. The vector angle and magnitude difference are calculated as the first set of indicators for evaluating the uniformity of load distribution. The relative error between the predicted total load of the system and the target total load is calculated as the second indicator for evaluating the consistency of system load. A comprehensive decision function is established, with the optimization objective being to minimize the first set of indicators and the second set of indicators. During the high-pressure process, the comprehensive decision function is dynamically solved based on the real-time values ​​of the first set of indicators and the second set of indicators to obtain a set of desired adjustment parameters for each spring that optimizes the current optimization objective. The desired adjustment parameters are mapped as independent compensation control commands to control the action of the actuators of the high-pressure equipment.

[0055] Furthermore, the comparison module 14 is used to perform the following method: In the comprehensive decision function, the optimization weight set for the first set of indicators for evaluating the uniformity of load distribution is higher than the optimization weight set for the second indicator for evaluating the consistency of system load.

[0056] Furthermore, the adjustment module 15 is used to perform the following method: The compensation control command is analyzed to obtain the desired residual plastic deformation and desired stiffness change direction for each spring. Based on the desired stiffness change direction, if stiffness needs to be reduced, after the spring completes the basic high-pressure stroke, the creep-induced rate is continued to be applied at a rate lower than the basic loading rate until the actual residual strain increment reaches the desired residual strain increment. If stiffness needs to be increased, after the spring completes the basic high-pressure stroke, the actuator is controlled to perform a stress relaxation cycle that includes partial unloading and a specific pressure holding stage.

[0057] Furthermore, the adjustment module 15 is used to perform the following method: The independent compensation control command is decoded to extract the compensation mode identifier and compensation amount parameter. If the compensation mode identifier corresponds to the first type of compensation mode, the compensation amount parameter is directly mapped to the expected residual plastic deformation amount, and the expected stiffness change direction is determined to require a reduction in stiffness. If the compensation mode identifier corresponds to the second type of compensation mode, the target residual stress distribution state is calculated based on the compensation amount parameter, and the expected residual plastic deformation amount and unloading control parameter are calculated based on the target residual stress distribution state. At the same time, the expected stiffness change direction is determined to require an increase in stiffness.

[0058] Furthermore, the adjustment module 15 is used to perform the following method: The actuator of the high-pressure equipment is controlled to return to the preset unloading stroke from the end position of the basic high-pressure stroke in the unloading direction; the pressure is kept constant at the end position of the preset unloading stroke, and the specific pressure holding stage is entered and the pressure holding time is continued for a preset time; after the specific pressure holding stage ends, the actuator is controlled to reload to the set safe position.

[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An adaptive load distribution control method with spring-forced processing, characterized in that, The method includes: Set the target total system load and the target load distribution ratio among the springs of the target safety valve spring assembly under the design preload state of the safety valve; Multiple springs belonging to the same safety valve are subjected to synchronous high-pressure treatment, and the real-time pressure value and compression displacement of each spring are collected in real time during the high-pressure loading process. Based on the real-time pressure value and the compression displacement, the predicted load value contributed by each spring under the design preload state is calculated in real time if the target safety valve spring assembly is directly assembled into the target valve under the current high pressure state, as well as the predicted total system load of the entire spring assembly. The predicted load value is compared with the target load allocation ratio, and the predicted total system load is compared with the target total system load. Based on the results of the comparison, an independent compensation control command is generated for each spring. According to the compensation control command, the end point of each spring is differentially adjusted during the high-pressure process, so that the predicted load value approaches the target load distribution ratio and the predicted total load of the system approaches the target total load. The predicted load value is compared with the target load allocation ratio, and the predicted total system load is compared with the target total system load. Based on the comparison results, an independent compensation control command is generated for each spring, including: The predicted load values ​​are converted into predicted load distribution vectors according to the proportional relationship, and compared with the target distribution vector determined by the target load allocation ratio. The vector angle and magnitude difference are calculated as the first set of indicators for evaluating the uniformity of load distribution. The relative error between the predicted total load of the system and the target total load is calculated as a second indicator for evaluating the consistency of system load. A comprehensive decision function is established, wherein the comprehensive decision function takes minimizing the first set of indicators and the second indicator as the optimization objective; During the high-pressure process, the comprehensive decision function is dynamically solved based on the real-time values ​​of the first set of indicators and the second set of indicators to obtain a set of desired adjustment parameters for each spring that optimizes the current optimization objective. The desired adjustment parameters are mapped to independent compensation control commands that control the actions of the actuators in the high-pressure equipment.

2. The adaptive load distribution control method for spring-forced processing as described in claim 1, characterized in that, Based on the real-time pressure value and the compression displacement, the predicted load value contributed by each spring under the designed preload state is calculated in real time under the current high pressure state, if the target safety valve spring assembly is directly assembled into the target valve, as well as the predicted total system load of the entire spring assembly, including: During the high-pressure loading process, the real-time force-displacement data stream, consisting of the real-time pressure value of each spring and the corresponding compression displacement, is input into a pre-built virtual assembly coupling model that includes the nonlinear stiffness of the valve disc-spring contact. The virtual assembly coupling model dynamically identifies and outputs the instantaneous equivalent stiffness and residual deformation trend of each spring under the current high pressure state based on the input real-time force-displacement data stream. Based on the instantaneous equivalent stiffness and residual deformation trend of each spring, the virtual assembly coupling model continues to solve for the virtual assembly compression of each spring required to achieve the design preload state under the geometric constraints of the target valve. Based on the virtual assembly compression and the instantaneous equivalent stiffness, the predicted load value contributed by each spring under the design preload is calculated, and the predicted load values ​​of all springs are summed to obtain the total predicted system load of the entire spring assembly.

3. The adaptive load distribution control method for spring-forced processing as described in claim 2, characterized in that, The virtual assembly coupling model is an iterative solution model built based on the displacement compatibility equation and the force balance equation; The displacement coordination equation is used to characterize that the sum of the compression displacements of each spring under the valve disc constraint is a constant value, and the force balance equation is used to characterize that the sum of the predicted load values ​​of each spring is equal to the resultant force on the valve disc.

4. The adaptive load distribution control method for spring-forced processing as described in claim 3, characterized in that, The instantaneous equivalent stiffness is obtained by calculating the rate of change of the real-time pressure value near the current sampling point relative to the compression displacement. The residual deformation trend is obtained by extrapolating and predicting the change in the free length of the spring after unloading by analyzing the real-time force-displacement data stream of the current and historical sampling points.

5. The adaptive load distribution control method for spring-forced processing as described in claim 1, characterized in that, In the comprehensive decision function, the optimization weight set for the first set of indicators for evaluating the uniformity of load distribution is higher than the optimization weight set for the second indicator for evaluating the consistency of system load.

6. The adaptive load distribution control method for spring-forced processing as described in claim 1, characterized in that, Based on the compensation control command, the endpoint of the high-pressure loading of each spring is adjusted differentially during the high-pressure process, including: The compensation control command is analyzed to obtain the desired residual plastic deformation and desired stiffness change direction for each spring. According to the desired stiffness change direction, if it is necessary to reduce the stiffness, after the spring completes the basic strong compression stroke, it continues to be loaded at a creep induction rate lower than the basic loading rate until the actual residual strain increment reaches the desired residual strain increment. If increased stiffness is required, after the spring completes its basic high-pressure stroke, the actuator is controlled to perform a stress relaxation cycle that includes partial unloading and a specific pressure-holding phase.

7. The adaptive load distribution control method for spring-forced processing as described in claim 6, characterized in that, Parsing the compensation control command to obtain the desired residual plastic deformation and desired stiffness change direction for each spring includes: The independent compensation control command is decoded to extract the compensation mode identifier and compensation amount parameter; If the compensation mode identifier corresponds to the first type of compensation mode, then the compensation amount parameter is directly mapped to the expected residual plastic deformation amount, and the expected stiffness change direction is determined as the stiffness that needs to be reduced. If the compensation mode identifier corresponds to the second type of compensation mode, the target residual stress distribution state is calculated based on the compensation amount parameter, and the expected residual plastic deformation amount and unloading control parameters are calculated based on the target residual stress distribution state. At the same time, the direction of the expected stiffness change is determined as the stiffness that needs to be increased.

8. The adaptive load distribution control method for spring-forced processing as described in claim 6, characterized in that, If increased stiffness is required, after the spring completes its basic high-pressure stroke, the actuator is controlled to perform a stress relaxation cycle that includes partial unloading and a specific pressure-holding phase, including: The actuator of the high-pressure equipment is controlled to return to the preset unloading stroke from the end position of the basic high-pressure stroke in the unloading direction; The pressure remains constant at the end of the preset unloading stroke, and the specific pressure holding stage continues for a preset pressure holding time. After the specific pressure holding phase ends, the actuator is controlled to reload to the set safe position.

9. An adaptive load distribution control system with spring-forced compression, characterized in that, The system for implementing the adaptive load distribution control method for spring-forced processing according to any one of claims 1-8, the system comprising: Target setting module: Sets the target total system load and the target load distribution ratio between each spring when the safety valve spring assembly is in the safety valve design preload state; Data acquisition module: synchronously applies high pressure to multiple springs belonging to the same safety valve, and during the high pressure loading process, collects the real-time pressure value and compression displacement of each spring. Calculation module: Based on the real-time pressure value and the compression displacement, calculate in real time the predicted load value that each spring will contribute under the design preload state if the target safety valve spring assembly is directly assembled into the target valve under the current high pressure state, as well as the predicted total system load of the entire spring assembly; Comparison module: performs a dual comparison between the predicted load value and the target load allocation ratio, and between the predicted total system load and the target total system load, and generates an independent compensation control command for each spring based on the dual comparison results; Adjustment module: Based on the compensation control command, the adjustment module differentiates the end point of each spring during the high-pressure process, so that the predicted load value approaches the target load distribution ratio and the predicted total load of the system approaches the target total load.

Images