A method and system for continuous processing of dynamic characteristic curve of pressure reducing valve based on double closed loop control

Abstract

This invention provides a method and system for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, relating to the field of fluid transmission technology. The method includes: installing the valve in a test circuit and setting the rated inlet pressure, rated outlet pressure, and rated flow rate required for the test; continuously adjusting and locking the outlet flow rate of the pressure reducing valve under test to the rated flow rate based on the rated flow rate, while maintaining a constant outlet flow rate; under the condition of a constant outlet flow rate, continuously and smoothly reducing the inlet pressure of the pressure reducing valve under test from the rated inlet pressure at a preset linear rate, and synchronously acquiring the instantaneous values ​​of each inlet pressure and the corresponding instantaneous value of the outlet pressure at a preset high-frequency sampling rate during the pressure reduction process to obtain an initial pressure characteristic point set. This invention realizes the construction of a constant flow rate - continuously variable inlet pressure pressure characteristic test condition and a constant inlet pressure - continuously variable flow rate flow characteristic test condition.

Description

Technical Field

[0001] This invention relates to the field of fluid transmission technology, and in particular to a method and system for continuously processing the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control. Background Technology

[0002] As a core pressure regulating component in fluid transmission and industrial control systems, pressure reducing valves are widely used in various industrial scenarios with stringent requirements for pressure control accuracy and operational stability. Their dynamic performance, such as pressure characteristics and flow characteristics, directly determines the operational reliability and control accuracy of the entire control system. Currently, the industry still generally adopts the traditional manual point-by-point testing mode for testing the dynamic characteristics of pressure reducing valves, which has technical limitations and application pain points.

[0003] For example, when professional hydraulic component manufacturers conduct batch performance testing on pilot-operated pressure reducing valves used in construction machinery, they often use traditional industry-standard testing methods. When testing pressure characteristics, they rely on manual adjustment of the inlet pressure regulating valve and the outlet throttle valve. They first fix the outlet flow rate and then adjust the inlet pressure step by step. After each pressure adjustment, they wait for the pressure to stabilize and then manually record the data. Only a small number of discrete data points can be collected per test. Because manual adjustment makes it difficult to achieve continuous, linear, and stable changes in inlet pressure, and lacks an independent decoupled dual-closed-loop control mechanism, it is difficult to completely lock the outlet flow rate constant during dynamic changes in inlet pressure. This results in severe flow coupling interference in the collected discrete data points. The fitted pressure characteristic curve deviates from the actual pressure supply response characteristics of the pressure reducing valve, leading to large dispersion in the flow characteristic curve, poor consistency in repeated tests of the same valve, and a cumbersome and time-consuming process for testing the full dynamic characteristics of a single valve. Batch testing efficiency is extremely low, frequently resulting in misjudgments of qualified products and missed detections of unqualified products. This brings quality and safety risks to downstream OEMs' supporting systems, such as inaccurate pressure regulation, abnormal actuator operation, and insufficient safety redundancy in overall machine operation. Summary of the Invention

[0004] This invention provides a method and system for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, enabling the construction of pressure characteristic test conditions with constant flow rate and continuously variable inlet pressure, and flow characteristic test conditions with constant inlet pressure and continuously variable flow rate.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, a method for continuously processing the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, the method comprising: Step 1: Install the valve in the test circuit and set the rated inlet pressure, rated outlet pressure and rated flow rate required for the test; Based on the rated flow rate, continuously adjust and lock the outlet flow rate of the pressure reducing valve under test to the rated flow rate, and maintain the outlet flow rate of the pressure reducing valve under test constant. Step 2: Under the condition that the outlet flow rate remains constant, the inlet pressure of the pressure reducing valve under test is continuously and smoothly reduced from the rated inlet pressure at a preset linear change rate. During the pressure reduction process, the instantaneous values ​​of each inlet pressure and the corresponding instantaneous values ​​of the outlet pressure are collected synchronously at a preset high-frequency sampling rate to obtain the initial pressure characteristic point set. Step 3: Reconstruct continuous curves from the initial pressure characteristic point set to obtain an initial population containing multiple candidate curve parameter combinations; use the weighted sum of squared geometric distances between each candidate curve and each discrete point in the initial pressure characteristic point set as the fitness evaluation index, and iteratively update the search direction of each candidate curve parameter combination in the population to obtain the iterated pressure characteristic curve. Step 4: Extract a target inlet pressure value from the pressure characteristic curve within the steady-state operating range; lock the inlet pressure of the pressure reducing valve under test as the target inlet pressure value; Step 5: Under the condition that the inlet pressure remains constant, the outlet flow rate of the pressure reducing valve under test is gradually increased from zero to the rated flow rate in a preset continuous variation mode. During the flow rate change process, the set value of each outlet flow rate and the corresponding stable value of the outlet pressure are collected simultaneously to obtain the flow characteristic curve.

[0006] Secondly, a continuous processing system for the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control includes: The test module is used to install the valve in the test circuit and set the rated inlet pressure, rated outlet pressure and rated flow required for the test; based on the rated flow, the outlet flow of the pressure reducing valve under test is continuously adjusted and locked to the rated flow, and the outlet flow of the pressure reducing valve under test is kept constant. The acquisition module is used to continuously and smoothly reduce the inlet pressure of the pressure reducing valve under test from the rated inlet pressure at a preset linear rate while keeping the outlet flow constant. During the pressure reduction process, it synchronously acquires the instantaneous values ​​of each inlet pressure and the corresponding instantaneous values ​​of the outlet pressure at a preset high-frequency sampling rate to obtain the initial pressure characteristic point set. The calculation module is used to reconstruct continuous curves from the initial pressure characteristic point set to obtain an initial population containing multiple candidate curve parameter combinations. The weighted sum of squared geometric distances between each candidate curve and each discrete point in the initial pressure characteristic point set is used as the fitness evaluation index. The search direction of each candidate curve parameter combination in the population is iteratively updated to obtain the iterated pressure characteristic curve. The extraction module is used to extract a target inlet pressure value from the pressure characteristic curve within the steady-state operating range; and to lock the inlet pressure of the pressure reducing valve under test as the target inlet pressure value. The processing module is used to gradually increase the outlet flow rate of the pressure reducing valve under test from zero to the rated flow rate in a preset continuous variation mode while keeping the inlet pressure constant. During the flow rate change process, the module simultaneously collects the set value of each outlet flow rate and the corresponding stable value of the outlet pressure to obtain the flow characteristic curve.

[0007] Thirdly, a computing device includes: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0008] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0009] The above-described solution of the present invention has at least the following beneficial effects: By employing a dual-closed-loop control architecture combining independently decoupled inlet pressure closed-loop control and outlet flow closed-loop control, coupled with a pressure characteristic data acquisition method that features continuous linear adjustment of inlet pressure under constant flow conditions and high-frequency synchronous sampling, an adaptive continuous curve reconstruction algorithm based on population iteration and gradient optimization of discrete point sets, and a complete flow characteristic testing process design based on the identification of the steady-state operating region of the pressure characteristic curve for constant inlet pressure locking and continuous smooth adjustment of outlet flow, this effectively overcomes the problems of discrete test data and omission of key operating conditions caused by the reliance on manual step-by-step operation in the dynamic characteristic testing of pressure reducing valves. It also solves the problem of lacking independently decoupled control... This approach avoids the core defects of pressure and flow coupling interference caused by the control mechanism, difficulty in accurately and stably constructing test conditions, and serious distortion of characteristic curves and the actual pressure supply response characteristics of the pressure reducing valve. It avoids the industry pain points of cumbersome test procedures, low efficiency, poor consistency and reproducibility of results, and high risk of product misjudgment and missed detection caused by manual operation. It achieves accurate decoupling construction of test conditions for the two core dynamic indicators of pressure and flow characteristics of pressure reducing valve, obtains high-density continuous test data covering the entire working range, and generates high-precision continuous smooth characteristic curves that fit the actual regulating performance of pressure reducing valve, thus achieving automated, continuous and standardized processing of the entire test process. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating a method for continuously processing the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, as provided in an embodiment of the present invention.

[0011] Figure 2 This is a schematic diagram of a continuous processing system for the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, provided by an embodiment of the present invention.

[0012] Figure 3 This is a schematic diagram of the raw data from the pressure characteristic test.

[0013] Figure 4 This is a schematic diagram showing the pressure characteristic fitting curve and the identification of the steady-state working region.

[0014] Figure 5 This is a schematic diagram of the flow characteristic curve. Detailed Implementation

[0015] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0016] like Figure 1 As shown in the figure, an embodiment of the present invention proposes a method for continuously processing the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control. The method includes the following steps: Step 1: Install the valve in the test circuit and set the rated inlet pressure, rated outlet pressure and rated flow rate required for the test; Based on the rated flow rate, continuously adjust and lock the outlet flow rate of the pressure reducing valve under test to the rated flow rate, and maintain the outlet flow rate of the pressure reducing valve under test constant. Step 2: Under the condition that the outlet flow rate remains constant, the inlet pressure of the pressure reducing valve under test is continuously and smoothly reduced from the rated inlet pressure at a preset linear change rate. During the pressure reduction process, the instantaneous values ​​of each inlet pressure and the corresponding instantaneous values ​​of the outlet pressure are collected synchronously at a preset high-frequency sampling rate to obtain the initial pressure characteristic point set. Step 3: Reconstruct continuous curves from the initial pressure characteristic point set to obtain an initial population containing multiple candidate curve parameter combinations; use the weighted sum of squared geometric distances between each candidate curve and each discrete point in the initial pressure characteristic point set as the fitness evaluation index, and iteratively update the search direction of each candidate curve parameter combination in the population to obtain the iterated pressure characteristic curve. Step 4: Extract a target inlet pressure value from the pressure characteristic curve within the steady-state operating range; lock the inlet pressure of the pressure reducing valve under test as the target inlet pressure value; Step 5: Under the condition that the inlet pressure remains constant, the outlet flow rate of the pressure reducing valve under test is gradually increased from zero to the rated flow rate in a preset continuous variation mode. During the flow rate change process, the set value of each outlet flow rate and the corresponding stable value of the outlet pressure are collected simultaneously to obtain the flow characteristic curve.

[0017] In this embodiment of the invention, the technical means of connecting the pressure reducing valve under test to the test circuit and setting rated operating parameters, continuously adjusting and locking the outlet flow of the pressure reducing valve under test to the rated flow through closed-loop control of the outlet flow to construct a constant flow benchmark operating condition, is complemented by the technical means of continuously and steadily reducing the inlet pressure of the pressure reducing valve under test from the rated value at a preset linear change rate under constant outlet flow conditions, and synchronously collecting the instantaneous values ​​of inlet and outlet pressures at a preset high-frequency sampling rate to construct an initial pressure characteristic point set. Combined with the technical means of continuously reconstructing the initial pressure characteristic point set into curves, generating an initial population containing multiple sets of candidate curve parameter combinations, and iteratively optimizing the high-precision pressure characteristic curve by repeatedly iterating and optimizing the weighted sum of squares of the geometric distances between candidate curves and discrete points as the fitness evaluation index, the target inlet pressure value within the steady-state operating region is extracted from the pressure characteristic curve and the inlet pressure of the pressure reducing valve under test is locked to the target value to construct a constant inlet pressure. The benchmark operating condition technology, ultimately coupled with a closed-loop control solution that gradually increases the outlet flow rate from zero to the rated flow rate under constant inlet pressure, and simultaneously collects flow and pressure data to generate a flow characteristic curve, effectively overcomes the problems of discrete test data and omission of key operating conditions caused by manual point-by-point step-by-step adjustment in the dynamic characteristic test of pressure reducing valves. It solves the core technical defects of traditional test methods, such as pressure and flow coupling interference, difficulty in maintaining stable test conditions, and serious distortion of the fitted characteristic curve with the actual pressure supply response characteristics of the pressure reducing valve, caused by the lack of an independent decoupling control mechanism. It achieves accurate decoupling and stable construction of the pressure characteristic test condition of constant flow rate - continuously variable inlet pressure and the flow characteristic test condition of constant inlet pressure - continuously variable flow rate, obtains high-density continuous test data covering the entire operating range of the pressure reducing valve, and generates a high-precision continuous smooth characteristic curve that closely matches the actual dynamic adjustment performance of the pressure reducing valve.

[0018] In a preferred embodiment of the present invention, step 1 above may include: Step 1.1: Connect the pressure reducing valve under test in series to the main flow path of the test circuit, and obtain the reference control parameter set through the rated inlet pressure, rated outlet pressure, and rated flow rate. Specifically, it includes: rigidly connect the inlet end of the pressure reducing valve under test to the outlet end of the inlet pressure control unit of the test circuit through a high-pressure sealing pipeline, and rigidly connect the outlet end of the pressure reducing valve under test to the inlet end of the flow control unit of the test circuit through a high-pressure sealing pipeline to complete the series connection of the pressure reducing valve under test in the main flow path of the test circuit. After the connection is completed, conduct an airtight pressure-holding test on the installation and connection parts of the pressure reducing valve under test. The pressure-holding pressure is set to the rated inlet pressure of the pressure reducing valve under test, and the pressure-holding duration is set to 30 seconds. If the pressure drop value during the pressure-holding process does not exceed 0.05 MPa, the airtightness test is qualified to ensure that there is no leakage in the test circuit. Through the upper computer interaction interface of the test system, input the three core rated working parameters of the rated inlet pressure, rated outlet pressure, and rated flow rate corresponding to the pressure reducing valve under test. After receiving the input of the three core rated working parameters, retrieve the built-in control parameter matching database of the system, and match six basic control parameters corresponding to the rated working parameters of the pressure reducing valve under test, namely the flow closed-loop control proportional coefficient, flow closed-loop control integral coefficient, flow closed-loop control differential coefficient, steady-state flow tolerance threshold, feed-forward compensation coefficient, and reference valve opening value. Package the six basic control parameters in an orderly manner to generate the reference control parameter set corresponding to the pressure reducing valve under test. The steady-state flow tolerance threshold is set to ±0.2%.

[0019] Step 1.2: Based on the reference control parameter set, start the flow closed-loop control loop located downstream of the pressure reducing valve under test, and obtain the flow deviation signal by capturing the instantaneous flow feedback value in the downstream pipeline in real time and performing dynamic deviation calculation between the instantaneous flow feedback value and the rated flow rate. Specifically, it includes: send the generated reference control parameter set to the lower computer execution unit corresponding to the flow closed-loop control loop. After receiving the reference control parameter set, the lower computer execution unit starts the flow closed-loop control loop located downstream of the pressure reducing valve under test. After the loop is started, use the high-precision mass flow controller installed in the downstream pipeline of the pressure reducing valve under test to capture the fluid instantaneous flow feedback value in the downstream pipeline in real time at a sampling frequency of 100 Hz. Perform dynamic deviation calculation between the instantaneous flow feedback value at the same sampling moment and the pre-input rated flow rate. The calculation process uses the following formula: ; where is the flow deviation signal at the sampling moment, is the rated flow rate of the pressure reducing valve under test, is the instantaneous flow feedback value captured at the sampling moment. Send the flow deviation signal calculated at each sampling moment to the lower computer execution unit of the flow closed-loop control loop in real time to complete the real-time generation and transmission of the flow deviation signal.

[0020] Step 1.3: Based on the flow deviation signal, a regulating valve opening compensation command is obtained, driving the downstream precision regulating valve to perform continuous opening adjustment, causing the instantaneous flow feedback value to converge towards the rated flow until the instantaneous flow feedback value falls within the preset steady-state flow tolerance threshold range. Specifically, this includes: After receiving the real-time flow deviation signal, the lower-level execution unit of the flow closed-loop control loop combines the flow closed-loop control proportional coefficient, flow closed-loop control integral coefficient, and flow closed-loop control derivative coefficient from the reference control parameter set, and adopts an incremental proportional-integral-derivative control algorithm. ; in, For the first Within each sampling period, the incremental value of the opening compensation of the downstream precision control valve is the final output of the algorithm and is directly used as the core control parameter of the control valve opening compensation command. The unit is the percentage of the control valve opening. The proportional coefficient for the flow closed-loop control is a pre-matched and tuned fixed control parameter in the reference control parameter set. It is dimensionless and used to adjust the control gain of the proportional link. The integral coefficient of the flow closed-loop control in the reference control parameter set is a fixed control parameter that is pre-matched and tuned. It is dimensionless and used to eliminate static deviations in flow regulation. The differential coefficients of the flow closed-loop control are the reference control parameter set. They are fixed control parameters that are pre-matched and tuned, and are dimensionless. They are used to suppress dynamic overshoot and fluctuations during the flow regulation process. For the first The flow deviation signal calculated at each sampling time is completely consistent with the flow deviation signal defined in step 1.2; For the first The flow deviation signal corresponding to each sampling time, that is, the flow deviation calculation result of the previous sampling period; For the first The flow deviation signal corresponding to each sampling time, that is, the flow deviation calculation result of the first two sampling periods.

[0021] The flow deviation signal is processed to generate a corresponding control valve opening compensation command. The lower-level execution unit sends the generated control valve opening compensation command to the downstream precision control valve in real time, driving the valve's actuator to perform continuous opening adjustment actions. By changing the flow cross-sectional area of ​​the control valve, the flow resistance of the downstream pipeline is adjusted, causing the instantaneous flow feedback value in the pipeline to continuously converge towards the rated flow. During the adjustment process, the instantaneous flow feedback value at each sampling moment is continuously judged for steady state. When the instantaneous flow feedback values ​​captured within 15 consecutive sampling cycles all fall within the preset steady-state flow tolerance threshold range, the flow adjustment is determined to have reached a steady state.

[0022] Step 1.4: Based on the instantaneous flow feedback value falling within the steady-state flow tolerance threshold range, lock the current opening position of the downstream precision control valve. By performing real-time feedforward compensation for pipeline flow resistance disturbances, the outlet flow through the pressure reducing valve under test is stably anchored at the rated flow, obtaining a constant flow reference condition for continuous pressure regulation. Specifically, this includes: after completing the steady-state flow determination, immediately sending an opening lock command to the lower-level execution unit. The standard for steady-state flow determination is that the instantaneous flow feedback values ​​captured within 15 consecutive sampling periods all fall within the preset steady-state flow tolerance threshold range, which is set to ±0.2%. After receiving the opening lock command, the lower-level execution unit immediately latches the current opening position of the downstream precision control valve, generates an opening latch reference value, and simultaneously blocks all opening adjustment command channels within the lower-level execution unit except for the feedforward compensation command. It prohibits the control valve actuator from receiving opening adjustment commands related to flow closed-loop proportional-integral-derivative control, retaining only the sole write permission for the feedforward compensation command.

[0023] The built-in steady-state flow maintenance algorithm is activated synchronously. This algorithm is a feedforward compensation algorithm based on the flow characteristics of a hydraulic thin-walled throttling orifice. It can achieve hysteresis-free pre-compensation for flow disturbances caused by inlet pressure fluctuations, solving the industry pain points of traditional testing processes where the outlet flow cannot be kept constant when the inlet pressure changes dynamically, and where there is flow coupling interference in the test data. The algorithm's sampling frequency is consistent with the inlet pressure data acquisition frequency, set to 100 Hz, corresponding to a sampling period of 0.01 seconds, which perfectly matches the sampling period of the previous flow closed-loop control.

[0024] After the steady-state flow maintenance algorithm is activated, a high-precision inlet pressure sensor installed upstream of the pressure reducing valve under test collects the real-time inlet pressure value at a sampling frequency of 100 Hz. The difference between the real-time inlet pressure value collected in each sampling cycle and the inlet pressure reference value corresponding to the opening lock-in moment is calculated to obtain the inlet pressure change. The calculation formula is as follows: ; in For the first The change in inlet pressure corresponding to each sampling period For the first The real-time inlet pressure value of the pressure reducing valve under test, collected in each sampling cycle. The inlet pressure reference value corresponding to the opening lock moment is the measured fixed value of the inlet pressure collected when the flow rate reaches the steady-state judgment condition.

[0025] The algorithm constructs a feedforward compensation model based on the pre-calibrated throttling flow characteristics of the control valve. The valve orifice flow of the downstream precision control valve conforms to the commonly used thin-walled throttling orifice flow characteristic equation in the field of hydraulic transmission, providing a physical calculation basis for the feedforward compensation model. The equation is as follows: ; in, This refers to the instantaneous flow rate through the downstream precision regulating valve, i.e., the outlet flow rate through the pressure reducing valve being measured. The flow coefficient at the valve orifice is a fixed constant calibrated at the factory and is determined by the valve orifice structure and the viscosity of the test medium. The flow cross-sectional area of ​​the regulating valve orifice is used to determine the regulating valve opening. A single-valued linear function, To regulate the pressure difference between the inlet and outlet of the valve, The fluid density of the test medium is denoted as , and is a fixed constant.

[0026] When the valve is locked, the outlet pressure of the pressure reducing valve is essentially constant. The change in the pressure difference between the inlet and outlet of the regulating valve is entirely caused by the fluctuation of the inlet pressure of the pressure reducing valve. Based on the control objective of constant flow, the algorithm calculates the corresponding feedforward compensation value of the regulating valve opening through a feedforward compensation model. The calculation formula is as follows: ; in For the first The feedforward compensation value of the control valve opening corresponding to each sampling period is the final output control quantity of the algorithm. The feedforward compensation gain coefficient is a fixed constant pre-calibrated using the valve orifice flow characteristics. The calibration reference is that the flow fluctuation amplitude after compensation does not exceed ±0.2% of the rated flow. For the first The change in inlet pressure corresponding to each sampling period is completely consistent with the definition of the aforementioned formula.

[0027] After the algorithm calculates the opening feedforward compensation value for each sampling period, it immediately encapsulates this compensation value into a feedforward compensation command and sends it to the actuator of the downstream precision control valve. Upon receiving the feedforward compensation command, the control valve actuator uses the opening latched reference value as a base and superimposes the opening feedforward compensation value to complete real-time opening fine-tuning. The final formula for calculating the real-time opening command output to the control valve is as follows: ; in For the first The final real-time opening command is sent to the precision control valve every sampling cycle. The control valve opening latch-up reference value, which is a fixed constant, is stored at the opening lock-in moment. For the first The opening feedforward compensation value is calculated for each sampling period. This feedforward compensation process can offset the pipeline flow disturbance caused by fluctuations in pipeline gas source pressure and continuous adjustment of inlet pressure in real time, so that the outlet flow of the pressure reducing valve under test is continuously and stably anchored at the rated flow. The flow fluctuation amplitude is controlled within the steady-state flow tolerance threshold range of ±0.2% throughout the process, completely eliminating the flow coupling interference during the dynamic change of inlet pressure, and completing the construction of the constant flow reference condition required for continuous pressure adjustment.

[0028] In this embodiment of the invention, the following techniques are employed: connecting the pressure reducing valve under test in series to the main flow path of the test circuit; generating a set of reference control parameters based on the rated inlet pressure, rated outlet pressure, and rated flow rate; in conjunction with the technique of starting the downstream flow closed-loop control circuit of the pressure reducing valve under test based on the reference control parameter set; capturing the instantaneous flow feedback value in the downstream pipeline in real time and dynamically calculating the flow deviation signal by comparing it with the rated flow rate; generating a regulating valve opening compensation command based on the flow deviation signal; driving the downstream precision regulating valve to perform continuous opening adjustment actions to converge the instantaneous flow feedback value to the rated flow rate within the preset steady-state flow tolerance threshold range; and finally, locking the current opening position of the downstream precision regulating valve and activating the steady-state flow maintenance algorithm to perform real-time feedforward compensation for pipeline flow resistance disturbances. The full-process flow closed-loop control scheme effectively overcomes the problems of low flow setting accuracy and poor steady-state maintenance caused by the reliance on manual adjustment of the outlet throttle valve in traditional pressure reducing valve pressure characteristic testing. It solves the core defects of traditional testing methods, such as easy flow fluctuation and difficulty in stabilizing the rated flow during dynamic changes in inlet pressure, due to the lack of an independent flow closed-loop control mechanism. It avoids the industry pain point of pressure characteristic test data distortion caused by the coupling interference of flow fluctuation and pressure regulation. Thus, it realizes high-precision continuous adjustment and stable locking of the outlet flow of the pressure reducing valve under test, and constructs a constant flow benchmark condition that is not affected by subsequent continuous adjustment of inlet pressure. It eliminates the influence of flow coupling factors on pressure characteristic testing and ensures the stability, consistency and reproducibility of the pressure characteristic test condition.

[0029] In a preferred embodiment of the present invention, step 2 above may include: Step 2.1: Based on the constant flow reference condition, activate the inlet pressure closed-loop control loop located upstream of the pressure reducing valve under test, and construct a target pressure control trajectory with the rated inlet pressure as the starting reference value and a preset linear change rate for decreasing operations. Specifically, this includes: confirming that the constant flow reference condition is in a stable operating state, with the confirmation standard being that the outlet flow fluctuation amplitude is within the steady-state flow tolerance threshold range for thirty consecutive sampling periods, and there are no abnormal disturbance signals. A start command is sent to the lower-level execution unit. After receiving the command, the lower-level execution unit activates the inlet pressure closed-loop control loop located upstream of the pressure reducing valve under test, constructing a target pressure control trajectory with the rated inlet pressure as the starting reference value and a preset linear change rate for decreasing operations. The linear change rate is set to 0.5 MPa per minute, and the termination value of the target pressure control trajectory is 10% of the rated inlet pressure. The target pressure control trajectory is calculated using the following formula: ; in for The target import pressure value at any given time. The rated inlet pressure of the pressure reducing valve under test. The linear rate of change of inlet pressure. This refers to the continuous operating time of the inlet pressure regulation process. The target pressure control trajectory is a continuous linear function with no abrupt changes. The rate of pressure change remains constant throughout the process, ensuring a smooth and shock-free inlet pressure regulation process. The pressure regulation process stops when the calculated target inlet pressure value drops to the preset termination value.

[0030] Step 2.2: Drive the upstream high-precision fast-response regulating valve to strictly follow the target pressure control trajectory and output throttling and unloading actions, continuously and smoothly regulating the upstream air supply pressure. This achieves a continuous and stable reduction in the inlet pressure of the pressure reducing valve from its rated inlet pressure. Specifically, the lower-level execution unit discretizes the generated target pressure control trajectory into time-by-time target pressure commands matching the sampling frequency. The sampling frequency is set to 100 Hz, consistent with the sampling frequency of the flow acquisition stage. The lower-level execution unit sends the time-by-time target pressure commands to the upstream high-precision fast-response regulating valve in real time, driving the valve core actuator to strictly follow the time-by-time target pressure commands and output continuous throttling and unloading actions. The high-precision fast-response regulating valve adopts a direct-acting electro-hydraulic servo structure with a response time of no more than 5 milliseconds, enabling micron-level precise control of the valve core opening. By changing the throttling gap between the valve core and the valve seat, continuous and smooth throttling regulation is achieved on the high-pressure fluid output from the upstream air supply source. During the adjustment process, the inlet pressure sensor installed on the upstream side of the pressure reducing valve under test collects the actual value of the inlet pressure in real time and feeds it back to the lower-level execution unit in real time, forming a complete closed-loop control link for the inlet pressure. This ensures that the inlet pressure of the pressure reducing valve under test starts from the rated inlet pressure and decreases continuously and smoothly along the target pressure control trajectory, with the pressure fluctuation amplitude not exceeding ±0.1% of the target pressure value throughout the process.

[0031] Step 2.3: During the complete cycle of continuous and stable pressure reduction at the inlet of the pressure-reducing valve under test, parallel data acquisition is performed on the upstream inlet pressure sensing node and the downstream outlet pressure sensing node at a preset high-frequency sampling rate to obtain the instantaneous inlet and outlet pressure values ​​in the time series dimension. Specifically, this includes triggering a synchronous data acquisition command when the inlet pressure closed-loop control loop starts and begins to execute the pressure reduction action. The synchronous data acquisition command is sent to the acquisition units corresponding to the upstream inlet pressure sensing node and the downstream outlet pressure sensing node. After receiving the command, the two acquisition units synchronously execute the data acquisition operation at a preset high-frequency sampling rate. The high-frequency sampling rate is set to 1000 Hz, which is much higher than the execution frequency of the closed-loop control loop, ensuring complete capture of the dynamic response details during the pressure change process. During the acquisition process, the two acquisition units are driven by the same clock source to ensure the synchronization of data acquisition. A corresponding timestamp is generated at each sampling moment to obtain the instantaneous inlet and outlet pressure values ​​in the time series dimension. Each set of synchronously acquired pressure data is assigned a unique time sequence identifier. All the instantaneous values ​​of inlet and outlet pressures are stored in the built-in temporary data buffer according to the sampling time sequence until the inlet pressure regulation process reaches the preset termination value, thus completing the parallel data acquisition within the complete cycle.

[0032] Step 2.4 involves performing timestamp alignment calibration and high-frequency noise filtering on the parallel-captured instantaneous inlet and outlet pressure values. Each calibrated instantaneous inlet pressure value at each sampling moment is mapped and paired with its corresponding instantaneous outlet pressure value, resulting in multiple pressure response data pairs. All pressure response data pairs are then sequentially packaged according to the acquisition time sequence to construct an initial pressure characteristic point set. Specifically, this includes retrieving the instantaneous inlet and outlet pressure values ​​acquired within the complete cycle from the temporary data buffer and performing timestamp alignment calibration. The timestamp alignment calibration uses a clock synchronization interpolation algorithm based on a unified clock source from the two sensor nodes. For pressure data with slight offsets at the sampling time, linear interpolation is used to map the data onto a unified standard sampling time axis, eliminating time synchronization errors caused by hardware acquisition delays. After calibration, each standard sampling moment corresponds to a unique instantaneous inlet and outlet pressure value. After calibration, high-frequency noise filtering is performed on the time-axis aligned pressure data. The filtering uses a moving average filtering algorithm with a sliding window length of five consecutive sampling points. For each standard sampling time, the arithmetic mean of the pressure data at the two sampling points before and after the current sampling point is taken as the filtered pressure data, filtering out high-frequency spike noise caused by pipeline fluid disturbances and electrical signal interference during the acquisition process. After filtering, the instantaneous inlet pressure value after calibration filtering at each standard sampling time is mapped and paired with the corresponding instantaneous outlet pressure value to obtain multiple pressure response data pairs. All pressure response data pairs are sequentially encapsulated in ascending order of the standard sampling time axis, removing unstable data from the pressure regulation start and end stages, retaining valid data pairs from the complete linear pressure reduction process, and constructing the initial pressure characteristic point set.

[0033] In this embodiment of the invention, the technical means of activating the closed-loop control loop of the upstream inlet pressure of the pressure reducing valve under test based on the established constant flow reference condition, constructing a target pressure control trajectory with the rated inlet pressure as the starting reference value and performing a decreasing operation at a preset linear change rate, are combined with the technical means of driving the upstream high-precision fast-response regulating valve to strictly follow the target pressure control trajectory and output throttling and unloading action, and continuously and smoothly adjusting the upstream air supply pressure to achieve a continuous and stable reduction of the inlet pressure of the pressure reducing valve under test from the rated inlet pressure. This is combined with the technical means of parallel data acquisition of upstream and downstream pressure sensing nodes at a preset high-frequency sampling rate within the complete cycle of continuous inlet pressure reduction, obtaining the instantaneous values ​​of inlet and outlet pressures in the time series dimension, and then performing timestamp alignment calibration and high-frequency noise filtering on the parallel captured pressure data, completing the one-to-one mapping and pairing of sampled data, and encapsulating it according to the time sequence to construct an initial pressure characteristic point set—a full-process pressure closed-loop control and data acquisition process. This integrated solution effectively overcomes the problems of discontinuous and unstable pressure changes and omission of key operating conditions caused by the reliance on manual, step-by-step adjustment of inlet pressure in traditional pressure-reducing valve pressure characteristic testing. It solves the core technical defects of traditional testing methods, such as difficulty in achieving continuous linear scanning of inlet pressure under constant flow conditions, asynchronous sampling data, and high dispersion. It avoids industry pain points such as unstable manual adjustment rhythm, high-frequency noise in sampling data, and test data distortion caused by inaccurate data pairing, which makes it difficult to truly reflect the pressure supply response characteristics of the pressure-reducing valve. As a result, it achieves high-precision, continuous, linear, and stable adjustment of the inlet pressure of the pressure-reducing valve under test, obtains high-density, time-accurately aligned, and low-noise pressure response data covering the entire test range, and constructs a high-quality and highly complete initial pressure characteristic point set. It completely eliminates test errors caused by factors such as discontinuous pressure adjustment, asynchronous data, and noise interference, ensuring the authenticity, accuracy, and consistency of pressure characteristic test data.

[0034] In a preferred embodiment of the present invention, step 3 above may include: Step 3.1: Construct a parameterized mathematical model characterizing the pressure-reducing valve's supply response based on the initial pressure characteristic point set. Within the preset physical parameter boundary constraints, randomly generate multiple sets of model parameters, encapsulate them to obtain an initial population containing multiple candidate curve parameter combinations. Specifically, retrieve the fully encapsulated initial pressure characteristic point set from the data storage unit. This initial pressure characteristic point set contains multiple sets of one-to-one corresponding instantaneous inlet and outlet pressure values. All data points undergo timestamp alignment and filtering. Based on the initial pressure characteristic point set, construct a parameterized mathematical model characterizing the pressure-reducing valve's supply response. This parameterized mathematical model is a third-order nonlinear polynomial model, used to accurately fit the nonlinear mapping relationship between the inlet and outlet pressures of the pressure-reducing valve, adapting to the pressure regulation response characteristics of pilot-operated and direct-acting pressure-reducing valves. The model expression uses the following formula: ; in This is the calculated outlet pressure value for the pressure reducing valve. This refers to the inlet pressure input value of the pressure reducing valve. The coefficient of the cubic term, The coefficient of the quadratic term, The coefficient of the linear term, For constant terms, , , , The four items are the model parameters to be solved, which are the core components of the candidate curve parameter combination.

[0035] Preset physical parameter boundary constraints are set. These constraints are determined based on the numerical range of the rated operating parameters and the initial pressure characteristic point set of the pressure reducing valve under test. The constraint range is set as follows: The value range is from -0.005 to 0.005. The value range is from -0.1 to 0.1. The value range is from 0 to 1. The value range is from 0 to 1.2 times the rated outlet pressure of the pressure reducing valve under test. Within the above physical parameter boundary constraints, multiple sets of model parameters are randomly generated using the Latin hypercube sampling method. The Latin hypercube sampling method can ensure that the generated parameters are uniformly distributed within the constraint interval, avoiding the problem of incomplete coverage of the optimization range caused by parameter clustering. The number of model parameter sets generated by sampling is set to 100 sets, and each set of parameters contains one-to-one corresponding parameters. , , , Four values ​​are used to independently and orderly encapsulate 100 sets of model parameters, thus constructing an initial population containing 100 candidate curve parameter combinations.

[0036] Step 3.2 involves iterating through each candidate curve parameter combination in the initial population and substituting them into the parameterized mathematical model to obtain the corresponding candidate fitting curve. The geometric distance between continuous interpolation points on each candidate fitting curve and discrete points in the initial pressure characteristic point set is calculated. A weighted summation of the squares of these geometric distances is performed using a preset working condition priority weighting coefficient to obtain the fitness evaluation index for each candidate curve parameter combination. Specifically, this includes: iterating through each candidate curve parameter combination in the initial population, substituting each candidate curve parameter combination into the parameterized mathematical model to obtain the candidate fitting curve corresponding to that set of parameters. Within the inlet pressure value coverage range of the initial pressure characteristic point set, continuous interpolation points are generated with a step size of 0.01 MPa. Each continuous interpolation point corresponds to a unique inlet pressure value. The inlet pressure values ​​of the interpolation points are substituted into the candidate fitting curve to calculate the outlet pressure fitting value corresponding to each interpolation point. The two-dimensional Euclidean geometric distance between each discrete point in the initial pressure characteristic point set and the corresponding interpolation point on the candidate fitting curve is calculated using the following formula: ; in The geometric distance is the distance to the i-th discrete point in the initial pressure characteristic point set. For the first Measured inlet pressure values ​​at discrete points For the candidate fitted curve and the first The inlet pressure value corresponding to each discrete point. For the first Measured values ​​of outlet pressure at discrete points The fitted value of the outlet pressure is the interpolation point on the candidate fitted curve.

[0037] By combining the preset operating condition priority weighting coefficients, a weighted summation of the squared geometric distances corresponding to each discrete point is performed to obtain the fitness evaluation index corresponding to the parameter combination of the candidate curves. The operating condition priority weighting coefficients are set based on the rated operating conditions of the pressure reducing valve under test, specifically: for discrete points with inlet pressure within ±10% of the rated inlet pressure, the weighting coefficient is set to 1.5; for discrete points in the low-pressure section with inlet pressure between 10% and 30% of the rated inlet pressure, the weighting coefficient is set to 1.2; and for discrete points in other intervals, the weighting coefficient is set to 1.0. The fitness evaluation index is calculated using the following formula: ; in This is a fitness evaluation index for candidate curve parameter combinations. This represents the total number of discrete points in the initial pressure characteristic point set. For the first The priority weight coefficients for the working conditions corresponding to each discrete point. For the first The geometric distance between discrete points. The smaller the value of the fitness evaluation index, the higher the fit between the fitted curve corresponding to the parameter combination of the candidate curve and the initial pressure characteristic point set, and the better the fitting effect.

[0038] Step 3.3: Based on the distribution of fitness evaluation indicators, identify the final guiding parameter combination of indicators in the current population. According to the preset gradient optimization rules and parameter perturbation strategy, perform parameter orientation shifting and cross-recombination on the remaining candidate curve parameter combinations in the population to complete the iterative update of the search direction of each candidate curve parameter combination. Specifically, this includes: retrieving the fitness evaluation indicators corresponding to all candidate curve parameter combinations in the initial population and analyzing the distribution of fitness evaluation indicators. Sort all fitness evaluation indicators in ascending order of values. After sorting, extract the fitness evaluation indicator with the smallest value and identify the candidate curve parameter combination corresponding to this indicator as the guiding parameter combination in the current population. The guiding parameter combination is the final parameter combination of the fitting effect in the current iteration cycle, providing a convergence direction for the iterative update of the remaining parameter combinations. Calculate the position vector difference between the remaining non-guiding candidate curve parameter combinations and the guiding parameter combination in the multidimensional parameter space. The multidimensional parameter space is a four-dimensional parameter space constructed with four model parameters a, b, c, and d as coordinate axes. Each candidate curve parameter combination corresponds to a unique position vector in the four-dimensional parameter space. The position vector difference is calculated using the following formula: ; in For the first The position vector difference corresponding to the parameter combinations of each non-guided candidate curve This is the position vector of the guide parameter combination in the four-dimensional parameter space. For the first The position vector of a combination of non-guided candidate curve parameters in the four-dimensional parameter space.

[0039] By combining the preset gradient optimization rules, a convergent gradient vector pointing to the guide parameter combination is obtained, thus determining the basic convergence direction of each candidate curve parameter combination. The gradient optimization rule adopts the stochastic gradient descent algorithm, which can improve the computational efficiency of iterative optimization while ensuring convergence accuracy and avoiding getting trapped in local finalities. The learning rate of the algorithm is set to 0.05. The convergent gradient vector is calculated using the following formula: ; in For the first The convergent gradient vector corresponding to the parameter combination of each unguided candidate curve The learning rate for the stochastic gradient descent algorithm is... For the first The position vector difference corresponds to the parameter combination of each unguided candidate curve. The direction of the convergence gradient vector is the basic convergence direction of this set of candidate curve parameter combinations. A preset parameter perturbation strategy is used to obtain a random perturbation factor conforming to a preset distribution law. The parameter perturbation strategy uses a normal distribution random number generation method, which can retain global exploration capability during iteration and avoid premature convergence of parameters. The preset distribution law is a normal distribution with a mean of 0 and a variance of 0.1. The generated random perturbation factor and the convergence gradient vector are vector superimposed to obtain a comprehensive search direction vector that combines global exploration and local fine-tuning capabilities. The calculation process uses the following formula: ; in For the first The comprehensive search direction vector corresponding to the combination of parameters of each non-guided candidate curve. This is the convergent gradient vector corresponding to this set of parameters. The four-dimensional random perturbation factor corresponds to this set of parameters. Each dimension of the random perturbation factor conforms to a normal distribution with a mean of 0 and a variance of 0.1. A directional offset update within the parameter domain is performed on each candidate curve parameter combination in the population along the comprehensive search direction vector to obtain the offset parameter nodes. A feature dimension cross-recombination operation is performed between the offset parameter nodes. The cross-recombination operation uses a single-point cross-over method with a cross-over probability of 0.7. One dimension of the four-dimensional parameters is randomly selected as the cross-over dimension, and the values ​​of the two sets of parameter nodes in the cross-over dimension are interchanged to complete the cross-recombination. Boundary checks are performed on the cross-recombined parameters to ensure that all parameters are within the preset physical parameter boundary constraints. Parameters exceeding the boundary are truncated. Finally, the updated candidate parameter combinations for the next round of iteration calculation are obtained, completing the iterative update of the search direction for each candidate curve parameter combination.

[0040] Step 3.4 involves iteratively updating the fitness evaluation index and search direction in a loop until the fluctuation difference of the fitness evaluation index in adjacent iterations converges to a preset threshold range. The final parameter combination of the index within the convergence period is extracted and substituted into the parameterized mathematical model for continuous reconstruction and smooth interpolation of all data points to obtain the iteratively updated stress characteristic curve. Specifically, after completing one round of iterative updates to the search direction, the fitness evaluation index calculation and search direction update steps are iteratively executed for the updated candidate parameter combination population. After each round of iteration, the fluctuation difference of the final fitness evaluation index of the population in the current iteration period and the previous iteration period is calculated, and it is determined whether the fluctuation difference has converged to a preset threshold range. The preset threshold range is set to 1. e-6， The maximum number of iterations is set to 1000 to avoid infinite loops in the iteration process.

[0041] When the fluctuation difference of the final fitness evaluation index between adjacent iteration cycles is less than or equal to 1 for three consecutive iteration cycles. e -6 The iteration loop terminates when the number of iterations reaches 1000. The candidate curve parameter combination with the smallest fitness evaluation index value within the convergence period at the termination of the iteration is extracted as the final parameter combination. This set of parameters is the final model parameter for the fitting effect throughout the entire iteration process.

[0042] Substituting the final parameter combination into the parametric mathematical model, continuous reconstruction of all data points is performed within the inlet pressure value coverage range of the initial pressure characteristic point set, using a step size of 0.005 MPa. The fitted outlet pressure value corresponding to each inlet pressure step size point across the entire range is calculated. Cubic spline interpolation is used to smoothly interpolate the reconstructed discrete fitting points. Cubic spline interpolation ensures that the fitted curve is continuously differentiable across the entire range, without abrupt inflection points, perfectly replicating the continuous variation characteristics of the pressure reducing valve's supply pressure response. The smoothed interpolated complete curve is then encapsulated and stored to obtain the iterative pressure characteristic curve.

[0043] In this embodiment of the invention, a technique is employed to construct a parameterized mathematical model of the pressure-reducing valve's supply response based on an initial pressure characteristic point set, and to generate an initial population of multiple candidate curve parameter combinations within preset physical parameter boundary constraints. This is complemented by a fitness verification method using the weighted sum of squared geometric distances between the fitted curve and discrete points as the fitness evaluation index. This is combined with a parameter-oriented offset and cross-recombination iterative optimization scheme based on gradient optimization rules and parameter perturbation strategies. Finally, a full-process adaptive fitting scheme is completed through iterative iteration until the index converges and the final parameters are extracted to achieve continuous curve reconstruction and smooth interpolation. Therefore, this effectively overcomes the problems of low accuracy and difficulty in adapting to the nonlinear supply response characteristics of pressure-reducing valves in traditional fitting methods. It solves the core defects of traditional methods, such as difficulty in filtering sampling noise, insufficient fitting weights for key operating conditions, and susceptibility to local finalities leading to curve distortion. It avoids the pain points of strong subjectivity and poor consistency in manual fitting, thereby achieving adaptive high-precision continuous curve reconstruction of discrete sampling point sets, effectively filtering noise interference, ensuring fitting accuracy for key rated operating conditions, avoiding fitting into local finalities, and generating a continuous and smooth pressure characteristic curve that closely matches the actual supply pressure characteristics of the pressure-reducing valve.

[0044] In a preferred embodiment of the present invention, step 3.3 above may include: Step 3.31 involves numerically sorting and selecting the extreme values ​​of the fitness evaluation indices for all candidate curve parameter combinations within the current population. The candidate curve parameter combination with the smallest fitness evaluation index is then extracted as the guiding parameter combination. Specifically, this includes retrieving all candidate curve parameter combinations within the current iteration period from the iterative calculation cache unit, along with the corresponding fitness evaluation index for each candidate curve parameter combination. The fitness evaluation index is a quantitative value characterizing the degree of fit between the candidate fitted curve and the initial pressure characteristic point set; a smaller value indicates a better fit. All fitness evaluation indices are sorted in ascending order using a bubble sort algorithm to ensure that all indices are arranged in ascending order. After sorting, the association mapping between each indices and the corresponding candidate curve parameter combinations is simultaneously retained. Extreme value selection is then performed on the sorted fitness evaluation indices. The extreme value selection rule is to extract the smallest fitness evaluation index at the top of the sorted queue and lock the candidate curve parameter combination associated with that index. The locked candidate curve parameter combination is marked as the guiding parameter combination, which is the final parameter set representing the fitting effect within the current iteration period.

[0045] Step 3.32: Calculate the position vector difference between the remaining non-guided candidate curve parameter combinations and the guided parameter combination in the multidimensional parameter space. Combine this with a pre-defined gradient optimization rule to obtain the convergence gradient vector pointing to the guided parameter combination, thus determining the basic convergence direction for each candidate curve parameter combination. Specifically, this involves constructing a four-dimensional multidimensional parameter space using the four undetermined parameters of the parameterized mathematical model as coordinate axes. The four coordinate axes correspond to the cubic coefficient *a*, quadratic coefficient *b*, linear coefficient *c*, and constant term *d*, respectively. Each candidate curve parameter combination corresponds to a unique position vector in the multidimensional parameter space, and the four dimensions of the position vector correspond one-to-one with the four coefficient values ​​of that parameter combination. Extract the position vector of the guided parameter combination in the multidimensional parameter space, and extract the position vectors of the remaining non-guided candidate curve parameter combinations in the multidimensional parameter space. Calculate the position vector difference between each non-guided candidate curve parameter combination and the guided parameter combination using the following formula: ; in For the first The position vector difference corresponding to the parameter combinations of the group of non-guided candidate curves. To guide the position vector of the parameter combination in the multidimensional parameter space, For the first The position vector of the combination of parameters of the group of non-guided candidate curves in the multidimensional parameter space.

[0046] Based on the pre-defined gradient optimization rule, the convergent gradient vector pointing from the guided parameter combination to the unguided candidate curve parameter combination is calculated, thus determining the basic convergence direction of each candidate curve parameter combination. The gradient optimization rule employs the stochastic gradient descent algorithm, which improves iterative computation efficiency while maintaining convergence accuracy and avoids getting trapped in local final solutions. The algorithm's fixed learning rate is set to 0.05. The convergent gradient vector is calculated using the following formula: ; in For the first The convergent gradient vector corresponding to the combination of parameters of the group of unguided candidate curves. The learning rate for the stochastic gradient descent algorithm is... For the first The position vector difference corresponding to the parameter combinations of the group of unguided candidate curves. The spatial orientation of the convergent gradient vector is the basic convergence direction for the iterative update of the parameter combinations of this group of candidate curves.

[0047] Step 3.33: A preset parameter perturbation strategy is invoked to obtain a random perturbation factor conforming to a preset distribution pattern. This random perturbation factor is then superimposed with the convergent gradient vector to obtain a comprehensive search direction vector that combines global exploration and local fine-tuning capabilities. Specifically, this includes: invoking the preset parameter perturbation strategy to generate a random perturbation factor conforming to a preset distribution pattern. The parameter perturbation strategy uses a normal distribution random number generation method, which can retain the global exploration capability of parameter optimization during iteration and avoid premature convergence problems. The preset distribution pattern is a standard normal distribution with a mean of 0 and a variance of 0.1. For each set of non-guided candidate curve parameter combinations, a four-dimensional random perturbation factor corresponding to the four-dimensional multi-dimensional parameter space is generated. The value of each dimension of the random perturbation factor independently conforms to a normal distribution with a mean of 0 and a variance of 0.1. The generated four-dimensional random perturbation factor is then superimposed with the convergent gradient vector corresponding to the same set of parameters to obtain a comprehensive search direction vector that combines global exploration and local fine-tuning capabilities. The calculation process uses the following formula: ; in For the first The comprehensive search direction vector corresponding to the combination of parameters of the group of non-guided candidate curves. For the first The convergent gradient vector corresponding to the combination of parameters of the group of unguided candidate curves. For the first The four-dimensional random perturbation factor corresponds to the combination of parameters of the group of unguided candidate curves. The comprehensive search direction vector not only retains the basic direction of convergence to the final fitted parameters, but also expands the exploration range of parameter optimization through random perturbation, thus balancing the local convergence accuracy and global optimization capability of the iterative process.

[0048] Step 3.34: Perform directional offset updates within the parameter domain for each candidate curve parameter combination in the population along the comprehensive search direction vector, and perform feature dimension cross-recombination operations between the offset parameter nodes to obtain the updated candidate parameter combinations for the next round of iteration calculation. This completes the iterative update of the search direction for each candidate curve parameter combination. Specifically, this includes: performing directional offset updates within the parameter domain for the position of each group of parameters in the multidimensional parameter space along the comprehensive search direction vector corresponding to each group of non-guided candidate curve parameter combinations. The step size of the directional offset update is consistent with the magnitude of the comprehensive search direction vector to obtain the offset parameter nodes. Perform physical parameter boundary checks on all offset parameter nodes to ensure that each offset parameter value is within the preset physical parameter boundary constraint range. The preset physical parameter boundary constraints are: value range of a: -0.005 to 0.005; value range of b: -0.1 to 0.1; value range of c: 0 to 1; value range of d: 0 to 1.2 times the rated outlet pressure of the pressure reducing valve under test. Perform boundary truncation processing on parameter values ​​that exceed the boundary range, and correct the excess values ​​to the critical value of the corresponding boundary.

[0049] After boundary verification, a cross-recombination operation is performed between the offset parameter nodes. This operation uses a single-point cross-recombination method with a cross-probability of 0.7. The cross-recombination process involves randomly selecting two sets of offset parameter nodes as pairs, generating random numbers between 0 and 1. When the random number is less than or equal to 0.7, the cross-recombination operation is triggered. One dimension of the four-dimensional parameters is randomly selected as the cross-dimensional value, and the values ​​of the two pairs of paired parameter nodes are swapped in this cross-dimensional relationship. When the random number is greater than 0.7, the original values ​​of the paired parameter nodes are retained without adjustment. Boundary verification is then performed again on all parameter nodes that have undergone the cross-recombination operation to ensure that all parameter values ​​conform to the physical parameter boundary constraints. This results in updated candidate parameter combinations for the next iteration, completing the iterative update of the search direction for each candidate curve parameter combination within the current iteration cycle.

[0050] In this embodiment of the invention, the technical means of numerically sorting and extreme value screening of the fitness evaluation index of all candidate curve parameter combinations in the current population, and extracting the candidate curve parameter combinations with the final fitness as the guiding parameter combination, are complemented by the technical means of calculating the position vector difference between the non-guided candidate parameter combination and the guiding parameter combination in the multidimensional parameter space, and obtaining the convergence gradient vector by combining the preset gradient optimization rules to determine the basic convergence direction. Furthermore, the technical means of generating a random perturbation factor conforming to a preset distribution law by calling a preset parameter perturbation strategy, and superimposing it with the convergence gradient vector to obtain a comprehensive search direction vector that has both global exploration and local fine-tuning capabilities, are then executed along the comprehensive search direction. This full-process parameter optimization scheme, which uses directional offset updates and feature dimension cross-recombination to obtain updated candidate parameter combinations to complete the iterative update of the search direction, effectively overcomes the problems of blind convergence direction and difficulty in balancing global exploration and local convergence accuracy in traditional curve fitting parameter iterative optimization algorithms. It solves the core technical defects of traditional optimization methods, such as being prone to getting trapped in local final solutions, low iterative convergence efficiency, and poor stability of parameter optimization results. It avoids the pain points of traditional algorithms, such as premature parameter convergence and difficulty in balancing fitting accuracy in all working conditions and key areas. Thus, it achieves accurate orientation and efficient iterative optimization of the candidate curve parameter combination search direction, perfectly balancing the global exploration capability and local fine-tuning capability of parameter optimization.

[0051] In a preferred embodiment of the present invention, step 4 above may include: Step 4.1: Analyze the pressure characteristic curve of the supply pressure effect after iteration, calculate the first-order rate of change of the outlet pressure with respect to the inlet pressure in each continuous segment of the curve, and screen out the continuous intervals where the absolute value of the first-order rate of change is continuously lower than the preset steady-state threshold. These continuous intervals are defined as the steady-state operating region. Specifically, this includes: retrieving the pressure characteristic curve after iterative optimization from the data storage unit. The pressure characteristic curve is a continuously differentiable function characterizing the mapping relationship between the outlet pressure and the inlet pressure of the pressure reducing valve. The horizontal axis of the function represents the inlet pressure, and the vertical axis represents the outlet pressure. Discretize the entire interval of the pressure characteristic curve into multiple continuous small segments with an inlet pressure step size of 0.01 MPa. Each small segment corresponds to a unique inlet pressure interval and outlet pressure interval. Calculate the first-order rate of change of the outlet pressure with respect to the inlet pressure in each small segment. The first-order rate of change is the instantaneous rate of change of the outlet pressure with respect to the inlet pressure, used to characterize the sensitivity of the pressure reducing valve's outlet pressure to fluctuations in the inlet pressure. The calculation process uses the following formula: ; in For the first The first-order rate of change of export pressure with respect to import pressure corresponds to each tiny segment. The horizontal axis of the pressure characteristic curve represents the inlet pressure. The vertical axis of the pressure characteristic curve represents the outlet pressure. For the first The midpoint value of the inlet pressure in a small segment.

[0052] The preset steady-state threshold is set to 0.05. The first-order rate of change of all micro-segments is iterated, and micro-segments with an absolute value of the first-order rate of change less than or equal to 0.05 are selected. Continuity verification is performed on the selected micro-segments. The continuity verification rule is that when the number of consecutive adjacent micro-segments meeting the requirements is greater than or equal to 200, and the total span of the corresponding inlet pressure is greater than or equal to 20% of the rated inlet pressure, the interval formed by these consecutive adjacent micro-segments is defined as the steady-state operating region. Within the steady-state operating region, the outlet pressure of the pressure reducing valve basically does not fluctuate with changes in the inlet pressure, meeting the normal operating conditions requirements of the pressure reducing valve.

[0053] Step 4.2: Within the steady-state operating range, extract the corresponding abscissa inlet pressure data according to the preset outlet pressure condition optimization principle, and determine the abscissa inlet pressure data as the target inlet pressure value. Specifically, this includes: locking the defined steady-state operating range and retrieving all data points of the pressure characteristic curve within this range. The preset outlet pressure condition optimization principle prioritizes selecting the data point within the steady-state operating range where the deviation between the outlet pressure value and the rated outlet pressure of the pressure reducing valve under test is the smallest. The abscissa inlet pressure value corresponding to this data point is the target inlet pressure value to be extracted. Traverse all data points within the steady-state operating range and calculate the absolute deviation between the outlet pressure value and the rated outlet pressure for each data point. The calculation process uses the following formula: ; in For the first in the steady-state working region The absolute deviation of the export pressure corresponding to each data point For the first The export pressure values ​​for each data point. The rated outlet pressure of the pressure reducing valve under test is used. The data point with the smallest absolute deviation is selected, and the corresponding inlet pressure data on the horizontal axis is extracted. This inlet pressure data is then determined as the target inlet pressure value. If multiple data points have the same absolute deviation value and are all minimum values, the inlet pressure corresponding to the data point whose inlet pressure value is closest to the rated inlet pressure is selected as the target inlet pressure value.

[0054] Step 4.3: The target inlet pressure value is sent as the new pressure setting benchmark to the inlet pressure closed-loop control loop located upstream of the pressure reducing valve under test, terminating the previous linear pressure reduction control logic and obtaining a constant pressure maintenance command. Specifically, this includes: using the determined target inlet pressure value as the new pressure setting benchmark, encapsulating it into a standard control command, and sending it to the lower-level execution unit corresponding to the inlet pressure closed-loop control loop located upstream of the pressure reducing valve under test. A control logic termination command is sent to the lower-level execution unit to terminate the previously executed linear pressure reduction control logic, clear the target pressure control trajectory cache data corresponding to the linear pressure reduction control logic, and prohibit the lower-level execution unit from continuing to execute linear pressure reduction-related adjustment actions. After receiving the new pressure setting benchmark and the control logic termination command, the lower-level execution unit uses the target inlet pressure value as the sole control target and generates a constant pressure maintenance command. The constant pressure maintenance command is the core execution command of the closed-loop control loop, used to drive the actuator to complete the constant locking adjustment of the inlet pressure.

[0055] Step 4.4: Based on the constant pressure maintenance command, the upstream high-precision fast-response regulating valve is driven to perform fine-tuning compensation actions, which offset the fluctuations in the pipeline gas source in real time. This causes the inlet pressure of the pressure reducing valve under test to dynamically converge and stabilize at the target inlet pressure value, completing the inlet pressure locking of the pressure reducing valve under test to obtain a constant inlet pressure reference condition. Specifically, the lower-level execution unit sends the generated constant pressure maintenance command to the upstream high-precision fast-response regulating valve in real time, driving the valve core actuator of the regulating valve to perform continuous fine-tuning compensation actions. The high-precision fast-response regulating valve adopts a direct-acting electro-hydraulic servo structure with a response time of no more than 5 milliseconds, enabling micron-level precise adjustment of the valve core opening. The pressure output of the upstream pipeline is adjusted by changing the throttling gap. During the adjustment process, the inlet pressure sensor installed upstream of the pressure reducing valve under test collects the actual inlet pressure value at the inlet end of the pressure reducing valve under test in real time at a sampling frequency of 100 Hz, and feeds back the collected actual pressure value to the lower-level execution unit in real time, forming a complete closed-loop control link. The lower-level execution unit calculates the deviation between the actual inlet pressure and the target inlet pressure in real time, and uses a proportional-integral-derivative (PID) control algorithm to generate corresponding opening compensation commands. This drives the regulating valve to perform real-time fine-tuning, offsetting inlet pressure disturbances caused by fluctuations in pipeline gas source pressure and changes in fluid flow resistance. When the deviation between the actual inlet pressure and the target inlet pressure is less than or equal to ±0.1% of the target inlet pressure within 30 consecutive sampling periods, the inlet pressure is considered locked. The current stable operating condition is defined as the constant inlet pressure reference condition, which provides a undisturbed pressure reference environment for subsequent flow characteristic tests.

[0056] In this embodiment of the invention, the following technical means are employed: analyzing the iterative pressure characteristic curve, calculating the first-order rate of change of outlet pressure to inlet pressure in each continuous segment of the curve, and screening and defining the continuous interval where the absolute value of the first-order rate of change is continuously lower than a preset steady-state threshold as the steady-state working region. This is complemented by a technique for extracting and determining the target inlet pressure value within the steady-state working region based on a preset outlet pressure condition optimization principle. Furthermore, this is combined with techniques for sending the target inlet pressure value to the inlet pressure closed-loop control loop, terminating the linear pressure reduction logic, and generating a constant pressure maintenance command. Finally, the entire process is implemented by driving an upstream high-precision regulating valve to perform fine-tuning compensation, real-time offsetting of pipeline gas source fluctuations to lock in the inlet pressure, and constructing a constant inlet pressure benchmark condition. This solution effectively overcomes the problems of relying on manual experience to select the pressure benchmark in traditional flow characteristic testing, making it difficult to accurately identify the true steady-state operating range of the pressure reducing valve. It solves the core defects of manually setting the benchmark deviating from the rated steady-state operating condition, low inlet pressure locking accuracy, and susceptibility to gas source fluctuations. It avoids the industry pain points of inaccurate test results and poor consistency in repeated tests of the same valve caused by improper selection of the benchmark operating condition. In this way, it achieves accurate identification of the steady-state operating range of the pressure reducing valve and scientific selection of the target inlet pressure value, completes high-precision and stable locking of the inlet pressure, constructs an undisturbed constant inlet pressure benchmark operating condition, eliminates the interference of inlet pressure fluctuations on flow characteristic testing, and ensures the consistency between the test operating condition and the actual operating condition of the pressure reducing valve.

[0057] In a preferred embodiment of the present invention, step 5 above may include: Step 5.1: Based on the constant inlet pressure reference condition, activate the flow closed-loop control loop located downstream of the pressure reducing valve under test. This generates a continuous ramp-type flow ramp control command sequence with zero flow as the starting threshold and rated flow as the ending threshold. Specifically, this includes: confirming that the constant inlet pressure reference condition is in a stable operating state. The confirmation standard is that within thirty consecutive sampling periods, the deviation between the actual inlet pressure value and the target inlet pressure value of the pressure reducing valve under test does not exceed ±0.1% of the target inlet pressure value, and there are no abnormal pressure disturbance signals. A start command is sent to the lower-level execution unit. After receiving the command, the lower-level execution unit activates the flow closed-loop control loop located downstream of the pressure reducing valve under test. This loop is the same independent control link as the flow closed-loop control loop in the previous pressure characteristic test phase, enabling high-precision closed-loop regulation of the outlet flow. It is completely decoupled from the inlet pressure closed-loop control loop, avoiding coupling interference between pressure and flow. A continuous ramp-type flow rate ramp-up control command sequence is constructed, starting with zero flow rate as the threshold and ending with the rated flow rate of the pressure reducing valve under test as the threshold. The flow rate ramp-up process adopts a linear change law, with the change rate set to 20% of the rated flow rate per minute to ensure a smooth and shock-free flow rate change without abrupt changes. The flow rate ramp-up control command sequence is calculated using the following formula: ; in for The target outflow value at any given time. The rated flow rate of the pressure reducing valve under test. The rate of linear change of the outlet flow rate is set to 0.2 per minute. This refers to the continuous running time of the flow ramp-up process. When the calculated target outlet flow rate reaches the rated flow rate, the flow ramp-up adjustment process stops. The time step of the command sequence is kept consistent with the sampling frequency of the flow closed-loop control loop. The sampling frequency is set to 100 Hz to ensure that each control cycle corresponds to a unique target flow rate command.

[0058] Step 5.2: Following the flow ramp control command sequence, the downstream precision regulating valve is driven to gradually increase its opening. This continuously decreases the flow resistance in the downstream pipeline, achieving a smooth, gradual increase in the outlet flow rate through the pressure reducing valve from zero to the rated flow rate. Specifically, the lower-level execution unit discretizes the generated continuous ramp flow control command sequence into time-by-time target flow commands that match the sampling frequency. Each control cycle corresponds to one target flow command, with no command intervals or abrupt changes. The lower-level execution unit sends the time-by-time target flow commands to the downstream precision regulating valve in real time, driving the valve core actuator to strictly follow the time-by-time target flow commands and perform a gradual increase in opening. The precision regulating valve uses a high-precision electro-hydraulic proportional throttle valve with micron-level valve core positioning accuracy, ensuring no jamming throughout the entire process and enabling continuous and smooth adjustment of the flow cross-sectional area. The regulating valve continuously reduces the flow resistance of the downstream pipeline by increasing the valve core opening. The pipeline flow resistance is inversely proportional to the flow cross-sectional area of ​​the regulating valve; increasing the opening increases the flow cross-sectional area and decreases the pipeline flow resistance. Under constant inlet pressure, the outlet flow rate through the pressure reducing valve gradually increases as the pipeline flow resistance decreases. During the regulation process, a high-precision mass flow controller installed downstream of the pressure reducing valve collects the actual outlet flow rate in real time at a sampling frequency of 100 Hz and feeds it back to the lower-level execution unit in real time, forming a complete flow closed-loop control link. This ensures that the outlet flow rate starts from zero and gradually and smoothly increases along the flow ramp control command sequence, with the total flow fluctuation amplitude not exceeding ±0.2% of the target flow rate value, and without any sudden flow changes or shocks.

[0059] Step 5.3: During the dynamic full cycle of gradually and smoothly increasing outlet flow, the outlet flow setpoint of the current flow closed-loop control loop is captured in real time through flow setpoint capture commands and pressure steady-state monitoring commands. The dynamic feedback data from the downstream outlet pressure sensing node is continuously tracked. Specifically, this includes triggering a synchronous monitoring command simultaneously with the start of the flow closed-loop control loop and the commencement of flow ramp-up actions. The synchronous monitoring command includes both flow setpoint capture and pressure steady-state monitoring commands. The flow setpoint capture command is sent to the lower-level execution unit of the flow closed-loop control loop. After receiving the command, the lower-level execution unit captures the corresponding outlet flow setpoint of the current flow closed-loop control loop in real time for each control cycle during the dynamic full cycle of gradually and smoothly increasing outlet flow. The outlet flow setpoint is completely consistent with the time-by-time target flow command of the flow ramp-up control command sequence, serving as the flow control target for each control cycle. The steady-state pressure monitoring command is sent to the acquisition unit corresponding to the downstream outlet pressure sensing node. Upon receiving the command, the acquisition unit continuously tracks the dynamic feedback data from the downstream outlet pressure sensing node at a high-frequency sampling frequency of 1000 Hz. The outlet pressure sensing node uses a high-precision pressure sensor with a measurement accuracy of no less than ±0.05%, accurately capturing minute dynamic changes in outlet pressure. A unique time sequence identifier is assigned to the outlet flow setpoint captured for each control cycle. Simultaneously, the dynamic feedback data of outlet pressure collected within the same time sequence window is associated with and stored with this time sequence identifier in a temporary data buffer until the flow ramp-up process reaches the rated flow termination value, completing the capture of the flow setpoint and the tracking of dynamic outlet pressure data throughout the entire dynamic cycle.

[0060] Step 5.4: When the fluctuation amplitude of the outlet pressure dynamic feedback data converges to the steady-state pressure tolerance range within the preset time window, a steady-state locking signal is triggered. The average value of the outlet pressure dynamic feedback data within the preset time window is recorded as the corresponding stable outlet pressure value. All outlet flow setpoints and the steadily locked outlet pressure stable values ​​are mapped in two-dimensional coordinates and continuously smoothed in ascending order of flow rate to construct a flow characteristic curve. Specifically, this includes: retrieving the associated outlet flow setpoints and outlet pressure dynamic feedback data stored in the temporary data cache throughout the entire cycle; matching the outlet flow setpoint corresponding to each time sequence identifier with the outlet pressure dynamic feedback data within the corresponding time window. The preset time window is set to 1 second, and the steady-state pressure tolerance range is set to ±0.3% of the average value of the outlet pressure dynamic feedback data within the corresponding time window. For the outlet pressure dynamic feedback data within each time window, its fluctuation amplitude is calculated. The fluctuation amplitude is the difference between the maximum and minimum pressure values ​​within the time window, calculated using the following formula: ; in This represents the amplitude of the export pressure fluctuation within the corresponding time window. This represents the maximum value of the dynamic feedback data for outlet pressure within this time window. This represents the minimum value of the dynamic feedback data for outlet pressure within this time window.

[0061] When the calculated fluctuation amplitude converges to within the steady-state pressure tolerance range within a preset 1-second time window, a steady-state locking signal is triggered. The arithmetic mean of all outlet pressure dynamic feedback data within this time window is calculated, and this mean is recorded as the stable outlet pressure value corresponding to the current outlet flow rate setpoint. All outlet flow rate setpoints are iterated throughout the entire cycle, and the stable outlet pressure value corresponding to each flow rate setpoint is matched and recorded. Unstable data from the initial and final stages of flow rate ramp-up are removed, and matching data pairs within the effective flow range are retained. All effective outlet flow rate setpoints and the stable outlet pressure values ​​locked in steady state are mapped onto a two-dimensional coordinate system in ascending order of flow rate, with outlet flow rate as the x-axis and stable outlet pressure value as the y-axis, constructing a discrete data point set. A cubic spline interpolation algorithm is used to continuously and smoothly fit the discrete data point set. The cubic spline interpolation algorithm ensures that the fitted curve is continuously differentiable throughout the entire flow range, without abrupt inflection points, accurately restoring the continuous response characteristics of the pressure reducing valve outlet pressure with flow rate changes. Finally, the fitted curve is encapsulated and stored, constructing the flow characteristic curve of the pressure reducing valve under test.

[0062] In this embodiment of the invention, the technical means of activating the downstream flow closed-loop control loop of the pressure reducing valve under constant inlet pressure reference condition and generating a continuous ramp-type flow ramp control command sequence with zero flow as the starting threshold and rated flow as the ending threshold, are combined with the technical means of driving the downstream precision regulating valve to perform a gradual increase in opening degree to achieve a smooth increase in outlet flow from zero to rated flow. Furthermore, the technical means of capturing the outlet flow setpoint in real time throughout the flow change cycle and continuously tracking the dynamic feedback data of outlet pressure are combined with the technique of triggering steady-state locking when the pressure fluctuation amplitude converges to the steady-state tolerance range, recording the pressure average, and completing data mapping and continuous smooth fitting to construct the flow characteristics. The full-process solution for the curve effectively overcomes the problems of discontinuous flow changes, discrete test data, and omission of key operating conditions caused by the reliance on manual point-by-point adjustment in traditional flow characteristic testing. It solves the core defects of characteristic curve distortion caused by strong subjectivity of manual steady-state judgment, inaccurate data pairing, and difficulty in achieving continuous flow scanning under constant inlet pressure. It avoids the industry pain points of cumbersome testing process, low efficiency, and poor consistency of repeated testing of the same valve. As a result, it achieves high-precision continuous smooth adjustment of outlet flow and full operating condition coverage, accurately and synchronously collects flow data and steady-state pressure data, eliminates subjective errors of manual operation, and generates a continuous and smooth flow characteristic curve that fits the actual flow response characteristics of the pressure reducing valve.

[0063] like Figure 2As shown, embodiments of the present invention also provide a continuous processing system for the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, comprising: The test module is used to install the valve in the test circuit and set the rated inlet pressure, rated outlet pressure and rated flow required for the test; based on the rated flow, the outlet flow of the pressure reducing valve under test is continuously adjusted and locked to the rated flow, and the outlet flow of the pressure reducing valve under test is kept constant. The acquisition module is used to continuously and smoothly reduce the inlet pressure of the pressure reducing valve under test from the rated inlet pressure at a preset linear rate while keeping the outlet flow constant. During the pressure reduction process, it synchronously acquires the instantaneous values ​​of each inlet pressure and the corresponding instantaneous values ​​of the outlet pressure at a preset high-frequency sampling rate to obtain the initial pressure characteristic point set. The calculation module is used to reconstruct continuous curves from the initial pressure characteristic point set to obtain an initial population containing multiple candidate curve parameter combinations. The weighted sum of squared geometric distances between each candidate curve and each discrete point in the initial pressure characteristic point set is used as the fitness evaluation index. The search direction of each candidate curve parameter combination in the population is iteratively updated to obtain the iterated pressure characteristic curve. The extraction module is used to extract a target inlet pressure value from the pressure characteristic curve within the steady-state operating range; and to lock the inlet pressure of the pressure reducing valve under test as the target inlet pressure value. The processing module is used to gradually increase the outlet flow rate of the pressure reducing valve under test from zero to the rated flow rate in a preset continuous variation mode while keeping the inlet pressure constant. During the flow rate change process, the module simultaneously collects the set value of each outlet flow rate and the corresponding stable value of the outlet pressure to obtain the flow characteristic curve.

[0064] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0065] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0066] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0067] Experimental example: Test object and test parameter settings Tested pressure reducing valve: A certain model of pilot-operated pneumatic pressure reducing valve used in industrial automation production lines.

[0068] Rated operating parameters: Rated inlet pressure: 1.0 MPa Rated outlet pressure: 0.7 MPa Rated flow rate: 100 Standard Limits per Minute (SLM) Experimental Procedures and Data Analysis Step 1, establish a constant flow baseline: Calculate flow deviation The system uses a PID algorithm to drive the downstream precision control valve to change its opening, thereby adjusting the pipeline flow resistance. After approximately 5 seconds of adjustment, the flow rate stabilizes within the range of 100 ± 0.2 SLM (steady-state flow tolerance threshold ± 0.2%). At this point, the system locks the current opening of the downstream control valve and activates the feedforward compensation algorithm to handle flow disturbances that may be caused by subsequent changes in inlet pressure. Thus, the constant flow baseline condition is established, preparing for pressure characteristic testing.

[0069] Step 2, Pressure Characteristic Testing and Data Acquisition: The upstream high-precision, fast-response control valve strictly follows this trajectory, causing the inlet pressure Pin of the measured valve to continuously and smoothly decrease to 0.2 MPa. During this process, the system simultaneously acquires the instantaneous values ​​of the inlet and outlet pressures at a high-frequency sampling rate of 1000 Hz. Approximately 9600 pairs of timestamp-aligned inlet and outlet pressure data are collected, forming the initial pressure characteristic point set.

[0070] Figure 2 The X-axis represents the inlet pressure and the Y-axis represents the outlet pressure. The blue dashed line shows the inlet pressure linearly decreasing from 1.0 MPa to 0.2 MPa. The red scatter plot represents the outlet pressure values ​​collected simultaneously, forming a dense scatter plot. It can be observed that the outlet pressure is not a straight line; it changes gradually when the inlet pressure is high and low, but changes more rapidly in the middle range.

[0071] Step 3, Reconstruction and Fitting of Pressure Characteristic Curves: Assume the pressure characteristics of the pressure reducing valve can be described by a third-order polynomial: ,parameter[ , , , The algorithm randomly generates 100 sets of candidate parameters. The weighted sum of squared geometric distances between the curve plotted for each set of parameters and all discrete points is used as the fitness evaluation index. By simulating gradient descent and parameter perturbation strategies, the parameters are continuously updated iteratively to find the parameter combination that optimizes fitness. After iterative optimization, the optimal parameter combination is obtained, and a continuous and smooth pressure characteristic curve is generated.

[0072] Figure 3 The background scatter plot (gray dots) represents the original discrete data, and the red solid line represents the continuous pressure characteristic curve obtained through optimization algorithm fitting. The green shaded area represents the calculation of the first derivative at each point on the curve; continuous intervals where the absolute value of the rate of change is less than a preset threshold (e.g., 0.05) are determined as the steady-state operating region. In this example, this region is approximately... =Between 0.5 MPa and 0.9 MPa. Within this range, the outlet pressure is minimally affected by fluctuations in the inlet pressure, and the pressure-regulating valve exhibits good pressure-stabilizing performance.

[0073] Step 4, extract and lock the target pressure: The target value for the upstream pressure closed-loop control was switched from a linear decreasing trajectory to a constant 0.52 MPa. The upstream regulating valve responded quickly, stabilizing the inlet pressure within the range of 0.52 ± 0.001 MPa. At this point, the constant inlet pressure baseline condition was established, preparing for flow characteristic testing.

[0074] Step 5, Flow Characteristic Testing and Curve Generation: The downstream precision control valve performs a gradual opening action, increasing the flow rate. The flow rate increases continuously and smoothly. Throughout the entire flow rate change process, the system records each flow rate setpoint. And monitor export pressure in real time. The system detects that at a certain flow setpoint, the fluctuation range of the outlet pressure is less than the steady-state tolerance within a 1-second time window, indicating that the pressure has reached steady state, and records the steady-state pressure value. Each flow setpoint corresponds to a steady-state outlet pressure value. All data points are arranged in ascending order of flow rate and then smoothed to obtain the flow characteristic curve.

[0075] Figure 4 The X-axis represents the outlet flow rate, and the Y-axis represents the outlet pressure. The blue curve is a continuous and smooth flow characteristic curve. When the flow rate is low, the outlet pressure remains very stable. As the flow rate increases to near the rated capacity, the outlet pressure decreases to a certain extent, which is consistent with the typical characteristics of a pressure reducing valve.

[0076] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for continuously processing the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, characterized in that, The method includes: Step 1: Install the valve in the test circuit and set the rated inlet pressure, rated outlet pressure and rated flow rate required for the test; Based on the rated flow rate, continuously adjust and lock the outlet flow rate of the pressure reducing valve under test to the rated flow rate, and maintain the outlet flow rate of the pressure reducing valve under test constant. Step 2: Under the condition that the outlet flow rate remains constant, the inlet pressure of the pressure reducing valve under test is continuously and smoothly reduced from the rated inlet pressure at a preset linear change rate. During the pressure reduction process, the instantaneous values ​​of each inlet pressure and the corresponding instantaneous values ​​of the outlet pressure are collected synchronously at a preset high-frequency sampling rate to obtain the initial pressure characteristic point set. Step 3: Reconstruct continuous curves from the initial pressure characteristic point set to obtain an initial population containing multiple candidate curve parameter combinations; use the weighted sum of squared geometric distances between each candidate curve and each discrete point in the initial pressure characteristic point set as the fitness evaluation index, and iteratively update the search direction of each candidate curve parameter combination in the population to obtain the iterated pressure characteristic curve. Step 4: Extract a target inlet pressure value from the pressure characteristic curve within the steady-state operating range; lock the inlet pressure of the pressure reducing valve under test as the target inlet pressure value; Step 5: Under the condition that the inlet pressure remains constant, the outlet flow rate of the pressure reducing valve under test is gradually increased from zero to the rated flow rate in a preset continuous variation mode. During the flow rate change process, the set value of each outlet flow rate and the corresponding stable value of the outlet pressure are collected simultaneously to obtain the flow characteristic curve.

2. The method for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control according to claim 1, characterized in that, Step 1: Install the valve in the test circuit and set the rated inlet pressure, rated outlet pressure, and rated flow rate required for the test; Based on the rated flow rate, the outlet flow rate through the pressure reducing valve under test is continuously adjusted and locked to the rated flow rate, and the outlet flow rate of the pressure reducing valve under test is kept constant, including: The pressure reducing valve under test is connected in series to the main passage of the test circuit, and the reference control parameter set is obtained by using the rated inlet pressure, rated outlet pressure and rated flow rate; Based on the reference control parameter set, the flow closed-loop control loop located downstream of the pressure reducing valve under test is started. The instantaneous flow feedback value in the downstream pipeline is captured in real time, and the dynamic deviation between the instantaneous flow feedback value and the rated flow is calculated to obtain the flow deviation signal. Based on the flow deviation signal, the regulating valve opening compensation command is obtained, which drives the downstream precision regulating valve to perform continuous opening adjustment action, so that the instantaneous flow feedback value converges to the rated flow until the instantaneous flow feedback value falls within the preset steady-state flow tolerance threshold range. Based on the instantaneous flow feedback value falling within the steady-state flow tolerance threshold range, the opening position of the current downstream precision regulating valve is locked. By real-time feedforward compensation of pipeline flow resistance disturbance, the outlet flow of the pressure reducing valve under test is stably anchored at the rated flow, thus obtaining the constant flow reference condition for continuous pressure regulation.

3. The method for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control according to claim 2, characterized in that, Step 2: Under the condition that the outlet flow rate remains constant, the inlet pressure of the pressure reducing valve under test is continuously and smoothly reduced from the rated inlet pressure at a preset linear rate. During the pressure reduction process, the instantaneous values ​​of each inlet pressure and the corresponding instantaneous values ​​of the outlet pressure are synchronously collected at a preset high-frequency sampling rate to obtain the initial pressure characteristic point set, including: Based on the constant flow reference condition, the inlet pressure closed-loop control loop located upstream of the pressure reducing valve under test is activated to construct the target pressure control trajectory with the rated inlet pressure as the starting reference value and the decreasing operation performed at a preset linear change rate. The upstream high-precision fast-response regulating valve is driven to strictly follow the target pressure control trajectory and output throttling and unloading actions to continuously and smoothly regulate the upstream air supply pressure, so as to achieve a continuous and stable reduction of the inlet pressure of the pressure reducing valve under test from the rated inlet pressure. During the complete cycle of continuous and stable reduction of the inlet pressure of the pressure reducing valve under test, parallel data capture is performed on the upstream inlet pressure sensing node and the downstream outlet pressure sensing node at a preset high-frequency sampling rate to obtain the instantaneous values ​​of the inlet pressure and the outlet pressure in the time series dimension. The instantaneous inlet and outlet pressure values ​​captured in parallel are calibrated with timestamp alignment and high-frequency noise filtering. The calibrated instantaneous inlet pressure value and the corresponding instantaneous outlet pressure value at each sampling time are mapped and paired one by one to obtain multiple pressure response data pairs. All pressure response data pairs are encapsulated in an orderly manner according to the acquisition time sequence to construct the initial pressure characteristic point set.

4. The method for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control according to claim 3, characterized in that, Step 3: Reconstruct continuous curves from the initial pressure characteristic point set to obtain an initial population containing multiple candidate curve parameter combinations; use the weighted sum of squared geometric distances between each candidate curve and each discrete point in the initial pressure characteristic point set as the fitness evaluation index, and iteratively update the search direction of each candidate curve parameter combination in the population to obtain the iterated pressure characteristic curves, including: A parameterized mathematical model characterizing the pressure supply response of a pressure reducing valve is constructed based on an initial pressure characteristic point set. Multiple sets of model parameters are randomly generated within the preset physical parameter boundary constraints, and encapsulated to obtain an initial population containing multiple candidate curve parameter combinations. The candidate curve parameter combinations in the initial population are traversed and substituted into the parameterized mathematical model to obtain the corresponding candidate fitting curves. The geometric distance between the continuous interpolation points on each candidate fitting curve and the discrete points in the initial pressure characteristic point set is calculated. The geometric distance is weighted and squared by the preset working condition priority weight coefficient to obtain the fitness evaluation index of each candidate curve parameter combination. Based on the distribution trend of fitness evaluation indicators, the final guiding parameter combination of indicators in the current population is identified. According to the preset gradient optimization rules and parameter perturbation strategy, parameter orientation shift and cross-recombination are performed on the other candidate curve parameter combinations in the population to complete the iterative update of the search direction of each candidate curve parameter combination. The fitness evaluation index calculation and search direction are iteratively updated repeatedly until the fluctuation difference of the fitness evaluation index in adjacent iterations converges to the preset threshold range. The final parameter combination of the index within the convergence cycle is extracted and substituted into the parameterized mathematical model to continuously reconstruct and smoothly interpolate all data points to obtain the pressure characteristic curve after iteration.

5. The method for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control according to claim 4, characterized in that, Based on the distribution of fitness evaluation indicators, the final guiding parameter combinations of the indicators in the current population are identified. According to the preset gradient optimization rules and parameter perturbation strategy, parameter orientation shifting and crossover recombination are performed on the remaining candidate curve parameter combinations in the population to complete the iterative update of the search direction for each candidate curve parameter combination, including: The fitness evaluation index of all candidate curve parameter combinations in the current population is numerically sorted and extreme value screened, and the candidate curve parameter combination with the smallest fitness evaluation index is extracted as the guiding parameter combination. Calculate the position vector difference between the other non-guided candidate curve parameter combinations and the guided parameter combination in the multidimensional parameter space, and obtain the convergence gradient vector pointing to the guided parameter combination by combining the preset gradient optimization rules, and determine the basic convergence direction of each candidate curve parameter combination; The preset parameter perturbation strategy is called to obtain a random perturbation factor that conforms to the preset distribution law. The random perturbation factor and the convergent gradient vector are vector superimposed to obtain a comprehensive search direction vector that has both global exploration and local fine-tuning capabilities. The parameter domain of each candidate curve parameter combination in the population is updated by directional offset along the comprehensive search direction vector, and the feature dimension cross-recombination operation is performed between the offset parameter nodes to obtain the updated candidate parameter combination for the next round of iteration calculation, thus completing the iterative update of the search direction of each candidate curve parameter combination.

6. The method for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control according to claim 5, characterized in that, Step 4: Extract a target inlet pressure value from the pressure characteristic curve within the steady-state operating range; lock the inlet pressure of the pressure reducing valve under test as the target inlet pressure value, including: After analyzing the pressure characteristic curve of the supply pressure effect after iteration, calculate the first-order rate of change of the outlet pressure to the inlet pressure in each continuous segment of the curve, screen out the continuous intervals where the absolute value of the first-order rate of change is continuously lower than the preset steady-state threshold, and define the continuous intervals where the absolute value is continuously lower than the preset steady-state threshold as the steady-state working region. Within the steady-state operating range, the corresponding horizontal axis inlet pressure data is extracted according to the preset outlet pressure condition optimization principle, and the horizontal axis inlet pressure data is determined as the target inlet pressure value. The target inlet pressure value is used as the new pressure setting benchmark and sent to the inlet pressure closed-loop control loop located upstream of the pressure reducing valve under test, terminating the previous linear pressure reduction control logic and obtaining a pressure constant maintenance command. Based on the constant pressure maintenance command, the upstream high-precision fast-response regulating valve performs fine-tuning compensation action to offset the fluctuation of pipeline gas source in real time, so that the inlet pressure of the pressure reducing valve under test dynamically converges and stabilizes at the target inlet pressure value, thereby completing the inlet pressure locking of the pressure reducing valve under test and obtaining the constant inlet pressure reference condition.

7. The method for continuous processing of the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control according to claim 6, characterized in that, Step 5: Under the condition of keeping the inlet pressure constant, gradually increase the outlet flow rate through the pressure reducing valve under test from zero to the rated flow rate in a preset continuous variation mode, and simultaneously collect the setpoint values ​​of each outlet flow rate and the corresponding stable value of the outlet pressure during the flow rate change process to obtain the flow characteristic curve, including: Based on the constant inlet pressure reference condition, the flow closed-loop control loop located downstream of the pressure reducing valve under test is activated to obtain a continuous ramp flow ramp control command sequence with zero flow as the starting threshold and rated flow as the ending threshold. According to the flow ramp control command sequence, the downstream precision regulating valve is driven to perform a gradual increase in opening, and the flow resistance of the downstream pipeline is continuously reduced to achieve a gradual and smooth increase in the outlet flow of the pressure reducing valve under test from zero to the rated flow. During the dynamic full cycle of gradually and smoothly increasing outlet flow, the outlet flow setpoint of the current flow closed-loop control loop is captured in real time through flow setpoint capture command and pressure steady-state monitoring command, and the dynamic feedback data of the downstream outlet pressure sensing node is continuously tracked. When the fluctuation amplitude of the outlet pressure dynamic feedback data converges to the steady-state pressure tolerance range within the preset time window, the steady-state locking signal is triggered. The average value of the outlet pressure dynamic feedback data within the preset time window is recorded as the corresponding stable outlet pressure value. All outlet flow setpoints and the stable outlet pressure values ​​locked in steady state are mapped in two dimensions and continuously smoothed in order of increasing flow rate to construct the flow characteristic curve.

8. A continuous processing system for the dynamic characteristic curve of a pressure reducing valve based on dual closed-loop control, wherein the system implements the method as described in any one of claims 1 to 7, characterized in that, include: The test module is used to install the valve in the test circuit and set the rated inlet pressure, rated outlet pressure and rated flow required for the test; Based on the rated flow rate, the outlet flow rate of the pressure reducing valve under test is continuously adjusted and locked to the rated flow rate, and the outlet flow rate of the pressure reducing valve under test is kept constant. The acquisition module is used to continuously and smoothly reduce the inlet pressure of the pressure reducing valve under test from the rated inlet pressure at a preset linear rate while keeping the outlet flow constant. During the pressure reduction process, it synchronously acquires the instantaneous values ​​of each inlet pressure and the corresponding instantaneous values ​​of the outlet pressure at a preset high-frequency sampling rate to obtain the initial pressure characteristic point set. The calculation module is used to reconstruct continuous curves from the initial pressure characteristic point set to obtain an initial population containing multiple candidate curve parameter combinations. The weighted sum of squared geometric distances between each candidate curve and each discrete point in the initial pressure characteristic point set is used as the fitness evaluation index. The search direction of each candidate curve parameter combination in the population is iteratively updated to obtain the iterated pressure characteristic curve. The extraction module is used to extract a target inlet pressure value from the pressure characteristic curve within the steady-state operating range; and to lock the inlet pressure of the pressure reducing valve under test as the target inlet pressure value. The processing module is used to gradually increase the outlet flow rate of the pressure reducing valve under test from zero to the rated flow rate in a preset continuous variation mode while keeping the inlet pressure constant. During the flow rate change process, the module simultaneously collects the set value of each outlet flow rate and the corresponding stable value of the outlet pressure to obtain the flow characteristic curve.

9. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images