Active double-cavity cylindrical piston gas flow standard device and test method

Abstract

The invention relates to an active double-cavity cylindrical piston gas flow standard device and a test method. The device comprises a fixed seat, a balance cylinder body fixedly installed on one side of the fixed seat, a metering cylinder body installed on the other side of the fixed seat, a piston metering rod arranged in the balance cylinder body and the metering cylinder body, and a driving assembly arranged on the piston metering rod. The driving assembly is located between the balance cylinder body and the metering cylinder body. According to the steps of the testing method, the double-cavity cylindrical piston structure with the balance cavity and the metering cavity symmetrically arranged is adopted, and the problem that thrust is unbalanced due to uneven stress of a traditional single-cavity piston structure under the high-pressure working condition is solved; according to the method, an intelligent PID control strategy integrating genetic algorithm optimization, LSTM state recognition and fuzzy control is adopted, and a closed-loop regulation mechanism with parameter self-tuning, state self-judgment and disturbance self-adaption is constructed; the three mechanisms cooperate with one another, and the self-adaptability, stability and intelligent adjusting capacity of the system under micro-flow and high-pressure gas flow control scenes are improved.

Description

Technical Field

[0001] This invention relates to an active dual-chamber cylindrical piston gas flow standard device and testing method, belonging to the field of gas flow standard devices. Background Technology

[0002] The piston-type gas flow standard device belongs to the volumetric method and is based on the dynamic measurement principle. The core components of this standard device are the piston body and the piston cylinder. The piston moves in a uniform linear motion to discharge air from the piston cylinder and into the flow meter under test, thereby realizing the test of the flow meter under test.

[0003] In the field of gas flow standard devices, active piston gas flow standard devices drive piston movement through a drive unit, thereby continuously generating a stable and adjustable gas flow rate during the smooth movement of the piston. They are used to calibrate flow meters and other flow instruments, and are a common gas flow measurement tool. Traditional active piston gas flow standard devices are divided into cylindrical pistons and disc pistons, but both are atmospheric pressure devices. The cylindrical piston determines the standard volume through the outer surface of the piston rod, while the disc piston determines the standard volume through the piston cylinder. By using a servo motor to drive the piston rod or piston disc to move, the effective volume in the metering chamber is changed, and the gas is discharged, thus achieving precise control of the gas flow rate.

[0004] However, when measuring flow meters at high pressures (greater than 0.4 MPa), the inconsistent gas pressure at both ends of the piston rod causes a significant load on the drive unit due to the gas reaction force. This leads to increased power demand, shortened drive unit lifespan, and the presence of the reaction force can also cause instability in gas flow output, especially during calibration of very low flow rates (around 1 ml / min). Traditional devices struggle to meet the stability and accuracy requirements for high-pressure and low-flow measurements, failing to meet the needs of modern high-precision flow metering.

[0005] Chinese utility model patent CN206038117U discloses a dual-chamber piston-type gas flow standard device, which includes a cylinder with a long cylindrical sealed cavity structure and a disc-shaped piston. A bracket is connected to the outside of the cylinder, and a stepper motor is fixed to the bracket. The piston is connected to a long, cylindrical, T-shaped piston rod that extends out of the cylinder. A lead screw is threaded onto the piston rod and connected to the stepper motor. This solution performs measurements under normal pressure, and both cavities are used as metering chambers, making it unable to stably measure high pressure loss under high pressure (0.1–0.6 MPa). Summary of the Invention

[0006] To overcome the shortcomings of the prior art, the technical problem to be solved by the present invention is to provide an active dual-chamber cylindrical piston gas flow standard device and testing method that provides stable airflow, easy pressure regulation, high accuracy, good stability, and is suitable for high pressure and small flow range.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] The present invention relates to an active dual-chamber cylindrical piston gas flow standard device, comprising a fixed base, a balance cylinder fixedly installed on one side of the fixed base, a metering cylinder fixedly installed on the other side of the fixed base, a piston metering rod passing through the balance cylinder and the metering cylinder, and a drive assembly mounted on the piston metering rod and driving the piston metering rod to perform reciprocating linear motion; the drive assembly is located between the balance cylinder and the metering cylinder; piston rod sealing assemblies are provided at the corresponding ends of the balance cylinder and the metering cylinder and between them and the piston metering rod; sealing caps are installed at the outer ends of the balance cylinder and the metering cylinder; and gas guide pipes connecting the two chambers are installed on the side walls of the balance cylinder and the metering cylinder.

[0009] The piston rod sealing assembly of the present invention includes a piston rod sealing assembly A disposed between the balance cylinder and the piston metering rod, and a piston rod sealing assembly B disposed between the metering cylinder and the piston metering rod. The piston rod sealing assembly B includes a sealing ring cap with its end embedded in the port of the metering cylinder, a first sealing sleeve nested in a slot at the inner port of the sealing ring cap, a second sealing sleeve nested in a slot at the other port of the first sealing sleeve, and a third sealing sleeve nested in a slot at the other port of the second sealing sleeve. Cap-shaped sealing rings are provided between the sealing ring cap and the metering cylinder, between the first sealing sleeve and the metering cylinder, and between the second sealing sleeve and the metering cylinder. A first piston rod guide ring is provided in the inner cavity of the sealing ring cap, and a second piston rod guide ring is provided in the inner cavity of the third sealing sleeve. A lip-shaped dustproof ring is provided on the inner wall of the sealing ring cap. A first O-ring and a second O-ring are sequentially provided between the sealing ring cap and the inner wall of the metering cylinder, and a third O-ring is provided between the second and third sealing sleeves.

[0010] The device of the present invention is provided with pressure tapping holes on both the first sealing sleeve and the second sealing sleeve, and a pressure tapping connector is provided on the metering cylinder body to communicate with the corresponding pressure tapping holes. The two pressure tapping connectors are connected to the external pressure gauge and differential pressure gauge respectively.

[0011] The drive assembly of the device of the present invention includes a servo motor, a reducer connected to the drive shaft of the servo motor, a ball screw connected to the reducer via a coupling, and a T-shaped connecting block; the ball screw is parallel to the piston measuring rod; the lower side of the T-shaped connecting block is connected to the ball screw via a precision nut, and the upper side of the T-shaped connecting block engages with the piston measuring rod via a guide hole therein; the servo motor drives the ball screw to rotate via the reducer, and the rotational power of the ball screw is transmitted to the piston measuring rod via the T-shaped connecting block, and the piston measuring rod moves precisely along a predetermined trajectory under the guidance of the T-shaped connecting block.

[0012] The device of the present invention has a grating ruler connecting block installed on the upper side of the T-shaped connecting block, and the upper side of the grating ruler connecting block is connected to the grating ruler; the piston measuring rod is parallel to the grating ruler.

[0013] The testing method for the active dual-chamber cylindrical piston gas flow standard device of the present invention adopts the following steps:

[0014] S1. Supply air to the flow controller under test and set the target flow value;

[0015] S2. Start the servo motor to reset the piston metering rod to the starting position;

[0016] S3. Calculate the target piston speed based on the target flow rate and piston cross-sectional area, and set it as the initial setpoint for the PID controller;

[0017] S4. Close the balance valve between the piston balance chamber and the metering chamber. The servo motor drives the ball screw to rotate through the reducer. The rotational power of the ball screw is transmitted to the piston metering rod through the T-shaped connecting block. Under the guidance of the T-shaped connecting block, the piston metering rod is pushed into the metering cylinder along a predetermined trajectory.

[0018] S5. Real-time measurement of piston metering rod displacement and time, calculation of actual gas flow rate and comparison with target gas flow rate to obtain flow error ΔQ, input to PID controller for closed-loop regulation;

[0019] S6. The control system monitors the pressure and temperature of the gas in the metering chamber and the error calculated from the measured flow rate in real time to determine whether the dual-chamber cylindrical piston gas flow standard device is in a stable or disturbed state.

[0020] S7. When the dual-chamber cylindrical piston gas flow standard device is in a disturbed state, the fuzzy controller is automatically activated to adjust the output of the PID controller.

[0021] S8. Once the pressure difference and flow rate error have stabilized, the dual-chamber cylindrical piston gas flow standard device begins the formal flow rate calibration process.

[0022] The initial setpoint of the PID controller in method S3 of this invention is specifically as follows:

[0023] The initial settings are determined using a PID parameter design method based on a genetic algorithm (GA). This method constructs a multi-objective fitness function based on piston metering rod parameters and flow control data from historical calibration processes, and sets the proportional coefficient K of the PID controller accordingly. p Integral coefficient K i Differential coefficient K d The genetic code is used to encode individuals, and global optimization of the parameter set is achieved through an iterative evolution process.

[0024] The optimization objectives include: minimizing the system's steady-state error, minimizing differential pressure fluctuations, and minimizing the settling time;

[0025] Genetic algorithms (GA) iteratively obtain PID parameter combinations through population initialization, fitness evaluation, selection, crossover, and mutation operations, and use them as the initial controller configuration for the control system.

[0026] The specific operation of closed-loop regulation in method S5 of the present invention is as follows:

[0027] S51. Compare the measured actual flow rate with the target flow rate, and automatically adjust the servo motor speed according to the error to minimize the error;

[0028] S52. Error Calculation: Real-time acquisition of the displacement and movement time of the piston metering rod, calculation of the current instantaneous flow rate Q1, and comparison with the set target flow rate Q2 to obtain the error: ΔQ=Q2-Q1; Unit: ml / min;

[0029] S53. The flow error ΔQ is used as the input to the PID controller and enters the adjustment algorithm. The PID controller adjusts according to the following formula.

[0030]

[0031] In the formula, K p K i K d The proportional, integral, and differential coefficients are obtained from the genetic algorithm optimization; u(t) is the controller output; ∫ΔQdt is the integral of the flow error ΔQ along the time axis; It is the first derivative of the flow error ΔQ with respect to time;

[0032] S54. The output u(t) is used to adjust the speed of the servo motor in real time, thereby changing the pushing speed of the piston metering rod driven by the ball screw.

[0033] S55. After adjustment, measure the displacement and time in real time, recalculate the flow rate, and form a continuous feedback loop until the flow rate error ΔQ approaches 0 or enters the allowable error range.

[0034] The specific steps in S6 of the method of the present invention for determining whether the gas flow standard device of the dual-chamber cylindrical piston is in a stable or disturbed state are as follows:

[0035] A state recognition model based on Long Short-Term Memory Neural Network (LSTM) is used to learn and classify continuously collected data to determine the current state of the system.

[0036] When the pressure difference ΔP < 5Pa, the continuous stabilization time is 60s, and the flow error ΔQ enters the ±1% target value range, the system is determined to be in a stable state and enters an effective metering cycle.

[0037] When the pressure difference ΔP and the flow error ΔQ are still in the process of convergence, the parameter fluctuations gradually slow down, the system is judged to be in a transition state, and the system continues to adjust but does not perform metering.

[0038] If a sudden change in pressure difference, a sharp deviation in error, or a large jump in temperature occurs, the system determines it to be in a disturbance state and immediately activates the fuzzy control intervention module to correct the output of the PID controller.

[0039] The fuzzy controller in method S7 of this invention has a dual-input, three-output structure. Its input variables include: the pressure difference ΔP between the balance chamber and the metering chamber; and the flow error ΔQ between the piston output flow rate and the set flow rate of the flow meter under test. These two variables are further divided into seven fuzzy linguistic variables: NB, NM, NS, ZO, PS, PM, and PB. The membership function type is a symmetric triangular function. The pressure difference ΔP and the flow error ΔQ are adapted to the input space of the fuzzy controller, and a linear normalization method is used to map the physical quantities to the range [-3, +3]. The typical variation range of the pressure difference ΔP is ±5 Pa, and the normalization function is: The baseline setting for the flow error ΔQ is ±1% of the target flow value, and the corresponding normalization function is:

[0040]

[0041] The output variables are the adjustments to three parameters of the current PID controller: the proportional coefficient K. p Integral coefficient K i and differential coefficient K d ;

[0042] The fuzzy control rules consist of 49 rules. The fuzzy inference uses the Mamdani model, and the centroid method is used for defuzzification. The final output correction value is combined with the original PID parameters to control the speed of the servo motor.

[0043] The positive effects of this invention are as follows: By adopting a dual-chamber cylindrical piston structure with a symmetrical arrangement of the balance chamber and the metering chamber, this invention solves the problem of thrust imbalance caused by uneven force distribution in traditional single-chamber piston structures under high-pressure conditions; the device, in conjunction with the first piston rod guide ring and the second piston rod guide ring, effectively improves the linear guidance accuracy and service life of the dynamic seal during the movement of the piston metering rod, ensuring stable operation of the piston within the range of 0.1MPa to 0.6MPa, and the stability of gas flow output is better than 0.02%.

[0044] The sealing assembly of this invention adopts a combination of "static sealing + dynamic sealing" to reduce the overall mass of the piston and avoid the influence of gravity on the sealing performance. The pressure-temperature ratio change rate of the sealing assembly is better than 0.01%.

[0045] This invention integrates genetic algorithm optimization, LSTM state recognition, and fuzzy control to construct an intelligent PID control strategy, which establishes a closed-loop regulation mechanism with parameter self-tuning, state self-judgment, and disturbance self-adaptation. The three mechanisms work together to improve the system's adaptability, stability, and intelligent regulation capabilities in scenarios involving small flow rates and high-pressure gas flow control. Attached Figure Description

[0046] Appendix Figure 1 This is a schematic diagram of the structure of the present invention;

[0047] Appendix Figure 2 This is a schematic diagram of the drive component structure of the present invention;

[0048] Appendix Figure 3 This is a cross-sectional view of the sealing assembly of the present invention. Detailed Implementation

[0049] As attached Figure 1 As shown, the active dual-chamber cylindrical piston gas flow standard device of the present invention includes a fixed base 8, a balance cylinder 1 fixedly installed on one side of the fixed base 8, a metering cylinder 2 fixedly installed on the other side of the fixed base 8, a piston metering rod 3 passing through the balance cylinder 1 and the metering cylinder 2, and a drive assembly installed on the piston metering rod 3 and driving the piston metering rod 3 to perform reciprocating linear motion; the drive assembly is located between the balance cylinder 1 and the metering cylinder 2; the balance cylinder 1 and the metering cylinder 2 are symmetrically arranged; the fixed base 8 is installed on the base plate 7; piston rod sealing assemblies are provided between the corresponding ends of the balance cylinder 1 and the metering cylinder 2 and the piston metering rod 3; sealing caps 4 are installed on the outer ends of the balance cylinder 1 and the metering cylinder 2; the inner cavity of the balance cylinder 1 is used for buffering and constant pressure; the inner cavity of the metering cylinder 2 is used to output a stable gas volume flow rate; the two chambers are connected by a balance gas guide pipe 9 to ensure that the gas pressure is stable and the flow rate is balanced during the movement of the piston metering rod 3.

[0050] As attached Figure 1 ,3 As shown, the piston rod sealing assembly of the present invention includes a piston rod sealing assembly A5 disposed between the balance cylinder 1 and the piston metering rod 3, and a piston rod sealing assembly B6 disposed between the metering cylinder 2 and the piston metering rod 3; the piston rod sealing assembly B6 includes a sealing ring cover 61 with its end embedded in the port of the metering cylinder 2, a first sealing sleeve 62 nested in the inner port groove of the sealing ring cover 61, a second sealing sleeve 63 nested in the other port groove of the first sealing sleeve 62, and a third sealing sleeve 64 nested in the other port groove of the second sealing sleeve 63; the sealing ring cover 65... A cap-shaped sealing ring 65 is provided between the first sealing sleeve 62 and the metering cylinder 2, between the first sealing sleeve 62 and the metering cylinder 2, and between the second sealing sleeve 63 and the metering cylinder 2. A first piston rod guide ring 66 is provided in the inner cavity of the sealing ring cover 61, and a second piston rod guide ring 67 is provided in the inner cavity of the third sealing sleeve 64. A lip-shaped dustproof ring 70 is provided on the inner wall of the sealing ring cover 61. A first O-ring 68 and a second O-ring 69 are sequentially provided between the sealing ring cover 61 and the inner wall of the metering cylinder 2, and a third O-ring 71 is provided between the second sealing sleeve 63 and the third sealing sleeve 64. The sealing ring cover 61, the first sealing sleeve 62, the second sealing sleeve 63, and the third sealing sleeve 64 are arranged sequentially and interlocked. The first piston rod guide ring 66 and the second piston rod guide ring 67 provide stable support and guidance, reduce the impact and wear of the piston metering rod 3 on the seals during movement, improve the durability and service life of the seals, and the multiple O-rings ensure the sealing performance between the sealing kit and the inner wall of the piston cylinder. The difference between piston rod sealing assembly A5 and piston rod sealing assembly B6 is that piston rod sealing assembly A5 does not have a pressure tapping hole.

[0051] Pressure tapping holes 72 are provided on both the first sealing sleeve 62 and the second sealing sleeve 63. Pressure tapping connectors 73 are provided on the metering cylinder 2 and communicate with the corresponding pressure tapping holes 72. The two pressure tapping connectors 73 are connected to external pressure gauges and differential pressure gauges to monitor the dynamic sealing performance of the piston metering rod 3.

[0052] As attached Figure 2 As shown, the drive assembly of the present invention includes a servo motor 10, a reducer 11 connected to the drive shaft of the servo motor 10, a ball screw 13 connected to the reducer 11 via a coupling 12, and a T-shaped connecting block 14; the ball screw 13 is parallel to the piston metering rod 3; the reducer 11 converts the high-speed rotational motion of the servo motor 10 into a low-speed, high-torque rotational motion, and transmits it to the ball screw 12 via the coupling.

[0053] The lower side of the T-shaped connecting block 14 is connected to the ball screw 13 via a precision nut, and the upper side of the T-shaped connecting block 14 engages with the piston measuring rod 3 via a guide hole. The servo motor 10 drives the ball screw 13 to rotate via a reducer 11. The rotational power of the ball screw 13 is transmitted to the piston measuring rod 3 through the T-shaped connecting block 14, and the piston measuring rod 3 moves precisely along a predetermined trajectory under the guidance of the T-shaped connecting block 14. A grating ruler connecting block 15 is installed on the upper side of the T-shaped connecting block 14, and the upper side of the grating ruler connecting block 15 is connected to a grating ruler 16. The piston measuring rod 3 is parallel to the grating ruler 16. The grating ruler 16 measures the displacement of the piston measuring rod 3 in real time. Temperature, pressure, and humidity sensors are used to monitor the internal environmental parameters of the cavity. The collected parameters are transmitted to the host computer via a communication interface. The host computer processes, converts, and displays the collected displacement, time, temperature, and pressure data, and participates in the calculation and feedback adjustment of the closed-loop control logic to form a complete metering calculation and control system. The control system is used to realize piston motion control, flow calculation, state conversion and error analysis. The control system is based on the PID regulation algorithm of machine learning, and interacts with the host computer, drive components and measurement modules (i.e. modules that collect corresponding parameters through temperature sensors, pressure sensors and crystal oscillators) for data interaction and feedback control.

[0054] Specifically: The control system calculates the cross-sectional area of ​​the piston based on its diameter. Combined with the horizontal displacement measured in real time by the grating ruler and the time measured by the crystal oscillator (integrated on the circuit board controlling the piston's movement), the volume of gas discharged by the piston per unit time can be obtained, i.e., the original flow rate value. At the same time, the control system collects temperature and pressure data at both the dual-chamber cylindrical piston gas flow standard device and the flow meter under test in real time. Based on the state equation, the operating flow rate measured at the piston is converted to the operating state of the flow meter under test, and its indication error is further calculated to achieve high-precision verification and calibration.

[0055] The control system supports the switching of multiple piston modules, including a 0.02L cylindrical piston with a flow rate range of 1–20 mL / min, a 0.2L cylindrical piston with a flow rate range of 10–200 mL / min, and a 3L cylindrical piston with a flow rate range of 0.15–3 L / min, enabling the verification or calibration of milliliter-level micro gas flow meters.

[0056] The machine learning-based PID control algorithm includes: the optimal PID parameter combination obtained by the genetic algorithm (GA); a state recognition model based on the LSTM long short-term memory neural network to identify the state of the system during the verification process; when the system is identified as being in a disturbance state, fuzzy control is used to further correct the PID parameters; and a three-stage control mechanism of parameter optimization, state recognition, and parameter correction.

[0057] A test method for a dual-chamber cylindrical piston gas flow standard device comprises the following steps:

[0058] S1. Supply air to the flow controller under test (the flow meter that can control the flow rate) and set the target flow rate value;

[0059] S2. Start the servo motor 10 to reset the piston metering rod 3 to the starting position;

[0060] S3. Calculate the target piston speed based on the target flow rate and piston cross-sectional area, and set it as the initial setpoint for the PID controller;

[0061] During the system initialization phase, to adapt to the dynamic control requirements of different piston specifications under complex flow conditions, a PID parameter design method based on genetic algorithm (GA) optimization is used for initial setting values. This PID parameter design method is based on the piston metering rod parameters and flow control data from historical calibration processes, constructing a multi-objective fitness function, and setting the proportional coefficient K of the PID controller... p Integral coefficient K i Differential coefficient K d The genetic algorithm encodes PID parameters into genetic individuals and achieves global optimization of the parameter set through an iterative evolution process. Iterative evolution involves evaluating, selecting, crossovering, and mutating the encoded PID parameter individuals according to the fitness function during the execution of the genetic algorithm, generating a new generation of parameter combinations, and then performing multiple rounds of iteration until the optimal solution that satisfies the target fitness function is obtained. The optimal PID parameters are then output. The optimization objectives include minimizing the steady-state error of the system, minimizing pressure fluctuation, and minimizing the settling time. The genetic algorithm (GA) continuously iterates to obtain the optimal PID parameter combination through initialization of the population, fitness evaluation, selection, crossover, and mutation, and uses it as the initial controller configuration of the control system.

[0062] S4. Close the balance valve between the piston balance chamber and the metering chamber. The servo motor 10 drives the ball screw 13 to rotate through the reducer 11. The rotational power of the ball screw 13 is transmitted to the piston metering rod 3 through the T-shaped connecting block 14. Under the guidance of the T-shaped connecting block 14, the piston metering rod 3 is pushed into the metering cylinder 2 along a predetermined trajectory.

[0063] S5. Real-time measurement of piston metering rod 3 displacement and time, calculation of actual gas flow rate, comparison with target gas flow rate to obtain flow error ΔQ, input into PID controller for closed-loop regulation;

[0064] The specific operation of closed-loop regulation is as follows:

[0065] S51. Compare the measured actual flow rate with the target flow rate, and automatically adjust the servo motor speed according to the error to minimize the error;

[0066] S52. Error Calculation: Real-time acquisition of the displacement and movement time of the piston metering rod, calculation of the current instantaneous flow rate Q1, and comparison with the set target flow rate Q2 to obtain the error: ΔQ=Q2-Q1; Unit: ml / min;

[0067] S53. The flow error ΔQ is used as the input to the PID controller and enters the adjustment algorithm. The PID controller adjusts according to the following formula.

[0068]

[0069] In the formula, K p K i K d The proportional, integral, and differential coefficients obtained by the genetic algorithm optimization; u(t) is the controller output; ∫ΔQdt is the integral of the flow error ΔQ along the time axis, that is, the error value from the start time to the current time is accumulated over time; It is the first derivative of the flow error ΔQ with respect to time, i.e., the rate of change of the flow error;

[0070] S54. The output u(t) is used to adjust the speed of the servo motor in real time, thereby changing the pushing speed of the piston metering rod driven by the ball screw.

[0071] S55. After adjustment, measure displacement and time in real time, recalculate flow rate, and form a continuous feedback loop until the flow rate error ΔQ approaches 0 or enters the allowable error range.

[0072] S6. The control system monitors the pressure and temperature of the gas in the metering chamber and the error calculated from the measured flow rate in real time to determine whether the dual-chamber cylindrical piston gas flow standard device is in a stable or disturbed state.

[0073] A state recognition model based on Long Short-Term Memory (LSTM) neural network is used to learn and classify continuously collected data to determine the current state of the system, as follows:

[0074] When the pressure difference ΔP < 5Pa, the continuous stabilization time is 60s, and the flow error ΔQ enters the ±1% target value range, the system is determined to be in a stable state and enters an effective metering cycle.

[0075] When the pressure difference and flow error are still in the process of convergence, the parameter fluctuations gradually slow down, the system is judged to be in a transition state, and the system continues to adjust but does not perform measurement.

[0076] If a sudden change in differential pressure, a sharp deviation in error, or a large jump in temperature occurs, the system determines it to be a disturbance state and immediately activates the fuzzy control intervention module to correct the output of the PID controller.

[0077] S7. When the dual-chamber cylindrical piston gas flow standard device is in a disturbed state, the fuzzy controller is automatically activated to adjust the output of the PID controller.

[0078] The fuzzy controller has a dual-input, three-output structure. Its input variables include: the pressure difference ΔP between the balance chamber and the metering chamber; and the flow error ΔQ between the piston output flow rate and the set flow rate of the flow meter under test. These two variables are further divided into seven fuzzy linguistic variables: NB (large negative), NM (medium negative), NS (small negative), ZO (zero), PS (small positive), PM (medium positive), and PB (large positive). The membership function type is a symmetric triangular function. The pressure difference ΔP and the flow error ΔQ are adapted to the input space of the fuzzy controller, and a linear normalization method is used to map the physical quantities to the range [-3, +3]. The typical variation range of the pressure difference ΔP is ±5 Pa, and the normalization function is: The baseline setting for the flow error ΔQ is ±1% of the target flow value, and the corresponding normalization function is:

[0079] The output variables are the corrections to three parameters of the current PID controller: ΔKp, ΔKi, and ΔKd.

[0080] The fuzzy control rules consist of 49 rules. The fuzzy inference uses the Mamdani model, and the centroid method is used for defuzzification. The final output correction value is combined with the original PID parameters to control the servo motor speed.

[0081] The rules table is as follows:

[0082] Rule number ΔP ΔQ ΔKp Δki ΔKd R1 NB NB PM PS ZO R2 NB NM PM PS ZO R3 NB NS PS ZO ZO R4 NB ZO PS ZO ZO R5 NB PS ZO ZO PS R6 NB PM ZO ZO PS R7 NB PB ZO NS PM R8 NM NB PM PS ZO R9 NM NM PS ZO ZO R10 NM NS PS ZO ZO R11 NM ZO ZO ZO PS R12 NM PS ZO ZO PS R13 NM PM ZO NS PM R14 NM PB ZO NS PM R15 NS NB PS ZO ZO R16 NS NM PS ZO ZO R17 NS NS ZO ZO PS R18 NS ZO ZO ZO PS R19 NS PS ZO NS PM R20 NS PM ZO NS PM R21 NS PB NS NS PM R22 ZO NB PS ZO ZO R23 ZO NM ZO ZO PS R24 ZO NS ZO ZO PS R25 ZO ZO ZO NS PM R26 ZO PS ZO NS PM R27 ZO PM NS NS PM R28 ZO PB NS NS PM R29 PS NB ZO ZO PS R30 PS NM ZO ZO PS R31 PS NS ZO NS PM R32 PS ZO ZO NS PM R33 PS PS NS NS PM R34 PS PM NS NS PM R35 PS PB NS ZO PB R36 PM NB ZO ZO PS R37 PM NM ZO NS PM R38 PM NS ZO NS PM R39 PM ZO NS NS PM R40 PM PS NS NS PM R41 PM PM NS ZO PB R42 PM PB NS ZO PB R43 PB NB ZO NS PM R44 PB NM ZO NS PM R45 PB NS NS NS PM R46 PB ZO NS NS PM R47 PB PS NS ZO PB R48 PB PM NS ZO PB R49 PB PB NS ZO PB

[0083] S8. Once the pressure difference and flow rate error have stabilized, the dual-chamber cylindrical piston gas flow standard device begins formal flow rate calibration.

[0084] This invention relates to an active dual-chamber cylindrical piston gas flow standard device with a flow range of (1-20) mL / min for a 0.02L cylindrical piston, (10-200) mL / min for a 0.2L cylindrical piston, and (0.15-3) L / min for a 3L cylindrical piston. The adjustable pressure range is (0.1-0.6) MPa. This active dual-chamber cylindrical piston gas flow standard device solves the problem of thrust imbalance caused by uneven force distribution under high-pressure conditions in traditional single-chamber piston structures by employing a dual-chamber cylindrical piston structure with a symmetrical arrangement of the balance chamber and the metering chamber. The device, in conjunction with the first and second piston rod guide rings, effectively improves the linear guidance accuracy and service life of the dynamic seals during the piston metering rod movement, ensuring stable piston operation within the range of 0.1 MPa to 0.6 MPa, with a gas flow output stability better than 0.02%.

[0085] The sealing assembly of this invention adopts a combination of "static sealing + dynamic sealing" to reduce the overall mass of the piston and avoid the influence of gravity on the sealing performance. The pressure-temperature ratio change rate of the sealing assembly is better than 0.01%.

[0086] This invention integrates genetic algorithm optimization, LSTM state recognition, and fuzzy control to construct an intelligent PID control strategy, which establishes a closed-loop regulation mechanism with parameter self-tuning, state self-judgment, and disturbance self-adaptation. The three mechanisms work together to improve the system's adaptability, stability, and intelligent regulation capabilities in scenarios involving small flow rates and high-pressure gas flow control.

Claims

1. An active dual-chamber cylindrical piston gas flow standard device, characterized in that, It includes a fixed base (8), a balance cylinder (1) fixedly installed on one side of the fixed base (8), a metering cylinder (2) fixedly installed on the other side of the fixed base (8), a piston metering rod (3) passing through the balance cylinder (1) and the metering cylinder (2), and a drive assembly installed on the piston metering rod (3) and driving the piston metering rod (3) to perform reciprocating linear motion; the drive assembly is located between the balance cylinder (1) and the metering cylinder (2); Piston rod sealing assemblies are provided between the corresponding ends of the balance cylinder (1) and the metering cylinder (2) and the piston metering rod (3), and sealing caps (4) are installed on the outer ends of the balance cylinder (1) and the metering cylinder (2); An air guide pipe (9) connecting the two chambers is installed on the side wall of the balance cylinder (1) and the metering cylinder (2).

2. The active dual-chamber cylindrical piston gas flow standard device according to claim 1, characterized in that, The piston rod sealing assembly includes a piston rod sealing assembly A (5) disposed between the balance cylinder (1) and the piston metering rod (3) and a piston rod sealing assembly B (6) disposed between the metering cylinder (2) and the piston metering rod (3); The piston rod sealing assembly B (6) includes a sealing ring cover (61) with its end embedded in the port of the metering cylinder (2), a first sealing sleeve (62) nested in the inner port groove of the sealing ring cover (61), a second sealing sleeve (63) nested in the other port groove of the first sealing sleeve (62), and a third sealing sleeve (64) nested in the other port groove of the second sealing sleeve (63); a cap-shaped sealing ring (65) is provided between the sealing ring cover (61) and the metering cylinder (2), between the first sealing sleeve (62) and the metering cylinder (2), and between the second sealing sleeve (63) and the metering cylinder (2); A first piston rod guide ring (66) is provided in the inner cavity of the sealing ring cover (61), and a second piston rod guide ring (67) is provided in the inner cavity of the third sealing sleeve (64); a lip-shaped dustproof ring (70) is provided on the inner wall of the sealing ring cover (61); A first O-ring (68) and a second O-ring (69) are sequentially arranged between the sealing ring cover (61) and the inner wall of the metering cylinder (2), and a third O-ring (71) is arranged between the second sealing sleeve (63) and the third sealing sleeve (64).

3. The active dual-chamber cylindrical piston gas flow standard device according to claim 2, characterized in that, Pressure tapping holes (72) are provided on both the first sealing sleeve (62) and the second sealing sleeve (63). Pressure tapping connectors (73) communicating with the corresponding pressure tapping holes (72) are provided on the metering cylinder body (2). The two pressure tapping connectors (73) are connected to the external pressure gauge and differential pressure gauge respectively.

4. The active dual-chamber cylindrical piston gas flow standard device according to claim 2, characterized in that, The drive assembly includes a servo motor (10), a reducer (11) connected to the drive shaft of the servo motor (10), a ball screw (13) connected to the reducer (11) via a coupling (12), and a T-shaped connecting block (14); the ball screw (13) is parallel to the piston metering rod (3); The lower side of the T-shaped connecting block (14) is connected to the ball screw (13) through a precision nut, and the upper side of the T-shaped connecting block (14) is engaged with the piston measuring rod (3) through the guide hole inside it; the servo motor (10) drives the ball screw (13) to rotate through the reducer (11), and the rotational power of the ball screw (13) is transmitted to the piston measuring rod (3) through the T-shaped connecting block (14), and the piston measuring rod (3) moves precisely along a predetermined trajectory under the guidance of the T-shaped connecting block (14).

5. The active dual-chamber cylindrical piston gas flow standard device according to claim 4, characterized in that, A grating ruler connecting block (15) is installed on the upper side of the T-shaped connecting block (14), and the upper side of the grating ruler connecting block (15) is connected to the grating ruler (16); The piston measuring rod (3) is parallel to the grating ruler (16).

6. A test method for an active dual-chamber cylindrical piston gas flow standard device according to any one of claims 1-5, characterized in that, The following steps are used: S1. Supply air to the flow controller under test and set the target flow value; S2. Start the servo motor (10) to reset the piston metering rod (3) to the starting position; S3. Calculate the target piston speed based on the target flow rate and piston cross-sectional area, and set it as the initial setpoint for the PID controller; S4. Close the balance valve between the piston balance chamber and the metering chamber. The servo motor (10) drives the ball screw (13) to rotate through the reducer (11). The rotational power of the ball screw (13) is transmitted to the piston metering rod (3) through the T-shaped connecting block (14). Under the guidance of the T-shaped connecting block (14), the piston metering rod (3) moves into the metering cylinder (2) along a predetermined trajectory. S5. Measure the displacement and time of the piston metering rod (3) in real time, calculate the actual gas flow rate and compare it with the target gas flow rate to obtain the flow error ΔQ, and input it into the PID controller for closed-loop regulation; S6. The control system monitors the pressure and temperature of the gas in the metering chamber and the error calculated from the measured flow rate in real time to determine whether the dual-chamber cylindrical piston gas flow standard device is in a stable or disturbed state. S7. When the dual-chamber cylindrical piston gas flow standard device is in a disturbed state, the fuzzy controller is automatically activated to adjust the output of the PID controller. S8. Once the pressure difference and flow rate error have stabilized, the dual-chamber cylindrical piston gas flow standard device begins the formal flow rate calibration process.

7. The test method for an active dual-chamber cylindrical piston gas flow standard device according to claim 6, characterized in that, The initial settings of the PID controller in S3 are as follows: The initial settings are determined using a PID parameter design method based on a genetic algorithm (GA). This method is based on the parameters of the piston metering rod (3) and the flow control data from the historical calibration process. A multi-objective fitness function is constructed, and the proportional coefficient K of the PID controller is set. p Integral coefficient K i Differential coefficient K d The genetic code is used to encode individuals, and global optimization of the parameter set is achieved through an iterative evolution process. The optimization objectives include: minimizing the system's steady-state error, minimizing differential pressure fluctuations, and minimizing the settling time; Genetic algorithms (GA) iteratively obtain PID parameter combinations through population initialization, fitness evaluation, selection, crossover, and mutation operations, and use them as the initial controller configuration for the control system.

8. The test method for an active dual-chamber cylindrical piston gas flow standard device according to claim 7, characterized in that, The specific operation of closed-loop regulation in S5 is as follows: S51. Compare the measured actual flow rate with the target flow rate, and automatically adjust the servo motor speed according to the error to minimize the error; S52. Error Calculation: Real-time acquisition of the displacement and movement time of the piston metering rod, calculation of the current instantaneous flow rate Q1, and comparison with the set target flow rate Q2 to obtain the error: ΔQ=Q2-Q1; Unit: ml / min; S53. The flow error ΔQ is used as the input to the PID controller and enters the adjustment algorithm. The PID controller adjusts according to the following formula. In the formula, K p K i K d The proportional, integral, and differential coefficients are obtained from the genetic algorithm optimization; u(t) is the controller output; ∫ΔQdt is the integral of the flow error ΔQ along the time axis; It is the first derivative of the flow error ΔQ with respect to time; S54. The output u(t) is used to adjust the speed of the servo motor in real time, thereby changing the pushing speed of the piston metering rod driven by the ball screw. S55. After adjustment, measure the displacement and time in real time, recalculate the flow rate, and form a continuous feedback loop until the flow rate error ΔQ approaches 0 or enters the allowable error range.

9. The test method for an active dual-chamber cylindrical piston gas flow standard device according to claim 6, characterized in that, The specific steps in S6 for determining whether the dual-chamber cylindrical piston gas flow standard device is in a stable or disturbed state are as follows: A state recognition model based on Long Short-Term Memory (LSTM) neural network is used to learn and classify continuously collected data to determine the current state of the system. When the pressure difference ΔP < 5Pa, the continuous stabilization time is 60s, and the flow error ΔQ enters the ±1% target value range, the system is determined to be in a stable state and enters an effective metering cycle. When the pressure difference ΔP and the flow error ΔQ are still in the process of convergence, the parameter fluctuations gradually slow down, the system is judged to be in a transition state, and the system maintains regulation but does not perform measurement. If a sudden change in pressure difference, a sharp deviation in error, or a large jump in temperature occurs, the system determines it to be in a disturbance state and immediately activates the fuzzy control intervention module to correct the output of the PID controller.

10. The testing method for an active dual-chamber cylindrical piston gas flow standard device according to claim 6, characterized in that... The fuzzy controller in S7 has a dual-input, three-output structure. Its input variables include: the pressure difference ΔP between the balance chamber and the metering chamber; and the flow error ΔQ between the piston output flow rate and the set flow rate of the flowmeter under test. These two variables are further divided into seven fuzzy linguistic variables: NB, NM, NS, ZO, PS, PM, and PB. The membership function type is a symmetric triangular function. The pressure difference ΔP and the flow error ΔQ are adapted to the input space of the fuzzy controller, and a linear normalization method is used to map the physical quantities to the range [-3, +3]. The typical variation range of the pressure difference ΔP is ±5 Pa, and the normalization function is: The baseline setting for the flow error ΔQ is ±1% of the target flow value, and the corresponding normalization function is: The output variables are the adjustments to three parameters of the current PID controller: the proportional coefficient K. p Integral coefficient K i and differential coefficient K d ; The fuzzy control rules consist of 49 rules. The fuzzy inference uses the Mamdani model, and the centroid method is used for defuzzification. The final output correction value is combined with the original PID parameters to control the speed of the servo motor.

Images