Stable dewatering method of steam-water separator

Abstract

The invention provides a stable dewatering method for a steam-water separator, which is used for controlling a dewatering valve which is arranged on a dewatering pipeline of the steam-water separator and can continuously adjust the opening degree, and comprises the following steps: establishing a working condition dewatering dynamic baseline model; monitoring operation parameters in real time through a sensor according to a preset sampling period; based on the real-time value of the operating parameter, switching between at least two of the following control modes to determine the final drain valve position of the drain valve: when d (Pload) / dt is less than or equal to epsilon 1 and d (Delta P) / dt is less than or equal to epsilon 2, switching to a steady-state optimization mode, and when d (Pload) / dt is less than or equal to epsilon 1 and d (Delta P) / dt is less than or equal to epsilon 2; when d (Pload) / dtgt; [epsilon] 1 or d ([delta] P) / dtlt; and when epsilon is 3, switching to a transient suppression mode, executing a reverse pre-control strategy, and reducing the opening degree of the drain valve to a preset safe opening degree or completely closing the drain valve within a first preset time T1. According to the invention, stable, efficient and safe drainage of the steam-water separator under all working conditions is realized.

Description

Technical Field

[0001] This invention relates to a control method for a steam-water separator, specifically a stable drainage method for a steam-water separator, belonging to the field of thermal system control technology. Background Technology

[0002] In the thermal cycle systems of nuclear or thermal power plants, the steam-water separator reheater (MSR) is a key device connecting the high-pressure and low-pressure cylinders. Its function is to separate the exhaust steam from the high-pressure cylinder into steam and water, and then reheat the dried steam before sending it to the low-pressure cylinder to perform work. The separated saturated water (i.e., condensate) needs to be discharged promptly and stably through the condensate drainage system. In existing technologies, condensate drainage control in steam-water separator reheaters mostly uses a single-loop PID controller based on liquid level deviation to regulate the condensate valve. However, the condensate drainage process in a steam-water separator is a complex process with large lag, nonlinearity, and time-varying characteristics. Fixed control parameters only perform adequately at a specific design load point. When the unit load changes, especially under transient conditions such as load shedding and rapid power increases or decreases, the control performance deteriorates sharply, often resulting in severe control lag and overshoot. Furthermore, during drastic changes in unit load, the upstream and downstream pressures of the condensate drainage system experience instantaneous and drastic changes, which may lead to loss of condensate drainage driving force or even negative pressure, causing backflow of steam from the low-pressure cylinder through the condensate drain manifold or steam re-entrainment at the condensate drain outlet. Traditional control methods rely solely on the lag parameter of liquid level, which cannot anticipate or proactively avoid such risks, posing a threat to the safe and stable operation of the unit. Summary of the Invention

[0003] Based on the above background, the purpose of this invention is to provide a stable drainage method for a steam-water separator, which solves the technical problems of lag in drainage control, poor stability, and inability to effectively deal with transient risks in the prior art, and achieves stable, efficient and safe drainage of the steam-water separator under all operating conditions.

[0004] To achieve the above-mentioned objectives, the present invention provides the following technical solution:

[0005] A method for stabilizing the drainage of a steam-water separator, used to control a drainage valve installed on the drainage pipeline of the steam-water separator and capable of continuously adjusting its opening, the method comprising the following steps:

[0006] Establish a dynamic baseline model for drainage under operating conditions V b = f(P load , P in , P out The dynamic baseline model for condensate drainage under operating conditions collects the corresponding unit power P at multiple stable load points of the steam-water separator. load , hydrophobic inlet pressure P in Main drain manifold pressure P outAnd the reference drain valve position V used to maintain liquid level stability b The data is processed using multinomial fitting or neural network training algorithms to obtain V. b With P load P in P out The pre-defined functional relationship between them;

[0007] Operating parameters are monitored in real time by sensors at a preset sampling period, and the operating parameters include at least the current liquid level L. c , hydrophobic inlet pressure P in Main drain manifold pressure P out and the unit's electrical power P load The hydrophobic pressure gradient ΔP = P was calculated. in -P out ; and calculate the rate of change of the hydrophobic pressure gradient d(ΔP) / dt and the rate of change of the unit's electrical power d(P). load ) / dt;

[0008] Based on the real-time values ​​of the operating parameters, switching is performed between at least two control modes to determine the final drain valve position V of the drain valve. f :

[0009] When |d(P load When |d(ΔP) / dt|≤ε1 and |d(ΔP) / dt|≤ε2, switch to steady-state optimization mode, and the final condensate drain valve position V f Through formula V f =V ff + V pid Calculated; where ε1 and ε2 are preset stable operating condition thresholds, V ff V is the feedforward regulating valve position calculated using the aforementioned dynamic baseline model for drainage under operating conditions. pid Based on the current liquid level L c With liquid level set value L s The feedback regulating valve position is obtained by deviation calculation;

[0010] When |d(P load When d(ΔP) / dt|>ε1 or d(ΔP) / dt<ε3, switch to transient suppression mode and execute reverse pre-control strategy. The reverse pre-control strategy includes: reducing the opening degree of the drain valve to a preset safe opening degree or completely closing it within a first preset time T1; where ε3 is a preset re-entrainment risk threshold.

[0011] Preferably, the feedback regulating valve position V pid The gain scheduling PID algorithm is calculated and obtained through a gain scheduling PID algorithm, which specifically includes:

[0012] Establish a PID parameter scheduling table, wherein the PID parameter scheduling table will allocate the unit's electrical power P load Multiple different operating ranges and multiple sets of PID parameter groups (K) pre-optimized for each range p , K i , K d Establish corresponding relationships;

[0013] According to the real-time monitored unit power P load The PID parameter group (K) corresponding to the current operating condition is determined by querying the PID parameter scheduling table. p , K i , K d );

[0014] Based on the determined PID parameter set and the current liquid level L c With liquid level set value L s The deviation e(t) is calculated using the following formula to determine the feedback control valve position V. pid :

[0015]

[0016] Among them, K p K is the proportionality coefficient. i K is the integral coefficient. d is the differential coefficient.

[0017] By dynamically adjusting the PID parameters according to the load, the feedback control section can always operate under optimal or near-optimal parameter settings, overcoming the shortcomings of traditional fixed-parameter PID controllers in adapting to nonlinear and time-varying objects.

[0018] Preferably, the rate of change of the hydrophobic pressure gradient d(ΔP) / dt and the rate of change of the unit's electrical power d(P) are calculated. load Before ) / dt, it also includes:

[0019] The hydrophobic pressure gradient ΔP and the unit's electrical power P load Time series data are preprocessed using low-pass filters or moving average algorithms.

[0020] Preferably, the reverse pre-control strategy further includes:

[0021] According to the |d(P load The amplitude of d(ΔP) / dt| or d(ΔP) / dt is dynamically adjusted to the first preset time T1 and / or the preset safety opening, wherein an amplitude with a larger rate of change corresponds to a longer first preset time T1 and / or a smaller preset safety opening.

[0022] The purpose of this setup is to ensure that the intensity of the reverse pre-control is matched to the severity of the disturbances the system experiences. Stronger suppression measures are taken when facing severe disturbances, while milder measures are adopted when facing smaller disturbances.

[0023] As a preferred method, dynamic adjustment based on amplitude is achieved through the following relationship:

[0024]

[0025]

[0026] Among them, T 1b The base valve closing time, KD is the preset safety opening degree, KD b The basic safety opening is set, and k1 and k2 are preset positive gain coefficients.

[0027] Preferably, after the transient suppression mode executes the reverse pre-control strategy, the method further includes:

[0028] After the first preset time T1 ends, the hydrophobic pressure difference gradient ΔP is continuously monitored. When ΔP recovers to a level greater than the preset stable driving pressure threshold ε4, the control mode is switched from the transient suppression mode to the steady-state optimization mode.

[0029] Preferably, when switching the control mode from the transient suppression mode to the steady-state optimization mode, the final hydrophobic valve position V is adjusted within a preset transition time after the switch. f The rate of change of the output is limited.

[0030] Preferably, the final drain valve position V of the drain valve is determined. f Also includes:

[0031] When the current liquid level L is monitored c At the preset alarm time T a The internal liquid level remains consistently higher than the high alarm level L. H When the condensate pressure gradient ΔP is continuously less than the blockage diagnosis pressure threshold ε5, the system switches to pulse unblocking mode and executes a pulse unblocking strategy. The pulse unblocking strategy includes performing the following actions cyclically for a preset number of times N: completely closing the condensate valve and maintaining it for a second preset time T2; opening the condensate valve to the fully open position and maintaining it for a third preset time T3.

[0032] Preferably, the working condition hydrophobic dynamic baseline model is V. b =V p + V c , where V pThe physical model is calculated based on Bernoulli's equation and fluid dynamics formulas, using real-time monitored hydrophobic pressure gradient ΔP and a preset hydrophobic pipeline flow capacity coefficient. c The system is constructed based on a neural network model, with the unit's electrical power P as the key factor. load The compensation model is an input and a valve position compensation value as the output.

[0033] Compared with the prior art, the present invention has the following advantages:

[0034] This invention provides a stable drainage method for a steam-water separator. By combining feedforward control and feedback control, the feedforward component pre-compensates most of the disturbances caused by load changes, while the feedback component performs fine-tuning, thereby significantly improving the stability and response speed of the liquid level control. Furthermore, by introducing a predictive mechanism based on the rate of change of system parameters and a reverse pre-control strategy, this invention can identify and avoid the risk of steam re-entrainment under transient operating conditions. By switching control modes and parameters under different operating conditions, this invention can adapt to the nonlinear and time-varying characteristics of the steam-water separator drainage process, thus maintaining good control performance throughout the entire operating range from low load to full load and from steady state to transient state. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0036] Figure 1 This is a schematic flowchart of a stable hydrophobic method for a steam-water separator according to the present invention. Detailed Implementation

[0037] The technical solution of the present invention will be further described in detail below through specific embodiments and in conjunction with the accompanying drawings. It should be understood that the implementation of the present invention is not limited to the following embodiments, and any modifications and / or alterations made to the present invention will fall within the protection scope of the present invention.

[0038] In this invention, unless otherwise specified, all parts and percentages are by weight, and the equipment and raw materials used are commercially available or commonly used in the art. Unless otherwise specified, the methods in the following embodiments are conventional methods in the art. Unless otherwise specified, the components or equipment in the following embodiments are general standard parts or components known to those skilled in the art, and their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0039] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings. In this detailed description, numerous specific details are set forth to facilitate explanation and provide a thorough understanding of the embodiments of the present invention. However, one or more embodiments may be practiced by those skilled in the art without these specific details.

[0040] like Figure 1 As shown, an embodiment of the present invention discloses a stable drainage method for a steam-water separator, used to control a drainage valve installed on the drainage pipeline of the steam-water separator and capable of continuously adjusting its opening. The detailed steps of the method are as follows.

[0041] Step 1: Establish a dynamic baseline model for drainage under operating conditions.

[0042] This step is completed during the unit commissioning phase or the startup process after a planned shutdown. Its purpose is to establish a reference drain valve position V that can predict the required position under different operating conditions. b The model, specifically, employs a modeling method combining physical and data models. The expression for the dynamic baseline model of the hydrophobic conditions is as follows:

[0043] V b =V p + V c

[0044] Among them, V p It is the physical model part based on fluid dynamics, V c This is the compensation model part based on neural networks.

[0045] Physical Model V p The valve flow rate is calculated based on a simplified formula derived from Bernoulli's equation:

[0046]

[0047] Among them, C v V represents the flow capacity coefficient of the condensate pipes and valves, a constant calibrated experimentally; ΔP is the real-time monitored condensate pressure gradient, and ρ is the density of the condensate. p Its function is to provide a relatively rough prediction of the valve position based on physical laws.

[0048] The hydrophobic density ρ changes with the temperature and pressure inside the steam-water separator, which in turn affects the unit's electrical power P. load Relatedly, directly measuring the hydrophobic density ρ is quite difficult. Therefore, a compensation model V is introduced. c V c It is a feedforward neural network model, whose input is the unit's electrical power P. load The output is a response to V. pThe compensation value. During the offline modeling phase, the hydrophobic inlet pressure P at different stable load points was collected, with each 5% load representing a sampling point from 20% to 100% load. in Main drain manifold pressure P out and the unit's electrical power P load And the actual drain valve position V that can maintain a stable liquid level at this time. a Then, the neural network model is trained so that its output satisfies:

[0049]

[0050] This enables the neural network model to learn the unit's electrical power P. load The influence of hydrophobic density ρ and other unmodeled dynamics on the physical model V p Compensation is performed to obtain the dynamic baseline model V for hydrophobicity under operating conditions. b .

[0051] Step 2: Online Monitoring and Data Preprocessing

[0052] During normal unit operation, the controller collects operating parameters in real time through various sensors at a preset sampling period, including:

[0053] The current liquid level L is obtained through the differential pressure level gauge. c ;

[0054] The condensate inlet pressure P is obtained by a pressure transmitter installed before the condensate inlet. in ;

[0055] The pressure P of the main drain manifold is obtained by a pressure transmitter installed on the main drain manifold. out ;

[0056] Obtain the unit's electrical power P from the power plant's DCS system. load .

[0057] The controller first calculates the hydrophobic pressure gradient ΔP = P in -P out .

[0058] Then, to avoid the interference of measurement noise on subsequent differential calculations, a first-order low-pass filter is used to filter ΔP and P. load The time series data is processed. The discrete form of the filtering algorithm is:

[0059]

[0060] Where x(k) is the original sampled value at the current time, y(k) is the filtered output value at the current time, y(k-1) is the filtered output value at the previous time, and α is the filtering coefficient (0 < α < 1). The smaller α is, the smoother the filtering effect, but the slower the response. Through this step, the smoothed ΔP and P are obtained. load .

[0061] Finally, the controller obtains the rate of change through differential calculation based on the smoothed data:

[0062]

[0063]

[0064] Step 3: Multimodal switching and valve position calculation

[0065] Based on the key parameters obtained in step two, the controller makes real-time logical judgments and switches between different control modes.

[0066] Mode A, Steady-state optimization mode

[0067] The switching condition is that when the unit is running smoothly, i.e., |d(P load The system enters this mode when |d(ΔP) / dt|≤ε1 and |d(ΔP) / dt|≤ε2. ε1 and ε2 are stable operating condition thresholds set based on historical data and expert experience.

[0068] The control strategy employs a combined feedforward and feedback control, ultimately controlling the drain valve position V. f = V ff + V pid .

[0069] Feedforward regulating valve position V ff The working condition hydrophobic dynamic baseline model established in step one is directly called, i.e., V. ff = V b .

[0070] Feedback regulating valve position V pid A gain-based PID algorithm is used for fine-tuning. A PID parameter scheduling table is pre-stored within the controller, which establishes the unit's electrical power P. load Different intervals and the corresponding optimal PID parameter set (K) p , K i ,K d The mapping relationship between P and P. The controller determines the P based on real-time P. load Query this table to obtain the value (K) under the current operating condition. p , K i , K d Then, based on the determined PID parameter set and the current liquid level L... cWith liquid level set value L s The deviation e(t) is calculated using the following formula to determine the feedback control valve position V. pid :

[0071] ;

[0072] Among them, K p K is the proportionality coefficient. i K is the integral coefficient. d is the differential coefficient.

[0073] Mode B, transient suppression mode

[0074] The switching condition is when the unit load changes drastically (|d(P)). load When d(ΔP) / dt>ε1, or when there is a risk of rapid loss of hydrophobic driving force (d(ΔP) / dt<ε3), the system immediately switches to this mode. ε3 is the re-entrainment risk threshold.

[0075] The control strategy is to implement a reverse pre-control strategy. The core of this strategy is to actively reduce or close the valve the moment the risk of suction is detected, and use the upstream water to form a short-term high-pressure liquid column at the drain outlet, using its static pressure to resist the suction force downstream.

[0076] The valve closing strength is dynamically adaptive, according to |d(P) load The amplitude of d(ΔP) / dt| or d(ΔP) / dt dynamically adjusts the valve closing time, i.e., the first preset time T1 and / or the preset safety opening. A larger rate of change in amplitude corresponds to a longer first preset time T1 and / or a smaller preset safety opening. This means that the greater the disturbance, the longer the valve closing time, and the more tightly the valve is closed. Dynamic adjustment of the amplitude is achieved through the following relationship:

[0077]

[0078]

[0079] Among them, T 1b The base valve closing time, KD is the preset safety opening degree, KD b The basic safety opening is set, and k1 and k2 are preset positive gain coefficients.

[0080] After time T1 ends, the controller continuously monitors ΔP. ​​Once ΔP rises above the safe and stable driving pressure threshold ε4, it indicates that the condensate drainage conditions have been restored. At this point, the controller prepares to switch back to mode A. To avoid control shock, during a preset transition time after switching back to mode A, the final condensate drainage valve position V is adjusted. f A rate of change limit is applied to the output to ensure smooth valve operation and avoid secondary disturbances.

[0081] Mode C, Pulse unblocking mode

[0082] The switching condition is that when the controller detects the current liquid level L within the preset alarm time... c The liquid level remains above the high alarm level L. H However, the hydrophobic pressure gradient ΔP remained below the blockage diagnosis pressure threshold ε5, and the controller determined that a physical blockage had occurred.

[0083] The control strategy involves executing N pulse impacts. In each cycle, the controller first completely closes the drain valve, maintaining it for a second preset time T2 to accumulate pressure. Then, it opens the valve to 100% full open and maintains it for a third preset time T3, utilizing the water hammer effect of the high-pressure water flow to impact the blockage. After completing one cycle, the controller rechecks the blockage diagnosis conditions. If the blockage has been cleared, the pulse strategy is terminated early, and the controller switches back to mode A. If the blockage remains unresolved after N cycles, the controller issues an alarm.

[0084] Under steady-state conditions, this method achieves extremely high level control accuracy with minimal fluctuations. Under transient conditions, it can proactively and rapidly implement optimal defensive measures, effectively preventing catastrophic accidents such as steam re-entrainment and ensuring unit safety.

[0085] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A method for stabilizing drainage in a steam-water separator, used to control a drainage valve installed on the drainage pipeline of the steam-water separator and capable of continuously adjusting its opening, characterized in that: The method includes the following steps: Establish a dynamic baseline model for drainage under operating conditions V b = f(P load , P in , P out The dynamic baseline model for condensate drainage under operating conditions collects the corresponding unit power P at multiple stable load points of the steam-water separator. load , hydrophobic inlet pressure P in Main drain manifold pressure P out And the reference drain valve position V used to maintain liquid level stability b The data is processed using multinomial fitting or neural network training algorithms to obtain V. b With P load P in P out The pre-defined functional relationship between them; Operating parameters are monitored in real time by sensors at a preset sampling period, and the operating parameters include at least the current liquid level L. c , hydrophobic inlet pressure P in Main drain manifold pressure P out and the unit's electrical power P load The hydrophobic pressure gradient ΔP = P was calculated. in -P out ; and calculate the rate of change of the hydrophobic pressure gradient d(ΔP) / dt and the rate of change of the unit's electrical power d(P). load ) / dt; Based on the real-time values ​​of the operating parameters, switching is performed between at least two control modes to determine the final drain valve position V of the drain valve. f : When |d(P load When |d(ΔP) / dt|≤ε1 and |d(ΔP) / dt|≤ε2, switch to steady-state optimization mode, and the final condensate drain valve position V f Through formula V f =V ff + V pid Calculated; where ε1 and ε2 are preset stable operating condition thresholds, V ff V is the feedforward regulating valve position calculated using the aforementioned dynamic baseline model for drainage under operating conditions. pid Based on the current liquid level L c With liquid level set value L s The feedback regulating valve position is obtained by deviation calculation; When |d(P load When d(ΔP) / dt|>ε1 or d(ΔP) / dt<ε3, switch to transient suppression mode and execute reverse pre-control strategy. The reverse pre-control strategy includes: reducing the opening degree of the drain valve to a preset safe opening degree or completely closing it within a first preset time T1; where ε3 is a preset re-entrainment risk threshold.

2. The method for stabilizing hydrophobicity in a steam-water separator according to claim 1, characterized in that: The feedback regulating valve position V pid The gain scheduling PID algorithm is calculated and obtained through a gain scheduling PID algorithm, which specifically includes: Establish a PID parameter scheduling table, wherein the PID parameter scheduling table will allocate the unit's electrical power P load Multiple different operating ranges and multiple sets of PID parameter groups (K) pre-optimized for each range p , K i , K d Establish corresponding relationships; According to the real-time monitored unit power P load The PID parameter group (K) corresponding to the current operating condition is determined by querying the PID parameter scheduling table. p , K i , K d ); Based on the determined PID parameter set and the current liquid level L c With liquid level set value L s The deviation e(t) is calculated using the following formula to determine the feedback control valve position V. pid : ； Among them, K p K is the proportionality coefficient. i K is the integral coefficient. d is the differential coefficient.

3. The method for stabilizing hydrophobicity in a steam-water separator according to claim 1, characterized in that: The rate of change of hydrophobic pressure gradient d(ΔP) / dt and the rate of change of unit power d(P) were calculated. load Before ) / dt, it also includes: The hydrophobic pressure gradient ΔP and the unit's electrical power P load Time series data are preprocessed using low-pass filters or moving average algorithms.

4. The method for stabilizing hydrophobicity in a steam-water separator according to claim 1, characterized in that: The reverse pre-control strategy also includes: According to the |d(P load The amplitude of d(ΔP) / dt| or d(ΔP) / dt is dynamically adjusted to the first preset time T1 and / or the preset safety opening, wherein an amplitude with a larger rate of change corresponds to a longer first preset time T1 and / or a smaller preset safety opening.

5. The method for stabilizing hydrophobicity in a steam-water separator according to claim 4, characterized in that: Dynamic adjustment based on amplitude is achieved through the following formula: Among them, T 1b The base valve closing time, KD is the preset safety opening degree, KD b The basic safety opening is set, and k1 and k2 are preset positive gain coefficients.

6. The method for stabilizing hydrophobicity in a steam-water separator according to claim 1, characterized in that: After the transient suppression mode executes the reverse pre-control strategy, the following is also included: After the first preset time T1 ends, the hydrophobic pressure difference gradient ΔP is continuously monitored. When ΔP recovers to a level greater than the preset stable driving pressure threshold ε4, the control mode is switched from the transient suppression mode to the steady-state optimization mode.

7. The method for stabilizing hydrophobicity in a steam-water separator according to claim 6, characterized in that: When switching the control mode from the transient suppression mode to the steady-state optimization mode, the final hydrophobic valve position V is controlled within a preset transition time after the switch. f The rate of change of the output is limited.

8. The method for stabilizing hydrophobicity in a steam-water separator according to claim 1, characterized in that: Determine the final drain valve position V of the drain valve. f Also includes: When the current liquid level L is monitored c At the preset alarm time T a The internal liquid level remains consistently higher than the high alarm level L. H When the condensate pressure gradient ΔP is continuously less than the blockage diagnosis pressure threshold ε5, the system switches to pulse unblocking mode and executes a pulse unblocking strategy. The pulse unblocking strategy includes performing the following actions cyclically for a preset number of times N: completely closing the condensate valve and maintaining it for a second preset time T2; opening the condensate valve to the fully open position and maintaining it for a third preset time T3.

9. The method for stabilizing hydrophobicity in a steam-water separator according to claim 1, characterized in that: The dynamic baseline model for hydrophobicity under operating conditions is V. b =V p + V c , where V p The physical model is calculated based on Bernoulli's equation and fluid dynamics formulas, using real-time monitored hydrophobic pressure gradient ΔP and a preset hydrophobic pipeline flow capacity coefficient. c The system is constructed based on a neural network model, with the unit's electrical power P as the key factor. load The compensation model is an input and a valve position compensation value as the output.

Images