Liquid level prediction control method and system based on liquid supplement event causal time sequence identification

Abstract

The present application relates to the technical field of non-electric variable automatic control, and discloses a liquid level prediction control method and system based on liquid supplement event causal time sequence identification. A controller collects time sequence data of liquid supplement control instructions, pressure, flow and liquid level, sequentially identifies instruction change, pressure response, flow response and liquid level response time, and determines response delays in each stage; estimates liquid volume consumption rate according to an effective inflow interval-free section, and records the queue of in-transit liquid supplement amount that has passed through the flow detection position but has not yet formed a liquid level response according to the liquid level response delay and its version; predicts the future liquid level by combining the current liquid level, liquid volume consumption rate, queue in-transit liquid supplement amount and response delays, dynamically determines the advance execution time and target liquid supplement amount; updates or freezes parameters by comparing the predicted causal response chain with the actual causal response chain, and executes degraded liquid level control when the response chain is abnormal. The present application can reduce low liquid level, repeated liquid supplement and liquid level overshoot caused by liquid supplement lag.

Description

Technical Field

[0001] This invention relates to the field of non-electric variable automatic control technology, and in particular to a liquid level prediction control method and system based on causal timing identification of liquid replenishment events. Background Technology

[0002] Steam generators, electric steamers, heating water tanks, cleaning tanks, and liquid storage tanks continuously consume liquids during operation due to evaporation, discharge, leakage, or process usage. To maintain the liquid level in the liquid chamber within a preset operating range, electrically controlled liquid replenishment actuators such as liquid replenishment pumps, solenoid valves, and proportional valves are typically installed on the liquid replenishment channel connected to the liquid chamber. The controller then controls the opening, closing, or adjustment of the liquid replenishment output based on the liquid level detection results.

[0003] Some existing liquid level control methods use upper and lower liquid level thresholds for control. That is, liquid replenishment is initiated when the detected liquid level is lower than the lower liquid level threshold, and liquid replenishment is stopped when the detected liquid level reaches the upper liquid level threshold. This control method is relatively simple to implement, but it is prone to problems such as untimely liquid replenishment response or untimely liquid replenishment stoppage when the liquid replenishment channel is long, the liquid supply pressure is unstable, the response speed of the liquid replenishment actuator varies, or there is a lag in liquid level detection.

[0004] Specifically, after the controller issues a replenishment control command to start or increase the replenishment output, the replenishment process typically involves the sequential stages of: activation of the electronically controlled replenishment actuator, pressure build-up in the replenishment channel, effective liquid flow, and replenishment liquid entering the liquid chamber and causing a change in liquid level. Each of these stages has a different response delay, which may vary depending on the supply pressure, replenishment channel resistance, actuator performance, and equipment operating status. Some existing control methods typically treat this process as a single delayed process or use pre-set fixed delay parameters, making it difficult to accurately reflect the staged response characteristics of the actual replenishment process.

[0005] Furthermore, the flow detection device's detection of replenishment flow only indicates that the replenished liquid has passed the corresponding flow detection position. Due to factors such as the length of the replenishment channel, gas compression within the channel, the liquid delivery process, the inlet structure of the liquid storage chamber, and level detection filtering, some replenished liquid that has passed the flow detection position may not have yet entered the liquid storage chamber, or may not have formed a recognizable level response in the level detection result. During this period, if the controller continues to increase the replenishment output based solely on the currently detected level, it may repeatedly count the replenishment liquid that has not yet formed a level response in the replenishment demand, thereby causing an excessive replenishment volume and level overshoot.

[0006] The liquid consumption rate of liquid-consuming equipment may also vary with equipment load, operating temperature, steam output, or liquid discharge status. Control methods using fixed replenishment volume, fixed trigger level, or fixed response delay are insufficient to simultaneously adapt to changes in liquid consumption rate, fluctuations in supply pressure, increased resistance in the replenishment channel, and performance degradation of the electronically controlled replenishment actuators. This can lead to problems such as premature replenishment, delayed replenishment, insufficient replenishment, excessive replenishment, or frequent start-stop of the electronically controlled replenishment actuators.

[0007] When an anomaly occurs during the replenishment process, the liquid level failing to rise as expected may be caused by various factors, such as insufficient liquid supply, ineffective operation of the electronically controlled replenishment actuator, blockage of the replenishment channel, failure of replenishment liquid to enter the liquid chamber, liquid leakage, abnormal pressure detection, abnormal flow detection, or abnormal liquid level detection. Some existing control methods mainly determine the replenishment status based on whether the liquid level has reached a preset threshold, making it difficult to distinguish the replenishment stage where the anomaly occurs based on the sequential relationship between pressure response, flow response, and liquid level response. Therefore, it is not conducive to taking timely and targeted measures such as flow limiting, power reduction, or downgrading of liquid level control.

[0008] The aforementioned problems are particularly prominent in liquid-consuming equipment with heating loads, such as electric steamers and steam generators. Excessive liquid replenishment may cause the liquid level in the liquid chamber to be too high, steam output to fluctuate, and the operating temperature to drop rapidly; insufficient liquid replenishment or a delayed liquid replenishment response may cause the liquid level to drop below the safe operating range, increasing the risk of the heating element operating without water or dry burning.

[0009] Therefore, how to adaptively determine the response delay of each stage in the replenishment process based on the time-series response relationship between replenishment control commands, pressure information, actual replenishment flow rate and liquid level information, reasonably consider the replenishment liquid that has passed the flow detection position but has not yet formed a liquid level response, predict the future liquid level in combination with the liquid consumption status, and distinguish different abnormal stages in the replenishment process are still technical problems that need to be solved in the liquid level control of liquid consumption equipment. Summary of the Invention

[0010] In order to solve at least one of the above-mentioned technical problems in the prior art, the present invention provides a liquid level prediction and control method based on causal timing identification of replenishment events.

[0011] In order to solve at least one of the above-mentioned technical problems in the prior art, the present invention provides a liquid level prediction and control system based on causal timing identification of liquid replenishment events.

[0012] The first objective of this invention is achieved as follows:

[0013] A liquid level prediction and control method based on causal timing identification of replenishment events is applied to a liquid consumption device. The liquid consumption device includes a liquid storage chamber, a replenishment channel communicating with the liquid storage chamber, an electrically controlled replenishment actuator for adjusting the replenishment output of the replenishment channel, a pressure detection device for detecting the pressure on the supply or outlet side of the electrically controlled replenishment actuator, a flow detection device for detecting the actual replenishment flow rate through the replenishment channel, a liquid level detection device for detecting the liquid level in the liquid storage chamber, and a controller. The controller pre-stores the correspondence between the liquid level and the liquid volume in the liquid storage chamber. The method comprises:

[0014] S1. Collect time-series data of replenishment control command, pressure information, actual replenishment flow rate and liquid level information; when the replenishment control command switches from the closed state to the open state or the increase of the replenishment control command reaches the preset command change threshold, establish a replenishment event window, and identify the command change time, pressure response time, flow response time and liquid level response time in the replenishment event window;

[0015] S2. When the command change time, the pressure response time, the flow response time, and the liquid level response time occur sequentially, the pressure supply response delay, the flow response delay, and the liquid level response delay are determined based on the time difference between adjacent times; the liquid volume consumption rate is estimated based on the liquid level change in the no-effective-inflow interval, and the replenishment volume that has passed the flow detection position but has not yet formed a liquid level response in the liquid chamber is accumulated according to the liquid level response delay to obtain the replenishment volume in the queue;

[0016] S3. Based on the current liquid level, the correspondence between the liquid level and the liquid volume, the liquid volume consumption rate, the in-transit replenishment amount in the queue, and each response delay, determine the predicted liquid level in the prediction time domain; when the predicted liquid level reaches or falls below the dynamic trigger boundary determined based on the liquid volume consumption rate and each response delay, determine the advance execution time and target replenishment amount of the electronically controlled replenishment actuator. The target replenishment amount is determined based on the difference between the target liquid volume and the predicted liquid volume at the expected liquid level response time. The predicted liquid volume includes the replenishment amount in the in-transit replenishment amount in the queue that is expected to form a liquid level response before the expected liquid level response time.

[0017] S4. Control the electronically controlled liquid replenishment actuator according to the advance execution time and the target liquid replenishment volume, and construct a predicted causal response chain based on the instruction change time and each response delay; when the actual causal response chain after liquid replenishment is complete, the deviation between the actual causal response chain and the predicted causal response chain is within a preset allowable range and meets the preset parameter update conditions, update each response delay; when the actual causal response chain is complete, the deviation is within the preset allowable range but does not meet the preset parameter update conditions, freeze each response delay; when the actual causal response chain is incomplete or the deviation exceeds the preset allowable range, switch to degraded liquid level control based on upper and lower liquid level thresholds.

[0018] By establishing a replenishment event window when the replenishment control command changes to increase the replenishment output, and sequentially identifying the command change time, pressure response time, flow response time, and level response time, the replenishment process can be divided into different stages such as pressure establishment, liquid flow, and level formation response. This solves the problem that existing control methods have difficulty distinguishing each response stage and accurately determining the replenishment timing. At the same time, the liquid volume consumption rate is estimated based on the level change in the no-effective-inflow interval, and the replenishment volume that has passed the flow detection position but has not yet formed a level response is included in the replenishment volume in the queue for level prediction. This can reduce repeated replenishment and level overshoot caused by level response lag. Furthermore, by combining dynamic trigger boundaries to determine the advance execution time and target replenishment volume in advance, and by comparing the predicted causal response chain with the actual causal response chain to realize parameter updates, parameter freezing, and degraded level control, it is beneficial to improve the timeliness, accuracy, adaptability, and operational safety of level control.

[0019] The primary objective of this invention can also be achieved using the following technical measures:

[0020] Furthermore, the controller filters the liquid level information to obtain a filtered liquid level, and uses the filtered liquid level at the current moment as the current liquid level;

[0021] The command change time is the moment when the fluid replenishment control command switches from the off state to the on state, or the moment when the increase in the fluid replenishment control command reaches the preset command change threshold.

[0022] When the pressure detection element is located on the outlet side of the electrically controlled liquid replenishment actuator, the controller determines the increase in pressure information relative to the pressure baseline as the pressure change. When the pressure detection element is located on the supply side of the electrically controlled liquid replenishment actuator, the controller determines the decrease in pressure information relative to the pressure baseline as the pressure change. When using the absolute change in pressure information for response judgment, the absolute change is determined as the pressure change only when the direction of change in pressure information is consistent with the preset effective response direction of the corresponding detection position.

[0023] The pressure response time is the moment when the pressure change first reaches the preset pressure change threshold and then continues to reach the first holding time. The pressure baseline is determined based on the pressure information within the preset baseline time before the command change time.

[0024] The flow response time is the moment when the actual replenishment flow rate first reaches the preset effective flow rate threshold and then continues to reach the second holding time.

[0025] The liquid level response time is the moment when the positive deviation of the filtered liquid level from the reference liquid level extrapolated according to the liquid volume consumption rate first reaches the preset liquid level response threshold and then continues to reach the third holding time.

[0026] By filtering the liquid level information and determining the pressure response based on the direction of pressure change when the pressure detection device is located on the supply or outlet side of the electronically controlled liquid replenishment actuator, and by setting change thresholds and holding times for the pressure response, flow response, and liquid level response respectively, the impact of liquid level fluctuations, pressure pulsations, instantaneous flow disturbances, and detection noise on the response timing identification results can be reduced. This solves the problem in existing technologies where the liquid replenishment response timing is easily misjudged due to signal fluctuations or different detection positions. In particular, using the reference liquid level extrapolated according to the liquid volume consumption rate as the liquid level response judgment benchmark can also distinguish between the liquid level drop caused by normal liquid consumption and the liquid level rise actually formed by liquid replenishment, thereby improving the identification accuracy of each response timing and each response delay.

[0027] Furthermore, the pressure supply response delay is the difference between the pressure response time and the command change time, the flow response delay is the difference between the flow rate response time and the pressure response time, and the liquid level response delay is the difference between the liquid level response time and the flow rate response time.

[0028] A replenishment event is defined as a valid replenishment event only when the replenishment control command does not change again to meet the preset command change conditions within the replenishment event window, the integral value of the actual replenishment flow rate reaches the preset minimum replenishment volume, and the changes of one or more selected operating parameters characterizing the liquid consumption state are all within their respective preset stable ranges.

[0029] Before a preset number of effective replenishment events are obtained, the controller uses pre-stored initial response delay and initial liquid volume consumption rate to predict the liquid level, or uses the degraded liquid level control; after the preset number of effective replenishment events are obtained, liquid level prediction control based on updated parameters is enabled.

[0030] By determining the pressure supply response delay, flow response delay, and level response delay based on adjacent response times, and only identifying the corresponding replenishment event as a valid replenishment event when the replenishment control command remains stable, the actual replenishment volume reaches the preset requirement, and the liquid consumption state is relatively stable, the interference of continuous changes in replenishment commands, insufficient replenishment volume, or sudden changes in equipment load on the parameter identification results can be eliminated, thus solving the problem of response delay identification distortion caused by invalid replenishment events. At the same time, when the number of valid replenishment events is insufficient, initial parameters are used for level prediction or degraded level control is used. After accumulating enough valid replenishment events, level prediction control based on updated parameters is then enabled, which helps to balance the availability of the equipment in the initial operation phase and the accuracy of the subsequent control process.

[0031] Furthermore, the controller pre-stores an initial effective liquid replenishment gain for characterizing the correspondence between the liquid level change and the liquid volume change;

[0032] When the actual fluid replenishment flow rate decreases to a preset no-flow threshold and continues to reach a preset flow end holding time, the flow end time is determined.

[0033] The earlier of the time after the liquid level response time and the time when the filtered liquid level first meets the preset liquid level stability condition is determined as the initial event evaluation time; when no time is detected where the filtered liquid level meets the preset liquid level stability condition, the time after the liquid level response time and the preset evaluation time is determined as the initial event evaluation time; the later of the initial event evaluation time and the sum of the flow end time and the current liquid level response delay is determined as the event evaluation time;

[0034] Within the event evaluation interval from the flow response time to the event evaluation time, the actual replenishment flow rate is integrated to obtain the event replenishment volume. The event consumption volume is determined based on the liquid volume consumption rate and the duration of the event evaluation interval. The original liquid level increment is determined based on the difference between the filtered liquid level at the event evaluation time and the filtered liquid level at the flow response time.

[0035] The effective replenishment volume is obtained by subtracting the event consumption volume from the event replenishment volume; when the effective replenishment volume is greater than the preset minimum effective volume and the original liquid level increment is greater than the preset minimum liquid level increment, the ratio of the original liquid level increment to the effective replenishment volume is determined as the event effective replenishment gain.

[0036] The controller performs amplitude-limited weighted calculations on the pressure supply response delay, flow response delay, liquid level response delay, and effective liquid replenishment gain corresponding to the effective liquid replenishment event based on the event confidence of the corresponding effective liquid replenishment event, to obtain candidate pressure supply response delay, candidate flow response delay, candidate liquid level response delay, and candidate effective liquid replenishment gain.

[0037] When the preset parameter update conditions are met, the controller uses the candidate pressure supply response delay, the candidate flow response delay, the candidate liquid level response delay, and the candidate effective liquid replenishment gain to update the corresponding current parameters, and corrects the correspondence between the liquid level and the liquid volume according to the updated current effective liquid replenishment gain.

[0038] By determining the event evaluation time after the replenishment event ends, integrating the actual replenishment flow rate within the event evaluation interval, and deducting the liquid consumption volume within the corresponding time period, we can obtain an effective replenishment volume that more closely approximates the actual liquid entering the container and causing a liquid level change. This solves the problem that calculations based solely on the flow rate integral value or liquid level change are easily affected by continuous liquid consumption. Furthermore, we determine the effective replenishment gain based on the original liquid level increment and the effective replenishment volume, and perform amplitude-limited weighted calculations on each response delay and the effective replenishment gain in conjunction with the event confidence level. This allows us to appropriately increase the parameter correction amplitude when the replenishment event quality is high and decrease the parameter correction amplitude when the event quality is low, thereby gradually correcting the correspondence between liquid level and liquid volume and reducing parameter deviations caused by equipment manufacturing errors, differences in container structure, and long-term use.

[0039] Furthermore, within multiple effective inflow intervals where the actual replenishment flow rate is lower than the preset no flow threshold and continues to reach the preset minimum duration, candidate liquid volume consumption rates are determined based on the filter liquid level drop, interval duration, and current effective replenishment gain within the corresponding effective inflow interval, and the current liquid volume consumption rate is determined based on the median or truncated average of the multiple candidate liquid volume consumption rates.

[0040] When the liquid consumption device has an adjustable load or a liquid discharge process, the controller also acquires at least one of the following operating status parameters: device load power, operating temperature, or liquid discharge flow rate, and corrects the current liquid volume consumption rate based on the operating status parameter.

[0041] By calculating the candidate liquid volume consumption rate in multiple intervals without effective inflow and using the median or truncated average to determine the current liquid volume consumption rate, the impact of single liquid level disturbances, detection errors, or occasional liquid discharge on the consumption rate estimation results can be reduced, solving the problem of insufficient stability when estimating liquid consumption rate using a single interval. When the liquid consumption equipment has an adjustable load or is in the process of liquid discharge, the current liquid volume consumption rate is further corrected according to the equipment load power, operating temperature, or liquid discharge flow rate, so that the liquid volume consumption rate can change with the actual operating state, thereby improving the accuracy of future liquid level prediction and liquid replenishment calculation.

[0042] Furthermore, the controller uses the integral of the actual replenishment flow rate in each sampling period as the replenishment volume unit, and records the corresponding timestamp and the liquid level response delay version used at that time when each replenishment volume unit enters the replenishment volume queue.

[0043] The controller determines the expected liquid level response time of each replenishment volume unit based on the timestamp of each replenishment volume unit and the corresponding liquid level response delay version.

[0044] The replenishment volume unit that has not yet reached the corresponding expected liquid level response time at the current time is determined as the replenishment volume unit in the queue, and in the replenishment volume unit in the queue, the replenishment volume unit whose expected liquid level response time is within the prediction time domain is determined as the effective replenishment volume unit in the prediction time domain.

[0045] When the expected liquid level response time of the corresponding replenishment volume unit is reached at the current time, the corresponding replenishment volume unit is removed from the replenishment volume queue;

[0046] After the liquid level response delay is updated, the newly added liquid level volume unit entering the liquid replenishment volume queue adopts the updated liquid level response delay, while the liquid level volume unit that has already entered the liquid replenishment volume queue retains its recorded liquid level response delay version.

[0047] By dividing the actual replenishment flow rate integral in each sampling period into replenishment volume units, and recording the timestamp and the liquid level response delay version used when entering the queue for each replenishment volume unit, the expected liquid level response time can be determined unit by unit. It can accurately distinguish between replenishment volumes that have already formed a liquid level response, replenishment volumes that are still in the process of being transported, and replenishment volumes that can form a liquid level response within the prediction time domain. This solves the problems in the prior art where it is difficult to accurately track the replenishment volume in transit, and it is easy to be counted repeatedly or missed. At the same time, after the liquid level response delay is updated, the newly entered replenishment volume units adopt the updated parameters, while the replenishment volume units that have already entered the queue retain the original delay version. This can avoid the parameter update causing retrospective changes to the expected response time of the existing replenishment volume, thereby ensuring the continuity and consistency of the queue calculation process.

[0048] Furthermore, the dynamic trigger boundary is higher than the preset minimum safe liquid level, and the liquid level margin between the dynamic trigger boundary and the preset minimum safe liquid level is determined based on the liquid volume consumption rate, the pressure supply response delay, the flow response delay, the liquid level response delay, the preset safe time, and the current effective liquid replenishment gain.

[0049] The expected liquid level response time is determined based on the advance execution time, the pressure supply response delay, the flow response delay, and the liquid level response delay;

[0050] The target replenishment volume is determined based on the difference between the target liquid volume and the predicted liquid volume at the expected liquid level response time, as well as the expected liquid consumption within the expected replenishment execution time. The predicted liquid volume at the expected liquid level response time includes the expected time-domain effective replenishment volume that is expected to form a liquid level response before that time.

[0051] The estimated liquid replenishment execution time is determined based on the calibrated liquid replenishment flow rate corresponding to the current liquid replenishment output of the electronically controlled liquid replenishment actuator and the target liquid replenishment volume. The controller updates the target liquid replenishment volume at least once based on the determined estimated liquid replenishment execution time.

[0052] By determining the level margin between the dynamic trigger boundary and the preset minimum safe level based on the liquid volume consumption rate, pressure supply response delay, flow response delay, level response delay, preset safety time, and current effective replenishment gain, the replenishment trigger position can be made to change with the liquid consumption rate and replenishment response rate. This solves the problem of replenishing too late when the level threshold is fixed and the consumption rate is fast or the response delay is long. At the same time, when determining the target replenishment volume, the in-transit replenishment volume that can form a level response before the expected level response time and the liquid consumption within the expected replenishment execution time are taken into account. The target replenishment volume is updated according to the expected replenishment execution time, which can reduce the duplicate compensation and consumption omission in the calculation of the target replenishment volume, thereby reducing the probability of insufficient replenishment and excessive replenishment.

[0053] Furthermore, the predicted causal response chain includes the predicted pressure response time, the predicted flow response time, and the predicted liquid level response time. The predicted pressure response time is determined based on the sum of the command change time and the pressure supply response delay. The predicted flow response time is determined based on the sum of the predicted pressure response time and the flow response delay. The predicted liquid level response time is determined based on the sum of the predicted flow response time and the liquid level response delay.

[0054] The controller determines the overall confidence level of the model based on the event confidence of multiple recent effective fluid replenishment events, the dispersion of each response delay, the completeness of the actual causal response chain, and the status of each detection device.

[0055] The preset parameter update conditions include: the actual causal response chain is complete, the deviation between each actual response time and the corresponding expected response time is within the preset allowable range, and the overall confidence level of the model is not lower than the preset confidence threshold.

[0056] When the preset parameter update conditions are met, the controller uses the candidate pressure supply response delay, the candidate flow response delay, the candidate liquid level response delay, and the candidate effective liquid replenishment gain to update the corresponding current parameters;

[0057] When the overall confidence level of the model is lower than the preset confidence threshold, freeze each response delay, the current effective liquid replenishment gain, and the liquid volume consumption rate.

[0058] By constructing predicted pressure response time, predicted flow response time, and predicted liquid level response time based on each response delay, and by comprehensively determining the overall confidence of the model by integrating the event confidence of multiple recent effective liquid replenishment events, the dispersion of each response delay, the completeness of the actual causal response chain, and the status of each detection device, the reliability of the current liquid level prediction model can be comprehensively evaluated. This solves the problem of parameter drift that can easily occur when parameters are updated directly based on a single response deviation. Parameters are only updated when the actual causal response chain is complete, the response time deviation is within the allowable range, and the overall confidence of the model meets the preset requirements. When the overall confidence of the model is insufficient, relevant parameters are frozen, which helps to prevent abnormal events or detection failures from contaminating the current control model and improves the stability of the parameter self-learning process.

[0059] Furthermore, when the replenishment control command changes to increase the replenishment output, if no pressure response is detected within the maximum allowable pressure supply response time, it is determined that there is at least one of the following: insufficient liquid supply source, ineffective operation of the electronically controlled replenishment actuator, or abnormality of the pressure detection device.

[0060] When a pressure response is detected but no flow response is detected within the maximum permissible flow response time, it is determined that at least one of the following is present: blockage of the replenishment channel, abnormality of the downstream passage of the electronically controlled replenishment actuator, or abnormality of the flow detection device.

[0061] When a flow response is detected but no level response is detected within the maximum permissible level response time, it is determined that at least one of the following exists: replenishment not entering the liquid chamber, liquid leakage, or abnormality of the level detection device.

[0062] The controller acquires at least one of the following: position feedback of the electrically controlled liquid replenishment actuator, drive current, liquid chamber inlet temperature information, equipment operating temperature information, equipment weight information, or leakage detection information. Based on the acquired information, it further distinguishes the corresponding abnormalities and limits the maximum liquid replenishment output of the electrically controlled liquid replenishment actuator according to the abnormality type. When the liquid consumption equipment has an adjustable load, it also limits the equipment load power.

[0063] After switching to the degraded level control, the controller continues to calculate candidate response delay, candidate effective replenishment gain, and candidate liquid volume consumption rate in the background. When a preset number of effective replenishment events are obtained continuously, the dispersion of each candidate response delay returns to the preset recovery range, and the overall confidence of the model determined according to the candidate parameters is not lower than the preset recovery threshold, the parameter freeze is lifted and the level prediction control is gradually restored.

[0064] By separately determining whether a pressure response occurs within the maximum permissible pressure supply response time, a flow response occurs within the maximum permissible flow response time, and a level response occurs within the maximum permissible level response time, the system can initially differentiate between insufficient liquid supply, ineffective operation of electrically controlled liquid replenishment actuators, blockage of liquid replenishment channels, liquid leakage, and abnormalities of different detection devices based on the location of the interruption in the liquid replenishment causal response chain. This solves the problem of difficulty in identifying the abnormal link based solely on the absence of a rise in liquid level. Furthermore, by combining actuator position feedback, drive current, inlet temperature, operating temperature, equipment weight, or leakage detection information, the abnormality can be confirmed. By limiting the liquid replenishment output and equipment load power according to the type of abnormality, the safety risks caused by abnormal liquid replenishment and insufficient liquid heating can be reduced. During the degraded level control period, candidate parameters continue to be calculated, and liquid level predictive control is gradually restored after continuous receipt of credible liquid replenishment events. This also helps to avoid liquid level fluctuations caused by directly switching control modes after the fault is cleared.

[0065] The second objective of this invention is achieved as follows:

[0066] A liquid level prediction and control system based on causal timing identification of replenishment events is applied to a liquid consumption device. The liquid consumption device has a liquid storage chamber and a replenishment channel communicating with the liquid storage chamber. The liquid level in the liquid storage chamber has a preset correspondence with the liquid volume. The liquid level prediction and control system includes:

[0067] An electronically controlled liquid replenishment actuator is disposed in the liquid replenishment channel and is used to regulate the liquid replenishment output entering the liquid chamber;

[0068] A pressure detection device is used to detect the pressure information on the supply side or the outlet side of the electronically controlled liquid replenishment actuator.

[0069] A flow detection device is used to detect the actual replenishment flow rate through the replenishment channel;

[0070] A liquid level detection element is used to detect the liquid level information within the liquid-containing cavity; and

[0071] The controller is electrically connected to the electronically controlled liquid replenishment actuator, the pressure detection device, the flow detection device, and the liquid level detection device, respectively.

[0072] The controller is configured to:

[0073] When a change occurs in the replenishment control command to increase the replenishment output, a replenishment event window is established. The pressure supply response delay, flow response delay, and level response delay are determined based on the command change, pressure response, flow response, and level response within the replenishment event window.

[0074] The liquid volume consumption rate is estimated based on the liquid level change in the no-inflow interval, and the amount of liquid replenishment in the queue that has passed the flow detection position but has not yet formed a liquid level response in the liquid chamber is determined based on the actual replenishment flow rate and the liquid level response delay.

[0075] Based on the current liquid level, the correspondence between the liquid level and the liquid volume, the liquid volume consumption rate, the in-transit replenishment volume in the queue, and the predicted liquid level for each response delay, an advance execution time and a target replenishment volume are determined when the predicted liquid level reaches or falls below the dynamic trigger boundary. The electrically controlled replenishment actuator is then controlled according to the advance execution time and the target replenishment volume.

[0076] A predicted causal response chain is constructed based on each response delay. The predicted causal response chain is compared with the actual causal response chain after fluid replenishment. Each response delay is updated or frozen based on the comparison results and the overall confidence of the model. When the actual causal response chain is incomplete or the deviation between the actual causal response chain and the predicted causal response chain exceeds a preset allowable range, the system switches to degraded level control based on upper and lower level thresholds.

[0077] By constructing a liquid level predictive control system using an electronically controlled liquid replenishment actuator, pressure detection actuator, flow detection actuator, liquid level detection actuator, and controller, the controller can determine the response delay at each stage based on the liquid replenishment control command, pressure response, flow response, and liquid level response. It also predicts the future liquid level by combining the liquid volume consumption rate and the amount of liquid replenishment en route. This system addresses the problem at the system level that traditional liquid level control relies solely on the current liquid level and cannot fully consider the liquid replenishment response lag and the volume of liquid replenishment en route. Furthermore, by dynamically determining the advance execution time and target liquid replenishment amount, and implementing parameter updates, parameter freezing, or downgraded liquid level control based on the comparison between the predicted and actual causal response chains, liquid replenishment execution, state detection, model correction, and anomaly protection form a closed-loop control, which improves the system's control accuracy, environmental adaptability, and operational reliability.

[0078] The beneficial effects of this invention are as follows:

[0079] This invention collects time-series data of replenishment control commands, pressure information, actual replenishment flow rate, and liquid level information. It sequentially identifies the command change time, pressure response time, flow response time, and liquid level response time, and determines the pressure supply response delay, flow response delay, and liquid level response delay respectively. This clearly reflects the actual process from pressure establishment and liquid flow to liquid level response after the replenishment command is issued, avoiding the simplistic view of the entire replenishment process as a fixed delay. It is beneficial to accurately determine the replenishment timing based on the actual state of the equipment.

[0080] This invention records the replenishment volume that has passed the flow detection location but has not yet formed a liquid level response in the liquid chamber as the replenishment volume in the queue. It further distinguishes the replenishment volume that can form a liquid level response in the prediction time domain. When predicting the future liquid level and calculating the target replenishment volume, the corresponding replenishment volume in the queue is included, thereby avoiding the controller from repeatedly increasing the replenishment output because the current liquid level has not yet risen. This helps to reduce the probability of excessive replenishment, liquid level overshoot, and frequent start-stop of the electronically controlled replenishment actuator.

[0081] This invention estimates the liquid volume consumption rate based on the liquid level change within the no-inflow zone and can correct the liquid volume consumption rate based on equipment load power, operating temperature, or liquid discharge flow rate. Simultaneously, it progressively updates each response delay and effective replenishment gain based on effective replenishment events and uses the updated parameters to correct the correspondence between liquid level and liquid volume. This allows the dynamic trigger boundary, advance execution time, and target replenishment volume to be adjusted according to changes in supply pressure, pipeline resistance, equipment load, and actuator performance, thereby improving the stability of liquid level control under different operating conditions.

[0082] This invention compares the predicted causal response chain with the actual causal response chain, and identifies possible abnormal links based on the stages where pressure response, flow response, and level response are missing or deviated. It can distinguish between insufficient liquid supply, abnormality of electronically controlled liquid replenishment actuators, blockage of liquid replenishment channels, liquid leakage, and abnormality of detection devices. When the model's overall confidence level is lower than a preset confidence threshold or when the liquid replenishment process is abnormal, the relevant parameters are frozen, and the system switches to degraded liquid level control based on upper and lower liquid level thresholds. At the same time, the liquid replenishment output and equipment load power are limited. After the recovery conditions are met, the liquid level prediction control is gradually restored, thereby improving the fault tolerance and operational safety of liquid consumption equipment. Attached Figure Description

[0083] Figure 1 This is a schematic diagram showing the structure and connection relationship of the liquid consumption device and the liquid level prediction and control system in an embodiment of the present invention.

[0084] Figure 2 This is a schematic diagram showing the functional components and connection relationships of the controller in an embodiment of the present invention.

[0085] Figure 3 This is a schematic diagram illustrating the timing relationship between the replenishment event window and the replenishment control command, pressure response, flow response, and liquid level response in an embodiment of the present invention.

[0086] Figure 4 This is a schematic diagram illustrating the enqueueing, classification, summation, removal, and liquid level response delay version update process of the liquid replenishment volume unit in this embodiment of the invention.

[0087] Figure 5 This is a schematic diagram illustrating the principles of liquid level prediction, dynamic trigger boundary, advance execution time, expected liquid level response time, and target replenishment volume determination in an embodiment of the present invention.

[0088] Figure 6 This is a schematic diagram of the overall process of the liquid level prediction and control method in an embodiment of the present invention.

[0089] Figure 7 This is a flowchart illustrating the process of fluid replenishment anomaly identification, downgraded liquid level control, and liquid level prediction control recovery in an embodiment of the present invention.

[0090] Figure 1 Solid arrows indicate the direction of liquid or vapor flow, while dashed arrows indicate detection signals, execution feedback, control commands, or electrical power regulation relationships. Figure 2 The arrows in the diagram indicate the signal transmission, control, or electrical power transmission relationships between the components; Figure 3 and Figure 5 The horizontal arrows indicate the direction of time extension, the double arrows indicate the corresponding time interval or liquid level margin, and the guide lines indicate the curves or positions corresponding to the text descriptions. Figure 4 , Figure 6 and Figure 7 The arrows in the diagram indicate the direction of process execution. Detailed Implementation

[0091] The present invention will be further described below with reference to the accompanying drawings. It should be understood that the following embodiments are only for illustrating the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. Where there is no conflict, the technical features in the following embodiments can be combined with each other.

[0092] Example 1, as Figures 1 to 7 As shown, this embodiment provides a liquid level prediction and control method and system based on causal timing identification of liquid replenishment events. In this embodiment, the liquid consumption device 100 is a commercial steam generator, and the liquid used is water.

[0093] The liquid consumption device 100 includes a water tank 110, a liquid chamber 120, a liquid replenishment channel 130, an electrically controlled liquid replenishment actuator 140, a pressure detection device 150, a flow detection device 160, a liquid level detection device 170, a temperature detection device 180, a leakage detection device 190, a controller 200, a heating element 210, a power regulation circuit 220, a steam outlet 230, and a one-way valve 240.

[0094] The water supply tank 110 constitutes the liquid supply source in this embodiment, used to store water to be added to the liquid storage chamber 120. The liquid storage chamber 120 is a steam generating chamber, and the heating element 210 is disposed at the bottom of the liquid storage chamber 120 for heating the water in the liquid storage chamber 120. The steam outlet 230 is disposed at the upper part of the liquid storage chamber 120 for outputting the steam generated by heating.

[0095] The replenishment channel 130 connects the water replenishment tank 110 and the liquid storage chamber 120. Along the water flow direction within the replenishment channel 130, an electrically controlled replenishment actuator 140, a pressure detection element 150, a flow detection element 160, and a one-way valve 240 are sequentially arranged. In this embodiment, the flow detection position refers to the detection position of the flow detection element 160 within the replenishment channel 130.

[0096] The electrically controlled liquid replenishment actuator 140 is an adjustable-speed DC water replenishment pump. The controller 200 adjusts the liquid replenishment flow rate entering the liquid chamber 120 by changing the drive duty cycle of the electrically controlled liquid replenishment actuator 140.

[0097] A pressure sensor 150 is located on the outlet side of the electrically controlled liquid replenishment actuator 140 and is used to detect the pressure information between the outlet of the electrically controlled liquid replenishment actuator 140 and the flow sensor 160. The flow sensor 160 is used to detect the actual liquid replenishment flow rate through the liquid replenishment channel 130 and flowing into the liquid storage chamber 120. A one-way valve 240 is used to restrict the reverse flow of water or steam in the liquid storage chamber 120 along the liquid replenishment channel 130.

[0098] In other embodiments, the electrically controlled liquid replenishment actuator 140 can be a solenoid valve, a proportional valve, a variable speed AC water pump, or a pump-valve combination. The pressure detection device 150 can also be disposed on the liquid supply side of the electrically controlled liquid replenishment actuator 140 to detect the liquid supply pressure between the water replenishment tank 110 and the electrically controlled liquid replenishment actuator 140.

[0099] When the pressure detection element 150 is located on the outlet side of the electrically controlled liquid replenishment actuator 140, the controller 200 determines the increase in pressure relative to the pressure baseline as the pressure change. When the pressure detection element 150 is located on the supply side of the electrically controlled liquid replenishment actuator 140, the controller 200 determines the decrease in pressure relative to the pressure baseline as the pressure change. When using the absolute change in pressure information for response judgment, the controller 200 only determines the absolute change as the pressure change when the direction of pressure change is consistent with the preset effective response direction of the corresponding detection position.

[0100] A liquid level detector 170 is disposed on the side of the liquid-containing cavity 120 for continuously detecting the liquid level information within the liquid-containing cavity 120. In this embodiment, the liquid level detector 170 is a capacitive liquid level detector. In other embodiments, the liquid level detector 170 may also be a continuously output pressure-type liquid level detector or a float-type liquid level detector, or a multi-point array electrode-type liquid level detector or a photoelectric liquid level detector; the liquid level detector 170 used should be able to output continuous liquid level values ​​or have a liquid level resolution sufficient to identify a preset liquid level response threshold.

[0101] Temperature detection element 180 is located at the bottom of liquid chamber 120 and close to heating element 210 to detect the operating temperature of liquid chamber 120. Leakage detection element 190 is located in liquid replenishment channel 130 and water collection area below liquid chamber 120 to detect whether liquid leakage has occurred in liquid replenishment channel 130 or liquid chamber 120.

[0102] The power regulation circuit 220 is electrically connected to both the controller 200 and the heating element 210. The controller 200 can adjust the actual heating power of the heating element 210 through the power regulation circuit 220. The power regulation circuit 220 can be a relay power regulation circuit, a thyristor power regulation circuit, a pulse width modulation power regulation circuit, or a voltage regulation circuit.

[0103] The controller 200 is electrically connected to the electronically controlled liquid replenishment actuator 140, the pressure detection device 150, the flow detection device 160, the liquid level detection device 170, the temperature detection device 180, the leakage detection device 190, and the power regulation circuit 220.

[0104] In other embodiments, the electrically controlled fluid replenishment actuator 140 may be equipped with a position detection element, which is used to provide feedback to the controller 200 on the valve core position, pump operating status, or execution status of the electrically controlled fluid replenishment actuator 140.

[0105] An inlet temperature sensor can be installed at the liquid replenishment inlet of the liquid chamber 120 to detect the temperature of the replenishing liquid entering the liquid chamber 120. A weight sensor can also be installed below the liquid consumption device 100 or the water replenishment tank 110 to detect the overall weight of the liquid consumption device 100 or the weight of the remaining liquid in the water replenishment tank 110.

[0106] The position detection device, inlet temperature detection device, and weight detection device are electrically connected to the controller 200, and one or more of them can be selected according to the abnormal diagnosis requirements of the liquid consumption device 100.

[0107] The controller 200 includes an event construction unit 201, a timing identification unit 202, a consumption estimation unit 203, a liquid level prediction unit 204, a liquid replenishment decision unit 205, a consistency diagnosis unit 206, and a degradation control unit 207. Each of these units can be implemented by a control program running in the controller 200 and does not need to be configured as independent hardware.

[0108] Event construction unit 201 is used to detect changes in replenishment control commands and establish a replenishment event window. Timing identification unit 202 is used to identify pressure response time, flow response time, and level response time, and determine the pressure supply response delay, flow path response delay, and level response delay. Consumption estimation unit 203 is used to estimate the liquid volume consumption rate. Level prediction unit 204 is used to determine the amount of replenishment in the queue and predict the future level. Replenishment decision unit 205 is used to determine the advance execution time and target replenishment amount. Consistency diagnosis unit 206 is used to compare the predicted causal response chain with the actual causal response chain. Degradation control unit 207 is used to execute degraded level control when identified parameters are unreliable or the replenishment process is abnormal.

[0109] The liquid-containing chamber 120 has a uniform cross-section structure within the preset working liquid level range. The controller 200 pre-stores the initial correspondence between the liquid level and the liquid volume in the liquid-containing chamber 120. The initial correspondence can be pre-calculated based on the structural dimensions of the liquid-containing chamber 120, or it can be obtained through quantitative water injection calibration when the equipment leaves the factory.

[0110] In this embodiment, within the preset working liquid level range, for every 1 mL of water added to the liquid chamber 120, the liquid level rises by 0.05 mm. This 0.05 mm / mL is used as the initial effective replenishment gain to characterize the correspondence between the change in liquid level and the change in liquid volume.

[0111] In other embodiments, the liquid-containing cavity 120 can be a variable cross-section structure within a preset working liquid level range. For the variable cross-section liquid-containing cavity 120, the controller 200 can pre-store a lookup table, calibration curve, or piecewise function between liquid level and liquid volume, and determine the corresponding local liquid level-volume conversion parameters based on the current liquid level interval. The current effective liquid replenishment gain can be used to correct the local conversion parameters of the lookup table, calibration curve, or piecewise function within the current liquid level interval.

[0112] The controller 200 updates the current effective liquid replenishment gain based on the effective liquid replenishment event, and uses the updated current effective liquid replenishment gain as the liquid level and liquid volume conversion parameter of the liquid chamber 120 near the current working liquid level.

[0113] Among them, the initial effective replenishment gain is a pre-stored parameter used when the equipment is running for the first time or when there are not enough effective replenishment events; the event effective replenishment gain is a parameter sample determined based on a single effective replenishment event; the candidate effective replenishment gain is a parameter calculated based on the event effective replenishment gain, the current effective replenishment gain, the basic update coefficient, and the event confidence level, and is waiting for consistency verification; the current effective replenishment gain is a control parameter that actually participates in level prediction and replenishment decision-making after consistency verification.

[0114] The controller 200 collects the replenishment control command, pressure information, actual replenishment flow rate, liquid level information, and operating temperature in a 50 ms sampling period, and stores the data in the order of collection time. In other embodiments, the sampling period can be set from 10 ms to 200 ms depending on the response speed of each detection element.

[0115] The liquid level prediction and control process in this embodiment corresponds to steps S1 to S4 in claim 1. To facilitate a detailed explanation of each control step, steps S1 to S4 are further expanded into steps S101 to S107 below.

[0116] Specifically, step S101 corresponds to the response time identification and response delay determination process in steps S1 and S2; steps S102 to S104 correspond to the effective replenishment event screening, candidate identification parameter formation, liquid volume consumption rate estimation and queue in-transit replenishment volume determination process in step S2; step S105 corresponds to step S3; and steps S106 and S107 correspond to step S4.

[0117] S101. Establish a fluid replenishment event window and identify each response time.

[0118] In this embodiment, the fluid replenishment control command is a drive duty cycle command output by the controller 200 to the electronically controlled fluid replenishment actuator 140, and the preset command change threshold is 10 percentage points.

[0119] When the controller 200 switches the electronically controlled liquid replenishment actuator 140 from the off state to the on state, or increases the drive duty cycle of the electronically controlled liquid replenishment actuator 140 by 10 percentage points relative to the drive duty cycle before the change, the event construction unit 201 determines the moment when the liquid replenishment control command changes as the command change moment, and establishes a liquid replenishment event window starting from the command change moment.

[0120] The fluid resuscitation event window lasts until the event evaluation time of this fluid resuscitation event, and does not exceed the preset maximum event time. In this embodiment, the preset maximum event time is 30 seconds.

[0121] When the preset maximum event time has not been reached and the event evaluation time is still not determined, the controller 200 determines the current liquid replenishment event as an invalid liquid replenishment event, does not use the current liquid replenishment event to form candidate parameters for updating each response delay and the current effective liquid replenishment gain, and performs anomaly identification based on the missing response stage.

[0122] The event construction unit 201 determines the pressure baseline based on the average pressure information within 1 second prior to the instruction change time. In other embodiments, the median or truncated average value of the pressure information within 1 second can also be used to determine the pressure baseline.

[0123] The timing identification unit 202 starts detecting pressure information from the moment the command changes. When the pressure detection element 150 is set on the liquid outlet side of the electronically controlled liquid replenishment actuator 140, if the increase or absolute change of the pressure information relative to the pressure baseline reaches 5 kPa and is maintained continuously for 0.2 s, the moment when the condition is met first is determined as the pressure response moment.

[0124] In this embodiment, 5 kPa is the preset pressure change threshold, and 0.2 s is the first holding time.

[0125] When the pressure detection element 150 is set on the liquid supply side of the electronically controlled liquid replenishment actuator 140, if the decrease or absolute change of the pressure information relative to the pressure baseline reaches the preset pressure change threshold and continues to reach the first holding time, the moment when the condition is first met is determined as the pressure response moment.

[0126] The timing identification unit 202 simultaneously detects the actual replenishment flow rate. When the actual replenishment flow rate reaches 20 mL / min and is maintained continuously for 0.3 s, the moment when this condition is met first is determined as the flow response moment.

[0127] In this embodiment, 20 mL / min is the preset effective flow rate threshold, and 0.3 s is the second holding time.

[0128] The controller 200 performs a moving average filter on the liquid level information output by the liquid level detection device 170 to obtain the filtered liquid level. In this embodiment, the moving average is performed using the liquid level information from the most recent 10 sampling periods. In other embodiments, median filtering, low-pass filtering, or Kalman filtering can also be used to obtain the filtered liquid level.

[0129] In the process of liquid level prediction, dynamic trigger boundary judgment, target liquid replenishment calculation, degraded liquid level control and liquid level prediction control recovery in this embodiment, the current liquid level refers to the filtered liquid level at the current moment.

[0130] The controller 200 extrapolates the trend of the liquid level in the liquid chamber 120 decreasing over time under the condition that no effective liquid replenishment has occurred, based on the current liquid volume consumption rate, and determines the liquid level corresponding to this decreasing trend as the reference liquid level.

[0131] When the positive deviation of the filtered liquid level from the reference liquid level reaches 0.8 mm and remains continuously for 0.8 s, the timing identification unit 202 determines the moment when the condition is first met as the liquid level response moment.

[0132] In this embodiment, 0.8 mm is the preset liquid level response threshold, and 0.8 s is the third holding time.

[0133] For pressure response, flow response and liquid level response, the controller 200 confirms whether the corresponding signal continues to meet the threshold condition after the corresponding hold time expires; when it is confirmed that the condition is met, the sampling time when the corresponding signal first reaches the corresponding threshold is back-recorded as the corresponding response time, instead of recording the time when the hold time expires as the corresponding response time.

[0134] When the difference between two adjacent response times does not exceed one sampling period, the controller 200 determines that the two responses occur simultaneously within the sampling resolution and sets the stage response delay between them to zero. As long as the pressure response, flow response, and level response do not exhibit an inversion exceeding one sampling period, the corresponding replenishment event can still be determined to have a complete causal order. If any adjacent response exhibits an inversion exceeding one sampling period, the replenishment event is not used to update the corresponding response delay.

[0135] When the command change moment, pressure response moment, flow response moment, and liquid level response moment occur in sequence, the timing identification unit 202 determines the corresponding liquid replenishment process as a liquid replenishment event with a complete causal sequence.

[0136] The timing identification unit 202 determines the time difference between the pressure response time and the command change time as the pressure supply response delay, the time difference between the flow response time and the pressure response time as the flow flow response delay, and the time difference between the liquid level response time and the flow rate response time as the liquid level response delay.

[0137] When the equipment is first run, the controller 200 uses pre-stored initial pressure response delay, initial flow response delay, and initial liquid level response delay to predict the liquid level. In this embodiment, the initial pressure response delay is 0.25 s, the initial flow response delay is 0.35 s, and the initial liquid level response delay is 1.50 s.

[0138] The controller 200 also has a pre-stored initial liquid volume consumption rate. The initial liquid volume consumption rate can be obtained by conducting a no-replenishment operation test of the liquid consumption device 100 under rated operating conditions, or it can be predetermined based on the heating power of the heating element 210, the rated thermal efficiency of the device, the density of water, and the vaporization characteristics.

[0139] Before a preset number of ineffective inflow intervals are obtained, the controller 200 extrapolates the reference liquid level using the initial liquid volume consumption rate; after a preset number of ineffective inflow intervals are obtained, the updated current liquid volume consumption rate is used to extrapolate the reference liquid level.

[0140] S102, Screen valid fluid replacement events and generate candidate identification parameters.

[0141] Within the replenishment event window, if the replenishment control command does not switch from the off state to the on state again, the drive duty cycle does not increase by more than 10 percentage points again, the integral value of the actual replenishment flow rate reaches 20 mL, and the change in heating power of the heating element 210 does not exceed 5% of the heating power at the start of the event, the corresponding replenishment event is determined as a valid replenishment event.

[0142] In this embodiment, the increase in the drive duty cycle by 10 percentage points constitutes a change that satisfies the preset command change condition, 20 mL is the preset minimum replenishment volume, and the heating power of the heating element 210 is an operating parameter characterizing the liquid consumption state.

[0143] To avoid interference with event evaluation caused by the reduction of fluid replenishment control commands, it can also be required that within the fluid replenishment event window, the decrease in drive duty cycle relative to the drive duty cycle at the start of the event does not exceed a preset stable change limit.

[0144] The above conditions are used to eliminate interference from continuous changes in replenishment commands, insufficient replenishment volume, or sudden changes in equipment load on the identification results. In other embodiments, the stability of liquid consumption can also be determined based on operating temperature, steam output, or liquid discharge flow rate.

[0145] The controller 200 detects the end time of the flow rate during this replenishment process. When the actual replenishment flow rate drops below 2 mL / min and remains at that level for 0.5 s, the moment when this condition is met first is determined as the end time of the flow rate. Here, 2 mL / min is the preset no-flow threshold, and 0.5 s is the preset flow rate end holding time.

[0146] The controller 200 determines the earlier of the time 2 seconds after the liquid level response time and the time when the filtered liquid level first meets the preset liquid level stability condition as the initial event evaluation time; if no time when the filtered liquid level meets the preset liquid level stability condition is detected within the liquid replenishment event window, the time 2 seconds after the liquid level response time is used as the initial event evaluation time.

[0147] If the evaluation time of the preliminary event is earlier than the sum of the flow end time and the current liquid level response delay, the sum of the flow end time and the current liquid level response delay shall be determined as the event evaluation time; otherwise, the evaluation time of the preliminary event shall be determined as the event evaluation time.

[0148] When the absolute value of the rate of change of the filtered liquid level is no greater than 0.1 mm / s for 1 second, the controller 200 determines that the filtered liquid level meets the preset liquid level stability condition. In this embodiment, 2 seconds is the preset evaluation time.

[0149] The controller 200 integrates the actual replenishment flow rate from the flow response time to the event evaluation time according to the sampling period. The replenishment volume corresponding to each sampling period is the product of the actual replenishment flow rate within that sampling period and the duration of the sampling period. The sum of the replenishment volumes corresponding to all sampling periods is the event replenishment volume.

[0150] The controller 200 multiplies the current liquid volume consumption rate by the duration from the flow response time to the event evaluation time to obtain the event consumption volume.

[0151] The controller 200 subtracts the filtered liquid level at the time of the flow response from the filtered liquid level at the time of the event evaluation to obtain the original liquid level increment.

[0152] The controller 200 subtracts the event consumption volume from the event replenishment volume to obtain the effective replenishment volume. When the effective replenishment volume is greater than 10 mL and the original liquid level increment is greater than 0.5 mm, the original liquid level increment is divided by the effective replenishment volume to obtain the event effective replenishment gain corresponding to this effective replenishment event.

[0153] In this embodiment, 10 mL is the preset minimum effective volume, and 0.5 mm is the preset minimum liquid level increment.

[0154] When the effective replenishment volume is no more than 10 mL or the original liquid level increment is no more than 0.5 mm, the controller 200 does not use the current replenishment event to form candidate effective replenishment gain. However, when other conditions are met, the current replenishment event can still be used to form candidate pressure supply response delay, candidate flow response delay, and candidate liquid level response delay.

[0155] The controller 200 determines the event confidence level based on the stability of the pressure signal, flow signal, liquid level signal, and equipment load during a valid replenishment event. The signal stability is determined based on the residual fluctuation after removing the corresponding normal response trend, in order to avoid misjudging the expected pressure rise, flow rate rise, or liquid level rise during a replenishment event as signal instability.

[0156] For a pressure signal, after the pressure response is confirmed, the controller 200 selects a preset pressure stability evaluation interval. Based on low-pass filtering, moving average, or monotonic trend fitting results, it determines the pressure response trend. The difference between the actual pressure and the pressure response trend is taken as the pressure residual. The difference between the maximum and minimum values ​​of the pressure residual within the pressure stability evaluation interval is determined as the actual pressure fluctuation amplitude. The controller 200 divides the actual pressure fluctuation amplitude by a preset maximum allowable pressure residual fluctuation amplitude to obtain the pressure fluctuation ratio. When the pressure fluctuation ratio is greater than 1, it is limited to 1, and then 1 is subtracted from the pressure fluctuation ratio to obtain the pressure stability.

[0157] Flow stability and level stability are determined using the same method. After the flow response and level response are confirmed, the controller 200 selects the corresponding preset stability evaluation interval, removes the normal upward or stable trend of the actual replenishment flow rate and the normal upward trend of the filtered level relative to the extrapolated reference level, and then determines the flow stability and level stability based on the fluctuation amplitude of the obtained flow residual and level residual. Equipment load stability is determined based on the ratio of the change in heating power, steam output, or liquid discharge flow rate within the replenishment event window to the corresponding preset maximum allowable change. The value range for each stability is 0 to 1; the smaller the residual fluctuation or load change, the closer the corresponding stability is to 1. When a sufficiently long stability evaluation interval cannot be obtained, the corresponding stability is set to 0, and the relevant parameters are not updated using that replenishment event.

[0158] The controller 200 multiplies the pressure stability, flow stability, liquid level stability, and load stability by their respective preset weights, and then sums the results to obtain the event confidence level. The sum of each preset weight is 1. In this embodiment, the preset weights for the four stability parameters are all 0.25.

[0159] The controller 200 uses the pressure supply response delay, flow response delay, liquid level response delay and effective liquid replenishment gain determined by this effective liquid replenishment event as parameter samples to form candidate pressure supply response delay, candidate flow response delay, candidate liquid level response delay and candidate effective liquid replenishment gain, respectively.

[0160] For any candidate parameter to be formed, the controller 200 first calculates the difference between the current parameter sample and the corresponding current parameter, then multiplies the difference by the basic update coefficient and the event confidence in sequence, and adds the resulting correction amount to the corresponding current parameter to obtain the candidate parameter.

[0161] The baseline update coefficient can range from 0.1 to 0.3. The higher the event confidence, the greater the correction magnitude of the candidate parameters by the effective fluid resuscitation event; the lower the event confidence, the smaller the correction magnitude.

[0162] The controller 200 also presets physical allowable lower limits and physical allowable upper limits for pressure supply response delay, flow response delay, liquid level response delay, and effective liquid replenishment gain. When the calculated candidate parameter is lower than the corresponding physical allowable lower limit, the physical allowable lower limit is used; when the calculated candidate parameter is higher than the corresponding physical allowable upper limit, the physical allowable upper limit is used.

[0163] The candidate parameters generated in step S102 do not directly replace the current parameters currently involved in liquid level prediction and liquid replenishment decision-making. Only after step S106 confirms that the actual causal response chain is complete, the deviation between each actual response time and the corresponding expected response time is within a preset allowable range, and the overall confidence level of the model is not lower than a preset confidence threshold, will the controller 200 use the corresponding candidate parameters to update the current parameters.

[0164] Before three valid replenishment events are obtained, the controller 200 uses the initial response delay, the initial liquid volume consumption rate, and the initial valid replenishment gain to predict the liquid level, or adopts degraded liquid level control; after three valid replenishment events are obtained, liquid level prediction control based on updated parameters is enabled.

[0165] S103, Estimate the liquid volume consumption rate.

[0166] When the actual replenishment flow rate is less than 2 mL / min and remains below this level for more than 5 seconds, the consumption estimation unit 203 determines the corresponding time interval as a period with no effective inflow. Here, 5 seconds is the preset minimum duration.

[0167] When a no-effective-inflow interval is determined, the controller 200 also checks whether the electrically controlled liquid replenishment actuator 140 is in the off state, whether there is reverse flow in the liquid replenishment channel 130, whether the equipment is performing active liquid drainage, and whether one or more of the selected operating status parameters (heating power, operating temperature, steam output, or liquid discharge flow rate) are within their respective preset stable ranges, and whether there is an abnormal jump in the filter liquid level exceeding the preset liquid level jump threshold. Only when the electrically controlled liquid replenishment actuator 140 is in the off state, there is no reverse flow or active liquid drainage, the selected operating status parameters are all stable, and the filter liquid level does not have an abnormal jump, is the corresponding time interval determined as a no-effective-inflow interval.

[0168] For each interval with no effective inflow, the controller 200 subtracts the filter level at the end of the interval from the filter level at the beginning of the interval to obtain the amount of drop in filter level; then, the amount of drop in filter level is divided by the current effective replenishment gain to obtain the corresponding amount of reduction in liquid volume; finally, the amount of reduction in liquid volume is divided by the duration of the interval with no effective inflow to obtain the candidate liquid volume consumption rate.

[0169] When the drop in filtered liquid level is not greater than zero, or when the calculated candidate liquid volume consumption rate exceeds the preset physical allowable range, the controller 200 will determine the corresponding no effective inflow interval as an invalid interval and discard the corresponding candidate liquid volume consumption rate; if the preset number of effective candidate liquid volume consumption rates has not yet been obtained, the current liquid volume consumption rate or the initial liquid volume consumption rate will continue to be used, and the overall confidence of the model will be reduced.

[0170] The controller 200 stores the five most recent candidate liquid volume consumption rates and arranges the candidate liquid volume consumption rates according to their numerical values, taking the candidate liquid volume consumption rate in the middle position as the current liquid volume consumption rate.

[0171] In other implementations, the maximum and minimum values ​​among multiple candidate liquid volume consumption rates can be deleted, and the average value of the remaining candidate liquid volume consumption rates can be calculated as the current liquid volume consumption rate.

[0172] When the current heating power of the heating element 210 changes by more than 10% relative to the heating power when the candidate liquid volume consumption rate is obtained, the controller 200 corrects the liquid volume consumption rate based on the current heating power and the current operating temperature.

[0173] The controller 200 can pre-store a calibration table between heating power, operating temperature, and consumption rate correction coefficient. This calibration table is established through non-replenishment operation tests of the liquid-consuming device 100 at different heating powers and operating temperatures.

[0174] The controller 200 obtains the corresponding consumption rate correction coefficient from the calibration table based on the current heating power and current operating temperature, and multiplies the uncorrected liquid volume consumption rate with the consumption rate correction coefficient to obtain the corrected current liquid volume consumption rate.

[0175] When the liquid consumption device 100 has a measurable steam output flow rate or other liquid discharge flow rate, the controller 200 can also correct the current liquid volume consumption rate based on the steam output flow rate or liquid discharge flow rate.

[0176] When the no-effective-inflow interval used to determine the current liquid volume consumption rate does not include an active drainage process, the controller 200 converts the steam output flow rate or other liquid discharge flow rate into a liquid volume discharge rate per unit time, and adds the liquid volume discharge rate to the current liquid volume consumption rate determined based on the liquid level drop to obtain a corrected current liquid volume consumption rate; when the current liquid volume consumption rate has already been determined based on the liquid level drop that includes the corresponding drainage process, the liquid volume discharge rate is not repeatedly added.

[0177] S104. Determine the amount of fluid to be replenished in the queue.

[0178] The controller 200 integrates the actual replenishment flow rate within each sampling period to obtain the replenishment volume unit for the corresponding sampling period.

[0179] When each replenishment volume unit enters the replenishment volume queue, the controller 200 records the generation time, replenishment volume, and the liquid level response delay value used at that time. The recorded generation time constitutes the corresponding timestamp, and the recorded liquid level response delay value and its parameter update time identifier constitute the corresponding liquid level response delay version.

[0180] The controller 200 adds the generation time of the replenishment volume unit to the liquid level response delay used by the replenishment volume unit to obtain the expected liquid level response time of the replenishment volume unit.

[0181] If the expected liquid level response time of the corresponding replenishment volume unit has not yet been reached at the current time, the replenishment volume unit is retained in the replenishment volume queue and identified as a replenishment volume unit in the queue.

[0182] The controller 200 adds up the replenishment volumes of each queue's in-transit replenishment volume unit to obtain the in-transit replenishment amount of the queue.

[0183] In the queue of liquid replenishment volume units, when the expected liquid level response time of the corresponding liquid replenishment volume unit is between the current time and the end time of the prediction time domain, the liquid replenishment volume unit is determined as a valid liquid replenishment volume unit in the prediction time domain.

[0184] The controller 200 adds up the replenishment volumes of each predicted time-domain effective in-transit replenishment volume unit to obtain the predicted time-domain effective in-transit replenishment amount.

[0185] When the current time reaches or exceeds the expected liquid level response time of the corresponding replenishment volume unit, the controller 200 removes the corresponding replenishment volume unit from the replenishment volume queue and considers that the replenishment volume has been reflected in the current filter liquid level.

[0186] This embodiment uses an equivalent pure time delay method to describe the response process between the flow detection location and the liquid level detection result. Specifically, each replenishment volume unit is considered to have not yet formed a liquid level response before its expected liquid level response time, and is considered to have fully reflected in the current filtered liquid level after reaching the expected liquid level response time. This equivalent pure time delay method is suitable for operating states where the deviation between the actual response time and the expected response time is within a preset allowable range. When the deviation exceeds the preset allowable range, the controller 200 does not update the corresponding delay parameters and performs parameter freezing or downgraded liquid level control.

[0187] In other embodiments where the replenishment channel exhibits a clearly distributed response, the controller 200 can further divide a replenishment volume unit into multiple response sub-units, assigning each sub-unit an expected liquid level response time and a response weight sequentially distributed within a preset response interval, with the sum of all response weights being 1. After reaching the expected liquid level response time of each sub-unit, the replenishment volume unit is gradually included in the replenishment volume that has already formed a liquid level response according to the corresponding response weight. Each response sub-unit retains the liquid level response delay version recorded when the replenishment volume unit was enqueued.

[0188] When the liquid level response delay is updated, the newly added liquid level volume unit in the liquid level replenishment volume queue adopts the updated liquid level response delay; the liquid level volume unit that has already entered the liquid level replenishment volume queue continues to use the liquid level response delay version recorded when it entered the queue, and will not be changed due to subsequent liquid level response delay updates.

[0189] The preset capacity of the replenishment volume queue is determined based on the maximum allowable liquid level response time, sampling period, and maximum continuous replenishment time. If the number of replenishment volume units in the replenishment volume queue reaches the preset capacity, the controller 200 stops using newly added replenishment volume units to update the liquid level prediction model, freezes the current parameters, and switches to degraded liquid level control; after replenishment in the replenishment channel stops and the maximum allowable liquid level response time has elapsed, the replenishment volume queue is reinitialized.

[0190] S105. Predict future liquid levels and determine the advance execution time and target replenishment volume.

[0191] In this embodiment, the preset minimum safe liquid level is 30 mm, the lower liquid level threshold for downgraded liquid level control is 35 mm, the upper liquid level threshold is 55 mm, and the target liquid level for normal control is 50 mm.

[0192] The liquid level prediction unit 204 determines the current liquid volume based on the current filtered liquid level and the correspondence between the currently used liquid level and the liquid volume.

[0193] For any prediction time within the prediction time domain, the controller 200 determines the time length from the current time to the prediction time, and multiplies the current liquid volume consumption rate by the time length to obtain the expected liquid consumption during that time period.

[0194] The controller 200 subtracts the expected liquid consumption from the current liquid volume and adds the replenishment volume of each effective in-transit replenishment volume unit in the prediction time domain that is expected to form a liquid level response before the corresponding prediction time, to obtain the predicted liquid volume at the corresponding prediction time.

[0195] The controller 200 converts the predicted liquid volume into a predicted liquid level based on the current correspondence between the liquid level and the liquid volume.

[0196] The controller 200 adds the current pressure supply response delay, the current flow response delay, the current liquid level response delay, and the preset safety time to obtain the liquid replenishment safety advance time. In this embodiment, the preset safety time is 1 second.

[0197] The controller 200 multiplies the current liquid volume consumption rate by the replenishment safety advance time to obtain the expected liquid consumption volume within the replenishment safety advance time; then, it converts the expected liquid consumption volume into a liquid level margin based on the current effective replenishment gain; finally, it adds the liquid level margin to the preset minimum safe liquid level to obtain the dynamic trigger boundary.

[0198] Let the current time be t0, the predicted time be t, the current liquid volume be Vnow, the current liquid volume consumption rate be Rc, and the expected replenishment volume of each replenishment volume unit that will form a liquid level response before the predicted time t be ΔVi. Then, the predicted liquid volume Vpred(t) corresponding to the predicted time t satisfies: Vpred(t) = Vnow - Rc × (t - t0) + ΣΔVi. The controller 200 converts Vpred(t) into the predicted liquid level based on the current correspondence between liquid level and liquid volume.

[0199] Let the preset minimum safe liquid level be Hmin, the current effective liquid replenishment gain be Kh, and the current pressure supply response delay, current flow response delay, current liquid level response delay, and preset safe time be τp, τq, τh, and Ts, respectively. Then the dynamic trigger boundary Htr satisfies: Htr = Hmin + Kh × Rc × (τp + τq + τh + Ts). For the variable cross-section volume liquid chamber 120, the controller 200 first converts the expected liquid consumption volume into the liquid level margin corresponding to the current liquid level range according to a lookup table, calibration curve, or piecewise function.

[0200] The liquid level prediction unit 204 calculates the predicted liquid level at each prediction time in chronological order within the prediction time domain. When the predicted liquid level first reaches or falls below the dynamic trigger boundary, the liquid replenishment decision unit 205 determines the corresponding prediction time as the advance execution time.

[0201] In this embodiment, the prediction time domain starts from the current moment, and the prediction step size is the sampling period of the liquid level information or an integer multiple of the sampling period. The length of the prediction time domain is not less than the sum of the current pressure supply response delay, the current flow response delay, the current liquid level response delay, the preset safety time, and the preset maximum liquid replenishment execution time. The preset maximum liquid replenishment execution time is determined based on the ratio of the preset maximum liquid replenishment amount to the calibrated lower limit of the liquid replenishment flow rate. When no prediction moment occurs within the prediction time domain at which the predicted liquid level reaches or falls below the dynamic trigger boundary, the controller 200 does not initiate liquid replenishment in the current sampling period and recalculates the predicted liquid level and the dynamic trigger boundary in the next sampling period.

[0202] When the advance execution time is no later than the current time, the controller 200 immediately starts or increases the liquid replenishment output of the electronically controlled liquid replenishment actuator 140; when the advance execution time is later than the current time, the controller 200 starts or increases the liquid replenishment output of the electronically controlled liquid replenishment actuator 140 when the advance execution time is reached.

[0203] When the controller 200 issues a replenishment control command to start or increase the replenishment output at an advance execution time, the advance execution time is also used as the command change time for this replenishment event.

[0204] The replenishment decision unit 205 will sequentially add the current pressure supply response delay, the current flow response delay, and the current liquid level response delay to the advance execution time to obtain the expected liquid level response time for this replenishment.

[0205] like Figure 5 As shown, the dynamic trigger boundary is higher than the preset minimum safe liquid level. The moment when the predicted liquid level first reaches or falls below the dynamic trigger boundary is the advance execution moment. The time interval between the expected liquid level response moment and the advance execution moment is the sum of the current pressure supply response delay, the current flow response delay, and the current liquid level response delay.

[0206] The replenishment decision unit 205 determines the predicted liquid volume corresponding to the expected liquid level response time. The predicted liquid volume includes the total effective in-transit replenishment volume in the predicted time domain that is expected to form a liquid level response before the expected liquid level response time.

[0207] The controller 200 determines the target liquid volume corresponding to the target liquid level based on the correspondence between the target liquid level and the liquid volume.

[0208] The controller 200 determines the initial estimated replenishment execution time based on the calibrated replenishment flow rate corresponding to the current replenishment output of the electronically controlled replenishment actuator 140. The initial estimated replenishment execution time can be determined based on the ratio between the preset initial replenishment volume and the calibrated replenishment flow rate.

[0209] The controller 200 determines the expected liquid consumption within the initial expected liquid replenishment execution time based on the initial expected liquid replenishment execution time and the current liquid volume consumption rate; then it calculates the difference between the target liquid volume and the predicted liquid volume at the expected liquid level response time, and adds the expected liquid consumption to the difference to obtain the initial target liquid replenishment amount.

[0210] The controller 200 re-determines the expected liquid replenishment execution time based on the initial target liquid replenishment volume and the calibrated liquid replenishment flow rate, and then re-determines the expected liquid consumption and the target liquid replenishment volume based on the re-determined expected liquid replenishment execution time.

[0211] Let the target liquid volume be Vtar, the predicted liquid volume at the expected liquid level response time be Vpred(tr), the expected replenishment execution time be Tfill, and the calibrated replenishment flow rate be Qcal. Then the target replenishment quantity Vadd satisfies: Vadd = max{0, Vtar - Vpred(tr) + Rc × Tfill}, and Tfill = Vadd / Qcal. The controller 200 iteratively determines the target replenishment quantity and the expected replenishment execution time according to the above two relationships.

[0212] If the difference between two consecutive target replenishment volumes is not greater than the preset replenishment volume difference, the latter target replenishment volume is used as the final target replenishment volume. If the difference is still greater than the preset replenishment volume difference, updates can continue, but the number of updates cannot exceed the preset maximum number of updates. In this embodiment, the preset maximum number of updates is 3.

[0213] When the preset maximum number of updates is reached, if the difference between two adjacent target replenishment volumes is still greater than the preset replenishment volume difference, the target replenishment volume calculated last and limited by the preset minimum and maximum replenishment volumes will be used as the final target replenishment volume. When the calibrated replenishment flow rate is not greater than zero, division by zero occurs during the calculation process, or the calculation result exceeds the preset physical allowable range, the controller 200 will stop the current predictive replenishment decision and switch to degraded level control based on the upper and lower level thresholds.

[0214] Since the predicted liquid volume at the expected liquid level response time has already taken into account the in-transit replenishment volume that is expected to form a liquid level response before that time, the controller 200 no longer performs repeated compensation for this part of the in-transit replenishment volume.

[0215] When the calculated target replenishment volume is not greater than zero, the controller 200 does not start the replenishment; when the final target replenishment volume is greater than zero but less than the preset minimum replenishment volume, the controller 200 may not start the replenishment or perform replenishment according to the preset minimum replenishment volume; when the final target replenishment volume is greater than the preset maximum replenishment volume, the controller 200 limits the final target replenishment volume to the preset maximum replenishment volume.

[0216] like Figure 5 As shown, the predicted liquid level after taking into account the effective in-transit replenishment volume in the prediction time domain is used to represent the impact of the in-transit replenishment volume expected to form a liquid level response before the expected liquid level response time on the predicted liquid level; the predicted liquid level after executing the target replenishment is used to represent the expected liquid level change trend after the replenishment is executed according to the final target replenishment volume.

[0217] The controller 200 activates the electronically controlled liquid replenishment actuator 140 at the pre-execution time and accumulates the actual liquid replenishment volume based on the actual liquid replenishment flow detected by the flow detection device 160. When the actual liquid replenishment volume reaches the final target liquid replenishment volume, the controller 200 shuts down the electronically controlled liquid replenishment actuator 140 or reduces the drive duty cycle of the electronically controlled liquid replenishment actuator 140.

[0218] S106. Compare the predicted causal response chain with the actual causal response chain, and determine whether to update the parameters.

[0219] The consistency diagnostic unit 206 adds the current pressure response delay to the command change time of this replenishment event to obtain the expected pressure response time; adds the current flow response delay to the expected pressure response time to obtain the expected flow response time; and adds the current liquid level response delay to the expected flow response time to obtain the expected liquid level response time.

[0220] The predicted pressure response time, predicted flow rate response time, and predicted liquid level response time form a predicted causal response chain in chronological order.

[0221] The pressure response time, flow response time, and liquid level response time actually detected by the pressure detection device 150, flow detection device 160, and liquid level detection device 170 constitute an actual causal response chain in chronological order.

[0222] The controller 200 determines the overall confidence level of the model based on the average event confidence level of the most recent N valid fluid replenishment events, the stability of each response delay, the completeness of the actual causal response chain, and the status of each detection element.

[0223] Where N is the smaller of the number of currently obtained effective fluid resuscitation events and 5, and N is not less than 3. When the number of obtained effective fluid resuscitation events is 3 or 4, the overall confidence of the model is determined based on the most recent 3 or 4 effective fluid resuscitation events, respectively; when the number of obtained effective fluid resuscitation events is not less than 5, the overall confidence of the model is determined based on the most recent 5 effective fluid resuscitation events.

[0224] The controller 200 averages the event confidence scores of the most recent N valid fluid replenishment events to obtain the average event confidence score.

[0225] The controller 200 determines the average value and standard deviation of the pressure supply response delay, flow response delay and liquid level response delay in the most recent N valid liquid replenishment events, and uses the ratio of the standard deviation of each type of response delay to the larger of the corresponding average value and the preset minimum positive value as the dispersion of the corresponding response delay; when all samples of a certain type of response delay are zero, the dispersion of that type of response delay is set to zero to avoid division by zero caused by the average value being zero.

[0226] The controller 200 takes the maximum value among the three types of response delay dispersion as the comprehensive response delay dispersion, and divides the comprehensive response delay dispersion by the preset maximum allowable dispersion to obtain the dispersion ratio. When the dispersion ratio is greater than 1, the dispersion ratio is limited to 1; the controller 200 then subtracts the dispersion ratio from 1 to obtain the response delay stability.

[0227] When pressure response, flow response, and level response are all detected and occur in the order of command change, pressure response, flow response, and level response, the actual causal response chain completeness is 1; when one response stage is missing, the actual causal response chain completeness is 0.67; when two response stages are missing, the actual causal response chain completeness is 0.33; when none of the three response stages are detected, the actual causal response chain completeness is 0.

[0228] The controller 200 determines the reliability of the detection status based on the ratio of the number of detections that have not experienced any faults to the total number of detections involved in the judgment. When a detection is found to have a broken wire, a signal exceeding the limit, a fixed value for a continuously preset time, or a self-test failure, the corresponding detection is identified as an abnormal detection.

[0229] The controller 200 multiplies the average event confidence, response delay stability, actual causal response chain completeness, and detection status confidence by their respective preset weights, and then sums the results to obtain the overall model confidence. The sum of each preset weight is 1. In this embodiment, the preset weights for the four parameters are all 0.25.

[0230] In this embodiment, the preset confidence threshold is 0.6, and the preset allowable deviation between each actual response time and the corresponding expected response time is the larger of 20% of the current response delay and one sampling period.

[0231] When the actual causal response chain is complete, the deviation between each actual response time and the corresponding expected response time does not exceed the corresponding preset allowable deviation, the current fluid replenishment event meets the conditions for a valid fluid replenishment event, and the overall confidence level of the model is not less than 0.6, the preset parameter update conditions are determined to be met.

[0232] When the preset parameter update conditions are met, the controller 200 uses the candidate pressure supply response delay, candidate flow response delay and candidate liquid level response delay formed in step S102 and limited by the physical allowable range to update the current pressure supply response delay, current flow response delay and current liquid level response delay respectively.

[0233] When the current effective liquid replenishment event also generates candidate effective liquid replenishment gains, the controller 200 uses the candidate effective liquid replenishment gains to update the current effective liquid replenishment gains, and corrects the correspondence between the liquid level and the liquid volume in the liquid chamber 120 according to the updated current effective liquid replenishment gains.

[0234] When the actual causal response chain is complete and the deviation between each actual response time and the corresponding expected response time does not exceed the corresponding preset allowable deviation, but the overall confidence of the model is less than 0.6, the controller 200 does not use the candidate parameters formed by this liquid replenishment event to replace the corresponding current parameters, and freezes each response delay, current effective liquid replenishment gain and current liquid volume consumption rate currently involved in liquid level prediction and liquid replenishment decision.

[0235] When the actual causal response chain is incomplete, or when the deviation between any actual response time and the corresponding expected response time exceeds the corresponding preset allowable deviation, the controller 200 controls the liquid consumption device 100 to switch to degraded liquid level control based on the upper liquid level threshold and the lower liquid level threshold, and re-determines the overall confidence of the model according to the completeness of the actual causal response chain, the deviation between each actual response time and the corresponding expected response time, and the status of each detection element, and performs anomaly identification according to the response stage where the missing or deviation occurs.

[0236] Parameter freezing refers to not replacing the control parameters currently involved in liquid level prediction and liquid replenishment decision-making with newly calculated candidate parameters until the recovery conditions are met.

[0237] During the degraded liquid level control, the controller 200 continues to collect liquid replenishment event data in the background and continues to calculate candidate pressure supply response delay, candidate flow response delay, candidate liquid level response delay, candidate effective liquid replenishment gain, and candidate liquid volume consumption rate to determine whether the liquid consumption device 100 meets the conditions for restoring liquid level predictive control.

[0238] S107, Perform abnormality monitoring, downgrade liquid level control, and recovery judgment.

[0239] In this embodiment, the maximum permissible pressure supply response time is 1 s, the maximum permissible flow response time is 2 s, and the maximum permissible liquid level response time is 5 s.

[0240] When the controller 200 issues a replenishment control command to start or increase the replenishment output, and no pressure response is detected within the maximum allowable pressure supply response time, the controller 200 determines that there is at least one of the following: insufficient liquid supply source, ineffective operation of the electronically controlled replenishment actuator 140, or abnormality of the pressure detection device 150.

[0241] Insufficient liquid supply in this embodiment includes water shortage in the water tank 110, disconnection of the liquid supply side of the liquid supply channel 130, or idling of the electronically controlled liquid supply actuator 140.

[0242] The controller 200 further acquires the drive current of the electronically controlled liquid replenishment actuator 140. When the drive current is lower than the preset lower limit of the drive current, it is first determined that the electronically controlled liquid replenishment actuator 140 is not being driven normally; when the drive current is within the normal range but no pressure response is detected, it is determined that there is at least one of the following: insufficient liquid supply, the electronically controlled liquid replenishment actuator 140 is not operating effectively, or the pressure detection device 150 is abnormal.

[0243] When a pressure response is detected, but no flow response is detected within the maximum permissible flow response time, the controller 200 determines that at least one of the following is present: blockage of the replenishment channel 130, abnormality of the downstream passage of the electronically controlled replenishment actuator 140, or abnormality of the flow detection device 160.

[0244] When a flow response is detected, but no level response is detected within the maximum permissible level response time, the controller 200 determines that at least one of the following is true: the replenishment fluid has not entered the liquid chamber 120, the replenishment channel 130 has leaked liquid, or the level detection element 170 is malfunctioning.

[0245] The controller 200 further distinguishes the corresponding anomalies by combining at least one of the following: position feedback of the electronically controlled liquid replenishment actuator 140, drive current, liquid chamber inlet temperature information, equipment operating temperature information, equipment weight information, or leakage detection information.

[0246] When the controller 200 issues a replenishment control command to start or increase the replenishment output, and the position feedback of the electronically controlled replenishment actuator 140 does not change in accordance with the replenishment control command, the controller 200 determines that there is an abnormal risk that the electronically controlled replenishment actuator 140 has not acted or has not acted properly.

[0247] When the flow detection device 160 detects the actual replenishment flow rate, and the inlet temperature information of the liquid chamber does not show a temperature change corresponding to the replenishment liquid entering, the controller 200 determines that there is at least one of the following: the replenishment liquid has not actually entered the liquid chamber 120, the inlet temperature detection is abnormal, or the replenishment channel 130 has an abnormal flow diversion.

[0248] When the flow detection element 160 detects the actual replenishment volume, but the weight change of the liquid consumption device 100 or the water replenishment tank 110 does not correspond to the actual replenishment volume, the controller 200 determines that there is at least one of the following: leakage in the replenishment channel 130, detection deviation of the flow detection element 160, or lack of liquid in the water replenishment tank 110.

[0249] When the leak detection device 190 detects liquid, the controller 200 prioritizes determining that there is a leak in the liquid replenishment channel 130 or the liquid container 120; when the leak detection device 190 does not detect liquid, the actual liquid replenishment flow is normal, but the output of the liquid level detection device 170 does not change for a long time, the controller 200 determines that there is an abnormal risk in the liquid level detection device 170.

[0250] Upon detecting a fluid replenishment anomaly, the controller 200 limits the maximum fluid replenishment output of the electrically controlled fluid replenishment actuator 140 according to the type of anomaly. When the fluid consumption device 100 has an adjustable heating load, the controller 200 also limits the heating power of the heating element 210 via the power regulation circuit 220.

[0251] In the degraded liquid level control state, when the filtered liquid level is below 35 mm, the controller 200 starts the electronically controlled liquid replenishment actuator 140 with a drive duty cycle not exceeding 60% of the rated output; when the filtered liquid level reaches 55 mm, the controller 200 shuts down the electronically controlled liquid replenishment actuator 140.

[0252] In degraded liquid level control mode, the controller 200 limits the heating power of the heating element 210 to below 50% of the rated power through the power regulation circuit 220. When the filtered liquid level is lower than the preset minimum safe liquid level or the temperature detection element 180 detects a risk of dry burning due to water shortage, the controller 200 shuts down the heating element 210.

[0253] During degraded level control, the recovery condition is determined to be met when the controller 200 continuously receives three valid replenishment events, the dispersion of each candidate response delay is less than 15%, no fault is detected in each sensor, and the overall confidence level of the model determined based on the candidate parameters recovers to above 0.8. Here, 15% corresponds to a preset recovery range, and 0.8 corresponds to a preset recovery threshold.

[0254] Once the recovery conditions are met, the controller 200 unfreezes the parameters and uses the candidate response delay, candidate effective liquid replenishment gain, and candidate liquid volume consumption rate calculated in the background as parameters to be recovered.

[0255] The controller 200 gradually transitions from frozen parameters to parameters to be restored in three consecutive preset recovery phases, with each preset recovery phase lasting 5 seconds.

[0256] In the first preset recovery stage, one-third of the difference between the frozen parameter and the parameter to be recovered is added to the frozen parameter; in the second preset recovery stage, one-third of the difference is added again; at the end of the third preset recovery stage, the complete parameter to be recovered is used.

[0257] When restoring the heating power of the heating element 210, the controller 200 gradually increases the heating power of the heating element 210 at a power ramp rate not exceeding 10% per second of the rated power until it is restored to the target heating power required for the current working state.

[0258] Through the above control method, the controller 200 can determine the time required from the issuance of the liquid replenishment control command to pressure establishment, from pressure establishment to actual flow, and from actual flow to the liquid level response in the liquid chamber 120, respectively, based on the actual liquid replenishment process of the commercial steam generator.

[0259] The controller 200 can also record the replenishment volume that has passed the flow detection position but has not yet formed a liquid level response in the liquid chamber 120 in the replenishment volume queue, and include the replenishment volume that is expected to form a liquid level response when predicting the future liquid level and determining the target replenishment volume, thereby reducing repeated replenishment and liquid level overshoot caused by liquid level response lag.

[0260] Meanwhile, the controller 200 can preliminarily distinguish between insufficient liquid supply, abnormality of the electronically controlled liquid replenishment actuator 140, blockage of the liquid replenishment channel 130, liquid leakage, or abnormality of the detection device based on the location of the interruption of the pressure response, flow response, and liquid level response. When the prediction result or response process is unreliable, it adopts degraded liquid level control based on the upper and lower liquid level thresholds and heating power limitation measures, thereby improving the stability of the liquid level control of the liquid chamber 120 and the operational safety of the liquid consumption device 100.

Claims

1. A liquid level prediction and control method based on causal timing identification of replenishment events, applied to a liquid consumption device, wherein the liquid consumption device comprises a liquid-containing chamber, a replenishment channel communicating with the liquid-containing chamber, an electrically controlled replenishment actuator for adjusting the replenishment output of the replenishment channel, a pressure detection device for detecting the pressure on the supply side or outlet side of the electrically controlled replenishment actuator, a flow detection device for detecting the actual replenishment flow rate through the replenishment channel, a liquid level detection device for detecting the liquid level in the liquid-containing chamber, and a controller, wherein the controller pre-stores the correspondence between the liquid level in the liquid-containing chamber and the liquid volume, characterized in that... The method includes: S1. Collect time-series data of replenishment control command, pressure information, actual replenishment flow rate and liquid level information; when the replenishment control command switches from the closed state to the open state or the increase of the replenishment control command reaches the preset command change threshold, establish a replenishment event window, and identify the command change time, pressure response time, flow response time and liquid level response time in the replenishment event window; S2. When the command change time, the pressure response time, the flow response time, and the liquid level response time occur sequentially, the pressure supply response delay, the flow response delay, and the liquid level response delay are determined based on the time difference between adjacent times; the liquid volume consumption rate is estimated based on the liquid level change in the no-effective-inflow interval, and the replenishment volume that has passed the flow detection position but has not yet formed a liquid level response in the liquid chamber is accumulated according to the liquid level response delay to obtain the replenishment volume in the queue; S3. Based on the current liquid level, the correspondence between the liquid level and the liquid volume, the liquid volume consumption rate, the in-transit replenishment amount in the queue, and each response delay, determine the predicted liquid level in the prediction time domain; when the predicted liquid level reaches or falls below the dynamic trigger boundary determined based on the liquid volume consumption rate and each response delay, determine the advance execution time and target replenishment amount of the electronically controlled replenishment actuator. The target replenishment amount is determined based on the difference between the target liquid volume and the predicted liquid volume at the expected liquid level response time. The predicted liquid volume includes the replenishment amount in the in-transit replenishment amount in the queue that is expected to form a liquid level response before the expected liquid level response time. S4. Control the electronically controlled liquid replenishment actuator according to the advance execution time and the target liquid replenishment volume, and construct a predicted causal response chain based on the instruction change time and each response delay; when the actual causal response chain after liquid replenishment is complete, the deviation between the actual causal response chain and the predicted causal response chain is within a preset allowable range and meets the preset parameter update conditions, update each response delay; when the actual causal response chain is complete, the deviation is within the preset allowable range but does not meet the preset parameter update conditions, freeze each response delay; when the actual causal response chain is incomplete or the deviation exceeds the preset allowable range, switch to degraded liquid level control based on upper and lower liquid level thresholds.

2. The liquid level prediction and control method according to claim 1, characterized in that: The controller filters the liquid level information to obtain a filtered liquid level, and uses the filtered liquid level at the current moment as the current liquid level. The command change time is the moment when the fluid replenishment control command switches from the off state to the on state, or the moment when the increase in the fluid replenishment control command reaches the preset command change threshold. When the pressure detection element is located on the outlet side of the electrically controlled liquid replenishment actuator, the controller determines the increase in pressure information relative to the pressure baseline as the pressure change; when the pressure detection element is located on the supply side of the electrically controlled liquid replenishment actuator, the controller determines the decrease in pressure information relative to the pressure baseline as the pressure change. When using the absolute change in pressure information for response judgment, the absolute change in pressure information is determined as the pressure change only when the direction of change in pressure information is consistent with the preset effective response direction of the corresponding detection position. The pressure response time is the moment when the pressure change first reaches the preset pressure change threshold and then continues to reach the first holding time. The pressure baseline is determined based on the pressure information within the preset baseline time before the command change time. The flow response time is the moment when the actual replenishment flow rate first reaches the preset effective flow rate threshold and then continues to reach the second holding time. The liquid level response time is the moment when the positive deviation of the filtered liquid level from the reference liquid level extrapolated according to the liquid volume consumption rate first reaches the preset liquid level response threshold and then continues to reach the third holding time.

3. The liquid level prediction and control method according to claim 2, characterized in that: The pressure supply response delay is the difference between the pressure response time and the command change time; the flow response delay is the difference between the flow rate response time and the pressure response time; and the liquid level response delay is the difference between the liquid level response time and the flow rate response time. A replenishment event is defined as a valid replenishment event only when the replenishment control command does not change again to meet the preset command change conditions within the replenishment event window, the integral value of the actual replenishment flow rate reaches the preset minimum replenishment volume, and the changes of one or more selected operating parameters characterizing the liquid consumption state are all within their respective preset stable ranges. Before a preset number of effective replenishment events are obtained, the controller uses pre-stored initial response delay and initial liquid volume consumption rate to predict the liquid level, or uses the degraded liquid level control; after the preset number of effective replenishment events are obtained, liquid level prediction control based on updated parameters is enabled.

4. The liquid level prediction and control method according to claim 3, characterized in that: The controller has a pre-stored initial effective replenishment gain for characterizing the relationship between the change in liquid level and the change in liquid volume. When the actual fluid replenishment flow rate decreases to a preset no-flow threshold and continues to reach a preset flow end holding time, the flow end time is determined. The earlier of the time after the liquid level response time and the time when the filtered liquid level first meets the preset liquid level stability condition is determined as the initial event evaluation time; when no time is detected where the filtered liquid level meets the preset liquid level stability condition, the time after the liquid level response time and the preset evaluation time is determined as the initial event evaluation time; the later of the initial event evaluation time and the sum of the flow end time and the current liquid level response delay is determined as the event evaluation time; Within the event evaluation interval from the flow response time to the event evaluation time, the actual replenishment flow rate is integrated to obtain the event replenishment volume. The event consumption volume is determined based on the liquid volume consumption rate and the duration of the event evaluation interval. The original liquid level increment is determined based on the difference between the filtered liquid level at the event evaluation time and the filtered liquid level at the flow response time. The effective replenishment volume is obtained by subtracting the event consumption volume from the event replenishment volume; when the effective replenishment volume is greater than the preset minimum effective volume and the original liquid level increment is greater than the preset minimum liquid level increment, the ratio of the original liquid level increment to the effective replenishment volume is determined as the event effective replenishment gain. The controller performs amplitude-limited weighted calculations on the pressure supply response delay, flow response delay, liquid level response delay, and effective liquid replenishment gain corresponding to the effective liquid replenishment event based on the event confidence of the corresponding effective liquid replenishment event, to obtain candidate pressure supply response delay, candidate flow response delay, candidate liquid level response delay, and candidate effective liquid replenishment gain. When the preset parameter update conditions are met, the controller uses the candidate pressure supply response delay, the candidate flow response delay, the candidate liquid level response delay, and the candidate effective liquid replenishment gain to update the corresponding current parameters, and corrects the correspondence between the liquid level and the liquid volume according to the updated current effective liquid replenishment gain.

5. The liquid level prediction and control method according to claim 4, characterized in that: In multiple effective inflow intervals where the actual replenishment flow rate is lower than the preset no flow threshold and continues to reach the preset minimum duration, candidate liquid volume consumption rates are determined based on the filter liquid level drop, interval duration, and current effective replenishment gain in the corresponding effective inflow intervals. The current liquid volume consumption rate is determined based on the median or truncated average of the multiple candidate liquid volume consumption rates. When the liquid consumption device has an adjustable load or a liquid discharge process, the controller also acquires at least one of the following operating status parameters: device load power, operating temperature, or liquid discharge flow rate, and corrects the current liquid volume consumption rate based on the operating status parameter.

6. The liquid level prediction and control method according to claim 4, characterized in that: The controller uses the integral of the actual replenishment flow rate in each sampling period as the replenishment volume unit, and records the corresponding timestamp and the liquid level response delay version used at that time when each replenishment volume unit enters the replenishment volume queue. The controller determines the expected liquid level response time of each replenishment volume unit based on the timestamp of each replenishment volume unit and the corresponding liquid level response delay version. The replenishment volume unit that has not yet reached the corresponding expected liquid level response time at the current time is determined as the replenishment volume unit in the queue, and in the replenishment volume unit in the queue, the replenishment volume unit whose expected liquid level response time is within the prediction time domain is determined as the effective replenishment volume unit in the prediction time domain. When the expected liquid level response time of the corresponding replenishment volume unit is reached at the current time, the corresponding replenishment volume unit is removed from the replenishment volume queue; After the liquid level response delay is updated, the newly added liquid level volume unit entering the liquid replenishment volume queue adopts the updated liquid level response delay, while the liquid level volume unit that has already entered the liquid replenishment volume queue retains its recorded liquid level response delay version.

7. The liquid level prediction and control method according to claim 6, characterized in that: The dynamic trigger boundary is higher than the preset minimum safe liquid level. The liquid level margin between the dynamic trigger boundary and the preset minimum safe liquid level is determined based on the liquid volume consumption rate, the pressure supply response delay, the flow response delay, the liquid level response delay, the preset safe time, and the current effective liquid replenishment gain. The expected liquid level response time is determined based on the advance execution time, the pressure supply response delay, the flow response delay, and the liquid level response delay; The target replenishment volume is determined based on the difference between the target liquid volume and the predicted liquid volume at the expected liquid level response time, as well as the expected liquid consumption within the expected replenishment execution time. The predicted liquid volume at the expected liquid level response time includes the expected time-domain effective replenishment volume that is expected to form a liquid level response before that time. The estimated liquid replenishment execution time is determined based on the calibrated liquid replenishment flow rate corresponding to the current liquid replenishment output of the electronically controlled liquid replenishment actuator and the target liquid replenishment volume. The controller updates the target liquid replenishment volume at least once based on the determined estimated liquid replenishment execution time.

8. The liquid level prediction and control method according to claim 4, characterized in that: The predicted causal response chain includes the predicted pressure response time, the predicted flow response time, and the predicted liquid level response time. The predicted pressure response time is determined based on the sum of the command change time and the pressure supply response delay. The predicted flow response time is determined based on the sum of the predicted pressure response time and the flow response delay. The predicted liquid level response time is determined based on the sum of the predicted flow response time and the liquid level response delay. The controller determines the overall confidence level of the model based on the event confidence of multiple recent effective fluid replenishment events, the dispersion of each response delay, the completeness of the actual causal response chain, and the status of each detection device. The preset parameter update conditions include: the actual causal response chain is complete, the deviation between each actual response time and the corresponding expected response time is within the preset allowable range, and the overall confidence level of the model is not lower than the preset confidence threshold. When the preset parameter update conditions are met, the controller uses the candidate pressure supply response delay, the candidate flow response delay, the candidate liquid level response delay, and the candidate effective liquid replenishment gain to update the corresponding current parameters; When the overall confidence level of the model is lower than the preset confidence threshold, freeze each response delay, the current effective liquid replenishment gain, and the liquid volume consumption rate.

9. The liquid level prediction and control method according to claim 8, characterized in that: When the replenishment control command changes to increase the replenishment output, and no pressure response is detected within the maximum allowable pressure supply response time, it is determined that there is at least one of the following: insufficient liquid supply source, ineffective operation of the electronically controlled replenishment actuator, or abnormal pressure detection device. When a pressure response is detected but no flow response is detected within the maximum permissible flow response time, it is determined that at least one of the following is present: blockage of the replenishment channel, abnormality of the downstream passage of the electronically controlled replenishment actuator, or abnormality of the flow detection device. When a flow response is detected but no level response is detected within the maximum permissible level response time, it is determined that at least one of the following exists: replenishment not entering the liquid chamber, liquid leakage, or abnormality of the level detection device. The controller acquires at least one of the following: position feedback of the electrically controlled liquid replenishment actuator, drive current, liquid chamber inlet temperature information, equipment operating temperature information, equipment weight information, or leakage detection information. Based on the acquired information, it further distinguishes the corresponding abnormalities and limits the maximum liquid replenishment output of the electrically controlled liquid replenishment actuator according to the abnormality type. When the liquid consumption equipment has an adjustable load, it also limits the equipment load power. After switching to the degraded level control, the controller continues to calculate candidate response delay, candidate effective replenishment gain, and candidate liquid volume consumption rate in the background. When a preset number of effective replenishment events are obtained continuously, the dispersion of each candidate response delay returns to the preset recovery range, and the overall confidence of the model determined according to the candidate parameters is not lower than the preset recovery threshold, the parameter freeze is lifted and the level prediction control is gradually restored.

10. A liquid level prediction and control system based on causal timing identification of replenishment events, applied to a liquid consumption device, wherein the liquid consumption device has a liquid-containing chamber and a replenishment channel communicating with the liquid-containing chamber, and the liquid level in the liquid-containing chamber has a preset correspondence with the liquid volume, characterized in that, The liquid level prediction and control system includes: An electronically controlled liquid replenishment actuator is disposed in the liquid replenishment channel and is used to regulate the liquid replenishment output entering the liquid chamber; A pressure detection device is used to detect the pressure information on the supply side or the outlet side of the electronically controlled liquid replenishment actuator. A flow detection device is used to detect the actual replenishment flow rate through the replenishment channel; A liquid level detection element is used to detect the liquid level information within the liquid-containing cavity; and The controller is electrically connected to the electronically controlled liquid replenishment actuator, the pressure detection device, the flow detection device, and the liquid level detection device, respectively. The controller is configured to: When a change occurs in the replenishment control command to increase the replenishment output, a replenishment event window is established. The pressure supply response delay, flow response delay, and level response delay are determined based on the command change, pressure response, flow response, and level response within the replenishment event window. The liquid volume consumption rate is estimated based on the liquid level change in the no-inflow interval, and the amount of liquid replenishment in the queue that has passed the flow detection position but has not yet formed a liquid level response in the liquid chamber is determined based on the actual replenishment flow rate and the liquid level response delay. Based on the current liquid level, the correspondence between the liquid level and the liquid volume, the liquid volume consumption rate, the in-transit replenishment volume in the queue, and the predicted liquid level for each response delay, an advance execution time and a target replenishment volume are determined when the predicted liquid level reaches or falls below the dynamic trigger boundary. The electrically controlled replenishment actuator is then controlled according to the advance execution time and the target replenishment volume. A predicted causal response chain is constructed based on each response delay. The predicted causal response chain is compared with the actual causal response chain after fluid replenishment. Each response delay is updated or frozen based on the comparison results and the overall confidence of the model. When the actual causal response chain is incomplete or the deviation between the actual causal response chain and the predicted causal response chain exceeds a preset allowable range, the system switches to degraded level control based on upper and lower level thresholds.

Images