A control method for real-time prediction and partitioned and graded online optimization of ammonia leakage of an ammonia engine

Abstract

The application discloses a kind of ammonia leakage amount real-time prediction and partition grading online optimization control method of ammonia engine, comprising: S1: real-time acquisition ammonia engine operating state data and multiple potential leakage area sensor monitoring signals;S2: construct dynamic prediction model based on time sequence attention mechanism, S3: according to the predicted ammonia leakage and leakage risk level, combined with engine current operation setting, call the strategy library of partition grading optimization control;S4: through online optimization controller, execute the optimization action combination of adaptive selection from the strategy library, carry out collaborative adjustment to one or more variables in the fuel injection parameter, ignition timing, inlet and exhaust valve phase and EGR rate of ammonia engine.The application realizes the real-time, accurate prediction of ammonia leakage amount and leakage risk level in the future specified time window, so as to grasp ammonia combustion condition data in real time for subsequent regulation and control.

Description

Technical Field

[0001] This invention relates to the field of ammonia leakage control technology for ammonia engines, and in particular to a control method for real-time prediction and zoned-level online optimization of ammonia leakage in ammonia engines. Background Technology

[0002] In recent years, with the continuous development of green energy technologies, the application of ammonia as a zero-carbon fuel in power plants has received widespread attention, especially in the field of ammonia engines. However, in actual operation, there is a risk of leakage in the storage, supply, and combustion systems of ammonia fuel. Ammonia leakage not only directly affects engine efficiency and operational safety but may also cause serious harm to the environment and human health. Therefore, how to achieve real-time and accurate prediction of ammonia leakage and, based on this, quickly and adaptively optimize and control the engine's operating status has become a key technical challenge for the safe and efficient operation of ammonia engines.

[0003] Currently, existing technologies for engine leak monitoring primarily focus on post-event detection and alarms, or rely on fixed thresholds and simple models for judgment, lacking the ability to predict dynamic changes in leak volume in real time. Furthermore, traditional control strategies often employ static zoning or single-level optimization methods, which are ill-suited to the complex and ever-changing operating conditions of engines and the dynamic response requirements under different leak levels. This lagging and coarse-grained control approach makes it impossible to adjust operating parameters in a timely manner when leaks occur in their early stages or when their severity changes, making it difficult to maximize engine performance while ensuring safety, and also hindering the achievement of graded and refined control based on different leak areas and severity levels.

[0004] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a control method for real-time prediction and zoned-level online optimization of ammonia leakage in ammonia engines, thereby solving the problems in the background technology.

[0006] This invention provides a control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine, comprising the following steps:

[0007] S1: Real-time acquisition of ammonia engine operating status data and after-treatment sensor leakage monitoring signals;

[0008] S2: Based on the data obtained in step S1, perform model experiments, use the temperature difference before and after the catalytic unit caused by the exothermic chemical reaction to back-calculate the concentration of unburned NH3 at the inlet of the aftertreatment system, correlate the ammonia leakage with the temperature before and after the catalytic unit, and generate a prediction model.

[0009] S3: Based on the prediction model, the operating status is divided into multiple risk level intervals. Combined with the current engine operating settings, the strategy library for partitioned and hierarchical optimization control is called. The strategy library determines the risk level based on the leakage risk level and provides multiple subsets of optimization actions.

[0010] S4: Execute an optimized action combination adaptively selected from the policy library by means of an online optimization controller.

[0011] In one or more embodiments, step S1 specifically includes:

[0012] S11: Deploy a sensor network and aftertreatment catalytic unit in each critical leakage area of ​​the ammonia engine. The sensor network includes various ammonia gas sensing devices and temperature sensors. The aftertreatment catalytic unit includes a first-stage SCR catalytic unit, a first-stage ASC catalytic unit, and a second-stage SCR catalytic unit arranged in the exhaust gas flow sequence. The temperature sensors are installed at the inlet and outlet of each sub-catalytic unit to measure the temperature rise before and after catalysis. The ammonia gas sensing devices are used to measure ammonia gas data.

[0013] S12: The sensor network synchronously acquires the raw signals of all sensors at a fixed sampling frequency, and uses a digital filter to perform noise reduction processing on the signals, filtering out high-frequency noise caused by engine vibration and electromagnetic interference.

[0014] S13: Perform time-domain, frequency-domain, and joint time-frequency-domain analysis on the denoised signal to extract various primary features, including pressure fluctuation amplitude, abnormal concentration gradient, specific frequency component energy, and signal correlation.

[0015] In one or more embodiments, step S2 includes:

[0016] S21: Determine the chemical reaction formula for each catalytic unit, and determine the enthalpy change of the chemical reaction based on the chemical reaction formula;

[0017] S22: Perform model tests and conduct multiple trial runs. Based on sensor data, calculate the actual ammonia concentration change of each catalytic unit and obtain the original emissions. The calculation is performed using the following formula.

[0018] ,

[0019] ,

[0020] ,

[0021] in, This refers to the ammonia concentration difference. The label for the first-stage SCR indicates the ammonia leakage amount of the first-stage SCR catalytic unit, ASC indicates the ammonia leakage amount of the first-stage ASC catalytic unit, in indicates the data at the inlet of the corresponding catalytic unit, out indicates the data at the outlet of the corresponding catalytic unit, and original discharge indicates the amount of unburned ammonia emitted in the original discharge.

[0022] In one or more embodiments, step S2 further includes:

[0023] S23: Wait for the temperature sensor reading to stabilize, retrieve the sensor data, and determine the temperature rise value ΔT corresponding to each catalytic unit and each trial run;

[0024] S24: For each trial run calibration point, and by fitting the temperature rise value and the change in ammonia concentration, the prediction model is obtained. The prediction model is designed to obtain the temperature rise value and, based on the fitted function, obtain the ammonia leakage data.

[0025] The process of fitting the temperature rise and the change in ammonia concentration is achieved through the following formula:

[0026] ,

[0027] ,

[0028] in, It is the NH3 concentration difference in the reaction on the catalytic unit. It is the mapping relationship between the NH3 concentration difference in the reaction and variables such as temperature rise. It is the temperature difference between the outlet and inlet of the catalytic unit. The airspeed is calculated based on the exhaust mass flow rate and the size of the catalytic converter. These are other input variables that may affect the mapping relationship.

[0029] In one or more embodiments, step S2 further includes:

[0030] S25: Further calculate the mapping relationship of the remaining reducing components, including CO, HC, and H2, processed by the first-stage ASC catalytic unit using the following formula:

[0031] ,

[0032] in, This represents the relationship between the concentration of reducing components and the temperature rise. The temperature rise of the reducing component is The concentration of the reducing agent component. Inlet temperature, It is the airspeed calculated based on the exhaust mass flow rate and the size of the catalytic converter.

[0033] In one or more embodiments, step S2 further includes:

[0034] S25: After obtaining the prediction model, the prediction model is verified multiple times;

[0035] S26: After verification, remove the ammonia sensing device;

[0036] The ammonia sensing device is used for recycling to debug and generate models; removing the ammonia sensing device is used to reduce the high cost of ammonia sensors in the entire system.

[0037] In one or more embodiments, the specific threshold setting and dynamic adjustment process for determining the leakage risk level in step S3 is as follows:

[0038] S31: Establish a static threshold comparison table based on the engine's basic operating conditions. The static threshold comparison table is based on the original NH3 leakage concentration value. It pre-defines normal zones, attention zones, warning zones, and danger zones, and sets corresponding response measures for each zone's threshold.

[0039] The response measures for the normal zone are to implement conventional control strategies; for the zone of concern, to initiate fine-tuning and optimization of combustion parameters; for the zone of warning, to initiate injection strategy adjustments; and for the zone of danger, to adopt torque limiting and power reduction measures combined with safety protection measures.

[0040] S32: Execute the aforementioned countermeasures, receive real-time ambient temperature and pressure data, and perform real-time correction on the thresholds corresponding to the operating conditions in the static threshold table to eliminate environmental impacts;

[0041] S33: By combining historical engine operating data and leak maintenance records, the values ​​in the threshold table are updated regularly through offline statistical learning methods, so that the threshold settings can adapt to the long-term decline of engine performance and the aging trend of components.

[0042] Secondly, the present invention provides a control system for real-time prediction and zone-level online optimization of ammonia leakage in an ammonia engine. Based on the same concept as the aforementioned control method for real-time prediction and zone-level online optimization of ammonia leakage in an ammonia engine, the system includes multiple execution modules. These execution modules implement the aforementioned control method for real-time prediction and zone-level online optimization of ammonia leakage in an ammonia engine during operation.

[0043] Thirdly, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the aforementioned control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine.

[0044] Fourthly, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the aforementioned control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine.

[0045] Compared with the prior art, the present invention has the following technical advantages:

[0046] This invention achieves real-time and accurate prediction of ammonia leakage and its risk level by real-time acquisition, multi-domain analysis, and feature fusion of signals from a sensor network deployed in different areas of an ammonia engine, and by using a dynamically predicted model based on a constructed temporal attention mechanism. This avoids the cost and time lag associated with directly using ammonia sensors to obtain leakage data. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the overall process of a control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine according to the present invention.

[0048] Figure 2 This is a schematic diagram of the overall hardware architecture of a control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine according to the present invention.

[0049] Figure 3 This is a schematic diagram of the test and calibration of the mapping relationship between the catalytic unit temperature rise and NH3 concentration in the control method of real-time prediction and zoned and graded online optimization of ammonia leakage in an ammonia engine according to the present invention.

[0050] Figure 4 This is a flowchart of the real-time ammonia leakage prediction method of the control method for real-time prediction and zoned and graded online optimization of ammonia leakage in an ammonia engine according to the present invention. Detailed Implementation

[0051] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings, but it should be understood that the scope of protection of the present invention is not limited to the specific embodiments.

[0052] Unless otherwise expressly stated, throughout the specification and claims, the term "comprising" or its variations such as "including" or "comprises" shall be understood to include the stated elements or components without excluding other elements or other components.

[0053] Example 1:

[0054] This embodiment provides a control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine, including the following steps:

[0055] S1: Real-time acquisition of ammonia engine operating status data and after-treatment sensor leakage monitoring signals;

[0056] S2: Based on the data obtained in step S1, perform model experiments, use the temperature difference before and after the catalytic unit caused by the exothermic chemical reaction to back-calculate the concentration of unburned NH3 at the inlet of the aftertreatment system, correlate the ammonia leakage with the temperature before and after the catalytic unit, and generate a prediction model.

[0057] S3: Based on the prediction model, the operating status is divided into multiple risk level intervals. Combined with the current engine operating settings, the strategy library for partitioned and hierarchical optimization control is called. The strategy library determines the risk level based on the leakage risk level and provides multiple subsets of optimization actions.

[0058] S4: Execute an optimized action combination adaptively selected from the policy library by means of an online optimization controller.

[0059] Specifically, this implementation method is a comprehensive approach for real-time prediction and zoned-level online optimization control of ammonia leakage. The specific implementation process is as follows: First, engine operating status data and sensor monitoring signals from the aftertreatment system are collected in real time using various sensors in the engine and aftertreatment system. Based on the collected basic data, bench model calibration tests are conducted. Utilizing the characteristic that ammonia releases heat during a chemical reaction in the catalytic converter, resulting in a temperature difference before and after the catalytic converter, the concentration of unburned NH3 at the inlet of the aftertreatment system is calculated in reverse. A quantitative correlation between ammonia leakage and the temperature rise before and after the catalytic converter is established, ultimately generating an ammonia leakage prediction model that can be used for real-time calculations. Based on the original ammonia leakage concentration value output by the prediction model, the engine operating status is divided into multiple risk level intervals. Combining the current operating parameters of the engine, a pre-stored zoned-level optimization control strategy library is invoked. The strategy library can match the corresponding subset of optimization actions according to the current leakage risk level. Finally, through the online optimization controller, the optimal action combination adapted to the current operating conditions is adaptively selected from the strategy library and executed, completing the closed-loop control of the engine operating status.

[0060] This implementation method can achieve real-time prediction of unburned ammonia leakage without relying on ammonia sensing equipment, solving the industry pain points of existing ammonia sensors such as high cost, small range, slow response and easy failure. At the same time, with the partitioned and hierarchical optimized control architecture, the control strategy can be adjusted according to the degree of leakage. While ensuring the power and economy of operation, it also takes into account safety and emission requirements, realizing efficient, safe and clean operation of ammonia engine under all operating conditions.

[0061] In one preferred embodiment, step S1 specifically includes:

[0062] S11: Deploy a sensor network and aftertreatment catalytic unit in each critical leakage area of ​​the ammonia engine. The sensor network includes various ammonia gas sensing devices and temperature sensors. The aftertreatment catalytic unit includes a first-stage SCR catalytic unit, a first-stage ASC catalytic unit, and a second-stage SCR catalytic unit arranged in the exhaust gas flow sequence. The temperature sensors are installed at the inlet and outlet of each sub-catalytic unit to measure the temperature rise before and after catalysis. The ammonia gas sensing devices are used to measure ammonia gas data.

[0063] S12: The sensor network synchronously acquires the raw signals of all sensors at a fixed sampling frequency, and uses a digital filter to perform noise reduction processing on the signals, filtering out high-frequency noise caused by engine vibration and electromagnetic interference.

[0064] S13: Perform time-domain, frequency-domain, and joint time-frequency-domain analysis on the denoised signal to extract various primary features, including pressure fluctuation amplitude, abnormal concentration gradient, specific frequency component energy, and signal correlation.

[0065] Specifically, based on step S1, this implementation method refines the sensor signal acquisition and preprocessing process. First, the system hardware deployment is completed. In the aftertreatment system of the ammonia engine, the first-stage SCR catalytic unit, the ASC catalytic unit, and the second-stage SCR catalytic unit are arranged sequentially according to the exhaust gas flow direction. Temperature sensors are installed at the inlet and outlet of each catalytic unit to collect temperature data before and after the catalytic unit to calculate the temperature rise. At the same time, ammonia gas sensing equipment for calibration tests and operating condition sensors for collecting engine operating parameters are deployed to form a sensor network. All sensors synchronously acquire raw signals at a fixed sampling frequency, and the raw signals are denoised by digital filters to filter out high-frequency noise generated by engine mechanical vibration, exhaust pulsation, and vehicle electromagnetic interference, thereby improving signal stability. Time-domain, frequency-domain, and time-frequency-domain joint analyses are performed on the denoised signals to extract various primary features such as pressure fluctuation amplitude, abnormal concentration gradient, specific frequency component energy, and multi-signal correlation, providing basic feature inputs for subsequent model construction and leak determination.

[0066] This embodiment provides the necessary hardware foundation for ammonia concentration prediction based on temperature rise by arranging temperature sensors before and after the catalytic unit. Multi-dimensional signal denoising and feature extraction processing can effectively eliminate interference signals and improve the accuracy and reliability of input data. At the same time, the ammonia sensing device is only configured as an auxiliary device in the calibration stage and does not need to be installed and operated on the vehicle for a long time, which provides a prerequisite for reducing the system hardware cost in the future.

[0067] In one preferred embodiment, step S2 includes:

[0068] S21: Determine the chemical reaction formula for each catalytic unit, and determine the enthalpy change of the chemical reaction based on the chemical reaction formula;

[0069] S22: Perform model tests and conduct multiple trial runs. Based on sensor data, calculate the actual ammonia concentration change of each catalytic unit and obtain the original emissions. The calculation is performed using the following formula.

[0070] ,

[0071] ,

[0072] ,

[0073] in, This refers to the ammonia concentration difference. The label for the first-stage SCR indicates the ammonia leakage amount of the first-stage SCR catalytic unit, ASC indicates the ammonia leakage amount of the first-stage ASC catalytic unit, in indicates the data at the inlet of the corresponding catalytic unit, out indicates the data at the outlet of the corresponding catalytic unit, and original discharge indicates the amount of unburned ammonia emitted in the original discharge.

[0074] Specifically, for SCR or ASC, the higher the concentration of NH3 participating in the reaction, the more significant the exothermic reaction on that catalytic unit. NH3 in the engine exhaust reacts sequentially in the first-stage SCR and ASC, and is completely consumed after the ASC stage. The second-stage SCR is designed to handle excess NOx in the exhaust. The NH3 reacting in the second-stage SCR originates from aftertreatment ammonia injection, not from exhaust NH3 leakage. Therefore, the NH3 consumption in the second-stage SCR is not included in the calculation of exhaust NH3 leakage. Thus, the NH3 concentration in the exhaust is equal to the sum of the concentrations consumed in the first-stage SCR and ASC. This implementation clarifies the intrinsic relationship between ammonia concentration and exothermic reaction at the chemical reaction mechanism level. By summing the ammonia consumption of the two catalytic units, the total ammonia leakage in the exhaust is calculated, ensuring the accuracy of the ammonia leakage calculation in principle. It also avoids the hardware challenges of directly measuring high-concentration exhaust ammonia, providing accurate concentration baseline data for establishing a mapping model between temperature rise and ammonia concentration.

[0075] In a preferred embodiment, step S2 further includes:

[0076] S23: Wait for the temperature sensor reading to stabilize, retrieve the sensor data, and determine the temperature rise value ΔT corresponding to each catalytic unit and each trial run;

[0077] S24: For each trial run calibration point, and by fitting the temperature rise value and the change in ammonia concentration, the prediction model is obtained. The prediction model is designed to obtain the temperature rise value and, based on the fitted function, obtain the ammonia leakage data.

[0078] The process of fitting the temperature rise and the change in ammonia concentration is achieved through the following formula:

[0079] ,

[0080] ,

[0081] in, It is the NH3 concentration difference in the reaction on the catalytic unit. It is the mapping relationship between the NH3 concentration difference in the reaction and variables such as temperature rise. It is the temperature difference between the outlet and inlet of the catalytic unit. The airspeed is calculated based on the exhaust mass flow rate and the size of the catalytic converter. These are other input variables that may affect the mapping relationship.

[0082] Specifically, the temperature change in the post-treatment system depends not only on the exothermic reaction in the catalytic unit but also on the system's heat dissipation. This invention establishes a mapping model between the temperature rise of the catalytic unit and the NH3 concentration through bench calibration. and A schematic diagram of the model and calibration method is shown below. Figure 3 As shown, the exothermic reaction of NH3 in SCR and ASC is different, therefore the relationship between temperature rise and NH3 concentration was determined for SCR and ASC respectively.

[0083] like Figure 3 As shown, under standard bench test conditions, for each catalytic unit (SCR or ASC), test data under different conditions were collected through bench tests to develop and calibrate the mapping model. In the bench tests, simulated exhaust gases with different NH3 and NOx concentrations were introduced into the SCR and ASC respectively, while simultaneously varying the exhaust mass flow rate and the catalytic unit inlet temperature. Key input variables such as the inlet temperature of the catalytic unit under steady state were recorded. ), outlet temperature ( ), inlet NH3 concentration ( ) and outlet NH3 concentration ( During the test, the values ​​selected for each key input variable should cover the boundary range of the engine's normal operating conditions, and several calibration points should be selected within this boundary range to ensure the comprehensiveness of the test data. As a preferred approach, the SCR inlet temperature should cover the range of 180℃-550℃, and multiple calibration points should be selected within this range.

[0084] For each set of calibration operating points (inlet temperature) of a single catalytic unit Space velocity (SV), NH3 inlet concentration NOx inlet concentration Using the recorded temperature and NH3 concentration data, the temperature rise value ΔT of the catalytic unit (as shown in Equation 1) and the NH3 concentration consumption value of the catalytic unit (as shown in Equations 2 and 3) are recorded. Considering the heat capacity and thermal inertia of the catalytic unit, the response of the temperature rise signal is not instantaneous, but has a first-order hysteresis characteristic. Therefore, the calibration process needs to maintain a sufficient time (preferably 3–5 minutes) at each operating point to fully reach thermal equilibrium. Therefore, under each set of test operating conditions, the engine is kept running stably for a sufficient period of time until the catalytic unit reaches thermal equilibrium and the temperature sensor measurement value stabilizes before the sensor data for the corresponding operating condition is retrieved.

[0085] By using a large amount of calibration data, a functional relationship between NH3 concentration consumption and the temperature rise of the catalytic unit can be fitted for both SCR and ASC catalytic units.

[0086] This implementation refines the calibration and fitting process of the prediction model. The technical advantage of this implementation is that a quantitative mapping relationship between temperature rise and ammonia concentration is established through steady-state bench calibration. The ammonia sensor converted by the aftertreatment catalytic unit is discarded, and online estimation of ammonia concentration can be achieved without the need for additional installation. The model input parameters are all conventionally collected signals from the engine, which are compatible with the computing power of the on-board controller. At the same time, the model incorporates influencing factors such as airspeed and inlet temperature, which can adapt to different operating conditions and effectively improve the accuracy and applicability of ammonia leakage prediction results.

[0087] In a preferred embodiment, step S2 further includes:

[0088] S25: Further calculate the mapping relationship of the remaining reducing components, including CO, HC, and H2, processed by the first-stage ASC catalytic unit using the following formula:

[0089] ,

[0090] in, This represents the relationship between the concentration of reducing components and the temperature rise. The temperature rise of the reducing component is The concentration of the reducing agent component. Inlet temperature, It is the airspeed calculated based on the exhaust mass flow rate and the size of the catalytic converter.

[0091] Specifically, considering the presence of ignition fuel, the exhaust of ammonia-fueled engines contains combustible components such as CO, HC, and H2 in addition to NH3. These components have very low concentrations in the exhaust, therefore their contribution to the temperature rise caused by oxidation in SCR or ASC is minimal. To accurately separate the temperature rise component caused by NH3 oxidation, a simple quantitative relationship model is also established for other reducing components from the ignition fuel. Since the concentrations of other reducing components from the ignition fuel are very low, and the efficient oxidation of reducing components occurs on the oxidizing catalyst ASC, from the perspective of model universality, CO, HC, H2, etc., are uniformly classified as reducing components, and a correspondence between the concentration of reducing components and the temperature rise is established on the ASC. As shown in Equation 7. The correspondence is as follows: There are multiple ways to determine this value. It can be obtained from catalyst sample test results, from test data of the engine operating in a pure alternative fuel mode, or by estimation using a chemical reaction model. The concentration of the reducing component in Equation 7 ( The value is determined by estimating the engine combustion state. The estimation method involves using parameters such as engine speed, torque, and oxygen concentration measured by a wide-range oxygen sensor, and then referring to a pre-calibrated map. If, during development, it is found that the concentration of other reducing components resulting from the ignition fuel is very low, this method may not be used. The function.

[0092] This embodiment corrects the temperature rise of non-ammonia reducing components brought by the ignition fuel, thereby eliminating the interference of other exothermic oxidation reactions on the ammonia concentration estimation. It avoids the error of overestimating the ammonia concentration due to the high temperature rise caused by the oxidation of other combustible components, and further improves the accuracy of the ammonia leakage prediction results.

[0093] In a preferred embodiment, step S2 further includes:

[0094] S25: After obtaining the prediction model, the prediction model is verified multiple times;

[0095] S26: After verification, remove the ammonia sensing device;

[0096] The ammonia sensing device is used for recycling to debug and generate models; removing the ammonia sensing device is used to reduce the high cost of ammonia sensors in the entire system.

[0097] As a specific application of the aforementioned implementation method, the real-time prediction method for NH3 leakage during actual engine operation is achieved through the following steps:

[0098] Step 1: Prediction Model Calibration

[0099] like Figure 4 As shown, the volumes of the first-stage SCR and ASC in the catalytic unit need to be calibrated in advance. and ); and it is necessary to determine the mapping relationship between NH3 concentration and SCR temperature rise through bench tests. The mapping relationship between NH3 concentration and ASC temperature rise ( The mapping relationship between the concentration of unburned components and temperature rise. ), and a calibration map table for determining the concentration of unburned components in the ignition fuel by using engine speed, torque, and wide-range oxygen sensor concentration.

[0100] Step 2: Data Collection

[0101] The controller acquires the following signal at a fixed sampling period (typically 10–100 ms): First-stage SCR inlet temperature First-stage SCR outlet temperature ASC outlet temperature Engine speed, engine torque, ammonia injection mass flow rate ignition fuel injection mass flow rate Intake mass flow rate Wide-range oxygen sensor readings Input the data into the prediction model for calculation.

[0102] Step 3: Signal Preprocessing

[0103] The acquired temperature signal ( , and The temperature data is filtered to eliminate sensor noise and glitches caused by exhaust pulsation, resulting in filtered temperature data. , and Commonly used filtering algorithms include first-order low-pass filtering, moving average filtering, or Kalman filtering.

[0104] Step 4: Calculation of catalytic unit temperature rise and space velocity

[0105] Calculate real-time temperature rise for SCR .

[0106] For ASC, the real-time temperature rise is first calculated based on the signal from the temperature sensor. Secondly, corrections are made based on the real-time temperature rise for the ignition fuel. Since ammonia engines often require ignition fuel, the engine exhaust may contain a small amount of reducing components from the ignition fuel. These components are effectively converted at the ASC (Autonomous Smoke Sensor) and cause a certain amount of temperature increase. Based on the engine speed, torque, and wide-range oxygen sensor readings collected in step 2, and the calibration table determined in step 1, the real-time temperature rise of the unburned components is calculated. After deducting the temperature contribution of unburned fuel from the total temperature rise, the corrected temperature rise of ASC can be expressed as follows: The specific formula is as follows:

[0107] ;

[0108] ;

[0109] ;

[0110] Mapping relationship between NH3 concentration and SCR temperature rise ( ) and the mapping relationship between NH3 concentration and ASC temperature rise ( The air velocity of the catalytic converter is related to the space velocity of the catalytic converter under this operating condition. Therefore, the air velocities of the SCR and ASC need to be determined before predicting the leakage. According to the definition, the air velocity of the catalytic converter is calculated as shown in Equation 11. Under standard conditions, the air density is... The catalytic unit volume has been confirmed in step 1, and the exhaust mass flow rate is determined by the ammonia injection mass flow rate. ignition fuel injection mass flow rate Intake mass flow rate The calculation yields the following result. Based on the formula below, the space velocity of the catalytic unit under this operating condition is calculated for both SCR and ASC. and :

[0111] ;

[0112] ;

[0113] Step 5: Model Query and Leakage Prediction

[0114] For SCR, summarize the real-time temperature rise of the SCR unit calculated in step 4. Compared with current operating parameters (SCR inlet temperature) and SCR airspeed Input the calibrated mapping model between constant NH3 concentration and SCR temperature rise. The concentration of NH3 reacted on the SCR was obtained. .

[0115] Similarly, for ASC, summarize the corrected temperature rise of the ASC cell calculated in step 4. Compared with current operating parameters (ASC inlet temperature) and ASC airspeed Input the calibrated mapping model between constant NH3 concentration and ASC temperature rise. The concentration of NH3 reacted on the ASC was obtained. .

[0116] Finally, the predicted concentration of NH3 leaked from the original discharge is calculated using the following formula. .

[0117] ;

[0118] Step 6: Output the prediction results

[0119] The predicted concentration of NH3 leakage from the original discharge. The output is sent to the graded control module, and the ammonia leakage concentration (numerical value or trend graph) can be displayed in real time on the dashboard or monitoring interface, and an audible and visual alarm will be issued when the safety threshold is exceeded.

[0120] As a preferred embodiment, the specific threshold setting and dynamic adjustment process for determining the leakage risk level in step S3 is as follows:

[0121] S31: Establish a static threshold comparison table based on the engine's basic operating conditions. The static threshold comparison table is based on the original NH3 leakage concentration value. It pre-defines normal zones, attention zones, warning zones, and danger zones, and sets corresponding response measures for each zone's threshold.

[0122] The response measures for the normal zone are to implement conventional control strategies; for the zone of concern, to initiate fine-tuning and optimization of combustion parameters; for the zone of warning, to initiate injection strategy adjustments; and for the zone of danger, to adopt torque limiting and power reduction measures combined with safety protection measures.

[0123] S32: Execute the aforementioned countermeasures, receive real-time ambient temperature and pressure data, and perform real-time correction on the thresholds corresponding to the operating conditions in the static threshold table to eliminate environmental impacts;

[0124] S33: By combining historical engine operating data and leak maintenance records, the values ​​in the threshold table are updated regularly through offline statistical learning methods, so that the threshold settings can adapt to the long-term decline of engine performance and the aging trend of components.

[0125] Specifically,

[0126] when If the ammonia leak concentration is determined to be within a safe range and the engine is in a low ammonia leak state, the conventional control strategy is executed without additional intervention. The control strategy maintains the basic control parameters calibrated in the current MAP table, including ammonia injection quantity MAP, injection timing MAP, and ignition diesel injection MAP for ammonia-diesel engines.

[0127] when If the ammonia leakage concentration is determined to be within the concern zone, indicating a moderate ammonia leakage level, then fine-tuning and optimization of combustion parameters should be implemented. Typical measures analyzed for ammonia-diesel engines include:

[0128] (1) Combustion chamber oxygen concentration adjustment: By adjusting the opening of the intake throttle valve or the exhaust gas recirculation rate, the excess air coefficient (λ value) in the cylinder can be appropriately increased to promote the complete combustion of ammonia.

[0129] (2) Fine adjustment of ammonia injection timing (slight advance): Appropriately advance the ammonia injection timing to prolong the mixing time of ammonia and air and the combustion reaction time, thereby reducing the generation of unburned ammonia;

[0130] (3) Slight increase in diesel ignition amount (step-by-step increment): The amount of diesel ignition injection is increased slightly to improve the initial ignition energy in the combustion chamber and promote the stable ignition of ammonia.

[0131] The above adjustments should be implemented gradually within the preset adjustable range to avoid engine instability caused by sudden parameter changes. After each adjustment, wait a certain period of time (e.g., 5–10 seconds) and then reassess. If the leakage rate drops to the normal range, further adjustments are stopped; if it remains in the concern zone, the next level of adjustment is implemented until the effect converges or the adjustment limit is reached.

[0132] when When the ammonia leak concentration is determined to be within the warning zone, it indicates that the ammonia leak is already at a high level. Fine-tuning the combustion parameters alone is insufficient to resolve the issue; a more robust injection strategy adjustment should be initiated. For ammonia-diesel engines, specific measures include:

[0133] (1) Active reduction of ammonia injection volume: Under the premise of ensuring that the engine output power is not lower than the lower limit threshold (such as 70% of the rated power), actively reduce the ammonia injection volume to reduce the source of ammonia leakage from the source; the reduction range starts from 5% and gradually increases until the leakage concentration drops below E2 or reaches the maximum allowable reduction.

[0134] (2) Injection timing lag adjustment (at appropriate time): Under special circumstances, moderately delaying the ammonia injection timing may lead to an increase in cylinder temperature and NOx, but can suppress a high proportion of ammonia leakage; this strategy is only used when the above reduction strategies are ineffective, and the lag range is strictly limited;

[0135] (3) Limit the rate of increase of engine power: Limit the slope of the driver's accelerator pedal command to avoid a sharp increase in temperature due to sudden increase in load.

[0136] when This indicates that the ammonia leak has reached a hazardous level, and the ammonia concentration is determined to be within a dangerous area, posing a direct threat to personnel safety and the environment. Safety protection measures should be activated immediately. Specific measures include:

[0137] (1) Engine torque-limited and power-reduced operation: immediately limit the engine output power to below the preset safe power threshold (such as 40%–50% of the rated power) and forcibly reduce the ammonia injection volume;

[0138] (2) Issue audible and visual alarm signals: Issue emergency warnings to the driver or on-site operators through the vehicle instrument panel, alarm indicator lights and audible alarm;

[0139] (3) Record fault codes and store data: Store fault codes (Diagnostic Trouble Code, DTC) and engine operating parameters for a period of time before and after the incident in the controller for subsequent fault analysis and maintenance;

[0140] (4) Safety advice output: Provide suggested operation prompts in the alarm information, such as "Please stop the car immediately and check the ammonia injection system" and "Please stay away from the exhaust pipe area";

[0141] (5) Enter fail-safe mode: If the ammonia leak continues to maintain a high concentration, the controller will automatically switch the engine to pure diesel operation mode (only applicable to ammonia-diesel dual-fuel engines) or forced shutdown protection mode until the fault is cleared.

[0142] The zonal and hierarchical control strategy of this implementation can take differentiated intervention measures according to the severity of ammonia leakage, avoiding the impact of simply limiting torque on engine power and economy, and prioritizing personnel and equipment safety in the event of high concentration leakage, thus achieving a dynamic balance between safety and operational economy. The real-time environmental correction and periodic update mechanism of the threshold ensures that the control strategy is always adapted to the actual operating state of the engine, avoiding misjudgment or control failure caused by component aging or environmental changes, and ensuring the control effectiveness throughout the engine's entire life cycle.

[0143] Example 2:

[0144] This embodiment provides a control system for real-time prediction and zoning-level online optimization of ammonia leakage in an ammonia engine. Based on the same concept as the aforementioned control method for real-time prediction and zoning-level online optimization of ammonia leakage in an ammonia engine, it includes multiple execution modules. These execution modules implement the aforementioned control method for real-time prediction and zoning-level online optimization of ammonia leakage in an ammonia engine during operation.

[0145] Example 3:

[0146] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine.

[0147] Example 4:

[0148] This embodiment provides a non-transitory computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements a control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine.

[0149] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine, characterized in that, Includes the following steps: S1: Real-time acquisition of ammonia engine operating status data and after-treatment sensor leakage monitoring signals; S2: Based on the data obtained in step S1, perform model experiments, use the temperature difference before and after the catalytic unit caused by the exothermic chemical reaction to back-calculate the concentration of unburned NH3 at the inlet of the aftertreatment system, correlate the ammonia leakage with the temperature before and after the catalytic unit, and generate a prediction model. S3: Based on the prediction model, the operating status is divided into multiple risk level intervals. Combined with the current engine operating settings, the strategy library for partitioned and hierarchical optimization control is called. The strategy library determines the risk level based on the leakage risk level and provides multiple subsets of optimization actions. S4: Execute an optimized action combination adaptively selected from the policy library by means of an online optimization controller.

2. The control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in claim 1, characterized in that, Step S1 specifically includes: S11: Deploy a sensor network and aftertreatment catalytic unit in each critical leakage area of ​​the ammonia engine. The sensor network includes various ammonia gas sensing devices and temperature sensors. The aftertreatment catalytic unit includes a first-stage SCR catalytic unit, a first-stage ASC catalytic unit, and a second-stage SCR catalytic unit arranged in the exhaust gas flow sequence. The temperature sensors are installed at the inlet and outlet of each sub-catalytic unit to measure the temperature rise before and after catalysis. The ammonia gas sensing devices are used to measure ammonia gas data. S12: The sensor network synchronously acquires the raw signals of all sensors at a fixed sampling frequency, and uses a digital filter to perform noise reduction processing on the signals, filtering out high-frequency noise caused by engine vibration and electromagnetic interference. S13: Perform time-domain, frequency-domain, and joint time-frequency-domain analysis on the denoised signal to extract various primary features, including pressure fluctuation amplitude, abnormal concentration gradient, specific frequency component energy, and signal correlation.

3. The control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in claim 2, characterized in that, Step S2 includes: S21: Determine the chemical reaction formula for each catalytic unit, and determine the enthalpy change of the chemical reaction based on the chemical reaction formula; S22: Perform model tests and conduct multiple trial runs. Based on sensor data, calculate the actual ammonia concentration change of each catalytic unit and obtain the original emissions. The calculation is performed using the following formula. ， ， ， in, This refers to the ammonia concentration difference. The label for the first-stage SCR indicates the ammonia leakage amount of the first-stage SCR catalytic unit, ASC indicates the ammonia leakage amount of the first-stage ASC catalytic unit, in indicates the data at the inlet of the corresponding catalytic unit, out indicates the data at the outlet of the corresponding catalytic unit, and original discharge indicates the amount of unburned ammonia emitted in the original discharge.

4. The control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in claim 3, characterized in that, Step S2 further includes: S23: Wait for the temperature sensor reading to stabilize, retrieve the sensor data, and determine the temperature rise value ΔT corresponding to each catalytic unit and each trial run; S24: For each trial run calibration point, and by fitting the temperature rise value and the change in ammonia concentration, the prediction model is obtained. The prediction model is designed to obtain the temperature rise value and, based on the fitted function, obtain the ammonia leakage data. The process of fitting the temperature rise and the change in ammonia concentration is achieved through the following formula: ， ， in, It is the NH3 concentration difference in the reaction on the catalytic unit. It is the mapping relationship between the NH3 concentration difference in the reaction and variables such as temperature rise. It is the temperature difference between the outlet and inlet of the catalytic unit. The airspeed is calculated based on the exhaust mass flow rate and the size of the catalytic converter. These are other input variables that may affect the mapping relationship.

5. The control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in claim 4, characterized in that, Step S2 further includes: S25: Further calculate the mapping relationship of the remaining reducing components, including CO, HC, and H2, processed by the first-stage ASC catalytic unit using the following formula: ， in, This represents the relationship between the concentration of reducing components and the temperature rise. The temperature rise of the reducing component is The concentration of the reducing agent component. Inlet temperature, It is the airspeed calculated based on the exhaust mass flow rate and the size of the catalytic converter.

6. The control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in claim 5, characterized in that, Step S2 further includes: S25: After obtaining the prediction model, the prediction model is verified multiple times; S26: After verification, remove the ammonia sensing device; The ammonia sensing device is used for recycling to debug and generate models; removing the ammonia sensing device is used to reduce the high cost of ammonia sensors in the entire system.

7. The control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in claim 6, characterized in that, The specific threshold setting and dynamic adjustment process for determining the leakage risk level in step S3 is as follows: S31: Establish a static threshold comparison table based on the engine's basic operating conditions. The static threshold comparison table is based on the original NH3 leakage concentration value. It pre-defines normal zones, attention zones, warning zones, and danger zones, and sets corresponding response measures for each zone's threshold. The response measures for the normal zone are to implement conventional control strategies; for the zone of concern, to initiate fine-tuning and optimization of combustion parameters; for the zone of warning, to initiate injection strategy adjustments; and for the zone of danger, to adopt torque limiting and power reduction measures combined with safety protection measures. S32: Execute the aforementioned countermeasures, receive real-time ambient temperature and pressure data, and perform real-time correction on the thresholds corresponding to the operating conditions in the static threshold table to eliminate environmental impacts; S33: Combining historical engine operating data and leak maintenance records, the values ​​in the threshold table are updated regularly through offline statistical learning methods, so that the threshold settings can adapt to the long-term decline of engine performance and the aging trend of components.

8. A control system for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine, based on the same concept as the control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in any one of claims 1-6, characterized in that, It includes multiple execution modules, which implement the control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine during operation.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements a control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in any one of claims 1 to 6.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a control method for real-time prediction and zoned-level online optimization of ammonia leakage in an ammonia engine as described in any one of claims 1 to 6.

Images