Fuel cell engine control method and system

Abstract

The invention relates to the technical field of control systems, and discloses a fuel cell engine control method and system, and the method comprises the steps: carrying out the noise cleaning of a real-time parameter sequence of a fuel cell engine, and obtaining a standard parameter sequence of the fuel cell engine; on the basis of the standard parameter sequence, performing situation simulation on the running state of the fuel cell engine to obtain the change trend of the fuel cell engine; based on the change trend, performing dynamic trend coupling calibration on the standard parameter sequence to obtain an efficiency evaluation value of the fuel cell engine; on the basis of historical operation modal data of the fuel cell engine, tracing decision mapping is carried out on the efficiency evaluation value, and a control decision scheme of the fuel cell engine is obtained; performing coding mapping on the control decision scheme to obtain a control decision instruction of the fuel cell engine; the control efficiency of the fuel cell engine can be improved.

Description

Technical Field

[0001] This invention relates to the field of control system technology, and in particular to a fuel cell engine control method and system. Background Technology

[0002] As a core power unit in the new energy field, the operational stability and control precision of fuel cell engines directly affect the overall system's efficiency output. However, existing fuel cell engine control technologies have significant shortcomings in real-time parameter processing, making it difficult to effectively eliminate parameter noise caused by various interference factors during operation. This results in insufficient accuracy of the acquired real-time parameter sequences, failing to provide reliable data support for subsequent operational status analysis. This problem directly leads to biases in the judgment of engine operating status, making it difficult to accurately capture its dynamic change characteristics, thereby affecting the timeliness and adaptability of control strategies and hindering the improvement of overall control efficiency.

[0003] Meanwhile, existing technologies lack systematic design in the efficiency assessment and decision generation stages. When assessing engine efficiency, dynamic calibration is not fully integrated with changes in operating conditions, leading to discrepancies between the assessed efficiency and actual operating conditions. Furthermore, when formulating control decision schemes, the reference value of historical operating mode data is often ignored, failing to establish a precise mapping relationship between efficiency assessment and control strategies through causal analysis, resulting in insufficiently targeted control decision schemes. In addition, some control technologies lack rigorous logical integrity verification during the command generation stage, potentially leading to execution vulnerabilities in control commands, further reducing control effectiveness and failing to meet the actual requirements for efficient and stable operation of fuel cell engines. Therefore, improving the control efficiency of fuel cell engines has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a fuel cell engine control method and system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a fuel cell engine control method, comprising: S1. Noise cleaning is performed on the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine; S2. Based on the standard parameter sequence, perform situational simulation on the operating state of the fuel cell engine to obtain the changing trend of the fuel cell engine; S3. Based on the changing trend, perform dynamic trend coupling calibration on the standard parameter sequence to obtain the efficiency evaluation value of the fuel cell engine; S4. Based on the historical operating mode data of the fuel cell engine, perform a retrospective decision mapping on the efficiency evaluation value to obtain the control decision scheme of the fuel cell engine; S5. Encode and map the control decision scheme to obtain the control decision command of the fuel cell engine.

[0006] In a preferred embodiment, the noise cleaning of the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine includes: Obtain the real-time parameter sequence of the fuel cell engine; Anomalies in the real-time parameter sequence are detected to obtain the noise point identification sequence of the fuel cell engine; Based on the noise point identification sequence, the real-time parameter sequence is repaired to obtain the repaired parameter sequence of the fuel cell engine; By eliminating the high-frequency fluctuation parameters in the repair parameter sequence, the standard parameter sequence of the fuel cell engine is obtained.

[0007] In a preferred embodiment, the step of performing situational simulation of the operating state of the fuel cell engine based on the standard parameter sequence to obtain the changing trend of the fuel cell engine includes: Key parameters characterizing the operating state of the fuel cell engine are extracted from the standard parameter sequence to obtain the operating state parameters of the fuel cell engine; The operating status parameters are quantified by status indicators to obtain the real-time status vector of the operating status parameters; Based on a preset dynamic evolution strategy, the real-time state vector is forward simulated to obtain the predicted state of the fuel cell engine. The predicted state is fitted with a trend to obtain the changing trend of the fuel cell engine.

[0008] In a preferred embodiment, the step of performing trend fitting on the predicted state to obtain the changing trend of the fuel cell engine includes: Analyze the predicted state to generate prediction information for the predicted state; The predicted information is reconstructed in dimensions to obtain the predicted state vector of the predicted state; The predicted state vector is trend-quantized to obtain the comprehensive trend intensity value of the fuel cell engine, wherein the comprehensive trend intensity value is calculated using the following formula: ; In the formula, This represents the overall trend strength value. This indicates the number of predicted state vectors. Indicates the fuel cell engine's first A predicted state vector, This represents the preset reference state vector. Indicates the fuel cell engine's first Weight factors for each predicted state vector, The norm representing the difference between two vectors; Based on the sign and magnitude of the comprehensive trend strength value, the evolution direction and severity of the fuel cell engine performance are determined, and the change trend of the fuel cell engine is generated.

[0009] In a preferred embodiment, the step of performing dynamic trend coupling calibration on the standard parameter sequence based on the changing trend to obtain the efficiency evaluation value of the fuel cell engine includes: Based on the aforementioned trend, the degree of influence of key parameters in the standard parameter sequence is determined, and a calibration strategy for the key parameters is obtained. Based on the calibration strategy, the dynamic calibration weights of the key parameters are calculated, wherein the calculation formula for the dynamic calibration weights is as follows: ; In the formula, Indicates the first in the standard parameter sequence Dynamic calibration weights for key parameters, This represents the total number of key parameters in the standard parameter sequence. Indicates the first in the standard parameter sequence The trend strength values ​​corresponding to the key parameters Indicates the first in the standard parameter sequence The trend strength values ​​corresponding to the key parameters This represents the preset trend sensitivity coefficient. Represents an exponential function; Based on the dynamic calibration weights, the parameters of the standard parameter sequence are weighted and fused to obtain the calibration parameters of the fuel cell engine; The calibration parameters are normalized to obtain the efficiency evaluation value of the fuel cell engine.

[0010] In a preferred embodiment, the calibration weight is calculated using the following formula: ; In the formula, This indicates the calibration parameters of the fuel cell engine. This represents the total number of parameters in the standard parameter sequence. Indicates the first in the standard parameter sequence Dynamic calibration weights for each parameter, Indicates the first in the standard parameter sequence The values ​​of the parameters, This represents the weighted average of the parameters in the standard parameter sequence. This indicates the preset fusion adjustment coefficient.

[0011] In a preferred embodiment, the step of performing abductive decision mapping on the efficiency evaluation value based on the historical operating mode data of the fuel cell engine to obtain the control decision scheme of the fuel cell engine includes: Modal classification is performed on the historical operating mode data to obtain the historical operating modes of the fuel cell engine; By matching the efficiency evaluation value with the historical operating mode, the historical cases of the fuel cell engine are obtained; Based on the historical control strategies in the historical cases, the efficiency evaluation value is represented in a coordinated manner to obtain the preliminary control decision of the fuel cell engine; Based on the current constraints of the fuel cell engine, the preliminary control decision is modified to obtain a control decision scheme for the fuel cell engine.

[0012] In a preferred embodiment, the step of matching the efficiency assessment value with the historical operating modes to obtain historical cases of the fuel cell engine includes: Based on the efficiency evaluation value, obtain the real-time operating parameter vector of the fuel cell engine and the historical parameter vector of the historical operating mode; Calculate the similarity between the real-time running parameter vector and the historical parameter vector, wherein the formula for calculating the similarity is as follows: ; In the formula, This indicates that the real-time running parameter vector is related to the first... The similarity of historical operating patterns This represents the dimension of the parameter vector. The first element of the real-time running parameter vector represents the... One portion, Indicating the first in the historical operation mode The th historical parameter vector of the th One component; Historical cases with similarity higher than a preset matching threshold are used as historical cases of the fuel cell engine.

[0013] In a preferred embodiment, encoding and mapping the control decision scheme to obtain the control decision command for the fuel cell engine includes: The control decision scheme is analyzed in a structured manner to obtain the decision elements of the fuel cell engine; Based on the aforementioned decision elements, the control decision scheme is matched and mapped with the controller of the fuel cell engine to obtain the control logic unit of the fuel cell engine; According to the preset communication protocol, the control logic units are organized and packaged in sequence to generate the original control commands for the fuel cell engine; The original control commands are logically verified to obtain the control decision commands for the fuel cell engine.

[0014] To address the above problems, the present invention also provides a fuel cell engine control system, the system comprising: The data preprocessing module is used to clean the noise from the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine. The situation prediction module is used to simulate the operating state of the fuel cell engine based on the standard parameter sequence, and obtain the changing trend of the fuel cell engine. The performance evaluation and calibration module is used to perform dynamic trend coupling calibration on the standard parameter sequence based on the changing trend to obtain the efficiency evaluation value of the fuel cell engine. The intelligent decision-making module is used to perform a retrospective decision mapping on the efficiency evaluation value based on the historical operating mode data of the fuel cell engine, and obtain the control decision scheme of the fuel cell engine. The instruction generation module is used to encode and map the control decision scheme to obtain the control decision instructions for the fuel cell engine.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention cleans the real-time parameter sequence of a fuel cell engine by noise removal, and obtains a standard parameter sequence by detecting anomalies, repairing data, and eliminating high-frequency fluctuations. Then, it performs dynamic trend coupling calibration by combining the operating state change trend, and weights and normalizes the key parameters according to the dynamic calibration weights to accurately obtain the efficiency evaluation value. This avoids parameter noise interference and static evaluation deviation, ensures that the efficiency evaluation is close to the actual operating conditions, and provides high-quality data support for subsequent control decisions.

[0016] 2. This invention maps efficiency evaluation values ​​to historical operating modal data, matches similar historical cases, and corrects preliminary decisions based on current constraints to generate targeted control schemes. Subsequently, the control decision schemes are analyzed, controllers are matched, and the integrity of the logic is verified to generate reliable control commands, reducing decision bias and command execution loopholes. This makes the control decisions more suitable for the actual needs of the engine, effectively improving the control efficiency and operational stability of the fuel cell engine. Attached Figure Description

[0017] Figure 1 This is a schematic flowchart of a fuel cell engine control method according to an embodiment of the present invention; Figure 2 This is a functional block diagram of a fuel cell engine control system provided in an embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a fuel cell engine control method. The execution entity of this fuel cell engine control method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, the fuel cell engine control method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a fuel cell engine control method according to an embodiment of the present invention. In this embodiment, the fuel cell engine control method includes: S1. Noise cleaning is performed on the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine; In this embodiment of the invention, the step of noise cleaning the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine includes: Obtain the real-time parameter sequence of the fuel cell engine; Anomalies in the real-time parameter sequence are detected to obtain the noise point identification sequence of the fuel cell engine; Based on the noise point identification sequence, the real-time parameter sequence is repaired to obtain the repaired parameter sequence of the fuel cell engine; By eliminating the high-frequency fluctuation parameters in the repair parameter sequence, the standard parameter sequence of the fuel cell engine is obtained.

[0021] Specifically, the entire process revolves around the parameter processing of the fuel cell engine. First, its real-time parameter sequence is obtained, and then anomalies in the sequence are detected to obtain a noise point identification sequence.

[0022] Specifically, the real-time parameter sequence is repaired based on the noise point identification sequence to obtain the repaired parameter sequence, and finally the high-frequency fluctuation parameters of the repaired parameter sequence are eliminated to obtain the standard parameter sequence.

[0023] Furthermore, when acquiring real-time parameter sequences, temperature sensors, pressure sensors, current sensors, and voltage sensors need to be installed on key operating parts of the fuel cell engine. All sensors continuously collect real-time values ​​of the corresponding parameters at fixed time intervals.

[0024] Furthermore, during the data acquisition process, it is ensured that the parameter values ​​recorded at each time point completely correspond to the actual operating state of the engine. Then, all the acquired parameter values ​​are arranged in chronological order to form a real-time parameter sequence.

[0025] Furthermore, when detecting anomalies in the real-time parameter sequence, the reasonable range of values ​​for each parameter is determined based on the design and operation standards of the fuel cell engine and the parameter data accumulated under normal operation over a period of time.

[0026] Furthermore, each parameter value in the real-time parameter sequence is checked one by one. If a parameter value exceeds the corresponding reasonable range, the position of that value in the sequence is marked as "abnormal". If it is within the reasonable range, it is marked as "normal".

[0027] Furthermore, all markers are arranged in chronological order according to the real-time parameter sequence to obtain the noise point identifier sequence.

[0028] Furthermore, when performing data repair based on the noise point identification sequence, first find all positions marked as "abnormal" and their corresponding abnormal values ​​in the real-time parameter sequence by referring to the noise point identification sequence. For each abnormal value, find the normal value corresponding to the previous position marked as "normal" and the normal value corresponding to the next position marked as "normal" in the real-time parameter sequence.

[0029] Furthermore, the two normal values ​​are added together and divided by 2 to obtain the average value. This average value is then used to replace the corresponding abnormal values ​​in the real-time parameter sequence. Once all positions marked as "abnormal" have been replaced, the repaired parameter sequence is obtained.

[0030] Furthermore, when eliminating high-frequency fluctuation parameters in the repair parameter sequence, the size of the sliding window is first determined. This window needs to contain multiple consecutive parameter values. Starting from the first parameter value in the repair parameter sequence, all parameter values ​​covered by the sliding window are added together. The sum is divided by the number of parameter values ​​in the window to obtain the average value. This average value is used as the value at the corresponding position in the new sequence.

[0031] Furthermore, the sliding window is moved one position backward to cover the next consecutive parameter values ​​in the repair parameter sequence. The operation of calculating the average value is repeated until the sliding window moves to the last parameter value in the repair parameter sequence. All the calculated average values ​​are arranged in the order of calculation to obtain the standard parameter sequence.

[0032] In summary, by installing multiple sensors on key operating parts of a fuel cell engine and collecting parameter values ​​at fixed time intervals and sorting them by time, the real-time parameter sequence of the engine can be obtained.

[0033] In general, by determining the reasonable range of each parameter based on the design and operation standards and historical normal operation parameters of the fuel cell engine, and then checking the real-time parameter sequence values ​​one by one and marking the status, the noise point identification sequence of the engine can be obtained.

[0034] In summary, by comparing the abnormal values ​​in the real-time parameter sequence with the noise point identification sequence and replacing them with the average value of the normal values ​​before and after them, the repair parameter sequence of the engine can be obtained.

[0035] In summary, by setting a sliding window and successively calculating the average value of the repair parameter sequence within the window and arranging them in order, the standard parameter sequence of the engine can be obtained.

[0036] S2. Based on the standard parameter sequence, perform situational simulation on the operating state of the fuel cell engine to obtain the changing trend of the fuel cell engine; In this embodiment of the invention, the step of performing situational simulation of the operating state of the fuel cell engine based on the standard parameter sequence to obtain the changing trend of the fuel cell engine includes: Key parameters characterizing the operating state of the fuel cell engine are extracted from the standard parameter sequence to obtain the operating state parameters of the fuel cell engine; The operating status parameters are quantified by status indicators to obtain the real-time status vector of the operating status parameters; Based on a preset dynamic evolution strategy, the real-time state vector is forward simulated to obtain the predicted state of the fuel cell engine. The predicted state is fitted with a trend to obtain the changing trend of the fuel cell engine.

[0037] The step of performing trend fitting on the predicted state to obtain the changing trend of the fuel cell engine includes: Analyze the predicted state to generate prediction information for the predicted state; The predicted information is reconstructed in dimensions to obtain the predicted state vector of the predicted state; The predicted state vector is trend-quantized to obtain the comprehensive trend intensity value of the fuel cell engine, wherein the comprehensive trend intensity value is calculated using the following formula: ; In the formula, This represents the overall trend strength value. This indicates the number of predicted state vectors. Indicates the fuel cell engine's first A predicted state vector, This represents the preset reference state vector. Indicates the fuel cell engine's first Weight factors for each predicted state vector, The norm representing the difference between two vectors; Based on the sign and magnitude of the comprehensive trend strength value, the evolution direction and severity of the fuel cell engine performance are determined, and the change trend of the fuel cell engine is generated.

[0038] Specifically, the entire process revolves around the state analysis of the fuel cell engine. First, key parameters characterizing its operating state are extracted from the standard parameter sequence to obtain operating state parameters. Then, the operating state parameters are quantified by state indices to obtain the real-time state vector.

[0039] Specifically, the real-time state vector is forward-simulated based on a preset dynamic evolution strategy to obtain the predicted state, and finally the predicted state is trend-fitted to obtain the change trend.

[0040] Furthermore, when extracting key parameters, the four types of parameters that directly reflect the engine's chemical reaction efficiency, gas supply stability, energy output capability, and membrane module working status are first identified as key parameter types: reaction temperature, hydrogen intake pressure, output current, and electrolyte membrane humidity.

[0041] Furthermore, all values ​​corresponding to these four parameter types are identified one by one from the standard parameter sequence. According to the classification method of reaction temperature value group, hydrogen inlet pressure value group, output current value group, and electrolyte membrane humidity value group, the identified values ​​are assigned to the corresponding groups. After all the corresponding values ​​have been assigned, the set containing these four value groups is the operating status parameter of the fuel cell engine.

[0042] Furthermore, when quantifying the status indicators, a specific quantification rule is first formulated for each group of values ​​in the operating status parameters. For example, different temperatures in the reaction temperature group correspond to different quantification values, different pressures in the hydrogen inlet pressure group correspond to different quantification values, different currents in the output current group correspond to different quantification values, and different humidity levels in the electrolyte membrane humidity group correspond to different quantification values.

[0043] Furthermore, following these rules, the values ​​in each numerical group are converted into corresponding quantized values ​​one by one. Then, in a fixed order of reaction temperature quantized value, hydrogen inlet pressure quantized value, output current quantized value, and electrolyte membrane humidity quantized value, all the converted quantized values ​​are arranged in sequence. The resulting ordered combination of quantized values ​​is the real-time state vector of the operating state parameters.

[0044] Furthermore, when implementing forward simulation, the preset dynamic evolution strategy includes the time interval setting of the evolution step and the parameter change rules. The time interval is set to one hour as an evolution step, and the parameter change rules stipulate that the change amplitude of the quantized value of each parameter within each evolution step shall not exceed one unit. The direction of change is determined according to the change trend of the parameters in the previous three hours corresponding to the real-time state vector.

[0045] Furthermore, if the quantization value of a certain parameter continues to rise within the first three hours, the quantization value of that parameter will continue to rise within that evolution step; if it continues to fall, it will continue to fall; if there is no obvious trend, it will remain unchanged.

[0046] Furthermore, using the real-time state vector as initial data, the quantized values ​​of each parameter after the first evolution step are calculated according to the set evolution step size and change rules, forming the simulation vector at the first moment. Then, the simulation vector at the second moment is repeatedly calculated based on the simulation vector at the first moment, and so on, to continuously calculate the simulation vectors corresponding to the next six evolution steps. These simulation vectors together constitute the predicted state of the fuel cell engine.

[0047] Furthermore, when performing trend fitting, the evolution step time corresponding to each simulated vector in the predicted state is first used as the horizontal axis coordinate, and the horizontal axis coordinate is marked sequentially as one hour later, two hours later, three hours later, four hours later, five hours later, and six hours later.

[0048] Furthermore, using the quantized values ​​of each parameter in each simulation vector as the vertical axis coordinates, for the reaction temperature parameter, each horizontal axis coordinate and the corresponding quantized reaction temperature value are combined to form a coordinate point. These coordinate points are marked sequentially on the plane and then connected with straight lines to form a broken line. The overall trend of the broken line is observed to determine the changing trend of the reaction temperature parameter.

[0049] Furthermore, if the line continues to rise, it indicates an upward trend; if it continues to fall, it indicates a downward trend; and if there is no significant rise or fall, it indicates a stable trend. The same method is used to mark coordinate points, draw line graphs, and determine the trend of hydrogen inlet pressure, output current, and electrolyte membrane humidity parameters. Finally, the trend determination results of the four types of parameters are summarized to obtain the changing trend of the fuel cell engine.

[0050] Specifically, the entire process revolves around determining the overall trend and changing trend of the fuel cell engine. First, predictive information is generated from the predicted state, and then the predicted state vector is constructed.

[0051] Specifically, the comprehensive trend intensity value is calculated and its source, significance and trend are explained in conjunction with the formula. Finally, the direction and severity of performance evolution are determined based on the comprehensive trend intensity value to obtain the change trend.

[0052] Furthermore, when generating prediction information, the time stamp corresponding to each simulation vector in the prediction state is first extracted. The time stamps are one hour later, two hours later, three hours later, four hours later, five hours later, and six hours later, respectively. Then, for each simulation vector corresponding to the time stamp, the quantitative values ​​of reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity are extracted respectively.

[0053] Furthermore, each time marker is associated with and described in words with the corresponding four types of parameter quantification values, such as "After one hour, the reaction temperature quantification value is a certain value, the hydrogen inlet pressure quantification value is a certain value, the output current quantification value is a certain value, and the electrolyte membrane humidity quantification value is a certain value". All such text descriptions corresponding to the time markers are sorted and summarized in order, and the complete text set formed is the prediction information of the predicted state.

[0054] Furthermore, when constructing the predicted state vector, the reaction temperature quantization value, hydrogen inlet pressure quantization value, output current quantization value, and electrolyte membrane humidity quantization value corresponding to each time mark are extracted one by one from the textual description of the predicted information. The four types of quantization values ​​corresponding to each time mark are then combined into a sub-vector in the order of "reaction temperature quantization value, hydrogen inlet pressure quantization value, output current quantization value, and electrolyte membrane humidity quantization value".

[0055] Furthermore, by arranging the sub-vectors corresponding to all time markers in chronological order, the resulting ordered vector structure containing six sub-vectors is the predicted state vector.

[0056] Furthermore, when calculating the comprehensive trend strength value, the first and last sub-vectors are extracted from the predicted state vector. The first sub-vector corresponds to one hour later, and the last sub-vector corresponds to six hours later.

[0057] Furthermore, the quantization differences of the same parameters in the two types of sub-vectors are calculated separately. Specifically, the reaction temperature quantization difference is the reaction temperature quantization value of the last sub-vector minus the reaction temperature quantization value of the first sub-vector; the hydrogen inlet pressure quantization difference is the hydrogen inlet pressure quantization value of the last sub-vector minus the hydrogen inlet pressure quantization value of the first sub-vector; the output current quantization difference is the output current quantization value of the last sub-vector minus the output current quantization value of the first sub-vector; and the electrolyte membrane humidity quantization difference is the electrolyte membrane humidity quantization value of the last sub-vector minus the electrolyte membrane humidity quantization value of the first sub-vector.

[0058] Furthermore, the differences between the quantified values ​​of these four parameters are added together to obtain the total difference. Then, the total difference is divided by four to obtain the comprehensive trend strength value of the fuel cell engine.

[0059] Furthermore, regarding the formula for the comprehensive trend strength value, the comprehensive trend strength value... This is the final calculation result; the number of predicted state vectors. Determined by the actual number of predicted state vectors generated; the first Predicted state vectors The preset reference state vector is obtained by extracting each vector sequentially from the predicted state vector. It is a vector pre-set according to the ideal or standard operating state of the fuel cell engine; the first Weight factors of each predicted state vector Assigning weights based on the time or importance of each predicted state vector; for example, the closer the time, the greater the weight. The norm of the difference between two vectors. The result is obtained by calculating the difference between the corresponding values ​​of the two vectors and then performing Euclidean norm operations on these differences, which is the square root of the sum of the squares of the differences.

[0060] Furthermore, this formula is achieved by summing the difference norms between each predicted state vector and the reference state vector multiplied by the corresponding weighting factor, and then dividing by the number of predicted state vectors. The overall trend strength value is obtained by comprehensively measuring the overall deviation of all predicted state vectors from the reference state vector, which reflects the overall trend strength of the fuel cell engine.

[0061] Furthermore, when the predicted state vector as a whole approximates the reference state vector, The overall value is too small; after summing, divide by... Received A smaller value indicates a weaker overall trend and that the engine condition is closer to the reference condition.

[0062] Furthermore, when the predicted state vector as a whole deviates from the reference state vector, The overall value is too large; after summing, divide by... Received A larger value indicates a stronger overall trend and a greater deviation of the engine condition from the reference state. The larger the value, the more significant the deviation of the fuel cell engine's state from the reference state. The smaller the value, the less significant the deviation from the trend.

[0063] Furthermore, when determining the changing trend of the fuel cell engine, first determine the sign of the comprehensive trend strength value. If the comprehensive trend strength value is greater than zero, the evolution direction of the fuel cell engine performance is performance improvement.

[0064] Furthermore, if the overall trend strength value is less than zero, the evolution direction is performance degradation; if the overall trend strength value is equal to zero, the evolution direction is performance stability.

[0065] Furthermore, the absolute value of the overall trend strength is determined. If the absolute value is less than or equal to 1, the performance evolution is characterized by a slight change; if the absolute value is greater than 1 and less than or equal to 2, the change is moderate; and if the absolute value is greater than 2, the change is significant. Finally, the determined evolution direction and degree of intensity are combined and described in words to form a complete conclusion, thus obtaining the change trend of the fuel cell engine.

[0066] In summary, by identifying reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity as key parameter types, and by selecting corresponding values ​​from the standard parameter sequence and classifying them into appropriate groups, the operating status parameters of the fuel cell engine can be obtained.

[0067] In summary, by defining specific quantization rules for each numerical group, converting each value in the operating state parameter into a quantized value and arranging it in a fixed order, the real-time state vector of the operating state parameter can be obtained.

[0068] In summary, by using a preset dynamic evolution strategy that includes evolution step size and parameter change rules, and calculating the simulation vector for multiple future evolution steps using the real-time state vector as initial data, the predicted state of the fuel cell engine can be obtained.

[0069] In summary, by setting horizontal and vertical axis coordinates, marking coordinate points, drawing polylines, judging trends, and summarizing results for each parameter in the predicted state, the changing trend of the fuel cell engine can be obtained.

[0070] In summary, by extracting the time stamps of each simulation vector in the predicted state, combining them with the four types of parameter quantization values ​​of the corresponding simulation vectors for textual association description and summarization, the predicted information of the predicted state can be generated.

[0071] In summary, by extracting the quantized values ​​of the four types of parameters corresponding to each time marker from the prediction information and combining them into sub-vectors in a fixed order, and then arranging the sub-vectors in chronological order, the prediction state vector of the prediction state can be obtained.

[0072] In summary, by extracting the first and last sub-vectors from the predicted state vector, calculating the difference in quantized values ​​of the same parameters, summing them, and then averaging them, the comprehensive trend strength value of the fuel cell engine can be obtained.

[0073] In summary, by clarifying the sources of each parameter in the comprehensive trend strength value formula, explaining the significance of the formula in measuring the overall deviation of the predicted state vector from the reference state vector, and illustrating... The relationship between the value and the degree of deviation of the engine state clearly demonstrates the effect of the formula.

[0074] In summary, by determining the direction of performance evolution by judging the sign of the comprehensive trend strength value, determining the degree of evolution by judging its absolute value, and combining the two to form a textual conclusion, the changing trend of fuel cell engines can be obtained.

[0075] S3. Based on the changing trend, perform dynamic trend coupling calibration on the standard parameter sequence to obtain the efficiency evaluation value of the fuel cell engine; In this embodiment of the invention, the step of performing dynamic trend coupling calibration on the standard parameter sequence based on the changing trend to obtain the efficiency evaluation value of the fuel cell engine includes: Based on the aforementioned trend, the degree of influence of key parameters in the standard parameter sequence is determined, and a calibration strategy for the key parameters is obtained. Based on the calibration strategy, the dynamic calibration weights of the key parameters are calculated, wherein the calculation formula for the dynamic calibration weights is as follows: ; In the formula, Indicates the first in the standard parameter sequence Dynamic calibration weights for key parameters, This represents the total number of key parameters in the standard parameter sequence. Indicates the first in the standard parameter sequence The trend strength values ​​corresponding to the key parameters Indicates the first in the standard parameter sequence The trend strength values ​​corresponding to the key parameters This represents the preset trend sensitivity coefficient. Represents an exponential function; Based on the dynamic calibration weights, the parameters of the standard parameter sequence are weighted and fused to obtain the calibration parameters of the fuel cell engine; The calibration parameters are normalized to obtain the efficiency evaluation value of the fuel cell engine.

[0076] The formula for calculating the calibration weight is as follows: ; In the formula, This indicates the calibration parameters of the fuel cell engine. This represents the total number of parameters in the standard parameter sequence. Indicates the first in the standard parameter sequence Dynamic calibration weights for each parameter, Indicates the first in the standard parameter sequence The values ​​of the parameters, This represents the weighted average of the parameters in the standard parameter sequence. This indicates the preset fusion adjustment coefficient.

[0077] Specifically, the entire process revolves around the parameter calibration and efficiency evaluation of fuel cell engines, always taking the changing trend as the core basis, and successively advancing the determination of key parameter calibration strategies, the accurate calculation of dynamic calibration weights, and the scientific acquisition of calibration parameters, ultimately generating a quantifiable evaluation value of engine efficiency. Each step is deeply integrated with the analysis of changing trends and is gradually implemented through clear and explicit calculation logic or operating rules.

[0078] Specifically, based on the changing trends of fuel cell engines, a detailed analysis was conducted on several key parameters in the standard parameter sequence, including reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity. This analysis explored the close correlation between these parameters and the changing trends, determining how fluctuations in each key parameter would affect the trends. If a fluctuation in a certain key parameter caused a significant change in the changing trend, that key parameter was identified as requiring focused calibration.

[0079] Specifically, based on the degree of influence of each key parameter on the trend of change, specific calibration directions and adjustment range requirements are formulated for different key parameters. The calibration directions and adjustment range requirements corresponding to all key parameters are systematically organized and summarized to form a complete set of rules, which is the calibration strategy for key parameters.

[0080] Furthermore, when calculating the dynamic calibration weights of key parameters, on the one hand, the initial weights are assigned to each key parameter in strict accordance with the calibration direction and adjustment range requirements in the calibration strategy. The more urgent the calibration requirement, the higher the initial weight is set.

[0081] Furthermore, the initial weights are dynamically adjusted based on the proximity of the current values ​​of key parameters in the standard parameter sequence to the calibration target values. If the current value is close to the calibration target value, it indicates that the parameter is in a good state, and the weight is appropriately reduced; if the current value is far from the calibration target value, it indicates that the parameter is in a large deviation, and the weight is appropriately increased.

[0082] Furthermore, the dynamic calibration weights are calculated using a formula: first, the exponent of the trend intensity value corresponding to each key parameter is calculated. These trend intensity values ​​are obtained from the trend analysis process. The preset trend sensitivity coefficient is set based on the inherent characteristics of the fuel cell engine or long-term operating experience. The exponential function is calculated according to the exponential operation rules with the natural constant as the base.

[0083] Furthermore, the exponent of the trend strength value corresponding to each key parameter is divided by the sum of the exponents of the trend strength values ​​corresponding to all key parameters, so as to reasonably allocate the weight of subsequent calibration based on the trend strength of the key parameters.

[0084] Furthermore, the more significant the influence of trend strength on the changing trend of a key parameter, the greater its dynamic calibration weight; the greater the trend strength value of a key parameter, the larger the calculated index result, and its dynamic calibration weight will also increase accordingly, and the importance attached to the calibration of this parameter will be increased accordingly. After a series of operations including initial allocation, dynamic adjustment, and formula calculation, the final dynamic calibration weight of each key parameter is finally obtained.

[0085] Furthermore, when calculating the calibration parameters of the fuel cell engine, on the one hand, the value of each key parameter in the standard parameter sequence is accurately obtained, and on the other hand, the dynamic calibration weight corresponding to each key parameter is obtained. The value of each key parameter is multiplied by its corresponding dynamic calibration weight, and then the weighted values ​​of all key parameters are summed.

[0086] Furthermore, the calibration parameters are calculated using formulas. This process consists of two parts: The first part is the sum of the products of each parameter value and its dynamic calibration weight. The parameter values ​​are directly extracted from the standard parameter sequence, and the dynamic calibration weights are derived from the previously calculated dynamic calibration weights. The second part is the fusion adjustment coefficient multiplied by the square root of the weighted sum of squares of the deviations of each parameter value from the weighted average. The fusion adjustment coefficient is preset based on actual needs or industry experience. The weighted average is calculated by multiplying each parameter value by its dynamic calibration weight and then summing the results. The square root operation follows the mathematical rules for square root calculations, taking the square root of the internal sum of squares.

[0087] Furthermore, this calculation method comprehensively considers the impact of parameter weighting and fusion effects, as well as the degree of dispersion, on calibration. The greater the parameter dispersion, the larger the value of the second part, and the larger the final calibration parameters. Conversely, the larger the dynamic calibration weight or parameter value of a certain parameter, the greater the contribution of the first part to that parameter, causing the calibration parameters to tilt in the direction of that parameter's influence. The overall calibration parameters adjust with changes in the dynamic calibration weight, parameter value, and degree of dispersion, thus reflecting more precise calibration requirements. Through these two methods, the calibration parameters for the fuel cell engine are ultimately obtained.

[0088] Furthermore, when obtaining the efficiency evaluation value of the fuel cell engine, it is first clarified that the numerical range of the fuel cell engine efficiency evaluation is 0 to 100, which represents the interval from the lowest level to the highest level of efficiency.

[0089] Furthermore, by analyzing a large amount of historical data or through theoretical derivation, the maximum and minimum values ​​that the calibration parameters may take are found.

[0090] Furthermore, the minimum value is subtracted from the current value of the calibration parameter, and the result is divided by the difference between the maximum and minimum values. Finally, the calculation result is multiplied by 100 to map the calibration parameter to the range of 0 to 100. The resulting value is an evaluation value that can quantify the efficiency of the fuel cell engine.

[0091] In summary, by comprehensively analyzing the correlation between key parameters and their changing trends, and by formulating specific calibration directions and adjustment range requirements for each key parameter, a calibration strategy for the key parameters can be developed.

[0092] In summary, by combining calibration strategies to reasonably allocate and dynamically adjust initial weights, and by calculating the formula based on factors such as trend strength values, the dynamic calibration weights of key parameters can be obtained.

[0093] In summary, the calibration parameters for fuel cell engines can be obtained by weighted summation of key parameter values ​​and dynamic calibration weights, as well as by formula calculations that take into account factors such as parameter dispersion.

[0094] In summary, by determining the numerical range of efficiency assessment and performing scientific mapping calculations on the calibration parameters, the efficiency assessment value of the fuel cell engine can be obtained.

[0095] S4. Based on the historical operating mode data of the fuel cell engine, perform a retrospective decision mapping on the efficiency evaluation value to obtain the control decision scheme of the fuel cell engine; In this embodiment of the invention, the step of performing abductive decision mapping on the efficiency evaluation value based on the historical operating mode data of the fuel cell engine to obtain the control decision scheme of the fuel cell engine includes: Modal classification is performed on the historical operating mode data to obtain the historical operating modes of the fuel cell engine; By matching the efficiency evaluation value with the historical operating mode, the historical cases of the fuel cell engine are obtained; Based on the historical control strategies in the historical cases, the efficiency evaluation value is represented in a coordinated manner to obtain the preliminary control decision of the fuel cell engine; Based on the current constraints of the fuel cell engine, the preliminary control decision is modified to obtain a control decision scheme for the fuel cell engine.

[0096] The step of matching the efficiency evaluation value with the historical operating modes to obtain historical cases of the fuel cell engine includes: Based on the efficiency evaluation value, obtain the real-time operating parameter vector of the fuel cell engine and the historical parameter vector of the historical operating mode; Calculate the similarity between the real-time running parameter vector and the historical parameter vector, wherein the formula for calculating the similarity is as follows: ; In the formula, This indicates that the real-time running parameter vector is related to the first... The similarity of historical operating patterns This represents the dimension of the parameter vector. The first element of the real-time running parameter vector represents the... One portion, Indicating the first in the historical operation mode The th historical parameter vector of the th One component; Historical cases with similarity higher than a preset matching threshold are used as historical cases of the fuel cell engine.

[0097] Specifically, the entire process revolves around the generation of control decision schemes for fuel cell engines. First, historical operating mode data is classified to obtain historical operating modes. Then, the efficiency evaluation value is matched with the historical operating modes to obtain historical cases.

[0098] Specifically, the efficiency assessment value is coordinated and represented based on historical control strategies in historical cases to obtain a preliminary control decision. Finally, the preliminary control decision is modified for feasibility based on current constraints to obtain a control decision scheme.

[0099] Furthermore, when performing modal classification, we first sort out the historical operating modal data of the fuel cell engine. This data includes the reaction temperature, hydrogen inlet pressure, output current, electrolyte membrane humidity and corresponding efficiency data of different operating periods in the past. Then, we determine the two core dimensions of modal classification: the power output level divided according to the power range corresponding to the output current, and the chemical reaction stability divided according to the fluctuation range of the reaction temperature.

[0100] Furthermore, each set of data in the historical operating mode data is examined one by one. The corresponding power is calculated based on the output current and classified into the corresponding power range. At the same time, the data is classified into the corresponding stability level according to the reaction temperature fluctuation range. All data under the same power range and stability level combination are grouped into one category. Each category corresponds to one operating characteristic. The set of all categories with clear operating characteristics is the historical operating mode of the fuel cell engine.

[0101] Furthermore, when performing similarity matching, the average efficiency assessment value corresponding to each historical operating mode is first extracted. This value is calculated by averaging the efficiency values ​​of all historical data under this mode. Then, the current fuel cell engine efficiency assessment value is compared with the average efficiency assessment value of all historical operating modes, and the difference between the current efficiency assessment value and each average efficiency assessment value is calculated. The historical operating mode with the smallest difference is then identified.

[0102] Furthermore, examine the historical fluctuation range of key parameters such as reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity under this historical operating mode to see if they are consistent with the fluctuation range of key parameters in the current standard parameter sequence. If they are consistent, use all historical operating records and corresponding control measures under this historical operating mode as a reference case. This reference case is the historical case of the fuel cell engine.

[0103] Furthermore, when implementing coordinated representation, historical control strategies are extracted from historical cases. These strategies include the adjustment threshold of reaction temperature, the control range of hydrogen inlet pressure, the adjustment step size of output current, and the maintenance method of electrolyte membrane humidity.

[0104] Furthermore, the current efficiency assessment value is compared with the control strategy triggering conditions under the corresponding efficiency level in historical cases. If the current efficiency assessment value is lower than the average efficiency under the same mode in historical cases, the adjustment method to improve efficiency in historical control strategies is referred to to clarify the parameters that need to be adjusted, the direction of adjustment, and the target range after adjustment. These adjustment contents are organized into clear operation instructions, and the resulting set of operation instructions is the initial control decision of the fuel cell engine.

[0105] Furthermore, when making feasibility modifications, the current constraints of the fuel cell engine should be clarified first, including the maximum gas supply of the hydrogen supply system, the maximum heat dissipation capacity of the cooling system, the maximum current that the output load can withstand, and the maximum humidification capacity of the humidification device.

[0106] Furthermore, each operation command in the preliminary control decision is examined one by one to determine whether the parameter adjustment corresponding to the command is within the allowable range of the current constraints. If the parameter value required by a command exceeds the upper limit of the constraint or is lower than the lower limit of the constraint, the target value of the parameter adjustment is corrected to the limit value allowed by the constraints. After all operation commands are corrected, the complete set of operation commands that conforms to the current constraints is the control decision scheme of the fuel cell engine.

[0107] Specifically, the entire process revolves around matching historical cases of fuel cell engines. First, the real-time operating parameter vector and the historical parameter vector are determined. Then, the similarity between the two is calculated in two ways. Finally, historical cases are selected by combining the preset matching threshold.

[0108] Furthermore, when determining the parameter vector, the corresponding real-time operating time is determined based on the efficiency evaluation value. The values ​​of four key parameters—reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity—are extracted from the standard parameter sequence at that time and arranged in a fixed order of "reaction temperature value, hydrogen inlet pressure value, output current value, and electrolyte membrane humidity value" to form the real-time operating parameter vector of the fuel cell engine.

[0109] Furthermore, in the historical operating modes, the historical operating mode with the closest average efficiency assessment value to the current efficiency assessment value is identified. From all historical data under this historical operating mode, the reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity values ​​of each set of data are extracted. The average value of the corresponding parameters for each set of data is calculated, and these average values ​​are arranged in the same order as the real-time operating parameter vector to form the historical parameter vector of the historical operating mode.

[0110] Furthermore, when calculating similarity, on the one hand, the parameter values ​​at corresponding positions in the real-time operating parameter vector and the historical parameter vector are compared one by one, and the difference between each pair of corresponding parameter values ​​is calculated, such as the difference in reaction temperature, the difference in hydrogen inlet pressure, the difference in output current, and the difference in electrolyte membrane humidity. The absolute values ​​of these four parameter differences are added together to obtain the sum of the absolute values ​​of the differences. Then, the sum of the absolute values ​​of the differences is divided by the number of parameters to obtain the average difference. The smaller the average difference, the higher the similarity between the real-time operating parameter vector and the historical parameter vector. The magnitude of the average difference quantitatively reflects the similarity between the two.

[0111] Furthermore, on the other hand, similarity is calculated using the formula: the first similarity of the real-time running parameter vector. The first component is derived from the real-time operating parameter vector, which extracts key parameter values ​​such as reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity at corresponding moments from the standard parameter sequence of the fuel cell engine, and arranges them in a fixed order. The value at each position.

[0112] Furthermore, in the historical operating mode, the first The th historical parameter vector of the th The first component is derived from historical data extracted from historical operating modes, including reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity values ​​for each set of data. The average value of each set of parameters is calculated, and these values ​​are arranged in the same order to form a historical parameter vector. Then, the first component is selected... In the historical parameter vector, the first The value at each position.

[0113] Furthermore, the dimension of the parameter vector is determined by the number of key parameters contained in the real-time running parameter vector and the historical parameter vector. If it contains four key parameters, the dimension is four.

[0114] Furthermore, during the calculation, each component of the real-time running parameter vector is first multiplied by the corresponding component of the historical parameter vector, and all products are added together to obtain the numerator; then, the sum of squares of each component of the real-time running parameter vector is calculated and the square root is taken to obtain its modulus, the sum of squares of each component of the historical parameter vector is calculated and the square root is taken to obtain its modulus, and the two modulus are multiplied together to obtain the denominator.

[0115] Furthermore, dividing the numerator by the denominator yields a ratio that reflects the angle between the two vectors. The closer the ratio is to 1, the more aligned the directions of the two vectors are. The real-time running parameter vector and the... The higher the similarity between two historical operating patterns, the better.

[0116] Furthermore, when the real-time running parameter vector is compared with the first... When the corresponding component values ​​of two historical parameter vectors are closer and their trends are more consistent, the product of each component is larger, the sum of the numerators is also larger, and the product of the magnitudes of the two vectors is closer to the size of the numerator, the ratio is closer to 1, and the similarity is higher. Conversely, when the corresponding component values ​​are different and their trends are inconsistent, the sum of the numerators is smaller, the product of the magnitudes of the two vectors is larger than the numerator, the ratio is closer to 0, and the similarity is lower.

[0117] Furthermore, when determining the preset matching threshold, the average difference between the real-time running parameter vector and the historical parameter vector in multiple successful matching cases in history is analyzed, and the maximum value of the average difference in these successful cases is selected as the preset matching threshold.

[0118] Furthermore, the average difference between the currently calculated real-time running parameter vector and the historical parameter vector is compared with a preset matching threshold.

[0119] Furthermore, if the average difference is less than or equal to the preset matching threshold, it means that the similarity between the two is higher than the preset matching threshold. At this time, all historical operation records, corresponding control measures and operation effect data corresponding to the historical operation mode to which the historical parameter vector belongs are compiled into reference cases. If the average difference is greater than the preset matching threshold, the historical operation mode to which the historical parameter vector belongs is excluded, and the historical parameter vectors of other historical operation modes are compared until a historical operation mode with a similarity higher than the preset matching threshold is found. The reference cases finally compiled are the historical cases of fuel cell engines.

[0120] In summary, by analyzing historical operating modal data, determining classification dimensions, and categorizing the data, we can obtain the historical operating modes of fuel cell engines.

[0121] In summary, by extracting the average efficiency assessment value of historical operating modes, comparing the difference between the current efficiency and the average efficiency, and checking the range of parameter fluctuations, historical cases of fuel cell engines can be obtained.

[0122] In summary, by extracting historical control strategies, comparing current efficiency with historical triggering conditions, and organizing adjustment instructions, preliminary control decisions for fuel cell engines can be obtained.

[0123] In summary, by clarifying the current constraints and examining and revising the operational instructions in the initial control decision, a control decision scheme for the fuel cell engine can be obtained.

[0124] In summary, by determining the real-time running time and extracting key parameter values ​​and arranging them in a fixed order, and by extracting the average values ​​of parameters from historical running modes and arranging them in the same order, it is possible to obtain the real-time running parameter vector and the historical parameter vector.

[0125] In summary, the similarity between real-time running parameter vectors and historical parameter vectors can be measured by using two methods: difference averaging and vector formula calculation.

[0126] In summary, by analyzing historical successful cases to determine the preset matching threshold, and then comparing the current average difference with the threshold and filtering and organizing them, we can obtain historical cases of fuel cell engines.

[0127] S5. Encode and map the control decision scheme to obtain the control decision command of the fuel cell engine.

[0128] The step of encoding and mapping the control decision scheme to obtain the control decision command for the fuel cell engine includes: The control decision scheme is analyzed in a structured manner to obtain the decision elements of the fuel cell engine; Based on the aforementioned decision elements, the control decision scheme is matched and mapped with the controller of the fuel cell engine to obtain the control logic unit of the fuel cell engine; According to the preset communication protocol, the control logic units are organized and packaged in sequence to generate the original control commands for the fuel cell engine; The original control commands are logically verified to obtain the control decision commands for the fuel cell engine.

[0129] Specifically, the entire process revolves around the generation of control decision commands for the fuel cell engine. First, the control decision scheme is analyzed in a structured manner to obtain the decision elements.

[0130] Specifically, control logic units are obtained by matching and mapping control decision schemes with controllers based on decision elements.

[0131] Specifically, the control logic units are organized and encapsulated according to a preset communication protocol to generate original control instructions, and finally the logical integrity of the original control instructions is verified to obtain control decision instructions.

[0132] Furthermore, during the structured analysis, the contents of the control decision scheme are first sorted out, including the adjustment target values ​​of several key parameters such as reaction temperature, hydrogen inlet pressure, output current, and electrolyte membrane humidity, the execution sequence of each parameter adjustment, the components responsible for performing the adjustment operation, and the operation mode of each component.

[0133] Furthermore, these contents are categorized and broken down into "parameter adjustment target category, execution sequence category, execution component category, and operation mode category". Specific information items are organized under each category. The set of all categories and corresponding information items constitutes the decision elements of the fuel cell engine.

[0134] Furthermore, when performing matching mapping, the functional module division of the fuel cell engine controller should be clarified first, including the temperature control module responsible for processing reaction temperature adjustment commands, the pressure control module responsible for processing hydrogen inlet pressure adjustment commands, the current control module responsible for processing output current adjustment commands, and the humidity control module responsible for processing electrolyte membrane humidity adjustment commands. At the same time, the input interface requirements of each module should be clarified.

[0135] Furthermore, the "parameter adjustment target" information in the decision elements is mapped to the input target parameters of each control module, the "execution timing" information is mapped to the start-up order of each module, the "execution component" information is mapped to the hardware execution unit associated with each module, and the "operation mode" information is mapped to the specific control parameters of each module. These correspondences are integrated into a logical combination that can be recognized by the controller. The resulting logical combination is the control logic unit of the fuel cell engine.

[0136] Furthermore, when organizing and encapsulating, the preset communication protocol requirements are first determined, including the structure of the instructions, the length of each segment, and the encoding format of the data.

[0137] Furthermore, the information of each part in the control logic unit is filled into the corresponding fields according to the structure of the communication protocol: the instruction identifier field is filled with the exclusive identifier specified by the protocol, the module address field is filled with the address code of each control module, the control logic field is filled with the specific control parameters of each module, the timing field is filled with the execution order of each module, and the verification field generates verification information according to the calculation method specified by the protocol. All fields are concatenated in the order specified by the protocol to form the complete instruction, which is the original control instruction of the fuel cell engine.

[0138] Furthermore, when implementing logical integrity verification, first establish logical integrity verification standards, including verification content: check whether the original control instructions contain the address codes of all control modules, whether the control parameters of each module are complete, whether there are any conflicts in the execution timing, and whether the verification segment data matches the previous few segments of data.

[0139] Furthermore, the original control instructions are checked segment by segment according to the standard: first, the module address segment is checked to confirm that the address codes of all control modules are included and there are no duplicates; then the control logic segment is checked to confirm that the control parameters of each module are not missing; then the timing segment is checked to confirm that there are no conflicts in the execution order.

[0140] Furthermore, the verification segment is verified by recalculating the verification values ​​of the previous data segments and comparing them with the verification segment data in the instruction to confirm that the two are consistent. If all check items meet the standards and no logical omissions or conflicts are found, the original control instruction is the control decision instruction of the fuel cell engine that has passed the verification.

[0141] In summary, by sorting out the content of the control decision-making scheme and breaking down and organizing the information items by category, the decision-making elements of the fuel cell engine can be obtained.

[0142] In summary, by clearly defining the functional modules and input requirements of the controller, and integrating the decision elements and modules into a logical combination, the control logic unit of the fuel cell engine can be obtained.

[0143] In summary, by filling in and concatenating fields according to the structure and requirements of a preset communication protocol, the raw control commands for a fuel cell engine can be generated.

[0144] In summary, by establishing verification standards and checking the original control commands segment by segment, the control decision commands for the fuel cell engine can be obtained.

[0145] like Figure 2 The diagram shown is a functional block diagram of a fuel cell engine control system provided in an embodiment of the present invention.

[0146] The fuel cell engine control system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the fuel cell engine control system 100 may include a data preprocessing module 101, a situation prediction module 102, a performance evaluation and calibration module 103, an intelligent decision-making module 104, and an instruction generation module 105. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0147] In this embodiment, the functions of each module / unit are as follows: The data preprocessing module 101 is used to perform noise cleaning on the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine. The situation prediction module 102 is used to perform situation simulation on the operating state of the fuel cell engine based on the standard parameter sequence, and obtain the changing trend of the fuel cell engine. The performance evaluation and calibration module 103 is used to perform dynamic trend coupling calibration on the standard parameter sequence based on the changing trend to obtain the efficiency evaluation value of the fuel cell engine. The intelligent decision-making module 104 is used to perform a retrospective decision mapping on the efficiency evaluation value based on the historical operating mode data of the fuel cell engine, so as to obtain the control decision scheme of the fuel cell engine. The instruction generation module 105 is used to encode and map the control decision scheme to obtain the control decision instructions of the fuel cell engine.

[0148] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0149] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0150] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0151] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0152] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A fuel cell engine control method, characterized in that, The method includes: S1. Noise cleaning is performed on the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine; S2. Based on the standard parameter sequence, perform situational simulation on the operating state of the fuel cell engine to obtain the changing trend of the fuel cell engine; S3. Based on the changing trend, perform dynamic trend coupling calibration on the standard parameter sequence to obtain the efficiency evaluation value of the fuel cell engine; S4. Based on the historical operating mode data of the fuel cell engine, perform a retrospective decision mapping on the efficiency evaluation value to obtain the control decision scheme of the fuel cell engine; S5. Encode and map the control decision scheme to obtain the control decision command of the fuel cell engine.

2. The fuel cell engine control method as described in claim 1, characterized in that, The process of noise-cleaning the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine includes: Obtain the real-time parameter sequence of the fuel cell engine; Anomalies in the real-time parameter sequence are detected to obtain the noise point identification sequence of the fuel cell engine; Based on the noise point identification sequence, the real-time parameter sequence is repaired to obtain the repaired parameter sequence of the fuel cell engine; By eliminating the high-frequency fluctuation parameters in the repair parameter sequence, the standard parameter sequence of the fuel cell engine is obtained.

3. The fuel cell engine control method as described in claim 1, characterized in that, The process of simulating the operating state of the fuel cell engine based on the standard parameter sequence to obtain the changing trend of the fuel cell engine includes: Key parameters characterizing the operating state of the fuel cell engine are extracted from the standard parameter sequence to obtain the operating state parameters of the fuel cell engine; The operating status parameters are quantified by status indicators to obtain the real-time status vector of the operating status parameters; Based on a preset dynamic evolution strategy, the real-time state vector is forward simulated to obtain the predicted state of the fuel cell engine. The predicted state is fitted with a trend to obtain the changing trend of the fuel cell engine.

4. The fuel cell engine control method as described in claim 3, characterized in that, The step of performing trend fitting on the predicted state to obtain the changing trend of the fuel cell engine includes: Analyze the predicted state to generate prediction information for the predicted state; The predicted information is reconstructed in dimensions to obtain the predicted state vector of the predicted state; The predicted state vector is trend-quantized to obtain the comprehensive trend intensity value of the fuel cell engine, wherein the comprehensive trend intensity value is calculated using the following formula: ； In the formula, This represents the overall trend strength value. This indicates the number of predicted state vectors. Indicates the fuel cell engine's first A predicted state vector, This represents the preset reference state vector. Indicates the fuel cell engine's first Weight factors for each predicted state vector, The norm representing the difference between two vectors; Based on the sign and magnitude of the comprehensive trend strength value, the evolution direction and severity of the fuel cell engine performance are determined, and the change trend of the fuel cell engine is generated.

5. The fuel cell engine control method as described in claim 1, characterized in that, The process of performing dynamic trend coupling calibration on the standard parameter sequence based on the changing trend to obtain the efficiency evaluation value of the fuel cell engine includes: Based on the aforementioned trend, the degree of influence of key parameters in the standard parameter sequence is determined, and a calibration strategy for the key parameters is obtained. Based on the calibration strategy, the dynamic calibration weights of the key parameters are calculated, wherein the calculation formula for the dynamic calibration weights is as follows: ； In the formula, Indicates the first in the standard parameter sequence Dynamic calibration weights for key parameters, This represents the total number of key parameters in the standard parameter sequence. Indicates the first in the standard parameter sequence The trend strength values ​​corresponding to the key parameters Indicates the first in the standard parameter sequence The trend strength values ​​corresponding to the key parameters This represents the preset trend sensitivity coefficient. Represents an exponential function; Based on the dynamic calibration weights, the parameters of the standard parameter sequence are weighted and fused to obtain the calibration parameters of the fuel cell engine; The calibration parameters are normalized to obtain the efficiency evaluation value of the fuel cell engine.

6. The fuel cell engine control method as described in claim 5, characterized in that, The formula for calculating the calibration weight is as follows: ； In the formula, This indicates the calibration parameters of the fuel cell engine. This represents the total number of parameters in the standard parameter sequence. Indicates the first in the standard parameter sequence Dynamic calibration weights for each parameter, Indicates the first in the standard parameter sequence The values ​​of the parameters, This represents the weighted average of the parameters in the standard parameter sequence. This indicates the preset fusion adjustment coefficient.

7. The fuel cell engine control method as described in claim 1, characterized in that, The process of performing abductive decision mapping on the efficiency evaluation value based on the historical operating mode data of the fuel cell engine to obtain the control decision scheme for the fuel cell engine includes: Modal classification is performed on the historical operating mode data to obtain the historical operating modes of the fuel cell engine; By matching the efficiency evaluation value with the historical operating mode, the historical cases of the fuel cell engine are obtained; Based on the historical control strategies in the historical cases, the efficiency evaluation value is represented in a coordinated manner to obtain the preliminary control decision of the fuel cell engine; Based on the current constraints of the fuel cell engine, the preliminary control decision is modified to obtain a control decision scheme for the fuel cell engine.

8. The fuel cell engine control method as described in claim 7, characterized in that, The step of matching the efficiency evaluation value with the historical operating modes to obtain historical cases of the fuel cell engine includes: Based on the efficiency evaluation value, obtain the real-time operating parameter vector of the fuel cell engine and the historical parameter vector of the historical operating mode; Calculate the similarity between the real-time running parameter vector and the historical parameter vector, wherein the formula for calculating the similarity is as follows: ； In the formula, This indicates that the real-time running parameter vector is related to the first... The similarity of historical operating patterns This represents the dimension of the parameter vector. The first element of the real-time running parameter vector represents the... One portion, Indicating the first in the historical operation mode The th historical parameter vector of the th One component; Historical cases with similarity higher than a preset matching threshold are used as historical cases of the fuel cell engine.

9. The fuel cell engine control method as described in claim 1, characterized in that, The step of encoding and mapping the control decision scheme to obtain the control decision command for the fuel cell engine includes: The control decision scheme is analyzed in a structured manner to obtain the decision elements of the fuel cell engine; Based on the aforementioned decision elements, the control decision scheme is matched and mapped with the controller of the fuel cell engine to obtain the control logic unit of the fuel cell engine; According to the preset communication protocol, the control logic units are organized and packaged in sequence to generate the original control commands for the fuel cell engine; The original control commands are logically verified to obtain the control decision commands for the fuel cell engine.

10. A fuel cell engine control system, characterized in that, The system includes: The data preprocessing module is used to clean the noise from the real-time parameter sequence of the fuel cell engine to obtain the standard parameter sequence of the fuel cell engine. The situation prediction module is used to simulate the operating state of the fuel cell engine based on the standard parameter sequence, and obtain the changing trend of the fuel cell engine. The performance evaluation and calibration module is used to perform dynamic trend coupling calibration on the standard parameter sequence based on the changing trend to obtain the efficiency evaluation value of the fuel cell engine. The intelligent decision-making module is used to perform a retrospective decision mapping on the efficiency evaluation value based on the historical operating mode data of the fuel cell engine, and obtain the control decision scheme of the fuel cell engine. The instruction generation module is used to encode and map the control decision scheme to obtain the control decision instructions for the fuel cell engine.

Images