Fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data

Abstract

The invention relates to the technical field of fuel cell fault diagnosis, in particular to a fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data, which comprises the following steps: S1, setting a voltage dynamic list, a current dynamic list, a hydrogen dynamic list, a result dynamic list and a rolling window for judging hydrogen leakage; s2, collecting voltage data of the fuel cell, current data of the load and hydrogen concentration data in the environment in real time; s3, counting the number of data in the voltage dynamic list, if the number of the data reaches the preset capacity of the rolling window, executing the step S4, and otherwise, returning to the step S2; s4, the current state is judged according to the current dynamic list in the rolling window, a judgment result is given through the voltage dynamic list or / and the hydrogen dynamic list in the rolling window, and the judgment result is stored in a result dynamic list; s5, judging whether to trigger a fault alarm or not according to the number of the preset circulating windows and a preset alarm threshold value; according to the invention, multi-dimensional monitoring is carried out on hydrogen leakage based on multi-dimensional sensing data, the accuracy is high, and the reliability is high.

Description

Technical Field

[0001] The present invention relates to the technical field of fuel cell fault diagnosis, and particularly relates to a method for diagnosing hydrogen leakage of a fuel cell based on multi-dimensional sensing data. Background Art

[0002] The reasons for hydrogen leakage in fuel cells involve multiple aspects, including seal aging or damage, improper assembly, component failure, external impact or damage, material quality problems, improper operation, and long-term operation and wear. Among them, seal aging, component failure, external impact or damage, and material quality problems can be avoided by performing an integrity check on the stack before each startup; while slight improper assembly often occurs after startup and is only discovered when it develops into serious hydrogen leakage.

[0003] Currently, hydrogen leakage detection methods mainly rely on a single sensor (such as a hydrogen concentration sensor), and only judge whether hydrogen leakage occurs based on hydrogen concentration data, which has obvious limitations in detection accuracy and real-time performance and is difficult to meet the high requirements of modern industry for hydrogen leakage detection. At the same time, the detection accuracy is also easily affected by factors such as slow sensor response, insufficient sensitivity, or environmental interference. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: to overcome the deficiencies in the prior art and provide a method for diagnosing hydrogen leakage of a fuel cell that realizes multi-dimensional monitoring of hydrogen leakage based on multi-dimensional sensing data, has high accuracy and strong reliability.

[0005] The technical solution adopted by the present invention to solve its technical problems is: a method for diagnosing hydrogen leakage of a fuel cell based on multi-dimensional sensing data, including the following steps:

[0006] S1. Set a voltage dynamic list, a current dynamic list, a hydrogen dynamic list, a result dynamic list, and a rolling window for judging hydrogen leakage;

[0007] S2. Real-time collect the voltage data of the fuel cell, the current data of the load, and the hydrogen concentration data in the environment, where the voltage data is stored in the voltage dynamic list, the current data is stored in the current dynamic list, and the hydrogen concentration data is stored in the hydrogen dynamic list;

[0008] S3. Count the number of data in the voltage dynamic list. If the number of data in the voltage dynamic list reaches the preset capacity of the rolling window, then execute step S4, otherwise return to step S2;

[0009] S4. Determine the current status based on the dynamic current list within the rolling window. If it is determined that the current is in a constant state, the judgment result is given through the dynamic voltage list within the rolling window; otherwise, the judgment result is given through the dynamic voltage list and the dynamic hydrogen list within the rolling window. The judgment result is stored in the dynamic result list;

[0010] S5. Determine whether to trigger a fault alarm based on the preset number of cyclic windows and the preset alarm threshold.

[0011] Furthermore, the method for determining the current status within the rolling window in step S4 is as follows:

[0012] If the standard deviation of the current data in the dynamic current list ≤ the preset threshold A1, and the difference between the maximum current and the minimum current in the dynamic current list ≤ the preset threshold A2, then it is determined that the current is in a constant state; otherwise, it is determined that the current is in a non-constant state.

[0013] Furthermore, the method for determining hydrogen leakage when the current is in a constant state in step S4 is as follows:

[0014] If the average voltage in the dynamic voltage list ≥ the preset threshold V1, then it is determined that there is no hydrogen leakage, and the judgment result is marked as 0;

[0015] If the preset threshold V2 ≤ the average voltage in the dynamic voltage list < the preset threshold V1, then it is determined that there is a slight leakage, and the judgment result is marked as 1.

[0016] If the preset threshold V3 ≤ the average voltage in the dynamic voltage list < the preset threshold V2, then it is determined that there is a moderate leakage, and the judgment result is marked as 2;

[0017] If the average voltage in the dynamic voltage list < the preset threshold V3, then it is determined that there is a severe leakage, and the judgment result is marked as 3.

[0018] Furthermore, the method for determining hydrogen leakage when the current is in a non-constant state in step S4 is as follows:

[0019] If the average voltage in the dynamic voltage list ≥ the preset threshold V1, then it is determined that there is no hydrogen leakage, and the judgment result is marked as 0;

[0020] If the average voltage in the dynamic voltage list < the preset threshold V1, then determine whether hydrogen is leaked according to the magnitude relationship between the maximum hydrogen concentration value in the dynamic hydrogen list and the preset threshold P1: If the maximum hydrogen concentration in the dynamic hydrogen list ≥ the preset threshold P1, then it is determined that hydrogen is leaked, and the judgment result is marked as 1 or 2 or 3; otherwise, it is determined that there is no hydrogen leakage, and the judgment result is marked as 0, and at the same time, sensor calibration is triggered.

[0021] Furthermore, the sensor calibration method is as follows:

[0022] If the data deviation between the voltage sensor and the spare sensor is ≥ the preset threshold E1, it is determined that the voltage sensor has failed. After calibrating the voltage sensor, re-initialize the voltage dynamic list and continuously monitor three rolling windows to verify the repair result.

[0023] Further, the trigger conditions for the fault alarm in step S5 are as follows:

[0024] When the number of data in the result dynamic list is equal to the preset number of loop windows, it is determined whether to trigger a fault alarm based on the size of the number of non-zero judgment results in the result dynamic list and the preset alarm threshold: If the number of non-zero judgment results in the result dynamic list is ≥ the preset alarm threshold, trigger a fault alarm; otherwise, do not trigger a fault alarm.

[0025] Further, it also includes step S6, and the content of step S6 is as follows:

[0026] After the judgment of the fault alarm trigger is completed, remove the first value in the result dynamic list, and clear the voltage dynamic list, current dynamic list, and hydrogen dynamic list within the rolling window corresponding to the first value, and then return to step S2.

[0027] The beneficial effects of the present invention are:

[0028] (1) The present invention realizes multi-dimensional monitoring of hydrogen leakage through voltage data, current data, and hydrogen concentration data. Compared with single-dimensional monitoring relying only on hydrogen concentration data, the detection accuracy is higher and the reliability is stronger. Furthermore, it can timely and accurately detect hydrogen leakage, thereby avoiding energy waste and safety accidents.

[0029] (2) The present invention adopts different judgment rules under different current states. Compared with adopting a unified judgment rule, it is more reasonable and refined, and further improves the detection accuracy and reliability.

[0030] (3) In the scenario where the current is non-constant, the present invention combines hydrogen concentration data for auxiliary judgment, further improving the detection accuracy and reliability; at the same time, a sensor calibration process is set up to timely detect sensor failures, ensure the accuracy and reliability of data collection, and avoid false alarms and missed alarms. Description of the Drawings

[0031] The present invention will be further described below with reference to the drawings and embodiments.

[0032] Figure 1 is the flowchart of the present invention. Detailed Embodiments

[0033] The present invention will now be further described in conjunction with the accompanying drawings and preferred embodiments. These drawings are all simplified schematic diagrams, only schematically showing the basic structure of the present invention, so they only show the components related to the present invention.

[0034] As Figure 1 shown, a fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data includes the following steps:

[0035] S1. Set up a voltage dynamic list, a current dynamic list, a hydrogen dynamic list, a result dynamic list, and a rolling window for judging hydrogen leakage. Specifically, the capacity of the rolling window is preset in advance.

[0036] S2. Real-time collect the voltage data of the fuel cell, the current data of the load, and the hydrogen concentration data in the environment. Among them, the voltage data is synchronously stored in the voltage dynamic list according to the time stamp, the current data is synchronously stored in the current dynamic list according to the time stamp, and the hydrogen concentration data is synchronously stored in the hydrogen dynamic list according to the time stamp. Specifically, the voltage data is collected by a voltage sensor, the current data is collected by a current sensor, and the hydrogen concentration data is collected by a hydrogen sensor.

[0037] S3. Count the number of data in the voltage dynamic list. If the number of data in the voltage dynamic list reaches the preset capacity of the rolling window, then execute step S4, otherwise return to step S2.

[0038] S4. Judge the current state according to the current dynamic list within the rolling window. If it is determined that the current is in a constant state, then give a judgment result through the voltage dynamic list within the rolling window, otherwise give a judgment result through the voltage dynamic list and the hydrogen dynamic list within the rolling window; the judgment result is stored in the result dynamic list.

[0039] S5. Judge whether to trigger a fault alarm according to the preset number of cycle windows and the preset alarm threshold.

[0040] Through the voltage data, current data and hydrogen concentration data, multi-dimensional monitoring of hydrogen leakage is realized. Compared with the single-dimensional monitoring relying only on hydrogen concentration data, the detection accuracy is higher and the reliability is stronger. Furthermore, hydrogen leakage can be detected in a timely and accurate manner, thus avoiding energy waste and safety accidents. At the same time, through the preset number of cycle windows and the preset alarm threshold, it is judged whether to trigger a fault alarm according to the comprehensive comparison of all judgment results within a period of time, further improving the reliability.

[0041] The method for judging the current state within the rolling window in step S4 is as follows:

[0042] If the standard deviation of the current data in the current dynamic list ≤ the preset threshold A1, and the difference between the maximum current and the minimum current in the current dynamic list ≤ the preset threshold A2, it is determined that the current is in a constant state; otherwise, it is determined that the current is in a non-constant state. Specifically, the preset threshold A1 is 2% of the rated current, and the preset threshold A2 is 5% of the rated current.

[0043] By determining the current state within the rolling window and adopting different determination rules under different current states, compared with adopting a unified determination rule, it is more reasonable and refined, further improving the detection accuracy and reliability. By identifying the stability of the current data, a reliable basis is provided for subsequent determination of voltage dips and leaks.

[0044] The method for determining hydrogen leakage when the current is in a constant state in step S4 is as follows:

[0045] If the average voltage in the voltage dynamic list ≥ the preset threshold V1, it is determined that there is no hydrogen leakage, and the judgment result is marked as 0;

[0046] If the preset threshold V2 ≤ the average voltage in the voltage dynamic list < the preset threshold V1, it is determined that there is a mild leakage, and the judgment result is marked as 1.

[0047] If the preset threshold V3 ≤ the average voltage in the voltage dynamic list < the preset threshold V2, it is determined that there is a moderate leakage, and the judgment result is marked as 2;

[0048] If the average voltage in the voltage dynamic list < the preset threshold V3, it is determined that there is a severe leakage, and the judgment result is marked as 3.

[0049] Specifically, the preset threshold V1 is 95% of the rated voltage, the preset threshold V2 is 90% of the rated voltage, and the preset threshold V3 is 80% of the rated voltage.

[0050] The method for determining hydrogen leakage when the current is in a non-constant state in step S4 is as follows:

[0051] If the average voltage in the voltage dynamic list ≥ the preset threshold V1, it is determined that there is no hydrogen leakage, and the judgment result is marked as 0;

[0052] If the average voltage in the voltage dynamic list < the preset threshold V1, it is determined whether hydrogen leaks according to the magnitude relationship between the maximum hydrogen concentration value in the hydrogen dynamic list and the preset threshold P1: If the maximum hydrogen concentration in the hydrogen dynamic list ≥ the preset threshold P1, it is determined that hydrogen leaks, and the judgment result is marked as 1 or 2 or 3; otherwise, it is determined that there is no hydrogen leakage, and the judgment result is marked as 0, and the sensor calibration is triggered simultaneously. Specifically, the preset threshold P1 is 1000 ppm.

[0053] Under the non-constant current state, auxiliary determination is carried out by combining hydrogen concentration data, further improving the detection accuracy and reliability. At the same time, a sensor calibration process is set up to timely detect sensor failures, ensure the accuracy and reliability of data collection, and avoid false alarms and missed alarms.

[0054] The sensor calibration method is as follows:

[0055] If the data deviation between the voltage sensor and the spare sensor ≥ the preset threshold E1, it is determined that the voltage sensor has failed. After calibrating the voltage sensor, re-initialize the voltage dynamic list and continuously monitor three rolling windows to verify the repair result. Specifically, the preset threshold E1 is 5%.

[0056] By comparing the data of the voltage sensor and the spare sensor, it is judged whether the voltage sensor has a failure, ensuring the accuracy and reliability of data collection.

[0057] The trigger conditions for the fault alarm in step S5 are as follows:

[0058] When the number of data in the result dynamic list is equal to the preset number of loop windows, it is determined whether to trigger a fault alarm according to the size of the number of non-zero judgment results in the result dynamic list and the preset alarm threshold: if the number of non-zero judgment results in the result dynamic list ≥ the preset alarm threshold, trigger a fault alarm; otherwise, do not trigger a fault alarm.

[0059] This application also includes step S6, and the content of step S6 is as follows:

[0060] After the judgment of the trigger of the fault alarm is completed, the first value in the result dynamic list is removed, and the voltage dynamic list, current dynamic list, and hydrogen dynamic list in the rolling window corresponding to the first value are cleared, and then return to step S2.

[0061] This application not only considers the abnormal voltage situation caused by sensor failures, but also conducts a detailed decomposition detection on voltage data and hydrogen concentration data. At the same time, a complete hydrogen leakage diagnosis is realized by integrating different current states, with higher accuracy and stronger reliability.

[0062] The above embodiments are only used to illustrate the technical concept and characteristics of the present invention, and their purpose is to enable those familiar with this technology to understand the content of the present invention and implement it. It cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit and essence of the present invention should be covered within the protection scope of the present invention.

Claims

1. A fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data, characterized in that It includes the following steps: S1. Set up a voltage dynamic list, a current dynamic list, a hydrogen dynamic list, a result dynamic list, and a rolling window for judging hydrogen leakage; S2. Collect the voltage data of the fuel cell, the current data of the load, and the hydrogen concentration data in the environment in real time, where the voltage data is stored in the voltage dynamic list, the current data is stored in the current dynamic list, and the hydrogen concentration data is stored in the hydrogen dynamic list; S3. Count the number of data in the voltage dynamic list. If the number of data in the voltage dynamic list reaches the preset capacity of the rolling window, then execute step S4; otherwise, return to step S2; S4. Judge the current state according to the current dynamic list within the rolling window. If it is determined that the current is in a constant state, the judgment result is given through the voltage dynamic list within the rolling window; otherwise, the judgment result is given through the voltage dynamic list and the hydrogen dynamic list within the rolling window; the judgment result is stored in the result dynamic list; S5. Judge whether to trigger a fault alarm according to the preset number of cycle windows and the preset alarm threshold.

2. The fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data according to claim 1, wherein The method for judging the current state within the rolling window in step S4 is as follows: If the standard deviation of the current data in the current dynamic list ≤ the preset threshold A1, and the difference between the maximum current value and the minimum current value in the current dynamic list ≤ the preset threshold A2, then it is determined that the current is in a constant state; otherwise, it is determined that the current is in a non-constant state.

3. The fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data according to claim 1, wherein, The method for judging hydrogen leakage when the current is in a constant state in step S4 is as follows: If the average voltage in the voltage dynamic list ≥ the preset threshold V1, then it is determined that there is no hydrogen leakage, and the judgment result is marked as 0; If the preset threshold V2 ≤ the average voltage in the voltage dynamic list < the preset threshold V1, then it is determined that there is a mild leakage, and the judgment result is marked as 1. If the preset threshold V3 ≤ the average voltage in the voltage dynamic list < the preset threshold V2, then it is determined that there is a moderate leakage, and the judgment result is marked as 2; If the average voltage in the voltage dynamic list < the preset threshold V3, then it is determined that there is a severe leakage, and the judgment result is marked as 3.

4. The fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data according to claim 1, wherein, The method for judging hydrogen leakage when the current is in a non-constant state in step S4 is as follows: If the average voltage in the voltage dynamic list ≥ the preset threshold V1, then it is determined that there is no hydrogen leakage, and the judgment result is marked as 0; If the average voltage in the voltage dynamic list < the preset threshold V1, then judge whether hydrogen leaks according to the size of the maximum hydrogen concentration value in the hydrogen dynamic list and the preset threshold P1: If the maximum hydrogen concentration in the hydrogen dynamic list ≥ the preset threshold P1, then it is determined that hydrogen leaks, and the judgment result is marked as 1 or 2 or 3; otherwise, it is determined that there is no hydrogen leakage, and the judgment result is marked as 0, and at the same time, sensor calibration is triggered.

5. The fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data according to claim 4, characterized in that, The sensor calibration method is as follows: If the data deviation between the voltage sensor and the spare sensor ≥ the preset threshold E1, then it is determined that the voltage sensor fails. After calibrating the voltage sensor, re-initialize the voltage dynamic list and continuously monitor three rolling windows to verify the repair result.

6. The fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data according to claim 1, characterized in that The trigger condition for the fault alarm in step S5 is as follows: When the number of data in the result dynamic list is equal to the preset loop window number, it is determined whether to trigger a fault alarm according to the size of the number of non-zero judgment results in the result dynamic list and the preset alarm threshold: if the number of non-zero judgment results in the result dynamic list ≥ the preset alarm threshold, a fault alarm is triggered; otherwise, a fault alarm is not triggered.

7. The fuel cell hydrogen leakage diagnosis method based on multi-dimensional sensing data according to claim 1, characterized in that, It further includes step S6, and the content of the step S6 is as follows: After the judgment of the fault alarm trigger is completed, the first value in the result dynamic list is removed, and the voltage dynamic list, current dynamic list, and hydrogen dynamic list within the rolling window corresponding to the first value are cleared, and then return to step S2.

Images