阅读说明：本技术 燃料电池系统在线故障诊断方法及装置 (Online fault diagnosis method and device for fuel cell system ) 是由 洪吉超 徐晓明 孙旭东 胡松 赤骋 陈东方 于 2021-06-08 设计创作，主要内容包括：本发明公开了一种燃料电池系统在线故障诊断方法及装置,其中,方法包括：获取燃料电池的电堆测试数据和当前运行数据；将电堆测试数据和当前运行数据输入至预设的故障诊断模型,得到诊断数据；根据诊断数据识别燃料电池是否故障,并且在燃料电池故障时,发送故障信号至预设终端。该方法能够对燃料电池系统进行实时在线检测、准确评估其故障状态,诊断过程快速、高效。(The invention discloses a fuel cell system online fault diagnosis method and a device, wherein the method comprises the following steps: acquiring the test data and the current operation data of the fuel cell stack; inputting the test data and the current operation data of the electric pile into a preset fault diagnosis model to obtain diagnosis data; and identifying whether the fuel cell is in fault according to the diagnosis data, and sending a fault signal to a preset terminal when the fuel cell is in fault. The method can perform real-time online detection on the fuel cell system, accurately evaluate the fault state of the fuel cell system, and has a quick and efficient diagnosis process.)

1. An online fault diagnosis method for a fuel cell system, characterized by comprising the steps of:

acquiring the test data and the current operation data of the fuel cell stack;

inputting the galvanic pile test data and the current operation data into a preset fault diagnosis model to obtain diagnosis data; and

and identifying whether the fuel cell is in fault according to the diagnosis data, and sending a fault signal to a preset terminal when the fuel cell is in fault.

2. The method of claim 1, further comprising:

and optimizing the electric pile test data according to the current operation data of the fuel cell so as to obtain the electric pile test data for next diagnosis.

3. The method of claim 1, further comprising:

generating a training set according to the stack test data and the operation data of the fuel cell;

and training a machine learning model corresponding to the current type of the fuel cell by using the training set until the machine learning model reaches a preset condition, and obtaining the preset fault diagnosis model.

4. The method of claim 1, further comprising:

generating a fault repair action for the vehicle based on the diagnostic data

And controlling the vehicle driving execution component to execute the fault repairing action.

5. The method of claim 1, further comprising:

and pre-establishing a fault characterization parameter database to obtain the diagnosis data by using the fault characterization parameter database.

6. An online fault diagnosis device for a fuel cell system, comprising:

the acquisition module is used for acquiring the electric pile test data and the current operation data of the fuel cell;

the input module is used for inputting the galvanic pile test data and the current operation data into a preset fault diagnosis model to obtain diagnosis data; and

and the sending module is used for identifying whether the fuel cell has a fault according to the diagnosis data and sending a fault signal to a preset terminal when the fuel cell has the fault.

7. The apparatus of claim 6, further comprising:

and the optimization module is used for optimizing the electric pile test data according to the current operation data of the fuel cell so as to obtain the electric pile test data for next diagnosis.

8. The apparatus of claim 6, further comprising:

the first generation module is used for generating a training set according to the stack test data and the operation data of the fuel cell;

and the training module is used for training the machine learning model corresponding to the current type of the fuel cell by using the training set until the machine learning model reaches a preset condition, so as to obtain the preset fault diagnosis model.

9. The apparatus of claim 6, further comprising:

a second generation module for generating a fail-over action of the vehicle based on the diagnostic data

And the control module is used for controlling the vehicle driving execution component to execute the fault repairing action.

10. The apparatus of claim 6, further comprising:

and the establishing module is used for establishing a fault characterization parameter database in advance so as to obtain the diagnosis data by using the fault characterization parameter database.

Technical Field

The invention relates to the technical field of batteries, in particular to a method and a device for diagnosing online faults of a fuel cell system.

Background

In the related art, the fault diagnosis of the fuel cell system is generally performed from three aspects of models, data driving and experimental tests.

However, the fuel cell system failure diagnosis method in the related art has several problems: (1) most of the electric pile experimental data or real vehicle operation data are only considered for analysis, and the coupling use of the two data is not comprehensively considered; (2) establishing a single machine learning model, wherein models required by different types of fuel cells are not considered to be possibly different; (3) the diagnosis method is complex, the implementation cost is high, and the diagnosis method can not be used for online fault diagnosis and needs to be solved.

Disclosure of Invention

The present invention is directed to solving, at least to some extent, one of the technical problems in the related art.

Therefore, an object of the present invention is to provide an online fault diagnosis method for a fuel cell system, which can perform online real-time detection on the fuel cell system, accurately evaluate the fault state of the fuel cell system, and has a fast and efficient diagnosis process.

Another object of the present invention is to provide an online fault diagnosis apparatus for a fuel cell system.

In order to achieve the above object, an embodiment of the invention provides an online fault diagnosis method for a fuel cell system, which includes the following steps:

acquiring the test data and the current operation data of the fuel cell stack;

inputting the galvanic pile test data and the current operation data into a preset fault diagnosis model to obtain diagnosis data; and

and identifying whether the fuel cell is in fault according to the diagnosis data, and sending a fault signal to a preset terminal when the fuel cell is in fault.

According to the fuel cell system online fault diagnosis method, coupling use and analysis of the pile experiment data and real vehicle operation data are considered, accuracy of model input parameters is improved, models required by different types of fuel cells are considered to be possibly different through establishing a machine learning model, the most suitable model is automatically matched according to the type of the fuel cell, a single model or a multi-model fusion method can be used, the diagnosis method is simple and easy, implementation cost is low, and the method can be used for online fault diagnosis.

In addition, the online fault diagnosis method for the fuel cell system according to the above embodiment of the present invention may further have the following additional technical features:

further, in an embodiment of the present invention, the method further includes:

and optimizing the electric pile test data according to the current operation data of the fuel cell so as to obtain the electric pile test data for next diagnosis.

Further, in an embodiment of the present invention, the method further includes:

generating a training set according to the stack test data and the operation data of the fuel cell;

and training a machine learning model corresponding to the current type of the fuel cell by using the training set until the machine learning model reaches a preset condition, and obtaining the preset fault diagnosis model.

Further, in an embodiment of the present invention, the method further includes:

generating a fault repair action of the vehicle according to the diagnosis data;

and controlling the vehicle driving execution component to execute the fault repairing action.

Further, in an embodiment of the present invention, the method further includes:

and pre-establishing a fault characterization parameter database to obtain the diagnosis data by using the fault characterization parameter database.

In order to achieve the above object, according to another aspect of the present invention, an online fault diagnosis apparatus for a fuel cell system is provided, including:

the acquisition module is used for acquiring the electric pile test data and the current operation data of the fuel cell;

the input module is used for inputting the galvanic pile test data and the current operation data into a preset fault diagnosis model to obtain diagnosis data; and

and the sending module is used for identifying whether the fuel cell has a fault according to the diagnosis data and sending a fault signal to a preset terminal when the fuel cell has the fault.

The fuel cell system online fault diagnosis device provided by the embodiment of the invention considers the coupling use and analysis of the pile experimental data and the real vehicle operation data, improves the accuracy of the model input parameters, automatically matches the most suitable model according to the type of the fuel cell by establishing a machine learning model and considering the possible difference of models required by different types of fuel cells, can use a single model or a multi-model fusion method, has simple diagnosis method and low implementation cost, and can be used for online fault diagnosis.

In addition, the online fault diagnosis device for the fuel cell system according to the above embodiment of the present invention may further have the following additional technical features:

further, in an embodiment of the present invention, the method further includes:

and the optimization module is used for optimizing the electric pile test data according to the current operation data of the fuel cell so as to obtain the electric pile test data for next diagnosis.

Further, in an embodiment of the present invention, the method further includes:

the first generation module is used for generating a training set according to the stack test data and the operation data of the fuel cell;

and the training module is used for training the machine learning model corresponding to the current type of the fuel cell by using the training set until the machine learning model reaches a preset condition, so as to obtain the preset fault diagnosis model.

Further, in an embodiment of the present invention, the method further includes:

a second generation module for generating a fail-over action of the vehicle based on the diagnostic data

And the control module is used for controlling the vehicle driving execution component to execute the fault repairing action.

Further, in an embodiment of the present invention, the method further includes:

and the establishing module is used for establishing a fault characterization parameter database in advance so as to obtain the diagnosis data by using the fault characterization parameter database.

Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.

Drawings

The foregoing and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

fig. 1 is a flowchart of an online fault diagnosis method of a fuel cell system according to an embodiment of the present invention;

fig. 2 is a flowchart of an on-line fault diagnosis method of a fuel cell system according to an embodiment of the present invention;

FIG. 3 is a flow diagram of a design of a fault diagnosis algorithm according to one embodiment of the present invention;

fig. 4 is a block schematic diagram of an online fault diagnosis apparatus of a fuel cell system according to an embodiment of the present invention.

Detailed Description

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.

An online fault diagnosis method and apparatus of a fuel cell system according to an embodiment of the present invention will be described below with reference to the accompanying drawings, and first, an online fault diagnosis method of a fuel cell system according to an embodiment of the present invention will be described with reference to the accompanying drawings.

Fig. 1 is a flowchart of an online fault diagnosis method of a fuel cell system according to an embodiment of the present invention.

As shown in fig. 1, the fuel cell system online fault diagnosis method includes:

in step S101, stack test data and current operation data of the fuel cell are acquired.

It should be understood that the manner of acquiring the stack test data and the current operation data of the fuel cell may be the manner of acquiring in the related art, and details are not described herein to avoid redundancy.

In step S102, the stack test data and the current operation data are input to a preset fault diagnosis model, so as to obtain diagnosis data.

Specifically, the fuel cell real vehicle operation data and the galvanic pile test data can be used for performing a training data subset of machine learning, training a machine learning model, according to a training result, the real vehicle operation data and the galvanic pile test data are used as a verification data subset to verify the accuracy of the established model, namely, the precision of the model is verified, whether the accuracy of the established model meets an expected requirement is judged, if yes, a fault diagnosis model is established for analyzing and early warning the current state, and otherwise, the parameters of the machine learning model or the training method are adjusted to meet the requirement of the training result.

In step S103, it is identified whether the fuel cell is malfunctioning according to the diagnostic data, and a malfunction signal is transmitted to a preset terminal when the fuel cell is malfunctioning.

It is understood that the fault diagnosis results of the stack may include flooding, dry membrane, gas starvation, short circuit, and/or catalyst poisoning, etc.; the preset terminal can be a mobile phone, a tablet, a PC terminal and the like.

Further, in an embodiment of the present invention, the method further includes: and optimizing the electric pile test data according to the current operation data of the fuel cell to obtain the electric pile test data for next diagnosis, so that the accuracy of the electric pile test data is continuously improved.

Further, in an embodiment of the present invention, the method further includes: generating a training set according to the stack test data and the operation data of the fuel cell; and training a machine learning model corresponding to the current type of the fuel cell by using the training set until the machine learning model reaches a preset condition, so as to obtain a preset fault diagnosis model.

That is to say, according to different types of fuel cells, the embodiment of the application can adaptively select and establish a machine learning model, where the machine learning model includes a linear model, a kernel method and a support vector machine, a decision tree and Boosting, a neural network, and the like, where the neural network includes a fully-connected neural network, a convolutional neural network, and a cyclic neural network, and a machine learning model most suitable for a current scene can be obtained by using a mode of fusing one or more models.

It should be noted that, because different machine learning models have respective applicable scenarios, the embodiment of the present application may determine the model most suitable for the current scenario from seven angles, including the size of the training set (e.g., large, small), the dimensions of the feature space (e.g., high dimension, low dimension), whether the features are independent of each other (e.g., independent, dependent), whether the features are linear features (e.g., linear, nonlinear), the requirement on the fitting degree, the missing value ratio (e.g., large, small), and other requirements (e.g., performance, time, space).

Further, in an embodiment of the present invention, the method further includes: generating a fault repairing action of the vehicle according to the diagnosis data; and controlling the vehicle driving execution component to execute a fault repairing action.

That is to say, the embodiment of the present application can implement the repair of the fault according to the diagnostic data.

Further, in an embodiment of the present invention, the method further includes: and a fault characterization parameter database is established in advance so as to obtain the diagnosis data by utilizing the fault characterization parameter database.

Specifically, when the diagnostic data is obtained, a fault characterization parameter database is generally established, a model sample database is established, and redundancy design is performed by combining multi-model fusion.

The established fault characterization parameter database comprises data of three aspects of models, data driving, experimental tests and the like, and fault diagnosis methods under certain specific occasions, such as: fault diagnosis methods of multiple sets of high-power PEMFC systems, multiple fault diagnosis methods of PEMFC systems in complex environments, fault diagnosis methods considering PEMFC system aging and the like.

In order to further understand the online fault diagnosis method for the fuel cell system according to the embodiment of the present application, the following detailed description is provided with reference to specific embodiments.

As shown in fig. 2, the online fault diagnosis method for a fuel cell system includes the steps of:

(1) according to the fuel cell, the electric pile test data and the real vehicle operation data of the fuel cell system are used as data accumulation in the early stage of fault diagnosis, and meanwhile, the real vehicle operation data can optimize the electric pile test data and continuously improve the accuracy of the electric pile test data.

(2) According to different types of fuel cells, a machine learning model is adaptively selected and established, wherein the machine learning model comprises a linear model, a kernel method, a support vector machine, a decision tree, Boosting, a neural network and the like, the neural network comprises a fully-connected neural network, a convolutional neural network and a cyclic neural network, and the machine learning model which is most suitable for the current scene can be obtained by using one or a combination of several models.

(3) The fuel cell real vehicle operation data and the electric pile test data can be used for training a training data subset for machine learning to train a machine learning model.

(4) And (4) according to the training result in the step (3), using the real vehicle operation data and the galvanic pile test data as a verification data subset to verify the accuracy of the established model, namely verifying the model accuracy.

(5) Judging whether the accuracy of the model building reaches the expected requirement or not, if so, continuing the step (6); if not, continuing the step (7);

(6) and establishing a fault diagnosis model for analyzing and early warning the current state.

(7) And (5) if the training result does not meet the requirement, adjusting the parameters of the machine learning model or the training method to meet the requirement of the training result, and continuing to execute the step (5).

Therefore, the fuel cell system can be subjected to real-time online detection and accurate fault state assessment, the diagnosis process is rapid and efficient, and the method can be used for a fuel cell control system or fuel cell test equipment.

Further, as shown in fig. 3, fig. 3 shows an example of a method for diagnosing a fault of a fuel cell on line based on a rapid electrochemical impedance spectroscopy. After the impedance spectrum of the fuel cell is obtained through measurement, a fault diagnosis algorithm needs to be designed to apply the impedance spectrum information to fault diagnosis. The design process of the fault diagnosis algorithm is divided into an off-line scene and an on-line scene. In an off-line scene, firstly, a normal experiment is required to be carried out to carry out a gorgon fault experiment, impedance spectrum data of the fuel cell under different experimental conditions are measured to obtain a training data set, then, the data are analyzed, and characteristics suitable for being applied to fault diagnosis are selected from the impedance spectrum by combining a specific theory and priori knowledge. And then selecting a proper classification algorithm according to the selected characteristics and the practical application scene, and designing a fault classifier. In an online scene, the rapid impedance spectrum measuring system runs in real time and records the impedance spectrum of the fuel cell, and then the impedance spectrum data is extracted by a feature extraction algorithm to obtain features for fault classification, and then a fault classifier is used for identifying health state and fault type of the current electric pile.

According to the fuel cell system online fault diagnosis method provided by the embodiment of the invention, coupling use and analysis of stack experimental data and real vehicle operation data are considered, the accuracy of model input parameters is improved, models required by different types of fuel cells are considered to be possibly different by establishing a machine learning model, the most suitable model is automatically matched according to the type of the fuel cell, a single model or a multi-model fusion method can be used, the diagnosis method is simple and easy, the implementation cost is low, and the method can be used for online fault diagnosis.

Next, an on-line fault diagnosis apparatus of a fuel cell system proposed according to an embodiment of the present invention is described with reference to the drawings.

Fig. 4 is a block schematic diagram of an online fault diagnosis apparatus for a fuel cell system according to an embodiment of the present invention.

As shown in fig. 4, the fuel cell system online fault diagnosis device includes: an acquisition module 100, an input module 200 and a sending module 300.

The acquiring module 100 is used for acquiring stack test data and current operation data of the fuel cell;

the input module 200 is used for inputting the galvanic pile test data and the current operation data into a preset fault diagnosis model to obtain diagnosis data; and

the sending module 300 is configured to identify whether the fuel cell is faulty according to the diagnostic data, and send a fault signal to a preset terminal when the fuel cell is faulty.

Further, in an embodiment of the present invention, the method further includes:

and the optimization module is used for optimizing the electric pile test data according to the current operation data of the fuel cell so as to obtain the electric pile test data for next diagnosis.

Further, in an embodiment of the present invention, the method further includes:

the first generation module is used for generating a training set according to the stack test data and the operation data of the fuel cell;

and the training module is used for training the machine learning model corresponding to the current type of the fuel cell by using the training set until the machine learning model reaches a preset condition, so as to obtain a preset fault diagnosis model.

Further, in an embodiment of the present invention, the method further includes:

a second generation module for generating a fail-over action of the vehicle based on the diagnostic data

And the control module is used for controlling the vehicle driving execution component to execute the fault repairing action.

Further, in an embodiment of the present invention, the method further includes:

and the establishing module is used for establishing a fault characterization parameter database in advance so as to obtain the diagnosis data by utilizing the fault characterization parameter database.

It should be noted that the foregoing explanation of the embodiment of the online fault diagnosis method for a fuel cell system also applies to the online fault diagnosis device for a fuel cell system of this embodiment, and details are not repeated here.

According to the fuel cell system online fault diagnosis device provided by the embodiment of the invention, coupling use and analysis of stack experimental data and real vehicle operation data are considered, the accuracy of model input parameters is improved, models required by different types of fuel cells are considered to be possibly different by establishing a machine learning model, the most suitable model is automatically matched according to the type of the fuel cell, a single model or a multi-model fusion method can be used, the diagnosis method is simple and easy, the implementation cost is low, and the fuel cell system online fault diagnosis device can be used for online fault diagnosis.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.

In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.