A method, device and storage device for identifying the operating conditions of an RSOC system

Abstract

This invention discloses a method, device, and storage device for RSOC system operating condition identification, comprising the following steps: an offline training phase and an online operating condition identification phase. The offline training phase involves retrieving historical data on the stack impedance, hydrogen production flow rate, and temperature in the system, along with the corresponding operating condition types; training and obtaining a denoised sparse autoencoder that maximizes the fitness value; and establishing an operating condition mapping model. The online operating condition identification phase involves acquiring the latest data from the system; processing the new data using the denoised sparse autoencoder; and outputting the operating condition type from the processed data through the mapping model. This invention utilizes the characteristics of stack impedance, hydrogen production flow rate, and temperature, and employs a denoised sparse autoencoder optimized by a heuristic cross-search algorithm for dimensionality reduction and noise reduction. This reduces interference from external signals, provides more reliable data for subsequent operating condition identification, and improves the accuracy of RSOC system operating condition identification.

Description

Technical Field

[0001] This invention relates to the field of state identification of reversible solid oxide fuel cell systems, and more particularly to a method, device and storage device for identifying the operating conditions of an RSOC system. Background Technology

[0002] RSOC (Reversible Solid Oxide Fuel Cell) systems are new energy power generation devices that utilize electrochemical reactions to generate electricity and produce hydrogen. Due to their extremely high requirements for operating environment, control design, operation, and maintenance, and their characteristics of power generation, load variation, hydrogen production, and multi-condition switching, RSOC systems are a type of new energy power generation prone to degradation. The proper functioning of an RSOC system not only affects the quality of power generation / hydrogen production, but also, if its operating conditions are incorrectly identified and remedial measures are not taken, irreversible damage to the RSOC stack can occur, leading to economic losses. Furthermore, the high-temperature environment and poor hydrogen production management can cause health problems for operators. Therefore, RSOC system operating condition identification is crucial.

[0003] The existing methods are mainly gas leak detection, but it is difficult to accurately identify the actual condition of the RSOC system by relying solely on gas leak detection. Summary of the Invention

[0004] To address the problem of difficulty in determining the actual state of an RSOC system, this invention proposes a method for identifying the operating conditions of an RSOC system, which specifically includes the following steps:

[0005] Offline training phase and online operating condition identification phase;

[0006] The offline training phase includes the following steps:

[0007] S11. Retrieve historical data on stack impedance, hydrogen production flow rate, and temperature in the RSOC system. U and the corresponding operating conditions. G ;

[0008] S12. Train and obtain a noise-reduced sparse autoencoder that maximizes the fitness value;

[0009] S13. Establish a working condition mapping model;

[0010] The online operating condition identification phase includes the following steps:

[0011] S21. Obtain the latest data on stack impedance, hydrogen production flow rate, and temperature in the RSOC system;

[0012] S22. The new data is processed by the noise-reducing sparse autoencoder to obtain the processed data;

[0013] S23. Output the working condition type of the processed data through the mapping model.

[0014] A storage device that stores instructions and data for implementing an RSOC system operating condition identification method.

[0015] An RSOC system operating condition identification device includes: a processor and a storage device; the processor loads and executes instructions and data in the storage device to implement an RSOC system operating condition identification method.

[0016] The beneficial effects provided by this invention are: by utilizing the characteristics of stack impedance, hydrogen production flow rate and temperature, and by using a noise-reducing sparse autoencoder optimized by a heuristic cross-search algorithm to reduce dimensionality and noise, the interference of external signals can be reduced, providing more reliable data for subsequent operating condition identification. Furthermore, the operating condition identification of the RSOC system is achieved through a neural network, thereby improving the accuracy of RSOC system operating condition identification. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0018] Figure 2 This is a schematic diagram of the feature abstraction process;

[0019] Figure 3 This is a schematic diagram of the mapping model structure;

[0020] Figure 4 This is a diagram showing the operating condition identification results of the RSOC system based on this invention.

[0021] Figure 5 This is a schematic diagram of the hardware device working in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0023] Please refer to Figure 1 , Figure 1 This is a flowchart of the method of the present invention.

[0024] This invention provides a method for identifying the operating conditions of an RSOC system, comprising the following steps:

[0025] Offline training phase and online operating condition identification phase;

[0026] The offline training phase includes the following steps:

[0027] S11. Retrieve historical data on stack impedance, hydrogen production flow rate, and temperature in the RSOC system. U and the corresponding operating conditions.G ;

[0028] S12. Train and obtain a noise-reduced sparse autoencoder that maximizes the fitness value;

[0029] 1. It should be noted that the training process in step S12 is as follows:

[0030] S121. Extract the time-domain, frequency-domain, and time-frequency-domain features of historical data U. And through a noise-reducing sparse autoencoder, the features are... Dimensionality reduction and noise reduction are performed to obtain features. ;

[0031] S122, with features and characteristics The fitness value is the reciprocal of the sum of squares of the differences between the features. The parameters in the denoising sparse autoencoder are optimized by a heuristic cross-rescue algorithm to obtain a denoising sparse autoencoder that maximizes the fitness value.

[0032] In the above process, fitness value ,in, Features or The number of.

[0033] S13. Establish a working condition mapping model;

[0034] The process of establishing the working condition mapping model in step S13 is as follows:

[0035] S131. Based on the noise reduction sparse autoencoder, obtain the intermediate abstract features as features. ;

[0036] Based on the denoising sparse autoencoder, intermediate abstract features are obtained as features. ;like Figure 2 The leftmost one is the feature. The abstract features in the middle are features The rightmost one is the noise-reducing sparse autoencoder feature. ;

[0037] S132. Establishing features through neural networks and corresponding working condition types The mapping model between the two is obtained by using the immune cloning algorithm to optimize the parameters of the neural network.

[0038] like Figure 3 As shown, the leftmost layer is the input layer, whose dimension and features... The dimensions are the same, with the rightmost layer being the output layer, whose dimensions are the same as the working condition type. The dimensions are the same, with the middle 4 layers being hidden layers. The parameters of the neural network are optimized using the immune cloning algorithm. That is, the neural network optimized by the immune cloning algorithm is the mapping model.

[0039] The online operating condition identification phase includes the following steps:

[0040] S21. Obtain the latest data on stack impedance, hydrogen production flow rate, and temperature in the RSOC system;

[0041] S22. The new data is processed by the noise-reducing sparse autoencoder to obtain the processed data;

[0042] S23. Output the working condition type of the processed data through the mapping model.

[0043] As one example, the latest data on stack impedance, hydrogen production flow rate, and temperature in the RSOC system are obtained. Extract RSOC system data The time-domain, frequency-domain, and time-frequency-domain features are obtained, and the features are obtained through the noise-reducing sparse autoencoder described in step 3. ; Features As input to the neural network, the output of the neural network is the operating condition type. .

[0044] As one embodiment, the present invention uses 20 sets of test samples for testing. Figure 4 In the diagram, Y-axis coordinates 1, 2, and 3 represent three different operating conditions. 1 indicates RSOC system reactor damage, 2 indicates abnormal hydrogen production in the RSOC system, and 3 indicates auxiliary component failure in the RSOC system. Circles represent the actual classification results of the RSOC system; asterisks represent the predicted results of the system's operating condition identification. Figure 4 As can be seen, using 20 sets of experimental sample datasets for testing, the method proposed in this patent can control the accuracy of working condition identification at 95%. Furthermore, compared with traditional SVM and Bayesian classification algorithms, its advantages are obvious, as shown in Table 1.

[0045] Table 1. Comparison of accuracy rates for operating condition identification

[0046]

[0047] Please see Figure 5 , Figure 5 This is a schematic diagram of the hardware device in operation according to an embodiment of the present invention. The hardware device specifically includes: an RSOC system operating condition identification device 401, a processor 402, and a storage device 403.

[0048] An RSOC system operating condition identification device 401: The RSOC system operating condition identification device 401 implements the RSOC system operating condition identification method.

[0049] Processor 402: The processor 402 loads and executes the instructions and data in the storage device 403 to implement the RSOC system operating condition identification method.

[0050] Storage device 403: The storage device 403 stores instructions and data; the storage device 403 is used to implement the RSOC system operating condition identification method.

[0051] The beneficial effects of this invention are: by utilizing the characteristics of stack impedance, hydrogen production flow rate and temperature, and by using a noise-reducing sparse autoencoder optimized by a heuristic cross-search algorithm to reduce dimensionality and noise, the interference of external signals can be reduced, providing more reliable data for subsequent operating condition identification. Furthermore, the operating condition identification of the RSOC system is achieved through a neural network, thereby improving the accuracy of RSOC system operating condition identification.

[0052] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying the operating conditions of an RSOC system, characterized in that: Includes the following steps: Offline training phase and online operating condition identification phase; The offline training phase includes the following steps: S11. Retrieve historical data on stack impedance, hydrogen production flow rate, and temperature in the RSOC system. U and the corresponding operating conditions. G ; S12. Train and obtain a noise-reduced sparse autoencoder that maximizes the fitness value; In step S12, the training process is as follows: S121. Extract the time-domain, frequency-domain, and time-frequency-domain features of historical data U. And through a noise-reducing sparse autoencoder, the features are... Dimensionality reduction and noise reduction are performed to obtain features. ; S122, with features and characteristics The fitness value is the reciprocal of the sum of squares of the differences between the features. The parameters in the denoised sparse autoencoder are optimized by a heuristic cross-rescue algorithm to obtain a denoised sparse autoencoder that maximizes the fitness value. fitness value ,in, Features or The number of; S13. Establish a working condition mapping model; The online operating condition identification phase includes the following steps: S21. Obtain the latest data on stack impedance, hydrogen production flow rate, and temperature in the RSOC system; S22. The new data is processed by the noise-reducing sparse autoencoder to obtain the processed data; S23. Output the working condition type of the processed data through the mapping model.

2. The RSOC system operating condition identification method as described in claim 1, characterized in that: The process of establishing the working condition mapping model in step S13 is as follows: S131. Based on the noise reduction sparse autoencoder, obtain the intermediate abstract features as features. ; S132. Establishing features through neural networks and corresponding working condition types The mapping model between the two is obtained by optimizing the parameters of the neural network using the immune cloning algorithm.

3. A storage device, characterized in that: The storage device stores instructions and data to implement the RSOC system operating condition identification method according to any one of claims 1 to 2.

4. An RSOC system operating condition identification device, characterized in that: include: A processor and a storage device; the processor loads and executes instructions and data in the storage device to implement the RSOC system operating condition identification method according to any one of claims 1 to 2.

Images