Fuel cell residual life prediction method, device, equipment, medium and product

Abstract

The invention discloses a fuel cell residual life prediction method and device, equipment, a medium and a product, and relates to the field of life prediction, and the method comprises the steps: obtaining the actual current density of a fuel cell; according to the actual current density, performing prediction by using a hierarchical and segmented prediction model to obtain a predicted UI curve; the layered and segmented prediction model comprises an ARIMA model and an electrochemical domain segmented fitting model; the electrochemical domain segmentation fitting model is used for performing segmentation simulation on an activation domain, an ohmic domain and a polarization domain; the hierarchical and segmented prediction model is obtained through optimization by using an NSGA-II multi-parameter optimization algorithm; and determining the residual life of the fuel cell by taking an area defined by the predicted UI curve and a coordinate axis as a health index. According to the method, the accuracy, reliability, robustness and feasibility of prediction can be improved.

Description

Technical Field

[0001] This application relates to the field of life prediction, and in particular to a method, apparatus, equipment, medium, and product for predicting the remaining life of a fuel cell. Background Technology

[0002] The development and deployment of renewable energy not only reduces dependence on traditional energy sources but also plays a crucial role in promoting global decarbonization and achieving long-term sustainable development goals. Proton exchange membrane fuel cells (PEMFCs), as a recognized and forward-looking cornerstone technology for renewable hydrogen energy utilization, are renowned for their high power density, zero carbon footprint, and broad applicability, and have been incorporated into global national energy strategies. However, their high manufacturing costs and limited operational lifespan remain significant obstacles to the widespread adoption of hydrogen technology and the global expansion of the hydrogen economy. Estimating and extending the operational lifespan of PEMFCs, while further reducing their overall lifecycle expenditures, is essential for promoting the continued growth of the hydrogen economy and the widespread adoption of sustainable transportation solutions.

[0003] Remaining useful life (RUL) is a key element of the Prognostics and Health Management (PHM) framework. Predicted RUL provides a crucial foundation for informed decision-making regarding dynamic load allocation within the power system, mitigating the adverse effects of severe operating conditions on the lifespan of PEMFC systems. Furthermore, RUL is an important input to PEMFC operation optimization strategies, enabling the implementation of smart energy management frameworks that dynamically adjust critical operating parameters (such as current density, power output, and thermal regulation) to effectively mitigate the degradation dynamics of critical components and extend the overall system lifespan. Moreover, RUL prediction supports predictive maintenance (PdM) strategies by facilitating optimal maintenance scheduling and resource allocation, thereby preventing the risks associated with inefficiencies or delayed maintenance related to premature intervention, ultimately improving system reliability and operational availability. In summary, RUL prediction proactively manages and extends the lifespan of PEMFCs, optimizes lifecycle cost-effectiveness, reduces maintenance expenditures, and thus improves the long-term sustainability and economic viability of the system.

[0004] However, existing life prediction methods still suffer from low accuracy in predicting the remaining life of novel fuel cells. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, device, medium, and product for predicting the remaining life of a fuel cell, which can improve the accuracy, reliability, robustness, and feasibility of the prediction.

[0006] To achieve the above objectives, this application provides the following solution:

[0007] In a first aspect, this application provides a method for predicting the remaining lifespan of a fuel cell, including:

[0008] Obtain the actual current density of the fuel cell;

[0009] The predicted UI curve is obtained by using a hierarchical segmented prediction model based on the actual current density. The hierarchical segmented prediction model includes an ARIMA model and an electrochemical domain segmented fitting model. The electrochemical domain segmented fitting model is used to perform segmented simulations of the activation domain, ohmic domain, and polarization domain. The hierarchical segmented prediction model is optimized using the NSGA-II multi-parameter optimization algorithm.

[0010] The area enclosed by the predicted UI curve and the coordinate axes is used as a health indicator to determine the remaining lifespan of the fuel cell.

[0011] In one embodiment, the construction process of the hierarchical segmented prediction model specifically includes:

[0012] Collect time-series data of healthy batteries; the time-series data includes different actual current densities and corresponding actual voltages;

[0013] Based on the time series data, multiple ARIMA models and the corresponding prediction model parameters of the ARIMA models are obtained;

[0014] Based on different actual current densities, the ARIMA model is used to predict the voltage corresponding to different actual current densities.

[0015] Based on the predicted voltage values ​​corresponding to different actual current densities, the electrochemical domain piecewise fitting model is used to perform piecewise simulations of the activation domain, ohmic domain, and polarization domain, resulting in multiple fitted UI curves and the corresponding fitting parameters for the fitted UI curves; the values ​​at the connection points of different fitted UI curves are equal.

[0016] The prediction model parameters and the fitting parameters corresponding to the ARIMA model are optimized using the NSGA-II multi-parameter optimization algorithm to obtain the optimized parameters.

[0017] The hierarchical segmented prediction model is determined based on the optimized parameters, the ARIMA model, and the fitted UI curve. In one embodiment, the expression for the electrochemical domain segmented fitting model is:

[0018] y 活化域 = a1 log(x) + b1, x∈[0.1,0.3]

[0019]

[0020] Among them, y 活化域 The fitted curve for the activation domain, y 欧姆域 The fitted curve in the Ohmic domain, y 极化域 Let a1 be the fitting curve for the polarization domain, b1 be the first fitting parameter for the active domain, b1 be the second fitting parameter for the active domain, and x be the current density. Let a2 be the first fitting parameter for the ohmic domain, b2 be the second fitting parameter for the ohmic domain, a3 be the first fitting parameter for the polarization domain, b3 be the second fitting parameter for the polarization domain, and x be the current density. c1 x c2 These are the current density values ​​at the junctions of the active and ohmic regions, and the ohmic and polarization regions, respectively.

[0021] In one embodiment, the prediction objective of the NSGA-II multi-parameter optimization algorithm is:

[0022]

[0023] The fitting objective of the NSGA-II multi-parameter optimization algorithm is:

[0024]

[0025] The iterative process of the NSGA-II multi-parameter optimization algorithm is as follows:

[0026]

[0027] Among them, y true,j,i For the i-th true value in the j-th group, y fit,k,l Let y be the l-th fitted value in the k-th group. pred,j,i For the i-th predicted value in the j-th group, y pred,k,l For the l-th predicted value in the k-th group, The parameters are for predicting model parameters and fitting parameters. Crossover is the crossover operation in NSGA-II, and Mutation is the mutation operation in NSGA-II. These are two parent parameters in NSGA-II, y j To predict the target value, n j Let y be the length of the data in the j-th group. k To fit the target value, n k is the length of the data in the k-th group.

[0028] In one embodiment, the area enclosed by the predicted UI curve and the coordinate axes is used as a health indicator to determine the remaining lifespan of the fuel cell, specifically including:

[0029] The area enclosed by the predicted UI curve and the coordinate axis is used as a health indicator, and the lifespan indicator value is calculated.

[0030] The remaining lifespan of the fuel cell is determined based on the lifespan index value and the lifespan index at the start time.

[0031] In one embodiment, the health indicator is expressed as:

[0032]

[0033] HI 新 =A 活化域 +A 欧姆域 +A 极化域

[0034] Among them, A 活化域 A is the area of ​​the UI curve of the activation domain. 欧姆域 Let A be the area of ​​the UI curve in the Ohm domain. 极化域 For the area of ​​the UI curve in the planning domain, HI 新 The proposed new lifetime index is defined as follows: a1 is the first fitting parameter of the active domain fitting curve, b1 is the second fitting parameter of the active domain fitting curve, a2 is the first fitting parameter of the ohmic domain fitting curve, b2 is the second fitting parameter of the ohmic domain fitting curve, a3 is the first fitting parameter of the polarization domain fitting curve, b3 is the second fitting parameter of the polarization domain fitting curve, x is the current density, x1 is the lower limit of the active domain current density, x2 is the upper limit of the active domain current density, x3 is the lower limit of the ohmic domain current density, x4 is the upper limit of the ohmic domain current density, x5 is the lower limit of the polarization domain current density, and x6 is the upper limit of the polarization domain current density.

[0035] Secondly, this application provides a fuel cell remaining life prediction device, comprising:

[0036] The acquisition module is used to acquire the actual current density of the fuel cell;

[0037] The prediction module is used to predict the UI curve based on the actual current density using a hierarchical segmented prediction model. The hierarchical segmented prediction model includes an ARIMA model and an electrochemical domain segmented fitting model. The electrochemical domain segmented fitting model is used to perform segmented simulations of the activation domain, ohmic domain, and polarization domain. The hierarchical segmented prediction model is optimized using the NSGA-II multi-parameter optimization algorithm.

[0038] The determination module is used to determine the remaining lifespan of the fuel cell by using the area enclosed by the predicted UI curve and the coordinate axis as a health indicator.

[0039] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the fuel cell remaining life prediction method.

[0040] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the fuel cell remaining life prediction method.

[0041] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the fuel cell remaining life prediction method.

[0042] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0043] This application provides a method, apparatus, device, medium, and product for predicting the remaining life of a fuel cell. The method uses a hierarchical segmented prediction model based on the actual current density to obtain a predicted UI curve. The hierarchical segmented prediction model includes an ARIMA model and an electrochemical domain segmented fitting model. The electrochemical domain segmented fitting model is used to simulate the activation domain, ohmic domain, and polarization domain in segments. The hierarchical segmented prediction model is optimized using the NSGA-II multi-parameter optimization algorithm. The area enclosed by the predicted UI curve and the coordinate axes is used as a health indicator to determine the remaining life of the fuel cell. Based on the NSGA-II multi-parameter optimization algorithm, combined with the ARIMA model and the electrochemical domain segmented fitting model, the problem of discontinuous UI values ​​in both the time and electrochemical domains can be solved. Furthermore, the electrochemical domain segmented fitting model can characterize the UI characteristics of the activation domain, ohmic domain, and polarization domain as comprehensively as possible, thereby improving the accuracy, reliability, robustness, and feasibility of the prediction. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is an application environment diagram of a fuel cell remaining life prediction method according to an embodiment of this application;

[0046] Figure 2 A flowchart illustrating a method for predicting the remaining life of a fuel cell, provided in an embodiment of this application;

[0047] Figure 3 This is an overall diagram of the fuel cell remaining life prediction method;

[0048] Figure 4 This is a diagram illustrating the UI values ​​at the start time.

[0049] Figure 5The curves showing the voltage versus time under different current densities (from top to bottom: 0.1-1.9 A / cm) are shown. 2 Interval 0.2A / cm 2 )picture;

[0050] Figure 6 Time-series prediction plots of voltage values ​​under different current densities;

[0051] Figure 7 The figure shows the piecewise fitting results for the electrochemical domain;

[0052] Figure 8 This is a graph showing the new HI extraction results based on the UI curve at the initial moment;

[0053] Figure 9 A functional module schematic diagram of a fuel cell remaining life prediction device provided in another embodiment of this application;

[0054] Figure 10 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0055] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0056] Existing data-driven RUL prediction methods use voltage, resistance, relative power loss rate, and composite indicators as health indicators (HIs) for data-driven RUL prediction. However, these methods cannot fully capture the degradation dynamics of PEMFCs under low, medium, and high power output conditions after long-term operation. Therefore, these HIs fail to comprehensively reflect the degree of PEMFC lifetime degradation from a full-condition perspective. The UI curve of a proton exchange membrane fuel cell depicts the relationship between voltage and operating current density, detailing its performance characteristics and degradation trends across different load ranges. This makes it possible to conduct a more comprehensive and intuitive assessment of PEMFC aging from a multi-load operation perspective. However, the UI curve measurement method for PEMFCs is inherently discrete, capturing only the instantaneous current-voltage characteristics at a specific time point, and the corresponding current density is also sampled in a discontinuous manner.

[0057] To address the aforementioned issues, this application proposes a hierarchical and piecewise RUL prediction framework for PEMFC based on the Autoregressive Integrated Moving Average Model (ARIMA) and piecewise fitting. This framework combines multidimensional parameter optimization and utilizes a novel HI derived from the continuous UI curve across the entire time domain to characterize the degradation and decay mechanisms of PEMFC under different load conditions, while quantifying the associated energy loss.

[0058] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0059] The fuel cell remaining life prediction method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on another server. Terminal 102 can send the actual current density to be processed to server 104. After receiving the actual current density, server 104 uses a hierarchical segmented prediction model to predict the actual current density, obtaining a predicted UI curve. The area enclosed by the predicted UI curve and the coordinate axis is used as a health indicator to determine the remaining lifespan of the fuel cell. Server 104 can feed back the obtained remaining lifespan to terminal 102. Furthermore, in some embodiments, the fuel cell remaining lifespan prediction method can also be implemented independently by server 104 or terminal 102. For example, terminal 102 can directly predict the remaining lifespan of the fuel cell based on the actual current density to be processed, or server 104 can obtain the actual current density to be processed from the data storage system and predict the remaining lifespan of the fuel cell based on the actual current density to be processed.

[0060] The terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster composed of multiple servers, or it can be a cloud server.

[0061] In one exemplary embodiment, such as Figure 2 and Figure 3As shown, a method for predicting the remaining life of a fuel cell is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 201 to 203. Wherein:

[0062] Step 201: Obtain the actual current density of the fuel cell;

[0063] Step 202: Based on the actual current density, a hierarchical segmented prediction model is used to predict the UI curve; the hierarchical segmented prediction model includes an ARIMA model and an electrochemical domain segmented fitting model; the electrochemical domain segmented fitting model is used to perform segmented simulations of the activation domain, ohmic domain, and polarization domain; the hierarchical segmented prediction model is optimized using the NSGA-II multi-parameter optimization algorithm.

[0064] Step 203: Use the area enclosed by the predicted UI curve and the coordinate axis as a health indicator to determine the remaining lifespan of the fuel cell.

[0065] By implementing steps 201 to 203 above, based on the NSGA-II multi-parameter optimization algorithm, combined with the ARIMA model and the electrochemical domain piecewise fitting model, the problem of UI values ​​being discontinuous in both the time domain and the electrochemical domain can be solved. Furthermore, the electrochemical domain piecewise fitting model can characterize the UI properties of the activation domain, ohmic domain, and polarization domain as much as possible, thereby improving the accuracy, reliability, robustness, and feasibility of the prediction.

[0066] In an exemplary embodiment, the construction process of the hierarchical segmented prediction model specifically includes:

[0067] Time-series data of a healthy battery is collected; the time-series data includes different actual current densities and corresponding actual voltages; multiple ARIMA models and corresponding prediction model parameters are obtained based on the time-series data; predictions are made using the ARIMA models based on different actual current densities to obtain predicted voltage values ​​corresponding to different actual current densities; based on the predicted voltage values ​​corresponding to different actual current densities, a piecewise fitting model for the electrochemical domain is used to simulate the active domain, ohmic domain, and polarization domain in segments, resulting in multiple fitted UI curves and corresponding fitting parameters; the values ​​at the connection points of different fitted UI curves are equal; the prediction model parameters corresponding to the ARIMA models and the fitted parameters are optimized using the NSGA-II multi-parameter optimization algorithm to obtain optimized parameters; a hierarchical piecewise prediction model is determined based on the optimized parameters, the ARIMA models, and the fitted UI curves.

[0068] Specifically, the construction of the time-domain ARIMA model:

[0069] Starting from the healthy battery period (0 hours of operation) to 2200 hours, the UI curve of the fuel cell operating under steady-state cycle conditions was collected every 200 hours, such as... Figure 4 As shown in the figure, the UI value measured at the initial moment can only be measured in the range of 0.1-1.9 A / cm due to limitations of the measurement technology and equipment. 2 The interval is 0.2A / cm 2 The output voltage at different current densities is discontinuous. The UI values ​​measured every 200 hours from 0 to 2200 hours are extracted according to different current densities. Figure 5 Based on the time series data shown, 10 ARIMA models are established, each with three parameters: p, d, and q, for a total of 30 prediction model parameters, as shown in Equation (1). This model can be used to predict voltage values ​​under different current densities in time series, such as... Figure 6 As shown, the dashed lines represent predicted values.

[0070] ARIMA i =f(p i ,d i ,q i ), i = 1, 2, 3... 10 (1)

[0071] Among them, ARIMA i For the i-th ARIMA model, p i Let d be the order of the autoregressive term in the i-th ARIMA model. i Let q be the difference order of the i-th ARIMA model. i Let f(p) be the order of the moving average term in the i-th ARIMA model. i ,d i ,q i ) represents the i-th ARIMA model paradigm.

[0072] Construction of piecewise fitting model for electrochemical domain:

[0073] After obtaining the predicted voltage values ​​under different current densities, piecewise fitting was performed for the active region, ohmic region, and polarization region, with current densities corresponding to the active region, ohmic region, and polarization region ranging from 0.1 to 0.3 A / cm². 2 0.5-1.3A / cm 2 1.5-1.9A / cm 2 The fitting is performed sequentially using equations (2)-(4), and the fitting is constrained. The values ​​at the connection points of different fitted curves are equal, as shown in equations (5) and (6). There are a total of 6 fitting parameters, where y 活化域 y 欧姆域 y 极化域Let x represent the fitted curves for the activation region, ohmic region, and polarization region, respectively. a1, b1, a2, b2, a3, b3 are the fitting parameters for the fitted curves of the activation region, ohmic region, and polarization region, respectively. c1 x c2 These are the current density values ​​at the junctions of the active and ohmic regions, and the ohmic and polarization regions, respectively. The results are as follows: Figure 7 The electrochemical domain fitting results are shown, where the dashed lines represent the fitted values ​​for the non-UI curve segments.

[0074] In practical applications, the expression for the piecewise fitting model of the electrochemical domain is:

[0075] y 活化域 =a1 log(x)+b1,x∈[0.1,0.3] (2)

[0076]

[0077] Among them, y 活化域 The fitted curve for the activation domain, y 欧姆域 The fitted curve in the Ohmic domain, y 极化域 Let a1 be the fitting curve for the polarization domain, b1 be the first fitting parameter for the active domain, b1 be the second fitting parameter for the active domain, and x be the current density. Let a2 be the first fitting parameter for the ohmic domain, b2 be the second fitting parameter for the ohmic domain, a3 be the first fitting parameter for the polarization domain, b3 be the second fitting parameter for the polarization domain, and x be the current density. c1 x c2 These are the current density values ​​at the junctions of the active and ohmic regions, and the ohmic and polarization regions, respectively.

[0078] NSGA-II multi-parameter optimization:

[0079] For the above 30 prediction model parameters and 6 fitting parameters, totaling 36 parameters, NSGA-II is used for multi-parameter optimization. The objectives are shown in equations (7) and (8), where equation (7) is the prediction objective and equation (8) is the fitting objective, where y true,j,i For the i-th true value in the j-th group, y fit,k,l Let y be the l-th fitted value in the k-th group. pred,j,i For the i-th predicted value in the j-th group, y pred,k,l For the l-th predicted value in the k-th group, the parameter iteration process is shown in equation (9), where There are 36 prediction and fitting parameters. Crossover is the crossover operation in NSGA-II, and Mutation is the mutation operation in NSGA-II. For the two parent parameters in NSGA-II, the optimized parameters return the ARIMA model and the piecewise fitting model to construct the final hierarchical piecewise prediction model based on the UI curve.

[0080] The ARIMA model and the electrochemical domain piecewise fitting model are collectively referred to as the hierarchical piecewise prediction model. Hierarchical means performing ARIMA first and then piecewise fitting, while piecewise means fitting in segments. Following the data flow, ARIMA prediction is performed first, followed by piecewise fitting, thus completing the construction of the UI curve hierarchical piecewise prediction model. "Optimized parameter return" refers to inputting the optimized 30 ARIMA prediction model parameters and 6 piecewise fitting parameters back into these two types of models to complete the model update, resulting in the final hierarchical piecewise prediction model.

[0081] In practical applications, the prediction objective of the NSGA-II multi-parameter optimization algorithm is:

[0082]

[0083] The fitting objective of the NSGA-II multi-parameter optimization algorithm is:

[0084]

[0085] The iterative process of the NSGA-II multi-parameter optimization algorithm is as follows:

[0086]

[0087] Among them, y true,j,i For the i-th true value in the j-th group, y fit,k,l Let y be the l-th fitted value in the k-th group. pred,j,i For the i-th predicted value in the j-th group, y pred,k,l For the l-th predicted value in the k-th group, The parameters are for predicting model parameters and fitting parameters. Crossover is the crossover operation in NSGA-II, and Mutation is the mutation operation in NSGA-II. These are two parent parameters in NSGA-II, y j To predict the target value, n j Let y be the length of the data in the j-th group. k To fit the target value, n k is the length of the data in the k-th group.

[0088] In an exemplary embodiment, the area enclosed by the predicted UI curve and the coordinate axes is used as a health indicator to determine the remaining lifespan of the fuel cell, specifically including:

[0089] The area enclosed by the predicted UI curve and the coordinate axis is used as a health indicator, and the lifespan indicator value is calculated.

[0090] The remaining lifespan of the fuel cell is determined based on the lifespan index value and the lifespan index at the start time.

[0091] Specifically, the input of the hierarchical segmented prediction model is time, and the output is the UI curve. The UI curve reflects the degradation dynamics of the PEMFC under low, medium, and high power output states. The area enclosed by the UI curve obtained after prediction and the coordinate axes is used as the HI of the PEMFC, as shown in equations (10), (11), (12), and (13), where x1-x6 are 0.1, 0.4, 0.4, 1.4, 1.4, and 1.9 respectively. The new HI at the starting time is as follows: Figure 8 As shown, let HI at this moment be HI0, and HI0 be the HI at the initial moment. Based on the characteristics of fuel cell lifespan degradation, when the HI of the UI curve at a certain predicted moment... time The value is 90% of HI0, which is the end of the fuel cell's lifespan. This completes the fuel cell lifespan prediction.

[0092] In practical applications, the health indicator is expressed as follows:

[0093]

[0094] HI 新 =A 活化域 +A 欧姆域 +A 极化域 (13)

[0095] Among them, A 活化域 A is the area of ​​the UI curve of the activation domain. 欧姆域 Let A be the area of ​​the UI curve in the Ohm domain. 极化域 For the area of ​​the UI curve in the planning domain, HI 新 As the new lifetime index, a1 is the first fitting parameter of the active domain fitting curve, b1 is the second fitting parameter of the active domain fitting curve, a2 is the first fitting parameter of the ohmic domain fitting curve, b2 is the second fitting parameter of the ohmic domain fitting curve, a3 is the first fitting parameter of the polarization domain fitting curve, b3 is the second fitting parameter of the polarization domain fitting curve, x is the current density, x1 is the lower limit of the active domain current density, x2 is the upper limit of the active domain current density, x3 is the lower limit of the ohmic domain current density, x4 is the upper limit of the ohmic domain current density, x5 is the lower limit of the polarization domain current density, and x6 is the upper limit of the polarization domain current density.

[0096] Formulas (9), (10), and (11) are integrals of formulas (2), (3), and (4) in the corresponding regions to calculate the new HI. Based on the new HI, lifetime prediction is performed. This is unrelated to the segmented prediction of the UI curve. The UI curve prediction result is a three-segmented curve of formulas (2), (3), and (4).

[0097] This application has the following advantages:

[0098] Based on NAGA-II multi-objective parameter optimization, a hierarchical segmented prediction method for time-electrochemical UI curves combined with ARIMA is proposed. This method overcomes the problem that the UI values ​​obtained by experiments are discontinuous in both the time and electrochemical domains. Furthermore, the segmented fitting of the electrochemical domain can characterize the UI properties of the activation domain, ohmic domain, and polarization domain as much as possible.

[0099] A novel UI-based HI (Hyperion of Energy) model was developed to characterize the degradation and decay mechanisms of PEMFCs under different operating conditions, while quantifying the associated energy losses. Traditional HI models can only reflect the degradation under a single operating condition, while the new UI-based HI model can reflect the degradation characteristics of PEMFCs under different operating conditions.

[0100] The above methods can improve the accuracy, reliability, robustness, and feasibility of UI curve prediction.

[0101] Based on the same inventive concept, this application also provides a fuel cell remaining life prediction device for implementing the fuel cell remaining life prediction method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more embodiments of the fuel cell remaining life prediction device provided below can be found in the limitations of the fuel cell remaining life prediction method described above, and will not be repeated here.

[0102] In one exemplary embodiment, such as Figure 9 As shown, a fuel cell remaining life prediction device is provided, comprising:

[0103] The acquisition module is used to acquire the actual current density of the fuel cell;

[0104] The prediction module is used to predict the UI curve based on the actual current density using a hierarchical segmented prediction model. The hierarchical segmented prediction model includes an ARIMA model and an electrochemical domain segmented fitting model. The electrochemical domain segmented fitting model is used to perform segmented simulations of the activation domain, ohmic domain, and polarization domain. The hierarchical segmented prediction model is optimized using the NSGA-II multi-parameter optimization algorithm.

[0105] The determination module is used to determine the remaining lifespan of the fuel cell by using the area enclosed by the predicted UI curve and the coordinate axis as a health indicator.

[0106] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 10As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores fuel cell remaining life prediction data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a fuel cell remaining life prediction method.

[0107] Those skilled in the art will understand that Figure 10 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method embodiments.

[0108] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the above-described method embodiments.

[0109] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described method embodiments.

[0110] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0111] In this application, all actions to acquire signals, information, or data are carried out in compliance with the relevant data protection laws and policies of the country where the location is situated, and with the authorization granted by the owner of the relevant device.

[0112] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0113] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0114] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0115] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for predicting the remaining life of a fuel cell, characterized in that, The method for predicting the remaining lifespan of a fuel cell includes: Obtain the actual current density of the fuel cell; The predicted UI curve is obtained by using a hierarchical segmented prediction model based on the actual current density. The hierarchical segmented prediction model includes an ARIMA model and an electrochemical domain segmented fitting model. The electrochemical domain segmented fitting model is used to perform segmented simulations of the activation domain, ohmic domain, and polarization domain. The hierarchical segmented prediction model is optimized using the NSGA-II multi-parameter optimization algorithm. The area enclosed by the predicted UI curve and the coordinate axes is used as a health indicator to determine the remaining lifespan of the fuel cell.

2. The fuel cell remaining life prediction method according to claim 1, characterized in that, The construction process of the hierarchical and segmented prediction model specifically includes: Collect time-series data of healthy batteries; the time-series data includes different actual current densities and corresponding actual voltages; Based on the time series data, multiple ARIMA models and the corresponding prediction model parameters of the ARIMA models are obtained; Based on different actual current densities, the ARIMA model is used to predict the voltage corresponding to different actual current densities. Based on the predicted voltage values ​​corresponding to different actual current densities, the electrochemical domain piecewise fitting model is used to perform piecewise simulations of the activation domain, ohmic domain, and polarization domain, resulting in multiple fitted UI curves and the corresponding fitting parameters for the fitted UI curves; the values ​​at the connection points of different fitted UI curves are equal. The prediction model parameters and the fitting parameters corresponding to the ARIMA model are optimized using the NSGA-II multi-parameter optimization algorithm to obtain the optimized parameters. The hierarchical segmented prediction model is determined based on the optimization parameters, the ARIMA model, and the fitted UI curve.

3. The fuel cell remaining life prediction method according to claim 1, characterized in that, The expression for the piecewise fitting model of the electrochemical domain is: and 活化域 =a1log(x)+b1,x∈[0.1,0.3] Among them, y 活化域 The fitted curve for the activation domain, y 欧姆域 The fitted curve in the Ohmic domain, y 极化域 Let a1 be the fitting curve for the polarization domain, b1 be the first fitting parameter for the active domain, b1 be the second fitting parameter for the active domain, and x be the current density. Let a2 be the first fitting parameter for the ohmic domain, b2 be the second fitting parameter for the ohmic domain, a3 be the first fitting parameter for the polarization domain, b3 be the second fitting parameter for the polarization domain, and x be the current density. c1 x c2 These are the current density values ​​at the junctions of the active and ohmic regions, and the ohmic and polarization regions, respectively.

4. The fuel cell remaining life prediction method according to claim 1, characterized in that, The prediction objective of the NSGA-II multi-parameter optimization algorithm is: The fitting objective of the NSGA-II multi-parameter optimization algorithm is: The iterative process of the NSGA-II multi-parameter optimization algorithm is as follows: Among them, y true,j,i For the i-th true value in the j-th group, y fit,k,l Let y be the l-th fitted value in the k-th group. pred,j,i For the i-th predicted value in the j-th group, y pred,k,l For the l-th predicted value in the k-th group, For predicting model parameters and fitting parameters, Crossover is the crossover operation in NSGA-II, and Mutation is the mutation operation in NSGA-II. These are two parent parameters in NSGA-II, y j To predict the target value, n j Let y be the length of the data in the j-th group. k To fit the target value, n k is the length of the data in the k-th group.

5. The fuel cell remaining life prediction method according to claim 1, characterized in that, The area enclosed by the predicted UI curve and the coordinate axes is used as a health indicator to determine the remaining lifespan of the fuel cell, specifically including: The area enclosed by the predicted UI curve and the coordinate axis is used as a health indicator, and the lifespan indicator value is calculated. The remaining lifespan of the fuel cell is determined based on the lifespan index value and the lifespan index at the start time.

6. The fuel cell remaining life prediction method according to claim 1, characterized in that, The health indicators are expressed as follows: HI 新 =A 活化域 +A 欧姆域 +A 极化域 Among them, A 活化域 A is the area of ​​the UI curve of the activation domain. 欧姆域 Let A be the area of ​​the UI curve in the Ohm domain. 极化域 For the area of ​​the UI curve in the planning domain, HI 新 As the new lifetime index, a1 is the first fitting parameter of the active domain fitting curve, b1 is the second fitting parameter of the active domain fitting curve, a2 is the first fitting parameter of the ohmic domain fitting curve, b2 is the second fitting parameter of the ohmic domain fitting curve, a3 is the first fitting parameter of the polarization domain fitting curve, b3 is the second fitting parameter of the polarization domain fitting curve, x is the current density, x1 is the lower limit of the active domain current density, x2 is the upper limit of the active domain current density, x3 is the lower limit of the ohmic domain current density, x4 is the upper limit of the ohmic domain current density, x5 is the lower limit of the polarization domain current density, and x6 is the upper limit of the polarization domain current density.

7. A fuel cell remaining life prediction device, characterized in that, The fuel cell remaining life prediction device includes: The acquisition module is used to acquire the actual current density of the fuel cell; The prediction module is used to predict the UI curve based on the actual current density using a hierarchical segmented prediction model. The hierarchical segmented prediction model includes an ARIMA model and an electrochemical domain segmented fitting model. The electrochemical domain segmented fitting model is used to perform segmented simulations of the activation domain, ohmic domain, and polarization domain. The hierarchical segmented prediction model is optimized using the NSGA-II multi-parameter optimization algorithm. The determination module is used to determine the remaining lifespan of the fuel cell by using the area enclosed by the predicted UI curve and the coordinate axis as a health indicator.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the fuel cell remaining life prediction method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the fuel cell remaining life prediction method according to any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the fuel cell remaining life prediction method according to any one of claims 1-6.

Images