Digital energy storage system reliability evaluation method and device considering power change and data reconstruction error

Abstract

The invention provides a digital energy storage system reliability evaluation method and device considering power change and data reconstruction error, and the method comprises the steps: constructing a two-dimensional state probability model which represents the power size and power fluctuation intensity influence of an energy storage converter in a digital energy storage system, based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage converter, the energy storage converter failure rate under the target working condition is solved; determining a data reconstruction error based on the energy storage battery variable of the battery cluster in the digital energy storage system and the corresponding output feature vector, and calculating the thermal runaway probability of the battery cluster according to the data reconstruction error; and according to the fault rate of the energy storage converter and the thermal runaway probability of the battery cluster, carrying out operation reliability evaluation on the digital energy storage system to obtain an evaluation result. According to the method, the operation reliability evaluation of the digital energy storage system is realized, and the safe and stable operation of the battery system in a high-pressure and high-temperature environment is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of risk assessment and management, and in particular to a method and device for assessing the reliability of a digital energy storage system taking into account power variation and data reconstruction errors. Background Art

[0002] A traditional battery energy storage system (BESS) is a device used to store and release electrical energy, usually including an energy storage system, such as a battery pack or supercapacitor, and a corresponding power electronic converter. The energy storage system can be used to balance grid loads, fluctuations in renewable energy output, and reserve emergency power. It is designed to provide efficient and reliable energy storage solutions to meet power needs in different scenarios.

[0003] The power conversion system (PCS) is one of the key components in traditional battery energy storage systems. Its main function is to convert the electric energy stored in the energy storage system into AC or DC in a controllable manner and output it to the power grid or other power supply equipment. PCS plays a role in regulating the flow of electric energy and optimizing energy conversion efficiency in the energy storage system. It is usually equipped with advanced control system components (such as IGBT, Mosfet, diode, etc.), which can adapt to different power needs and provide stable power output.

[0004] However, due to the diversity and complexity of the operating environment and operating conditions, compared with general industrial PCS, PCS used in battery energy storage systems has higher random variability and instability in output power, which may cause the aluminum bonding wires and solder layers of each power device of the PCS to undergo long-term and frequent expansion and contraction, leading to the generation of unbalanced electrothermal stress, resulting in fatigue accumulation, and may also cause the aluminum bonding wires of the power devices to break and the solder layers to crack.

[0005] In addition, the main focus of traditional battery energy storage systems is the reliability of the battery itself, which often involves a time span of several hours or even dozens of hours. Battery thermal runaway is a serious problem in battery energy storage reliability, especially when the battery overheats and enters a self-sustaining reaction, which may cause the temperature to rise rapidly, leading to catastrophic consequences such as fire or explosion.

[0006] In order to solve the problem of thermal runaway of batteries, existing technologies have proposed methods based on rules, models and state estimation to conduct battery fault diagnosis and early warning research. Although certain results have been achieved, they still face challenges in practical applications, especially in high pressure and high temperature environments. The safe and stable operation of battery systems requires more reliable technical support.

[0007] Therefore, how to solve the problem that the PCS output power in the existing digital energy storage system has large random variability and high risk of battery thermal runaway, which leads to the inability of the energy storage system to operate safely and stably, is an important issue that needs to be urgently solved in the field of risk assessment and management. Summary of the invention

[0008] The present invention provides a reliability assessment method and device for a digital energy storage system taking into account power variation and data reconstruction error, so as to overcome the defects of the existing digital energy storage system, such as large random variation of PCS output power and high risk of battery thermal runaway, which lead to the inability of the energy storage system to operate safely and stably, and realize the safe and stable operation of the battery system under high pressure and high temperature environment.

[0009] On the one hand, the present invention provides a reliability assessment method for a digital energy storage system taking into account power changes and data reconstruction errors, comprising: constructing a two-dimensional state probability model that characterizes the influence of power size and power fluctuation intensity of an energy storage inverter in a digital energy storage system, and solving the energy storage inverter failure rate under target working conditions based on the two-dimensional state probability model, component thermal stress factors of the energy storage inverter, and component temperature cycle factors; determining the data reconstruction error based on energy storage battery variables of a battery cluster in the digital energy storage system and their corresponding output eigenvectors, and calculating the thermal runaway probability of the battery cluster based on the data reconstruction error; wherein the output eigenvector is obtained by data reconstruction based on the energy storage battery variables based on a long short-term memory neural network; and performing an operation reliability assessment on the digital energy storage system based on the energy storage inverter failure rate and the thermal runaway probability of the battery cluster to obtain an assessment result.

[0010] Furthermore, based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage converter, the step of solving the failure rate of the energy storage converter under the target working condition specifically includes: constructing a unified failure rate model according to the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor; solving the failure rate of the energy storage converter under the target working condition according to the unified failure rate model.

[0011] Furthermore, the energy storage battery variables include at least voltage, current, state of charge and temperature, and each energy storage battery variable is constructed with a corresponding thermal runaway warning basic model; accordingly, the data reconstruction error is determined based on the energy storage battery variables of the battery cluster in the digital energy storage system and their corresponding output feature vectors, including: obtaining different energy storage battery variables of the battery cluster, and extracting features of different energy storage battery variables based on a pre-trained long short-term memory neural network to obtain output feature vectors corresponding to different energy storage battery variables; based on the pre-trained thermal runaway warning basic model, according to different energy storage battery variables and their corresponding output feature vectors, predicting the data reconstruction errors of different energy storage battery variables.

[0012] Furthermore, the step of calculating the thermal runaway probability of the battery cluster based on the data reconstruction error specifically includes: comparing the data reconstruction error of different energy storage battery variables with a preset threshold to obtain multiple comparison results; determining the thermal runaway probability of the battery cluster based on the multiple comparison results; wherein the preset threshold is set according to the variable value when thermal runaway occurs in the battery cluster.

[0013] Furthermore, the unified failure rate model is defined as follows: ; in, represents the failure rate of the energy storage converter, Indicates the power status. Indicates the power fluctuation intensity state, Indicates the number of power divisions, Indicates the number of power fluctuation intensity divisions, represents a two-dimensional state probability model, Indicates that the energy storage converter is in power state Component thermal stress factor under Indicates that the energy storage converter is in a power fluctuation intensity state Component temperature cycle factor under , They represent the basic failure rate of components corresponding to the thermal stress factor of components and the temperature cycle factor of components, respectively. Indicates the impact of component manufacturing quality on energy storage converters, Indicates the impact of reliability quality management and control level on the life cycle of components of energy storage converters, represents the overstress contribution factor of the components of the energy storage converter. Further, the two-dimensional state probability model is defined as follows: ; The component thermal stress factor is defined as follows: ; The component temperature cycle factor is defined as follows: ; in, Indicates that the energy storage converter is in power state , power fluctuation intensity status The accumulated time under Indicates the operating cycle, represents a constant, represents the activation energy constant, Indicates the power status of the components of the energy storage converter The temperature parameters are as follows: Indicates the junction temperature fluctuation cycle time of the components of the energy storage converter, Indicates the power fluctuation intensity of the components of the energy storage converter The cumulative running time under Indicates the power fluctuation intensity of the components of the energy storage converter The number of junction temperature cycle fluctuations under Indicates the running status Junction temperature fluctuation cycle time, Indicates the reference temperature fluctuation amplitude, Indicates the power fluctuation intensity of the components of the energy storage converter The junction temperature fluctuation amplitude under It indicates the maximum value of the junction temperature fluctuation of the components of the energy storage converter in each operating state.

[0014] In a second aspect, the present invention further provides a digital energy storage system reliability assessment device taking into account power changes and data reconstruction errors, including: an energy storage inverter reliability assessment module, which is used to construct a two-dimensional state probability model that characterizes the influence of power size and power fluctuation intensity of the energy storage inverter in the digital energy storage system, and based on the two-dimensional state probability model, the thermal stress factor of the components of the energy storage inverter and the temperature cycle factor of the components, solve the energy storage inverter failure rate under the target working condition; a battery cluster reliability assessment module, which is used to determine the data reconstruction error based on the energy storage battery variables of the battery cluster in the digital energy storage system and their corresponding output characteristic vectors, and calculate the thermal runaway probability of the battery cluster according to the data reconstruction error; wherein the output characteristic vector is obtained by data reconstruction based on the energy storage battery variables based on a long short-term memory neural network; a digital energy storage system operation reliability assessment module, which is used to evaluate the operation reliability of the digital energy storage system according to the energy storage inverter failure rate and the thermal runaway probability of the battery cluster to obtain an evaluation result.

[0015] In a third aspect, the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for evaluating the reliability of a digital energy storage system taking into account power variation and data reconstruction error as described in any one of the above is implemented.

[0016] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a digital energy storage system reliability assessment method taking into account power variation and data reconstruction error as described in any one of the above.

[0017] In a fifth aspect, the present invention further provides a computer program product, comprising a computer program, which, when executed by a processor, implements a reliability assessment method for a digital energy storage system taking into account power variation and data reconstruction error as described in any one of the above.

[0018] The reliability assessment method of a digital energy storage system considering power change and data reconstruction error provided by the present invention constructs a two-dimensional state probability model that characterizes the influence of the power size and power fluctuation intensity of the energy storage converter in the digital energy storage system, and solves the energy storage converter failure rate under the target working condition based on the two-dimensional state probability model, the component thermal stress factor of the energy storage converter and the component temperature cycle factor, and then determines the data reconstruction error based on the energy storage battery variables of the battery cluster in the digital energy storage system and its corresponding output characteristic vector, and calculates the thermal runaway probability of the battery cluster according to the data reconstruction error; wherein the output characteristic vector is obtained by reconstructing the data according to the energy storage battery variables based on the long short-term memory neural network, so as to evaluate the operation reliability of the digital energy storage system according to the energy storage converter failure rate and the thermal runaway probability of the battery cluster, and obtain an evaluation result. The method realizes the operation reliability assessment of the digital energy storage system by dividing the operation state of the energy storage converter in the digital energy storage system into two-dimensional states, and considering the power size, power fluctuation intensity and data reconstruction error at the same time, and ensures the safe and stable operation of the battery system under high pressure and high temperature environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 It is a schematic diagram of the overall structure of the digital energy storage system provided by an embodiment of the present invention.

[0021] Figure 2 It is a flow chart of a method for evaluating the reliability of a digital energy storage system taking into account power variation and data reconstruction error provided by an embodiment of the present invention.

[0022] Figure 3 It is a schematic diagram of a framework for energy storage battery data reconstruction based on a bidirectional long short-term memory neural network provided in an embodiment of the present invention.

[0023] Figure 4 It is a schematic diagram of the overall process of a method for evaluating the reliability of a digital energy storage system taking into account power variation and data reconstruction error provided by an embodiment of the present invention.

[0024] Figure 5It is a structural schematic diagram of a digital energy storage system reliability assessment device taking into account power variation and data reconstruction error provided by an embodiment of the present invention.

[0025] Figure 6 It is a schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] It should be noted that due to the influence of the operating environment and operating conditions, compared with general industrial PCS, the output power of PCS for battery energy storage systems is more random and unstable. Studies have shown that due to the changes in junction temperature and fluctuation intensity, the aluminum bonding wires and solder layers of each power device in the PCS are subjected to long-term and frequent expansion and contraction, resulting in the generation of unbalanced electrothermal stress and fatigue accumulation, which is the root cause of the breakage of the aluminum bonding wires and cracking of the solder layers of power devices.

[0028] Since the operating power of the energy storage PCS is directly related to the junction temperature of the power device, the fluctuation intensity of the PCS operating power is related to the junction temperature fluctuation intensity of its components. Therefore, comprehensive consideration of the two key factors of power size and power fluctuation intensity can accurately reflect the impact of real-time power random changes on the size and fluctuation of the junction temperature of PCS components. By combining the actual energy storage system PCS operation data, the reliability of the energy storage system considering the impact of power operation changes is proposed, which is of great significance for formulating a reasonable energy storage system operation and maintenance strategy.

[0029] In addition, the main concern of battery energy storage systems is the reliability of the batteries themselves. However, the reliability of battery energy storage itself often involves a time span of several hours or even dozens of hours, so the battery energy storage data needs to be reconstructed and calculated.

[0030] At the same time, in order to solve the problem of thermal runaway of batteries, a data-driven battery energy storage reliability technology is proposed. By adopting advanced algorithms such as Bi-LSTM, BPP, CNN, etc., and considering long-term and short-term reconstruction data, the reliability assessment calculation process is enhanced.

[0031] Based on the above ideas, the present invention proposes a reliability assessment method for a digital energy storage system considering power changes and data reconstruction errors. Before describing in detail the reliability assessment method for a digital energy storage system considering power changes and data reconstruction errors provided by an embodiment of the present invention, the overall structure of the digital energy storage system will first be briefly described here.

[0032] Specifically, Figure 1 The overall structural diagram of the digital energy storage system provided by the embodiment of the present invention is shown.

[0033] like Figure 1 As shown in the figure, the container of the digital energy storage system consists of 4 battery clusters and 4 PCSs. The battery cluster of a digital energy storage system consists of several battery modules connected in a 3-in-14-in-series (3P14S) dynamically reconfigurable battery network. Among them, the battery module is a lithium iron phosphate battery, which is composed of 16 battery cells connected in series. Since the battery cells are fixedly connected to form a module, and the modules are connected in a reconfigurable manner, the battery energy storage system can achieve module-level charge and discharge balance and fault isolation control. The digital energy storage system integrates PCS, battery energy switch (BCB), battery energy hub (BMMB), battery energy adapter (BPIB), battery energy network card (BDB) and other equipment.

[0034] according to Figure 1 It can be observed that the main factors affecting the reliability of the digital energy storage system involve the reliability of the battery cluster and the reliability of the PCS, which constitute the key components of the system reliability. In view of this, the present invention takes these two key aspects as the starting point to comprehensively analyze the operational reliability of the digital energy storage system.

[0035] Figure 2 The figure shows a flow chart of a method for evaluating the reliability of a digital energy storage system taking into account power variation and data reconstruction error provided by an embodiment of the present invention. Figure 2 As shown, the method includes steps S210-S230, and steps S210-S230 and related steps will be described in detail below.

[0036] S210, constructing a two-dimensional state probability model that characterizes the influence of the power size and power fluctuation intensity of the energy storage converter in the digital energy storage system, and solving the energy storage converter failure rate under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage converter.

[0037] It is easy to understand that the energy storage converter (i.e. PCS mentioned above) will be affected by the power size and power fluctuation intensity during operation. These factors will cause its internal components to bear uneven electrical and thermal stress, which in turn affects the accuracy of its reliability assessment.

[0038] In this regard, this embodiment establishes a two-dimensional state probability model of the energy storage converter based on a multi-state probability analysis method to characterize the influence of the power size and power fluctuation intensity of the PCS, which is specifically defined as follows: Formula (1).

[0039] (1).

[0040] In formula (1), represents a two-dimensional state probability model, Indicates that the energy storage converter is in power state , power fluctuation intensity status The accumulated time under Indicates the operating cycle, Indicates the power status. Indicates the power fluctuation intensity state, Indicates the number of power divisions, Indicates the number of power fluctuation intensity divisions.

[0041] The power fluctuation intensity can be defined as follows (2).

[0042] (2).

[0043] In formula (2), Indicates the power fluctuation intensity, represents the current data point, represents the number of data points, Represents data points The power value, Indicates the average power value.

[0044] At the same time, in order to measure the influence of junction temperature and junction temperature fluctuation on the failure rate of PCS components, this embodiment uses thermal stress factor and temperature cycle factor as measurement factors.

[0045] In a specific embodiment, according to the FIDES reliability guidelines, the PCS is in a power level of state (i.e. power level state ) is defined as the following formula (3).

[0046] (3).

[0047] In formula (3), Indicates a constant. Different components have different corresponding values. represents the activation energy constant, Indicates the power status of the components of the energy storage converter The following temperature parameters.

[0048] In another specific embodiment, according to the FIDES reliability guideline, the PCS is in a state where the power fluctuation intensity is the first state (i.e., power fluctuation intensity state ) is defined as the component temperature cycle factor under the following formula (4).

[0049] (4).

[0050] In formula (4), Indicates the junction temperature fluctuation cycle time of the components of the energy storage converter, Indicates the power fluctuation intensity of the components of the energy storage converter The cumulative running time under Indicates the power fluctuation intensity of the components of the energy storage converter The number of junction temperature cycle fluctuations under Indicates the running status Junction temperature fluctuation cycle time, Indicates the reference temperature fluctuation amplitude, Indicates the power fluctuation intensity of the components of the energy storage converter The junction temperature fluctuation amplitude under It indicates the maximum value of the junction temperature fluctuation of the components of the energy storage converter in each operating state.

[0051] It should be noted that many parameters in formula (4) are obtained by analyzing the stress cycle fluctuation of junction temperature of components in PCS using the rain flow counting method.

[0052] Subsequently, based on the two-dimensional state probability model, component thermal stress factor and component temperature cycle factor constructed above, a unified failure rate model under specific working conditions is constructed. By solving the unified failure rate model, the failure rate of the energy storage converter under the target working conditions (specific power size state, specific power fluctuation intensity state) can be obtained.

[0053] Specifically, considering that the factors affecting PCS components include not only the real-time power size of the PCS but also the power fluctuation, this embodiment divides the PCS into two-dimensional states, combines the thermal stress factor and the temperature cycle factor, and constructs a unified failure rate model as shown in the following formula (5) according to the FIDES reliability guideline.

[0054] (5).

[0055] In formula (5), represents the failure rate of the energy storage converter, Indicates the power status. Indicates the power fluctuation intensity state, Indicates the number of power divisions, Indicates the number of power fluctuation intensity divisions, represents a two-dimensional state probability model, Indicates that the energy storage converter is in power state Component thermal stress factor under Indicates that the energy storage converter is in a power fluctuation intensity state Component temperature cycle factor under , They represent the basic failure rate of components corresponding to the thermal stress factor of components and the temperature cycle factor of components, respectively. Indicates the impact of component manufacturing quality on energy storage converters, Indicates the impact of reliability quality management and control level on the life cycle of components of energy storage converters, Represents the overstress contribution factor of the components of the energy storage converter.

[0056] Therefore, step S210 can solve and obtain the failure rate of the energy storage converter under the target working condition, and further, step S220 is executed.

[0057] S220, determining a data reconstruction error based on energy storage battery variables of a battery cluster in a digital energy storage system and their corresponding output feature vectors, and calculating a thermal runaway probability of the battery cluster based on the data reconstruction error; wherein the output feature vector is obtained by data reconstruction based on the energy storage battery variables based on a long short-term memory neural network.

[0058] It is easy to understand that the reliability of battery energy storage itself often involves a time span of several hours or even dozens of hours. Therefore, it is necessary to reconstruct and calculate the battery energy storage variables. The energy storage battery variables include at least voltage U, state of charge SOC, temperature T and current I.

[0059] In a specific embodiment, in order to realize the reconstruction calculation of battery energy storage variables, a corresponding thermal runaway warning basic model is constructed for each energy storage battery variable, that is, a corresponding thermal runaway warning basic model is constructed for voltage U, state of charge SOC, temperature T and current I. The thermal runaway warning basic model takes the input feature vector and the output feature vector as the model input, and takes the data reconstruction error between the input feature vector and the output feature vector as the model output. Among them, the input feature vector is obtained by standardizing the collected energy storage battery variables, and the output feature vector is obtained by data reconstruction based on the long short-term memory neural network according to the energy storage battery variables.

[0060] Specifically, since the energy storage battery variables have different dimensions and have large differences in values, in order to ensure the training accuracy of the thermal runaway warning basic model, the collected energy storage battery variables can be preprocessed first. For example, preferably, this embodiment uses the z-score standardization method to use the sample mean and standard deviation to perform data preprocessing, as shown in the following formula (6).

[0061] (6).

[0062] In formula (6), is the original energy storage battery variable, is the preprocessed energy storage battery variable, represents the sample mean, Represents the sample standard deviation. As one of the inputs of the corresponding thermal runaway warning basic model.

[0063] At the same time, the long short-term memory neural network is used to extract the features of the collected energy storage battery variables to obtain the output feature vector, which is used as the second input of the corresponding thermal runaway warning basic model.

[0064] Specifically, the thermal runaway process of the battery cluster in the digital energy storage system often involves a time span of several hours or even more than ten hours. Therefore, in the process of data reconstruction, the thermal runaway warning basic model needs to have the ability to process long-time span data. In order to effectively improve the feature mining capability of the thermal runaway warning basic model, this embodiment proposes a battery cluster data feature mining method based on a bidirectional long short-term memory neural network (Bi-LSTM), which can effectively capture the features of long-time span data. For example, a two-layer long short-term memory neural network is combined to obtain a basic model, and then the collected energy storage battery variables are used as the input of the basic model, and the output feature vector can be obtained through feature extraction.

[0065] Figure 3 FIG. 1 shows a schematic diagram of a framework for energy storage battery data reconstruction based on a bidirectional long short-term memory neural network provided by an embodiment of the present invention. Figure 3 As shown, for the energy storage battery variables Voltage , state of charge SOC, temperature T and current I are reconstructed to obtain the corresponding output feature vector ,Right now , , , .

[0066] The basic model obtained by combining two layers of long short-term memory neural network, its hidden layer output at time can be seen in the following equations (7)-(9).

[0067] (7).

[0068] (8).

[0069] (9).

[0070] In equations (7)-(9), Represents the operation process of the bidirectional long short-term memory neural network Bi-LSTM, for The forward hidden layer output at time t is given by Time input , , , The hidden layer output at the previous moment It is calculated, which takes into account the impact of past hidden layer outputs on current hidden layer outputs. for The reverse hidden layer output at time t is given by Time input , , , And the hidden layer output at the next moment The calculation is done, which takes into account the impact of future hidden layer outputs on the current hidden layer outputs. represents the updated hidden layer output, ,, Represents the bias constant.

[0071] After processing the energy storage battery variables to obtain the input feature vector and the output feature vector, the input feature vector and the output feature vector are input into the constructed thermal runaway warning basic model to obtain the output data reconstruction error. .

[0072] It should be noted that when training and optimizing the thermal runaway warning basic model, the optimization goal is to minimize the output data and the reconstruction error of the output data. For details, please refer to the following formula (10).

[0073] (10).

[0074] In formula (10), Represents input data With output data The reconstruction error, express and The dimension of express Middle Dimensions in The element value at the moment, express Middle Dimensions in The element value at the moment.

[0075] Whether it is the training and optimization stage of the thermal runaway warning basic model or the application stage, it is necessary to use a bidirectional long short-term memory neural network to reconstruct the energy storage battery data.

[0076] Subsequently, the data reconstruction error is compared with a preset threshold to obtain a comparison result, and then the thermal runaway probability of the battery cluster is obtained, which can be seen in the following formula (11).

[0077] (11).

[0078] In formula (11), It is the judgment result of the battery cluster. When the value is 1, it indicates that the battery cluster has a thermal runaway risk and the thermal runaway probability is 1; when the value is 0, it indicates that the battery cluster does not have a thermal runaway risk and the thermal runaway probability is 0. It is the judgment threshold of thermal runaway of battery cluster. is the total number of basic models, the total number of thermal runaway warning basic models and the total number of energy storage battery variables. For example, in one embodiment, the energy storage battery variables include voltage, current, state of charge and temperature, then The value is 4, the total number of basic models used for data reconstruction is 4, and the total number of thermal runaway warning basic models used to predict data reconstruction errors is also 4.

[0079] The stability of the basic model for thermal runaway warning is defined as whether it can accurately warn of battery thermal runaway in different test data sets. Since the parameters of the basic model for thermal runaway warning are greatly affected by the battery input sample set (voltage U, state of charge SOC, temperature T, current I, etc.), it is difficult to ensure the stability of the basic model for thermal runaway warning. The basic model for thermal runaway warning will be used many times in the thermal runaway warning of energy storage batteries. If its stability is insufficient, it will increase the probability of misjudgment and missed judgment in certain warning processes. Ensemble learning can enhance the stability of the model by integrating the judgment results of multiple basic models for thermal runaway warning to judge thermal runaway. The probability of thermal runaway can be obtained as follows (12).

[0080] (12).

[0081] In formula (12), Indicates the number of the basic model / current thermal runaway warning basic model / energy storage battery variable, Indicates the total number of basic model / current thermal runaway warning basic model / energy storage battery variables, Indicates Basic model / current thermal runaway warning basic model / thermal runaway judgment result (thermal runaway probability) of energy storage battery variables.

[0082] Therefore, step S220 can calculate the thermal runaway probability of the battery cluster, and further, step S230 is executed.

[0083] S230: According to the failure rate of the energy storage converter and the thermal runaway probability of the battery cluster, an operation reliability evaluation is performed on the digital energy storage system to obtain an evaluation result.

[0084] It is easy to understand that this embodiment considers the PCS power change and the battery cluster data reconstruction error to construct a digital energy storage system operation reliability assessment model, which can be specifically seen in the following formula (13).

[0085] (13).

[0086] In formula (13), is the failure probability of the digital energy storage system, i.e., the evaluation result of the operational reliability evaluation, Represents the probability of failure associated with auxiliary equipment.

[0087] According to formula (13), this embodiment not only considers the thermal runaway probability related to the battery cluster , and the failure probability associated with the PCS , also considering the failure probability associated with auxiliary equipment In the digital energy storage system, auxiliary equipment includes battery energy switch (BCB), battery energy hub (BMMB), battery energy adapter (BPIB) and battery energy network card (BDB). These auxiliary equipment play a vital role in the overall operation of the system, and their potential failure probability must be considered to ensure that the performance and reliability of the system reach the optimal level.

[0088] Thus, the reliability evaluation results of the digital energy storage system can be obtained.

[0089] It should be noted that the execution order of step S210 of solving the energy storage converter failure rate under the target operating condition and step S220 of calculating the thermal runaway probability can be either first or later, and is not specifically limited here.

[0090] In this embodiment, a two-dimensional state probability model is constructed to characterize the influence of the power size and power fluctuation intensity of the energy storage converter in the digital energy storage system, and based on the two-dimensional state probability model, the thermal stress factor of the components of the energy storage converter and the temperature cycle factor of the components, the failure rate of the energy storage converter under the target working condition is solved, and then the data reconstruction error is determined based on the energy storage battery variables of the battery cluster in the digital energy storage system and their corresponding output characteristic vectors, and the thermal runaway probability of the battery cluster is calculated according to the data reconstruction error; wherein, the output characteristic vector is obtained by data reconstruction based on the energy storage battery variables based on the long short-term memory neural network, so as to evaluate the operation reliability of the digital energy storage system according to the failure rate of the energy storage converter and the thermal runaway probability of the battery cluster, and obtain the evaluation result. This method realizes the operation reliability evaluation of the digital energy storage system by dividing the operation state of the energy storage converter in the digital energy storage system into two-dimensional states, and taking into account the power size, power fluctuation intensity and data reconstruction error, and ensures the safe and stable operation of the battery system under high pressure and high temperature environment.

[0091] In some other embodiments, Figure 4 The overall flow chart of the method for evaluating the reliability of a digital energy storage system taking into account power variation and data reconstruction error provided by an embodiment of the present invention is shown.

[0092] like Figure 4 As shown, first, the original data is collected, including but not limited to battery cluster data (energy storage battery variables), PCS real-time active power, and component junction temperature fluctuation information.

[0093] Then, based on the battery cluster data, step S220 is executed to evaluate the reliability of the battery cluster to obtain the thermal runaway probability of the battery cluster; based on the PCS real-time active power and component junction temperature fluctuation information, step S210 is executed to evaluate the PCS reliability to obtain the energy storage converter failure rate.

[0094] Specifically, in the reliability evaluation process of the battery cluster, the bidirectional long short-term memory neural network Bi-LSTM is used as the basic model. The data is reconstructed to obtain an output feature vector. The training sample set is obtained by preprocessing the original battery cluster data, and the training sample set includes a voltage sample set, a current sample set, a state of charge sample set, and a temperature sample set.

[0095] The bidirectional long short-term memory neural network Bi-LSTM is used to reconstruct data for the voltage sample set, current sample set, state of charge sample set and temperature sample set respectively to obtain the output feature vectors corresponding to the voltage, current, state of charge and temperature.

[0096] Figure 4 The termination condition in is , that is, the data reconstruction of the four energy storage battery variables, namely voltage, current, state of charge and temperature, has been completed.

[0097] Then, the training sample set and its corresponding input feature vectors are input into the integrated model (i.e., the thermal runaway warning basic model described above). For example, the voltage sample set and its corresponding thermal runaway warning basic model 1, the current sample set and its corresponding thermal runaway warning basic model 2, the state of charge sample set and its corresponding thermal runaway warning basic model 3, and the temperature sample set and its corresponding thermal runaway warning basic model 4 can obtain the data reconstruction errors corresponding to the voltage, current, state of charge and temperature.

[0098] Next, the data reconstruction errors of voltage, current, state of charge and temperature are compared with the preset thresholds to obtain the single thermal runaway probability corresponding to voltage, current, state of charge and temperature. Finally, the single thermal runaway probability corresponding to voltage, current, state of charge and temperature is integrated to obtain the final thermal runaway probability of the battery cluster. The battery cluster reliability assessment process can refer to the above formulas (6)-(12).

[0099] In the process of PCS reliability evaluation, the two-dimensional state probability model obtained by two-dimensional state division and the junction temperature information obtained by the rain flow counting method are input into the unified model of PCS component failure rate to obtain the PCS failure probability, that is, the energy storage converter failure rate. The PCS reliability evaluation process can refer to the above formulas (1)-(5).

[0100] Finally, based on the thermal runaway probability of the battery cluster and the PCS failure probability, the integrated failure probability of the PCS and battery cluster of the digital energy storage system is obtained, that is, the final evaluation result, which can be seen in formula (13) above.

[0101] Corresponding to the digital energy storage system reliability assessment method considering power variation and data reconstruction error described in the above embodiments, the present invention also proposes a digital energy storage system reliability assessment device considering power variation and data reconstruction error.

[0102] Specifically, Figure 5 A schematic diagram of the structure of a digital energy storage system reliability assessment device taking into account power variation and data reconstruction error provided by an embodiment of the present invention is shown.

[0103] like Figure 5As shown, the device includes: an energy storage converter reliability evaluation module 510, which is used to construct a two-dimensional state probability model that characterizes the influence of the power size and power fluctuation intensity of the energy storage converter in the digital energy storage system, and solves the energy storage converter failure rate under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage converter; a battery cluster reliability evaluation module 520, which is used to determine the data reconstruction error based on the energy storage battery variables of the battery cluster in the digital energy storage system and their corresponding output feature vectors, and calculate the thermal runaway probability of the battery cluster according to the data reconstruction error; wherein the output feature vector is obtained by data reconstruction based on the energy storage battery variables based on the long short-term memory neural network; a digital energy storage system operation reliability evaluation module 530, which is used to evaluate the operation reliability of the digital energy storage system according to the energy storage converter failure rate and the thermal runaway probability of the battery cluster to obtain an evaluation result.

[0104] In this embodiment, a two-dimensional state probability model characterizing the power size and power fluctuation intensity of the energy storage converter in the digital energy storage system is constructed by the energy storage converter reliability evaluation module 510, and based on the two-dimensional state probability model, the thermal stress factor of the components of the energy storage converter and the temperature cycle factor of the components, the failure rate of the energy storage converter under the target working condition is solved, and then the battery cluster reliability evaluation module 520 determines the data reconstruction error based on the energy storage battery variables of the battery cluster in the digital energy storage system and the corresponding output feature vector, and calculates the thermal runaway probability of the battery cluster according to the data reconstruction error; wherein the output feature vector is obtained by data reconstruction based on the energy storage battery variables based on the long short-term memory neural network, so that the digital energy storage system operation reliability evaluation module 530 evaluates the operation reliability of the digital energy storage system according to the energy storage converter failure rate and the thermal runaway probability of the battery cluster, and obtains the evaluation result. The device realizes the operation reliability evaluation of the digital energy storage system by dividing the operation state of the energy storage converter in the digital energy storage system into two-dimensional states, and taking into account the power size, power fluctuation intensity and data reconstruction error, and ensures the safe and stable operation of the battery system under high pressure and high temperature environment.

[0105] It should be noted that the digital energy storage system reliability assessment device considering power variation and data reconstruction error provided in the embodiment of the present invention and the digital energy storage system reliability assessment method considering power variation and data reconstruction error described in the above embodiments can correspond to each other and will not be repeated here.

[0106] Figure 6 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 6As shown, the electronic device may include: a processor (processor) 610 , a communication interface (Communications Interface) 620 , a memory (memory) 630 and a communication bus 640 , wherein the processor 610 , the communication interface 620 , and the memory 630 communicate with each other through the communication bus 640 . The processor 610 can call the logic instructions in the memory 630 to execute the reliability assessment method of the digital energy storage system considering power changes and data reconstruction errors, the method comprising: constructing a two-dimensional state probability model that characterizes the influence of the power size and power fluctuation intensity of the energy storage inverter in the digital energy storage system, and solving the energy storage inverter failure rate under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage inverter; determining the data reconstruction error based on the energy storage battery variables of the battery cluster in the digital energy storage system and their corresponding output characteristic vectors, and calculating the thermal runaway probability of the battery cluster according to the data reconstruction error; wherein the output characteristic vector is obtained by data reconstruction based on the energy storage battery variables based on the long short-term memory neural network; and performing an operation reliability assessment on the digital energy storage system according to the energy storage inverter failure rate and the thermal runaway probability of the battery cluster to obtain an assessment result.

[0107] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0108] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the digital energy storage system reliability assessment method considering power changes and data reconstruction errors provided by the above methods, the method comprising: constructing a two-dimensional state probability model that characterizes the influence of power size and power fluctuation intensity of the energy storage inverter in the digital energy storage system, and solving the energy storage inverter failure rate under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage inverter; determining the data reconstruction error based on the energy storage battery variables of the battery cluster in the digital energy storage system and their corresponding output characteristic vectors, and calculating the thermal runaway probability of the battery cluster based on the data reconstruction error; wherein the output characteristic vector is obtained by reconstructing data based on the energy storage battery variables based on a long short-term memory neural network; and performing an operation reliability assessment on the digital energy storage system based on the energy storage inverter failure rate and the thermal runaway probability of the battery cluster to obtain an assessment result.

[0109] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the processor executes the digital energy storage system reliability assessment method considering power changes and data reconstruction errors provided by the above methods, the method comprising: constructing a two-dimensional state probability model that characterizes the influence of power size and power fluctuation intensity of the energy storage inverter in the digital energy storage system, and solving the energy storage inverter failure rate under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage inverter; determining the data reconstruction error based on the energy storage battery variables of the battery cluster in the digital energy storage system and their corresponding output characteristic vectors, and calculating the thermal runaway probability of the battery cluster based on the data reconstruction error; wherein the output characteristic vector is obtained by data reconstruction based on the energy storage battery variables based on the long short-term memory neural network; and performing an operation reliability assessment on the digital energy storage system based on the energy storage inverter failure rate and the thermal runaway probability of the battery cluster to obtain an assessment result.

[0110] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0111] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A reliability assessment method for a digital energy storage system considering power variation and data reconstruction error, characterized in that: include: Construct a two-dimensional state probability model that characterizes the influence of the power size and power fluctuation intensity of the energy storage converter in the digital energy storage system, and solve the failure rate of the energy storage converter under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage converter; Determine a data reconstruction error based on energy storage battery variables of a battery cluster in a digital energy storage system and their corresponding output feature vectors, and calculate a thermal runaway probability of the battery cluster based on the data reconstruction error; wherein the output feature vector is obtained by data reconstruction based on the energy storage battery variables based on a long short-term memory neural network; According to the failure rate of the energy storage converter and the thermal runaway probability of the battery cluster, an operational reliability assessment is performed on the digital energy storage system to obtain an assessment result.

2. The method for evaluating the reliability of a digital energy storage system considering power variation and data reconstruction error according to claim 1, characterized in that: The step of solving the failure rate of the energy storage converter under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage converter specifically includes: Constructing a unified failure rate model based on the two-dimensional state probability model, component thermal stress factors, and component temperature cycle factors; According to the unified failure rate model, the failure rate of the energy storage converter under the target working condition is solved.

3. The method for evaluating the reliability of a digital energy storage system considering power variation and data reconstruction error according to claim 1, characterized in that: The energy storage battery variables include at least voltage, current, state of charge and temperature, and each energy storage battery variable is constructed with a corresponding thermal runaway warning basic model; Accordingly, the data reconstruction error is determined based on the energy storage battery variables of the battery cluster in the digital energy storage system and the corresponding output feature vectors, including: Obtain different energy storage battery variables of the battery cluster, and extract features of the different energy storage battery variables based on a pre-trained long short-term memory neural network to obtain output feature vectors corresponding to the different energy storage battery variables; Based on the pre-trained thermal runaway warning basic model, the data reconstruction errors of different energy storage battery variables are predicted according to different energy storage battery variables and their corresponding output feature vectors.

4. The method for evaluating the reliability of a digital energy storage system considering power variation and data reconstruction error according to claim 3 is characterized in that: The step of calculating the thermal runaway probability of the battery cluster according to the data reconstruction error specifically includes: Comparing the data reconstruction errors of different energy storage battery variables with preset thresholds to obtain multiple comparison results; Determine the thermal runaway probability of the battery cluster based on multiple comparison results; The preset threshold is set according to a variable value when thermal runaway occurs in the battery cluster.

5. The method for evaluating the reliability of a digital energy storage system considering power variation and data reconstruction error according to claim 2, characterized in that: The unified failure rate model is defined as follows: ； in, represents the failure rate of the energy storage converter, Indicates the power status. Indicates the power fluctuation intensity state, Indicates the number of power divisions, Indicates the number of power fluctuation intensity divisions, represents a two-dimensional state probability model, Indicates that the energy storage converter is in power state Component thermal stress factor under Indicates that the energy storage converter is in a power fluctuation intensity state Component temperature cycle factor under , They represent the basic failure rate of components corresponding to the thermal stress factor of components and the temperature cycle factor of components, respectively. Indicates the impact of component manufacturing quality on energy storage converters, Indicates the impact of reliability quality management and control level on the life cycle of components of energy storage converters, Represents the overstress contribution factor of the components of the energy storage converter.

6. The method for evaluating the reliability of a digital energy storage system considering power variation and data reconstruction error according to claim 5 is characterized in that: The two-dimensional state probability model is defined as follows: ； The component thermal stress factor is defined as follows: ； The component temperature cycle factor is defined as follows: ； in, Indicates that the energy storage converter is in power state , power fluctuation intensity status The accumulated time under Indicates the operating cycle, represents a constant, represents the activation energy constant, Indicates the power status of the components of the energy storage converter The temperature parameters are as follows: Indicates the junction temperature fluctuation cycle time of the components of the energy storage converter, Indicates the power fluctuation intensity of the components of the energy storage converter The cumulative running time under Indicates the power fluctuation intensity of the components of the energy storage converter The number of junction temperature cycle fluctuations under Indicates the running status Junction temperature fluctuation cycle time, Indicates the reference temperature fluctuation amplitude, Indicates the power fluctuation intensity of the components of the energy storage converter The junction temperature fluctuation amplitude under It indicates the maximum value of the junction temperature fluctuation of the components of the energy storage converter in each operating state.

7. A reliability assessment device for a digital energy storage system taking into account power variation and data reconstruction error, characterized in that: include: The energy storage converter reliability evaluation module is used to construct a two-dimensional state probability model that characterizes the power size and power fluctuation intensity of the energy storage converter in the digital energy storage system, and solve the energy storage converter failure rate under the target working condition based on the two-dimensional state probability model, the component thermal stress factor and the component temperature cycle factor of the energy storage converter; A battery cluster reliability assessment module, used to determine a data reconstruction error based on energy storage battery variables of a battery cluster in a digital energy storage system and their corresponding output feature vectors, and calculate the thermal runaway probability of the battery cluster based on the data reconstruction error; wherein the output feature vector is obtained by data reconstruction based on the energy storage battery variables based on a long short-term memory neural network; The digital energy storage system operation reliability assessment module is used to assess the operation reliability of the digital energy storage system according to the energy storage converter failure rate and the thermal runaway probability of the battery cluster to obtain an assessment result.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the digital energy storage system reliability assessment method considering power variation and data reconstruction error is implemented as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for evaluating the reliability of a digital energy storage system taking into account power variation and data reconstruction error is implemented as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for evaluating the reliability of a digital energy storage system taking into account power variation and data reconstruction error is implemented as described in any one of claims 1 to 6.

Images