Risk early warning method and device for energy storage power station

Abstract

The invention belongs to the field of electric power, and discloses a risk early warning method and device for an energy storage power station, and the method comprises the steps: collecting the multi-source heterogeneous operation data of a battery cell level, a battery cluster level, a battery cabin level and a system level in the energy storage power station, and carrying out the preprocessing of the multi-source heterogeneous operation data; extracting a dynamic feature set related to battery fault evolution from the preprocessed data, and constructing a state vector used for representing a system safety state based on the dynamic feature set; constructing a system state space model based on a dynamic fault evolution process, and initializing model parameters based on test data; on the basis of the system state space model and the real-time operation data, predicting a future evolution path of the state vector, and calculating the probability that the evolution path enters a predefined fault area and the early warning time; and determining a current early warning level according to the probability and the early warning time, and triggering a linkage control strategy corresponding to the early warning level.

Description

Technical Field

[0001] This invention belongs to the field of power, and in particular relates to a risk warning method and device for energy storage power stations. Background Technology

[0002] With the rapid development of the new energy industry, electrochemical energy storage power stations, as an important means of energy storage and regulation, play a crucial role in grid peak shaving and new energy consumption. However, during long-term operation, energy storage batteries are affected by multiple factors, including electrochemical characteristics, environmental factors, and operating conditions, resulting in safety risks such as thermal runaway, internal short circuits, and overcharging / over-discharging. Once a malfunction occurs, it may lead to serious accidents such as fires and explosions, causing huge economic losses and safety hazards.

[0003] In related technologies, the safety monitoring of energy storage power stations mainly adopts threshold alarm methods, that is, an alarm is triggered when a certain parameter (such as voltage or temperature) exceeds a preset threshold. The limitations of this method are: first, the alarm is delayed, and when the parameter exceeds the threshold, the fault has often entered the acceleration stage, leaving a very short time window for emergency response; second, it lacks the ability to dynamically predict the fault evolution process and cannot achieve early warning; and third, the fusion of multi-source data is insufficient, making it difficult to accurately identify abnormal states under complex operating conditions. Summary of the Invention

[0004] In view of this, the present invention discloses a risk warning method and device for energy storage power stations, which can solve the shortcomings of related technologies.

[0005] To achieve the above objectives, the present invention discloses the following technical solution: According to a first aspect of the present invention, a risk early warning method for an energy storage power station is proposed, the method comprising: Collect multi-source heterogeneous operation data at the battery cell level, battery cluster level, battery compartment level and system level in the energy storage power station, and preprocess the multi-source heterogeneous operation data; Extract a dynamic feature set related to battery fault evolution from the preprocessed data, and construct a state vector to characterize the system's safety state based on the dynamic feature set; A system state-space model based on the dynamic fault evolution process is constructed, and the model parameters are initialized based on test data; wherein, the state-space model is used to describe the evolution law of the state vector over time; Based on the system state space model and real-time operating data, the future evolution path of the state vector is predicted, and the probability and warning time of the evolution path entering the predefined fault region are calculated. The current warning level is determined based on the probability and warning time, and the corresponding linkage control strategy is triggered.

[0006] According to a second aspect of the present invention, a risk warning device for an energy storage power station is provided, the device comprising: Data Acquisition Unit: Collects multi-source heterogeneous operation data at the individual battery cell level, battery cluster level, battery compartment level, and system level in the energy storage power station, and preprocesses the multi-source heterogeneous operation data; The first construction unit extracts a dynamic feature set related to battery fault evolution from the preprocessed data, and constructs a state vector to characterize the system's safety state based on the dynamic feature set. The second building unit: constructs a system state-space model based on the dynamic fault evolution process, and initializes the model parameters based on test data; wherein, the state-space model is used to describe the evolution law of the state vector over time; Prediction Unit: Based on the system state space model and real-time operating data, predicts the future evolution path of the state vector, and calculates the probability and warning time of the evolution path entering the predefined fault region; Triggering unit: Determines the current warning level based on the probability and warning time, and triggers the linkage control strategy corresponding to the warning level.

[0007] According to a third aspect of the present invention, an electronic device is provided, comprising: processor; Memory used to store processor-executable instructions; The processor implements the steps of the method as described in the first aspect by running the executable instructions.

[0008] According to a fourth aspect of the invention, a computer-readable storage medium is provided having computer instructions stored thereon that, when executed by a processor, implement the steps of the method as described in the first aspect.

[0009] As can be seen from the above technical solutions, the risk warning method for energy storage power stations disclosed in this invention is as follows: On the one hand, by tracking dynamic characteristics and predicting evolution paths, this method overcomes the lag inherent in traditional static threshold alarms, achieving proactive early warning and shifting the warning window from "after the fault occurs" to "during the fault evolution," or even "before the fault occurs." On the other hand, it finely divides fault evolution into multiple levels, achieving full coverage from early anomalies to emergency faults, and uses probability P and warning time TTA for quantitative evaluation, making the warning information more scientific and accurate. Furthermore, this method is a dynamic and adaptive intelligent system capable of learning fault evolution patterns and making predictions, rather than relying on fixed, isolated judgment rules, thereby improving the intelligence level of the early warning system. Attached Figure Description

[0010] Figure 1This is a flowchart of a risk warning method for an energy storage power station provided in an exemplary embodiment; Figure 2 This is a schematic structural diagram of a device provided in an exemplary embodiment; Figure 3 This is a block diagram of a risk warning device for an energy storage power station, provided as an exemplary embodiment. Detailed Implementation

[0011] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with one or more embodiments of the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of one or more embodiments of the present invention as detailed in the appended claims.

[0012] It should be noted that the steps of the corresponding methods in other embodiments are not necessarily performed in the order shown and described in this invention. In some other embodiments, the methods may include more or fewer steps than those described in this invention. Furthermore, a single step described in this invention may be broken down into multiple steps in other embodiments; and multiple steps described in this invention may be combined into a single step in other embodiments.

[0013] With the rapid development of the new energy industry, electrochemical energy storage power stations, as an important means of energy storage and regulation, play a crucial role in grid peak shaving and new energy consumption. However, during long-term operation, energy storage batteries are affected by multiple factors, including electrochemical characteristics, environmental factors, and operating conditions, resulting in safety risks such as thermal runaway, internal short circuits, and overcharging / over-discharging. Once a malfunction occurs, it may lead to serious accidents such as fires and explosions, causing huge economic losses and safety hazards.

[0014] In related technologies, the safety monitoring of energy storage power stations mainly adopts threshold alarm methods, that is, an alarm is triggered when a certain parameter (such as voltage or temperature) exceeds a preset threshold. The limitations of this method are: first, the alarm is delayed, and when the parameter exceeds the threshold, the fault has often entered the acceleration stage, leaving a very short time window for emergency response; second, it lacks the ability to dynamically predict the fault evolution process and cannot achieve early warning; and third, the fusion of multi-source data is insufficient, making it difficult to accurately identify abnormal states under complex operating conditions.

[0015] To address the shortcomings of related technologies, this invention proposes a risk warning method for energy storage power stations.

[0016] Figure 1This is a flowchart illustrating a risk warning method for an energy storage power station, as provided in an exemplary embodiment. Figure 1 As shown, the method may include the following steps: Step 101: Collect multi-source heterogeneous operation data at the battery cell level, battery cluster level, battery compartment level and system level in the energy storage power station, and preprocess the multi-source heterogeneous operation data.

[0017] The system collects multi-source heterogeneous operational data at the battery cell, battery cluster, battery compartment, and system levels within the energy storage power station through the Battery Management System (BMS), Environmental Monitoring System (EMS), and Energy Management System (EMS). Specifically, this includes: battery cell level data such as voltage, temperature, current, and internal resistance (sampling frequency 1Hz); battery cluster level data such as voltage, current, and temperature (sampling frequency 1Hz); battery compartment level data such as ambient temperature, smoke concentration, and hydrogen concentration (sampling frequency 0.1Hz); and system level data such as State of Charge (SOC), State of Health (SOH), and charge / discharge power (sampling frequency 0.1Hz).

[0018] By integrating monitoring data from different levels and types, a more complete picture of the system's safety status is constructed, avoiding missed or false alarms caused by the failure of a single data source or insufficient information, thereby improving the comprehensiveness and reliability of the early warning system. Furthermore, multi-source heterogeneous operational data enhances the ability to capture early, weak anomaly signals: for example, changes in battery internal resistance and consistency are often early signs of internal faults, preceding significant changes in macroscopic parameters such as voltage and temperature.

[0019] Step 102: Extract the dynamic feature set related to battery fault evolution from the preprocessed data, and construct a state vector to characterize the system safety state based on the dynamic feature set.

[0020] Step 103: Construct a system state-space model based on the dynamic fault evolution process, and initialize the model parameters based on the test data; wherein, the state-space model is used to describe the evolution law of the state vector over time.

[0021] Step 104: Based on the system state space model and real-time operating data, predict the future evolution path of the state vector, and calculate the probability and warning time of the evolution path entering the predefined fault region; Step 105: Determine the current warning level based on the probability and warning time, and trigger the linkage control strategy corresponding to the warning level. In this embodiment, on the one hand, by tracking dynamic features and predicting evolution paths, the lag inherent in traditional static threshold alarms is overcome, enabling proactive early warning. The warning window is moved forward from "after the fault occurs" to "during the fault evolution," or even "before the fault occurs." On the other hand, fault evolution is finely divided into multiple levels, achieving full coverage from early anomalies to emergency faults. Quantitative evaluation is performed using probability P and warning time time (TTA), making the warning information more scientific and accurate. Furthermore, this method is a dynamic and adaptive intelligent system capable of learning fault evolution patterns and making predictions, rather than relying on fixed, isolated judgment rules, thereby improving the intelligence level of the early warning system.

[0022] In one embodiment, the preprocessing of the multi-source heterogeneous operating data includes data cleaning and multi-source data time alignment. Specifically, data cleaning is performed first to remove invalid data that exceeds the normal range (such as voltage less than 2.5V or greater than 4.2V, temperature less than or equal to 20℃ or greater than 60℃). Missing data is filled using linear interpolation, and Kalman filtering is used for data smoothing. Then, multi-source data time alignment is performed: all data are given a unified timestamp, and the sampling time of the battery cell data (1Hz) is used as the reference time axis. Slowly changing data (such as ambient temperature and SOC) are time aligned using the nearest neighbor hold method. When the system state changes drastically (such as a sudden change in charging and discharging power exceeding 20%), forced alignment is triggered.

[0023] In this embodiment, common problems such as noise, missing data, and asynchrony in field data are effectively addressed, ensuring the accuracy and reliability of subsequent feature extraction and model prediction, and improving data quality. The proposed unified timescale and state-triggered resampling strategy solves the problem of synchronizing multi-source data in industrial fields, making the method more adaptable to complex real-world engineering environments.

[0024] In one embodiment, the dynamic feature set includes at least one of the following: First-order or second-order derivative characteristics, including voltage rate of change, temperature rate of change and their second derivative; Consistent degradation characteristics include the standard deviation or variance of cell voltage or temperature; Characteristics of performance degradation inflection point, including internal resistance change rate or capacity degradation rate; Pre-fault characteristics include the coupling relationship between the rate of temperature rise, the rate of voltage change, and the gas production rate and temperature rise under specific operating conditions. Extract a dynamic feature set related to battery fault evolution from the preprocessed data, including: The eigenvector z_1(t) of the first / second derivative mainly includes the first / second derivative of the voltage change rate dV / dt and d 2 V / dt 2The first / second derivative of the rate of temperature change, dT / dt, and d 2 T / dt 2 This represents the change in voltage and temperature per unit time and its acceleration, used to identify slow or accelerating trends in voltage and temperature.

[0025] Establish a consistent degradation characteristic z_2(t), including the standard deviation or variance σ of individual unit voltage and temperature. 2 (V), σ 2 (T) and the standard deviation or variance of the rate of change, etc., to quantify the risk of fault propagation.

[0026] Establish the performance degradation inflection point characteristic z_3(t), which mainly includes the rate of increase of internal resistance dR / dt and the abnormal degradation of capacity dQ / dt.

[0027] The precursor characteristics z_4(t) for different fault types include the rate of temperature rise, the rate of voltage change, and the coupling relationship between gas generation rate and temperature rise in a specific temperature range.

[0028] Construct a state vector based on multi-source data fusion, and define a security early warning feature vector Z(t): Z(t) =[z_1(t), z_2(t), z_3(t), z_4(t)]T.

[0029] In this embodiment, different fault types (such as internal short circuit and overcharge) exhibit different early-stage characteristics. This feature set provides a rich source of information for constructing highly discriminative state vectors, enhancing the model's ability to identify multiple fault types.

[0030] In one embodiment, the system state-space model is: ; Where Z(t+1) and Z(t) are the state vectors of the next time step and the current time step, respectively, A is the state transition matrix, B is the input matrix, u(t) is the input vector characterizing the system operating condition, and w(t) is the process noise; The state transition matrix A maintains multiple local models according to different fault evolution stages. The system activates the corresponding local model to make predictions based on the region where the current state vector Z(t) is located.

[0031] By combining the fault testing process and data, a standard system of characteristic parameters under different fault characteristics is established, as shown in Table 1 below. Different data features can be selected to establish state vectors according to different fault types.

[0032] A system state-space model based on the dynamic fault evolution process is constructed, and a discrete-time state-space model is used to describe the changes in the state vector: ; in, , Let the state vectors be the state vectors for the next and current time steps. The state transition matrix quantifies the coupling and transmission relationships between different fault characteristics. For the input matrix, This indicates the charging and discharging current operating condition. The process noise represents the uncertainty of the model.

[0033] Using test data under different fault types, including charging and discharging currents and state vectors calculated at different times, an online identification method is employed using recursive least squares. Since the state transition matrix is ​​time-varying, multiple local transition matrices are maintained based on the multiple hierarchical stages in the table. , ..., the system based on the current The corresponding local model is activated in the area to make predictions, and the model parameters are initialized based on the test data.

[0034] In this embodiment, on the one hand, the state-space model can quantify the coupling and transmission relationships between different fault characteristics, achieving higher prediction accuracy than simple relational models or single models. On the other hand, by employing multiple local transition matrices (A1, A2, ...) and activating them according to the current state, the model can adapt to the dynamic characteristics of different stages of fault evolution, improving prediction accuracy under different operating conditions.

[0035] In one embodiment, the predefined fault area is divided according to historical fault data or simulation data, and the fault evolution process is divided into multiple safety levels; the warning time is determined by calculating the time point when the predicted trajectory of the state vector first enters the boundary of the fault area; the probability is obtained by calculating the shortest distance between the predicted trajectory and the boundary of the fault area and mapping it using a probability function. As shown in Table 1, the safety levels include at least: early anomaly, progressive deterioration, accelerated danger, and emergency failure; the fault areas are defined for different fault types, including over-discharge, external short circuit, internal short circuit, overcharge, and overheating. Table 1. Characteristic parameter standard system under different fault characteristics

[0036] As shown in Table 2, based on historical fault data or simulation data, the evolution process of different fault types is precisely divided into four safety levels: Level 1 (early anomaly), Level 2 (gradual deterioration), Level 3 (accelerating danger), and Level 4 (emergency fault), and fault areas are defined and divided accordingly. Table 2. Safety levels and corresponding evolution paths in the failure evolution process.

[0037] The subscripts min and max represent the minimum and maximum preset values, d* / dt represents the first derivative, which is obtained by difference during calculation; d²* / dt² represents the second derivative, which is also obtained by difference during calculation; avg(*) represents averaging.

[0038] The state vectors at different times are calculated using the learned system state-space model. Based on actual data, predict the path of fault evolution and quantify the time it takes for a safety risk to enter the boundary of a certain fault region: Calculate the state vector to predict the trajectory The probability of entering each of the above-mentioned fault regions and time This can be achieved by calculating the distance between the predicted point and the boundary of the fault area using a probability function, and by using the time when the predicted trajectory first enters the boundary of the fault area as the warning time. ; ; in, This indicates entering the fault area. The probability, To predict the closest distance between the trajectory and the fault area, The threshold value set, The time series numbering for the predicted trajectory, For time step In one embodiment, the trigger linkage control strategy includes: sending an early warning message to the energy management system (EMS), the early warning message including at least an early warning level, a dominant abnormal parameter, and an early warning time (TTA); and executing a preset strategy corresponding to the early warning level, the preset strategy including prompting inspection, automatic power reduction operation, emergency disconnection of fault clusters, or activation of the fire ventilation system.

[0039] Establish a multi-level early warning criteria dynamic triggering and transfer mechanism, and provide early warning and linkage strategies: The fault area corresponding to the maximum probability value of the calculated fault area is taken as the warning area, and a warning is issued: {Warning level (I, II, III, IV), risk area (over-discharge, external short circuit, internal short circuit, overcharge, overheating), TTA (warning time), P (risk probability)}.

[0040] Once a certain level of warning condition is triggered, the system immediately generates warning information that includes the warning level, the dominant abnormal parameters (such as "accelerated voltage drop"), and the risk quantification value (such as the rate of change). This warning information directly links to the power plant's energy management system (EMS) to execute the preset control strategy corresponding to that level: Level 1 warning prompts inspection, Level 2 warning automatically reduces power output, and Level 3 warning urgently disconnects fault clusters and activates fire protection and ventilation.

[0041] For example: Determine the current warning level based on the probability P(S_i|t) and the warning time TTA: When P(S1|t)>0.3 and TTA>24 hours, a Level 1 warning (early anomaly) is triggered, sending a warning message to the Energy Management System (EMS) to remind inspection personnel to pay close attention; when P(S2|t)>0.5 and TTA>6 hours, a Level 2 warning (gradual deterioration) is triggered, automatically reducing power to 50% of the rated power; when P(S3|t)>0.7 and TTA>1 hour, a Level 3 warning (accelerating danger) is triggered, urgently cutting off the fault cluster and activating the fire ventilation system; when P(S4|t)>0.9, a Level 4 warning (emergency fault) is triggered, immediately cutting off the entire energy storage system and activating the station's fire protection system. The early warning information should include at least the warning level, the main abnormal parameters (such as the rate of temperature change, voltage consistency, etc.) and the warning time (TTA), so that operators can quickly locate the problem and take corresponding measures.

[0042] Figure 2 This is a schematic structural diagram of a device provided in an exemplary embodiment. Please refer to... Figure 2 At the hardware level, the device includes a processor 202, an internal bus 204, a network interface 206, memory 208, and non-volatile memory 210, and may also include other hardware required for its functions. One or more embodiments of the present invention can be implemented in software, for example, the processor 202 reads the corresponding computer program from the non-volatile memory 210 into memory 208 and then runs it. Of course, in addition to software implementation, one or more embodiments of the present invention do not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0043] Please refer to Figure 3 A risk warning device for energy storage power stations can be applied to, for example... Figure 3 The device shown, in order to implement the technical solution of the present invention, includes: The acquisition unit 301 is used to acquire multi-source heterogeneous operation data at the battery cell level, battery cluster level, battery compartment level and system level in the energy storage power station, and to preprocess the multi-source heterogeneous operation data. The first construction unit 302 is used to extract a dynamic feature set related to battery fault evolution from the preprocessed data, and construct a state vector to characterize the system safety state based on the dynamic feature set. The second construction unit 303 is used to construct a system state-space model based on the dynamic fault evolution process and initialize model parameters based on test data; wherein, the state-space model is used to describe the evolution law of the state vector over time; The prediction unit 304 is used to predict the future evolution path of the state vector based on the system state space model and real-time operating data, and to calculate the probability and warning time of the evolution path entering a predefined fault region. Triggering unit 305 is used to determine the current warning level based on the probability and warning time, and trigger the linkage control strategy corresponding to the warning level.

[0044] Optional, multi-source heterogeneous operating data includes: voltage, temperature, current, and internal resistance at the individual battery cell level; voltage, current, and temperature at the battery cluster level; ambient temperature, smoke concentration, and hydrogen concentration at the battery compartment level; and state of charge (SOC), state of health (SOH), and charge / discharge power at the system level. Optionally, the acquisition unit 301 is specifically used for: data cleaning and time alignment of multi-source data; The data cleaning includes: removing invalid data that exceeds the normal range, interpolating and filling in missing data, and smoothing the data using Kalman filtering or sliding window weighted average. The multi-source data time alignment includes: stamping all data with a unified timestamp, using the sampling time of high-frequency data as the reference time axis, employing the nearest neighbor hold method for slowly changing data, and triggering forced alignment when the system state changes drastically. Optionally, the dynamic feature set includes at least one of the following: First-order or second-order derivative characteristics, including voltage rate of change, temperature rate of change and their second derivative; Consistent degradation characteristics include the standard deviation or variance of cell voltage or temperature; Characteristics of performance degradation inflection point, including internal resistance change rate or capacity degradation rate; Pre-fault characteristics include the coupling relationship between the rate of temperature rise, the rate of voltage change, and the gas production rate and temperature rise under specific operating conditions. Optionally, the system state-space model is: ; Where Z(t+1) and Z(t) are the state vectors of the next time step and the current time step, respectively, A is the state transition matrix, B is the input matrix, u(t) is the input vector characterizing the system operating condition, and w(t) is the process noise; The state transition matrix A maintains multiple local models according to different fault evolution stages. The system activates the corresponding local model to make predictions based on the region where the current state vector Z(t) is located. Optional, The predefined fault zones are divided based on historical fault data or simulation data, and the fault evolution process is divided into multiple safety levels. The warning time is determined by calculating the time point when the state vector prediction trajectory first enters the boundary of the fault area. The probability is obtained by calculating the shortest distance between the predicted trajectory and the boundary of the fault area, and then mapping it using a probability function. Optionally, the safety levels include at least: early anomaly, gradual deterioration, accelerating danger, and emergency failure; The fault areas are defined for different fault types, including over-discharge, external short circuit, internal short circuit, overcharge, and overheating.

[0045] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, which can take the form of a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email sending and receiving device, game console, tablet computer, wearable device, or any combination of these devices.

[0046] In a typical configuration, a computer includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0047] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0048] Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0049] For any other form of computer-readable medium (or computer-readable storage medium) as described above, computer instructions may be stored thereon, which, when executed by a processor, implement one or more of the above embodiments, thereby realizing the technical solution of the present invention.

[0050] The present invention also proposes a computer program that, when executed by a processor, implements one or more of the embodiments described above, thereby realizing the technical solution of the present invention. This computer program may be specifically recorded on the above-described or other computer-readable media, and the present invention does not impose any limitations on this.

[0051] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0052] The foregoing has described specific embodiments of the invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0053] The terminology used in one or more embodiments of the invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” used in one or more embodiments of the invention and in the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0054] It should be understood that although the terms first, second, third, etc., may be used to describe various information in one or more embodiments of the present invention, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first information may also be referred to as second information without departing from the scope of one or more embodiments of the present invention, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0055] The above description is merely a preferred embodiment of one or more embodiments of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of the present invention should be included within the protection scope of one or more embodiments of the present invention.

Claims

1. A risk early warning method for an energy storage power station, characterized in that, include: Collect multi-source heterogeneous operation data at the battery cell level, battery cluster level, battery compartment level and system level in the energy storage power station, and preprocess the multi-source heterogeneous operation data; Extract a dynamic feature set related to battery fault evolution from the preprocessed data, and construct a state vector to characterize the system's safety state based on the dynamic feature set; A system state-space model based on the dynamic fault evolution process is constructed, and the model parameters are initialized based on test data; wherein, the state-space model is used to describe the evolution law of the state vector over time; Based on the system state-space model and real-time operating data, the future evolution path of the state vector is predicted, and the probability and warning time of the evolution path entering the predefined fault region are calculated. The predefined fault region is divided according to historical fault data or simulation data, and the fault evolution process is divided into multiple safety levels. The warning time is determined by calculating the time point when the predicted trajectory of the state vector first enters the boundary of the fault region. The probability is obtained by calculating the shortest distance between the predicted trajectory and the boundary of the fault region and mapping it using a probability function. The current warning level is determined based on the probability and warning time, and the corresponding linkage control strategy is triggered.

2. The method according to claim 1, characterized in that, Multi-source heterogeneous operation data includes: voltage, temperature, current, and internal resistance at the individual battery cell level; voltage, current, and temperature at the battery cluster level; ambient temperature, smoke concentration, and hydrogen concentration at the battery compartment level; and state of charge (SOC), state of health (SOH), and charge / discharge power at the system level.

3. The method according to claim 1, characterized in that, The preprocessing of the multi-source heterogeneous operating data includes data cleaning and multi-source data time alignment; The data cleaning includes: removing invalid data that exceeds the normal range, interpolating and filling in missing data, and smoothing the data using Kalman filtering or sliding window weighted average. The multi-source data time alignment includes: stamping all data with a unified timestamp, using the sampling time of high-frequency data as the reference time axis, employing the nearest neighbor hold method for slowly changing data, and triggering forced alignment when the system state changes drastically.

4. The method according to claim 1, characterized in that, The dynamic feature set includes at least one of the following: First-order or second-order derivative characteristics, including voltage rate of change, temperature rate of change and their second derivative; Consistent degradation characteristics include the standard deviation or variance of cell voltage or temperature; Characteristics of performance degradation inflection point, including internal resistance change rate or capacity degradation rate; Pre-fault characteristics include the coupling relationship between the rate of temperature rise, the rate of voltage change, and the gas production rate and temperature rise under specific operating conditions.

5. The method according to claim 1, characterized in that, The system state-space model is as follows: ； Where Z(t+1) and Z(t) are the state vectors of the next time step and the current time step, respectively, A is the state transition matrix, B is the input matrix, u(t) is the input vector characterizing the system operating condition, and w(t) is the process noise; The state transition matrix A maintains multiple local models according to different fault evolution stages. The system activates the corresponding local model to make predictions based on the region where the current state vector Z(t) is located.

6. The method according to claim 1, characterized in that, The safety levels include at least: early anomaly, gradual deterioration, accelerating danger, and emergency failure; The fault areas are defined for different fault types, including over-discharge, external short circuit, internal short circuit, overcharge, and overheating.

7. A risk early warning device for an energy storage power station, characterized in that, The device includes: Data Acquisition Unit: Collects multi-source heterogeneous operation data at the individual battery cell level, battery cluster level, battery compartment level, and system level in the energy storage power station, and preprocesses the multi-source heterogeneous operation data; The first construction unit extracts a dynamic feature set related to battery fault evolution from the preprocessed data, and constructs a state vector to characterize the system's safety state based on the dynamic feature set. The second building unit: constructs a system state-space model based on the dynamic fault evolution process, and initializes the model parameters based on test data; wherein, the state-space model is used to describe the evolution law of the state vector over time; Prediction Unit: Based on the system state-space model and real-time operating data, predicts the future evolution path of the state vector, and calculates the probability and warning time of the evolution path entering a predefined fault region; wherein, the predefined fault region is divided according to historical fault data or simulation data, and the fault evolution process is divided into multiple safety levels; the warning time is determined by calculating the time point when the predicted trajectory of the state vector first enters the boundary of the fault region; the probability is obtained by calculating the shortest distance between the predicted trajectory and the boundary of the fault region and mapping it using a probability function. Triggering unit: Determines the current warning level based on the probability and warning time, and triggers the linkage control strategy corresponding to the warning level.

8. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor implements the steps of the method as described in any one of claims 1-6 by running the executable instructions.

9. A computer-readable storage medium storing computer instructions thereon, characterized in that, When executed by the processor, this instruction implements the steps of the method as described in any one of claims 1-6.

Images