Battery running state monitoring and evaluating method and system, electronic equipment and storage medium

Abstract

The invention provides a battery running state monitoring and evaluating method and system, electronic equipment and a storage medium, and relates to the technical field of battery monitoring. The method comprises the following steps: collecting multi-mode monitoring data of a battery in a charging and discharging process; a multi-view encoder based on deep learning pre-processes the multi-modal monitoring data to obtain a multi-modal fusion feature vector; performing causal chain reasoning on the multi-modal fusion feature vector based on a preset fusion lead-acid storage battery knowledge graph to obtain a battery state causal chain, and performing state evaluation on the lead-acid storage battery based on the battery state causal chain to obtain a real-time state evaluation result; inputting the multi-modal fusion feature vector into a preset battery performance degradation trend estimation model to obtain a battery performance degradation trend; and comprehensively evaluating the state of the lead-acid storage battery based on the real-time state evaluation result and the battery performance degradation trend to obtain a comprehensive evaluation result. According to the method, the accuracy of the state evaluation result of the lead-acid storage battery is improved.

Description

Technical Field

[0001] The present application relates to the field of battery monitoring technology, and in particular to a battery operating status monitoring and evaluation method, system, electronic device and storage medium. Background Art

[0002] Lead-acid batteries are widely used in energy storage power stations, electric vehicles and other fields due to their low cost, high safety and long cycle life. In order to ensure the reliable operation of lead-acid batteries and extend their service life, their operating status needs to be monitored and evaluated in real time.

[0003] In the prior art, battery operation status monitoring and evaluation is mainly carried out by collecting monitoring data of a single mode such as voltage, current, and temperature, and evaluating the battery status based on a preset threshold or empirical model. However, the degradation process of lead-acid batteries is affected by many factors, such as charge and discharge rate, temperature, corrosion, etc. It is difficult to fully characterize the health status of the battery by relying solely on monitoring data of a single mode. In addition, most of the existing evaluation methods are based on simplified empirical models and lack in-depth consideration of the battery degradation mechanism, resulting in insufficient accuracy of the evaluation results. Summary of the invention

[0004] The present application provides a battery operating status monitoring and evaluation method, system, electronic device and storage medium, which are used to improve the accuracy of lead-acid battery status evaluation results.

[0005] In a first aspect, the present application provides a method for monitoring and evaluating a battery operation status, the method comprising: Collect multi-modal monitoring data of lead-acid batteries during the charging and discharging process; A multi-view encoder based on deep learning pre-processes the multi-modal monitoring data to obtain a multi-modal fusion feature vector; Based on the preset fusion lead-acid battery knowledge graph, causal chain reasoning is performed on the multimodal fusion feature vector to obtain a series of battery status causal chains, and the status of the lead-acid battery is evaluated based on the battery status causal chain to obtain real-time status evaluation results; Inputting the multimodal fusion feature vector into a preset battery performance degradation trend estimation model to obtain the battery performance degradation trend; Based on the real-time status evaluation results and the battery performance degradation trend, the status of the lead-acid battery is comprehensively evaluated to obtain a comprehensive evaluation result.

[0006] In the above technical solution, by collecting multimodal monitoring data of lead-acid batteries during the charging and discharging process, the multimodal monitoring data is preprocessed using a multi-view encoder based on deep learning to obtain a multimodal fusion feature vector. This method makes full use of the complementary information between different modal data, realizes effective data fusion through deep learning technology, and thus forms a feature representation that comprehensively describes the battery health status.

[0007] On this basis, this method uses causal chain reasoning to combine the multimodal fusion feature vector with the preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains. The introduction of the knowledge graph not only takes into account the complex causal relationship of the battery degradation process, but also integrates rich battery field knowledge, making the status assessment process more consistent with the physical mechanism of battery degradation. By evaluating the status of lead-acid batteries based on the battery status causal chain, the real-time status assessment results are more accurate and interpretable.

[0008] At the same time, this method inputs the multimodal fusion feature vector into the preset battery performance degradation trend estimation model to obtain the battery performance degradation trend. This model not only takes into account the historical degradation data, but also integrates the field knowledge of degradation mechanism battery state causal chain, which can more accurately characterize the dynamic characteristics and state transition laws of the degradation process. The resulting battery performance degradation trend has stronger long-term stability and extrapolation ability.

[0009] Finally, this method comprehensively considers the real-time status assessment results and the battery performance degradation trend, and comprehensively evaluates the status of the lead-acid battery to obtain a comprehensive assessment result. This multi-dimensional and multi-scale assessment method can fully reflect the health status of the battery and provide a more reliable basis for battery health management and maintenance decisions.

[0010] In a second aspect of the present application, a battery operation status monitoring and evaluation system is provided, the system comprising: A data acquisition module is used to collect multi-modal monitoring data of lead-acid batteries during the charging and discharging process; The feature extraction module is used for preprocessing the multimodal monitoring data based on the deep learning multi-view encoder to obtain the multimodal fusion feature vector; A real-time status assessment module is used to perform causal chain reasoning on the multimodal fusion feature vector based on the preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains, and to perform status assessment on the lead-acid battery based on the battery status causal chain to obtain a real-time status assessment result; A degradation trend prediction module is used to input the multimodal fusion feature vector into a preset battery performance degradation trend estimation model to obtain the battery performance degradation trend; The comprehensive evaluation module is used to comprehensively evaluate the state of the lead-acid battery based on the real-time state evaluation result and the battery performance degradation trend to obtain a comprehensive evaluation result.

[0011] In a third aspect of the present application, a computer storage medium is provided, wherein the computer storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the above method steps.

[0012] In the fourth aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the above method.

[0013] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. This application collects multimodal monitoring data of lead-acid batteries during the charging and discharging process, and uses a multi-view encoder based on deep learning to pre-process the multimodal monitoring data to obtain a multimodal fusion feature vector. This method makes full use of the complementary information between different modal data, and realizes effective data fusion through deep learning technology, thereby forming a feature representation that comprehensively describes the battery health status.

[0014] 2. This application uses causal chain reasoning on the multimodal fusion feature vector and the preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains. The introduction of the knowledge graph not only takes into account the complex causal relationship of the battery degradation process, but also integrates rich battery field knowledge, making the status assessment process more in line with the physical mechanism of battery degradation. By evaluating the status of lead-acid batteries based on the battery status causal chain, the real-time status assessment results obtained have higher accuracy and interpretability.

[0015] 3. This application comprehensively considers the real-time status assessment results and battery performance degradation trends, conducts a comprehensive assessment of the status of the lead-acid battery, and obtains a comprehensive assessment result. This multi-dimensional and multi-scale assessment method can fully reflect the health status of the battery and provide a more reliable basis for battery health management and maintenance decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a battery operating status monitoring and evaluation method provided in an embodiment of the present application; Figure 2 An architecture diagram of a battery operation status monitoring and evaluation system provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0017] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0018] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0019] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0020] In order to facilitate understanding of the method and system provided by the embodiments of the present application, before introducing the embodiments of the present application, the background of the embodiments of the present application is first introduced.

[0021] Lead-acid batteries have been widely used in energy storage power stations, electric vehicles and other fields due to their low cost, excellent safety performance, long cycle life and many other advantages. In order to ensure the reliable operation of lead-acid batteries and extend their service life, it is necessary to monitor and evaluate their operating status in real time.

[0022] At present, battery operation status monitoring and evaluation technology mainly collects monitoring data of a single mode such as voltage, current, and temperature, and evaluates the battery status based on a pre-set threshold or empirical model. However, the degradation process of lead-acid batteries is a complex physical and chemical process, which is affected by multiple factors such as charge and discharge rate, temperature, and corrosion. It is difficult to fully and accurately characterize the health status of the battery by relying solely on monitoring data of a single mode. In addition, most of the existing evaluation methods are based on simplified empirical models, lacking in-depth consideration and modeling of battery degradation mechanisms, resulting in the need to improve the accuracy and reliability of the evaluation results.

[0023] After the background introduction of the above content, those skilled in the art can understand the problems existing in the prior art. The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0024] Based on the above background technology, please refer to Figure 1 , Figure 1 A flow chart of a battery operation status monitoring and evaluation method provided in an embodiment of the present application. The system can be implemented by a computer program or can be run as an independent tool application. Specifically, in an embodiment of the present application, the method can be applied on a server, but can also be applied to electronic devices such as servers. A battery operation status monitoring and evaluation method includes the following steps: S101, collecting multimodal monitoring data of the lead-acid battery during the charging and discharging process; Specifically, during the charge and discharge process of the lead-acid battery, multiple sensors deployed on the battery synchronously collect multiple physical quantities such as the battery's voltage, current, temperature, and internal resistance to form time-synchronized multimodal monitoring data. Among them, the voltage sensor is used to measure the terminal voltage of the battery, the current sensor is used to measure the battery's charge and discharge current, the temperature sensor is used to measure the battery's surface temperature, and the internal resistance sensor is used to measure the battery's internal resistance. These sensors collect data in real time at a certain sampling frequency and transmit the data to the data acquisition unit for synchronization and storage.

[0025] By collecting multi-modal monitoring data, the state characteristics of lead-acid batteries can be reflected from different angles. For example, voltage data reflects the battery's state of charge and polarization, current data reflects the battery's charge and discharge rate and power characteristics, temperature data reflects the battery's thermal stability and heat dissipation performance, and internal resistance data reflects the battery's conductivity and health level. There is a certain correlation and complementarity between these different modal data, and fusion analysis can obtain more accurate and comprehensive battery status information.

[0026] S102, preprocessing the multimodal monitoring data based on a multi-view encoder based on deep learning to obtain a multimodal fusion feature vector; Specifically, the multimodal monitoring data is first standardized to unify the data of different modes into the same scale and range to eliminate the influence of dimensional differences. Then, the standardized multimodal monitoring data is input into the pre-trained deep learning multi-view encoder. The encoder consists of multiple sub-encoders, each of which corresponds to a data modality, such as voltage sub-encoder, current sub-encoder, temperature sub-encoder and internal resistance sub-encoder. Each sub-encoder uses a deep learning model such as a convolutional neural network (CNN) or a long short-term memory network (LSTM), and maps the original time series data to a low-dimensional feature space through multi-layer nonlinear transformation and feature extraction to obtain the feature vector of the modality.

[0027] After obtaining the feature vectors of each modality, the multi-view encoder performs weighted fusion of these feature vectors through the attention mechanism to obtain a multi-modal fusion feature vector. The attention mechanism can adaptively adjust the weights of different modal features, highlight features with strong expressiveness and large information content, and suppress redundant and noisy features, thereby obtaining a more refined and effective feature representation.

[0028] By using a deep learning multi-view encoder to pre-process multimodal monitoring data, we can fully explore the potential associations and complementary information between different modal data and learn more abstract and high-level feature representations. This multimodal fusion feature vector not only has lower dimensions and more concentrated information, but also has stronger semantic expression ability and noise robustness, which helps to improve the accuracy and efficiency of subsequent state assessment.

[0029] Based on the above embodiment, as an optional embodiment, a multi-view encoder based on deep learning pre-processes the multi-modal monitoring data to obtain a multi-modal fusion feature vector, including: S201, unifying the multimodal monitoring data to the same scale and distribution to obtain preprocessed multimodal monitoring data; Specifically, for each modality of monitoring data, first calculate its mean and standard deviation, and then use the standardization (Z-score) method to transform the data scale. Subtract the mean of the modality data from each data point, and then divide it by the standard deviation to obtain standardized data with a mean of 0 and a standard deviation of 1. In this way, the numerical range of different modality data is unified to the same scale, eliminating the impact of dimension.

[0030] Next, the standardized data is transformed. The distribution characteristics of each modality data, such as skewness and kurtosis, are observed to determine whether it conforms to the normal distribution. For data that deviates greatly from the normal distribution, methods such as Box-Cox transformation are used to correct the distribution so that it is as close to the normal distribution as possible. This distribution transformation can improve the statistical characteristics of the data and is conducive to the learning and generalization of deep learning models.

[0031] After completing the scale transformation and distribution transformation, the preprocessed multimodal monitoring data is obtained. These data are consistent in numerical range and distribution characteristics, and the heterogeneity between different modalities is significantly weakened. This unified data representation is not only conducive to subsequent feature extraction and fusion, but also can accelerate the training convergence of deep learning models and improve the learning effect and robustness of the model.

[0032] S202, performing multi-view feature extraction on the pre-processed multi-modal monitoring data to obtain a multi-modal fusion feature vector.

[0033] Specifically, for the preprocessed multimodal monitoring data, three parallel feature extractors are designed to extract features from the three perspectives of time domain, frequency domain and spatial domain. The time domain feature extractor uses the long short-term memory network (LSTM) to extract the dynamic change characteristics of the data through recursive learning of time series data, and characterize the law of battery status evolution over time. The frequency domain feature extractor uses the convolutional neural network (CNN) to extract the frequency domain features of the data through convolution and pooling operations on the spectrum graph, and capture the abnormal patterns of the battery status on the spectrum. The spatial domain feature extractor uses the graph convolutional network (GCN) to extract the spatial topological features of the data through convolution operations on the battery structure graph, and mine the association patterns between the components inside the battery.

[0034] In the feature extraction process, different feature extraction strategies are used for monitoring data of different modes. For one-dimensional time series data such as voltage and current, the time domain feature extractor is mainly used for processing. For frequency domain data such as impedance spectrum and EIS, the frequency domain feature extractor is mainly used for processing. For structured data such as battery structural parameters and material properties, the spatial domain feature extractor is mainly used for processing. Through targeted feature extraction, the characteristics of different modal data can be fully utilized to extract more effective and comprehensive state features.

[0035] After completing the multi-view feature extraction, the feature vectors of the three views are obtained. These feature vectors are combined in a certain splicing or fusion method to form a multimodal fusion feature vector. This feature vector integrates the feature information of the three views of time domain, frequency domain, and spatial domain, and comprehensively characterizes the battery state from multiple dimensions, with stronger representation and distinction capabilities. This multi-view fusion feature can effectively improve the accuracy of subsequent state evaluation and prediction, and enhance the generalization and robustness of the model.

[0036] S103, performing causal chain reasoning on the multimodal fusion feature vector based on the preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains, and performing status evaluation on the lead-acid battery based on the battery status causal chain to obtain a real-time status evaluation result; Specifically, a fused lead-acid battery knowledge graph is first constructed. The knowledge graph formally represents the key components, degradation mechanisms, health status, etc. of lead-acid batteries in the form of an ontology, and constructs causal reasoning rules through semantic associations between ontologies. For example, the knowledge graph contains causal reasoning rules such as "lead sulfate crystallization-> active material shedding-> capacity decay". Then, the multimodal fusion feature vector obtained in step S102 is mapped to the knowledge graph, and the nodes and relationships most relevant to the current feature vector are found through feature matching and semantic mapping.

[0037] Next, starting from the matched node, the preset causal reasoning rules are used to perform forward and backward reasoning to obtain a series of battery status causal chains. For example, based on the causal chain of "high temperature -> electrolyte evaporation -> resistance increase", it can be inferred that the current battery may be in a high temperature environment, and the electrolyte is evaporating, resulting in increased internal resistance. These causal chains reflect the health status and degradation mechanism of the battery from different angles, and have important diagnostic and predictive value.

[0038] Finally, based on the obtained battery status causal chain, the various health indicators of the battery are comprehensively analyzed to obtain real-time status assessment results. Specifically, according to the key nodes and paths in the causal chain, the current health status of the battery is judged, such as normal, mild degradation, moderate degradation or severe degradation, and the corresponding health metric score is given. At the same time, according to the degradation mechanism and trend reflected in the causal chain, the future degradation trend of the battery is predicted and the remaining service life is estimated.

[0039] Based on the above embodiment, as an optional embodiment, causal chain reasoning is performed on the multimodal fusion feature vector based on the preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains, including: S301, linking and matching the multimodal fusion feature vector with entities and attributes in a preset fusion lead-acid battery knowledge graph to obtain a mapping relationship network between the multimodal fusion feature vector and the preset fusion lead-acid battery knowledge graph; Specifically, we first build a fusion lead-acid battery knowledge graph, covering concepts, entities, attributes and relationships related to battery health status assessment. This knowledge graph organizes domain knowledge in the form of ontology, including knowledge entities and relationships in multiple aspects such as battery structure, materials, performance parameters, degradation mechanism, failure mode, etc. By extracting and integrating knowledge from domain literature, expert knowledge, experimental data, etc., we build a comprehensive, accurate and consistent knowledge representation and reasoning framework.

[0040] Then, each feature in the multimodal fusion feature vector is semantically linked to the entities and attributes in the knowledge graph. By calculating the similarity between the feature and the entity and attribute, the most relevant knowledge element for each feature is found, and the mapping relationship between the feature and the knowledge is established. For example, the capacity feature can be linked to entities such as "rated capacity" and "capacity attenuation" in the knowledge graph, and the internal resistance feature can be linked to entities such as "ohmic internal resistance" and "polarization internal resistance". Through this semantic link, the data-driven feature representation is connected to the knowledge-driven concept representation, realizing the fusion of the feature space and the knowledge space.

[0041] After completing the link matching of features and knowledge, a mapping relationship network of multimodal fusion feature vectors and fusion lead-acid battery knowledge graph is obtained. This mapping relationship network uses the knowledge graph as the skeleton and the feature vectors as nodes, and organically combines the two through semantic links. Each feature node establishes a semantic association with the relevant entities and attributes in the knowledge graph through links, reflecting the physical meaning and causal role of the feature in the battery health status. This feature link based on the knowledge graph provides important prior knowledge and constraints for subsequent causal reasoning, which helps to improve the accuracy and explainability of reasoning.

[0042] S302, performing causal chain path search based on the mapping relationship network to obtain a candidate causal chain set, and performing causal strength calculation on the causal chains in the candidate causal chain set to obtain the causal strength of each causal chain in the candidate causal chain set; Specifically, we first search for causal chain paths based on the mapping relationship network. Taking the feature node in the multimodal fusion feature vector as the starting point, we use graph search algorithms such as breadth-first search to find all possible paths from the starting node to the target state node in the knowledge graph. These paths represent the causal chains that the feature nodes may affect through the entities and relationships in the knowledge graph. By setting constraints such as path length and direction, we control the depth and breadth of the search and balance the integrity and complexity of the causal chain. During the search process, we score the paths using the relevance weights of the semantic links, and prioritize paths with high relevance. Finally, we get a set of candidate causal chains, each of which represents a possible causal influence path.

[0043] Then, for each causal chain in the set of candidate causal chains, its causal strength is calculated. Causal strength measures the degree and possibility of the influence of the causal chain on the state change, and it is necessary to comprehensively consider multiple factors such as the semantic relevance, knowledge reliability, and data support of the causal chain. First, the semantic relevance score of the entire causal chain is calculated using the semantic link weights in the causal chain. Secondly, the reliability of the domain knowledge on which the causal chain depends is evaluated by combining the confidence of entities and relationships in the knowledge graph. Thirdly, the causal chain is instantiated using multimodal monitoring data, and the support of the data for the causal chain is calculated. Finally, the causal strength value of each causal chain is obtained by combining the above factors and weighted summation and other methods. The higher the causal strength, the greater the influence of the causal chain on the state change and the higher the possibility of occurrence.

[0044] After completing the causal chain search and causal strength calculation, a set of candidate causal chains with strength attributes is obtained. This set reveals the possible causes and impact paths of state changes from multiple perspectives, providing rich causal basis for state assessment. By analyzing the strength rankings of different causal chains, the key influencing factors that are most likely to cause the current state can be identified, thereby guiding subsequent decision-making and intervention. At the same time, the semantic interpretability of the causal chain makes the evaluation results more transparent and credible, which helps experts analyze and verify the model output.

[0045] S303, sorting the causal chains based on causal strength, and selecting the first N causal chains with the largest causal strength as a series of battery status causal chains.

[0046] Specifically, first, all causal chains in the candidate causal chain set are sorted in descending order according to causal strength. Causal strength reflects the degree and possibility of each causal chain's influence on the state change. The greater the causal strength, the more likely the causal chain is to be the key cause of the current state. By sorting in descending order, the causal chains with high causal strength are placed in front, and the causal chains with low causal strength are placed in the back. This sorting method helps to quickly identify the most important and relevant causal chains, which facilitates the subsequent Top-N selection. Then, from the sorted causal chain set, the top N causal chains with the largest causal strength are selected as the battery state causal chain. The value of N can be set according to specific problems and requirements, and a smaller value is usually selected to control the number of selected causal chains. These N causal chains represent the cause paths that have the greatest impact on the current state of the battery and are most likely to cause state changes, reflecting the main influencing factors and key processes of state changes. By focusing on these key causal chains, the causal mechanism of state changes can be highly summarized and condensed, and the efficiency and accuracy of subsequent state evaluation and decision-making can be improved.

[0047] After completing the causal chain sorting and Top-N selection, a series of battery status causal chains are obtained. These causal chains are prioritized according to importance and relevance with causal strength as weight, representing the key influencing paths of battery status changes. These causal chains can be presented to experts or decision makers through visualization or natural language description to explain the main causes and influencing factors of the current battery status. The extraction and interpretation of this causal chain helps to enhance the comprehensibility and credibility of the status assessment results, and provide clear guidance and basis for subsequent maintenance decisions and intervention measures.

[0048] Based on the above embodiment, as an optional embodiment, before performing causal chain reasoning on the multimodal fusion feature vector based on the preset fusion lead-acid battery knowledge graph to obtain a series of battery state causal chains, it also includes: Acquire lead-acid battery knowledge text related to lead-acid batteries, and build a lead-acid battery ontology knowledge base based on the lead-acid battery knowledge text; Obtain sample data of lead-acid battery monitoring materials, and map the sample data of lead-acid battery monitoring materials to the ontology framework based on the lead-acid battery ontology knowledge base, forming a semantic network containing concept nodes, attribute edges, and relationship edges, and obtaining the preset fused lead-acid battery knowledge graph.

[0049] Specifically, knowledge texts related to lead-acid batteries are obtained, including textbooks, papers, patents, standards, reports and other document forms. Text mining and natural language processing are performed on the acquired knowledge texts to extract key concepts, attributes, relationships and other elements in the text. Ontology construction methods, such as ontology development 101 method and IDEF5 method, are used to organize the extracted knowledge elements into a hierarchical and structured concept system. On the basis of the concept system, the attribute associations and semantic relationships between concepts are defined to form a complete ontology diagram. The ontology diagram is formally represented as standard ontology languages ​​such as OWL and RDF and stored in the ontology database. Through the above process, an ontology knowledge base covering the core concepts in the field of lead-acid batteries is constructed, which describes the composition structure, material properties, working principle, failure mechanism and other aspects of lead-acid batteries, and forms a conceptual paradigm and semantic specification for describing lead-acid batteries.

[0050] Then, various monitoring materials and sample data of the lead-acid battery monitoring site are obtained, including multi-source heterogeneous data such as production logs, process parameters, online monitoring data, and energy spectrum images. These monitoring materials and sample data are semantically annotated and linked, and mapped to the conceptual framework of the ontology knowledge base. For structured parameters, logs and other data, they can be directly mapped to the corresponding data type attributes in the ontology. For unstructured text, image and other data, named entity recognition, relationship extraction and other technologies are used to identify the key concepts and relationships in the text and image, and link them to the corresponding concept nodes and relationship edges in the ontology knowledge base. Through semantic mapping and linking, the fusion of prior knowledge and data instances is realized, forming a semantic network containing concept nodes, attribute edges, and relationship edges. Among them, the concept nodes represent the core concepts of various physical quantities, components, materials, processes, etc. involved in battery health management; the attribute edges characterize the specific assignment of concept attributes; and the relationship edges characterize the semantic associations between concepts. This fused semantic network is the preset fused lead-acid battery knowledge graph.

[0051] Based on the above embodiment, as an optional embodiment, the state of the lead-acid battery is evaluated based on the battery state causal chain to obtain a real-time state evaluation result, including: S401, matching the multimodal monitoring data with the battery status causal chain to obtain key indicator values ​​in the multimodal monitoring data, and comparing and calculating the key indicator values ​​with corresponding state nodes in the battery status causal chain to obtain the degree of deviation between the key indicator values ​​and a preset state threshold; Specifically, first match the multimodal monitoring data with the battery status causal chain. Traverse each causal chain and identify the monitoring data types and features related to it. For example, the capacity attenuation causal chain may correspond to monitoring data such as voltage, current, and temperature, while the internal resistance increase causal chain may correspond to monitoring data such as voltage, temperature, and EIS. Through semantic mapping and data type matching, a correspondence between the state nodes in the causal chain and the features in the monitoring data is established. This matching process associates the abstract causal chain with the specific monitoring data, laying the foundation for subsequent indicator extraction and comparison calculation.

[0052] Then, the key indicator values ​​related to the causal chain are extracted from the matching monitoring data. For each state node, identify the key monitoring indicators that best reflect the state, such as capacity, internal resistance, voltage, etc. Utilize statistical analysis, signal processing and other methods to calculate the real-time values ​​or statistical values ​​of these key indicators from the original monitoring data. For example, the capacity value is extracted from the charge and discharge curve, the internal resistance value is extracted from the EIS spectrum, and the voltage mean and fluctuation range are extracted from the voltage time series. The extracted key indicator values ​​represent the actual performance of the battery at the corresponding state node and quantify the key characteristics of the battery state.

[0053] Next, the extracted key indicator values ​​are compared and calculated with the preset thresholds of the corresponding state nodes in the battery state causal chain. Each state node is set with a normal working range or threshold, which reflects the health level of the state. By comparing the size relationship between the key indicator value and the preset threshold, the degree to which the current battery state deviates from the normal level is judged. For example, if the capacity indicator value is lower than the preset capacity lower limit threshold, it indicates that the battery capacity is severely attenuated; if the internal resistance indicator value is higher than the preset internal resistance upper limit threshold, it indicates that the battery internal resistance has increased significantly. By calculating the difference or ratio between the key indicator value and the threshold, the indicator deviation, that is, the degree of abnormality of the battery at the corresponding state node, can be quantitatively characterized.

[0054] Based on the above embodiment, as an optional embodiment, the sample data of the lead-acid battery inspection materials is mapped to the ontology framework based on the lead-acid battery ontology knowledge base to form a semantic network including concept nodes, attribute edges, and relationship edges, and obtain a preset fused lead-acid battery knowledge graph, including: Preprocess the sample data of lead-acid battery monitoring information to obtain lead-acid battery monitoring data; Specifically, the sample data of the lead-acid battery monitoring data is cleaned. Check the noise data such as outliers, outliers, missing values, etc. in the data, use algorithms such as anomaly detection and isolation forest to identify and process outliers, use methods such as mean filling and regression interpolation to estimate and fill missing values, filter and correct low-quality and inconsistent data, and improve the validity and reliability of the data. At the same time, check the consistency of the data, identify conflicts and contradictions between data from different sources and in different formats, and use data fusion, knowledge reasoning and other technologies to eliminate data conflicts and ensure data consistency. Then, the cleaned monitoring data sample data is converted and standardized. Convert unstructured text, image and other data into structured attribute values, such as extracting keywords in the text as attribute labels and extracting regional features in the image as attribute vectors. Normalize and dimensionless the numerical attributes to eliminate the impact of differences in dimensions and numerical ranges. Encode the categorical attributes and convert the qualitative description into quantitative representation. Through attribute valueization and quantification, the structured representation of the original monitoring data sample data is achieved.

[0055] After completing data cleaning and conversion, semantic annotation is performed on the structured monitoring data. Based on the concepts, attributes, and relationships in the lead-acid battery ontology knowledge base, semantic annotation is performed on the attribute items of the monitoring data, and the attribute names are mapped to ontology concepts, and the attribute values ​​are mapped to concept instances to form annotation knowledge at the semantic level. Using technologies such as ontology reasoning and semantic similarity calculation, the semantic consistency between the annotation knowledge and the ontology knowledge base is checked, and ambiguous or conflicting annotation results are disambiguated and corrected. Through annotation, the monitoring data is given clear semantics to prepare for subsequent knowledge fusion.

[0056] Finally, the semantically annotated monitoring data is semantically linked and fused according to the ontology to form a semantic network containing concept nodes, attribute edges, and relationship edges. The annotated attribute items are linked to the ontology concept nodes, with the attribute name as the attribute edge and the attribute value as the weight of the attribute edge. For the annotated concept instances, they are linked to the corresponding concept nodes, and semantic associations are established with other concept instances based on the relationship edges defined in the ontology. Through the links, the scattered monitoring data are connected in series into a closely related semantic network. This fused semantic network is the preset fused lead-acid battery knowledge graph.

[0057] Based on the lead-acid battery ontology knowledge base, the lead-acid battery monitoring data is mapped, the entities in the lead-acid battery monitoring data are mapped to the corresponding ontology concept nodes in the ontology knowledge base, and the attributes in the structured data are mapped to the attribute edges of the ontology concept nodes; Specifically, the sample data of the lead-acid battery monitoring data are preprocessed. The sample data of the monitoring data usually comes from various links such as battery production, use, and maintenance, including structured parameters, log data, and unstructured text and image data. These heterogeneous data are uniformly cleaned, filtered, converted, and processed. For structured data, data quality is improved through methods such as data completion, outlier correction, and consistency verification. For unstructured data, text parsing, image segmentation, and other technologies are used to extract key information elements. On the basis of data cleaning, according to the data chart definition of the ontology knowledge base, the monitoring data is processed by data type conversion, data normalization, etc., so that it meets the data format requirements of the ontology mapping. At the same time, security measures such as data desensitization and access control are adopted to ensure the privacy and confidentiality of the sample data of the monitoring data during the processing process. Through a series of preprocessing operations, the dispersed and heterogeneous sample data of the monitoring data are converted into unified and standardized lead-acid battery monitoring data, providing a high-quality data foundation for ontology mapping. Then, based on the lead-acid battery ontology knowledge base, the lead-acid battery monitoring data is data mapped. Traverse each data item in the monitoring data and identify the entity type to which it belongs, such as battery, positive electrode, electrolyte, charging and discharging process, etc. Map the identified entities to the ontology concept nodes with the same semantic type in the ontology knowledge base. For example, map "lead-acid battery 12#" in the monitoring data to an instance of the "lead-acid battery" concept node in the ontology knowledge base. On the basis of entity mapping, extract the attribute information of the entity in the monitoring data and map it to the attribute edge of the corresponding concept node. For example, map the production date, rated capacity and other attributes of "lead-acid battery 12#" to the data type attributes of the "lead-acid battery" concept. Through entity mapping and attribute mapping, the key information elements in the monitoring data establish semantic associations with the concept system in the ontology knowledge base and are integrated into the semantic network of the knowledge base. On this basis, use the relationship types defined in the ontology knowledge base to identify the semantic relationships between the concept nodes in the monitoring data and form relationship edges. For example, identify the "inclusion" relationship between "lead-acid battery 12#" and "positive plate A", and the "participation" relationship between "lead-acid battery 12#" and "charge and discharge record X". Through the link of relationship edges, the entities in the monitoring data are associated to form a knowledge network with close association and rich semantics. So far, the semantic mapping of lead-acid battery monitoring data to the ontology knowledge base has been completed, forming a fusion semantic network containing concept nodes, attribute edges, and relationship edges, that is, the preset fusion lead-acid battery knowledge graph has been obtained.

[0058] Based on the lead-acid battery ontology knowledge base, the relationship extraction of lead-acid battery monitoring data is carried out to extract the semantic relationship between named entities in the lead-acid battery monitoring data, and the semantic relationship is mapped into the relationship edge between the ontology concept nodes; Specifically, based on the lead-acid battery ontology knowledge base, the lead-acid battery monitoring data is extracted. Scan each data item in the monitoring data to identify the named entities that appear therein, such as "lead-acid battery 12#", "positive plate A", "charge and discharge record X", etc. Then, analyze the context information between the named entities, such as entity co-occurrence, word order relationship, etc., and use natural language processing technologies such as syntactic analysis and dependency analysis to identify the semantic relationship between entities and form relationship knowledge in the form of triples, such as ("lead-acid battery 12#", "includes", "positive plate A"), ("lead-acid battery 12#", "participates", "charge and discharge record X"), etc. In order to improve the accuracy and coverage of relationship extraction, this embodiment adopts a multi-strategy relationship extraction method based on rules, pattern matching, and machine learning. The rule-based method customizes a batch of high-confidence relationship extraction rules for quickly and accurately extracting common explicit relationships. Based on the pattern matching method, a pattern library for relation extraction is constructed. The relation patterns in sentences are discovered through pattern matching, which is used to extract implicit relations and complex semantic relations. The machine learning-based method uses a neural network model with an attention mechanism to achieve relation extraction through end-to-end relation classification, which improves the generalization ability and robustness of relation extraction. Through multi-strategy relation extraction, rich relation knowledge is obtained from lead-acid battery monitoring data.

[0059] Then, the extracted semantic relations are mapped to the relationship edges between the ontology concept nodes. Each relationship triple is traversed, and its subject entity and object entity are respectively mapped to the constructed ontology concept nodes, and the relationship type between the two concept nodes is determined according to the predicate in the triple. For example, for the relationship triple of ("lead-acid battery 12#", "include", "positive plate A"), the graph definition of the ontology knowledge base is queried, and it is found that there is a "include" relationship between "lead-acid battery" and "positive plate", so a "include" type relationship edge is added between the "lead-acid battery 12#" node and the "positive plate A" node. For another example, for the relationship triple of ("lead-acid battery 12#", "participate", "charge and discharge record X"), the graph is queried to find that there is a "participate" relationship between "lead-acid battery" and "charge and discharge record", so a "participate" type relationship edge is added. Relationship mapping is performed on all extracted semantic relations, which are converted into relationship edges between ontology concept nodes and integrated into the existing semantic network containing concept nodes and attribute edges, ultimately forming a fused semantic network with closer associations and richer semantics, i.e., the preset fused lead-acid battery knowledge graph.

[0060] Based on ontology concept nodes, attribute edges and relationship edges, a preset fused lead-acid battery knowledge graph is constructed.

[0061] Specifically, the data model of the graph specifies the types of components of the graph and their logical relationships, and is the basis for constructing the graph. This embodiment uses a property graph as the data model of the graph. The property graph consists of nodes and edges, and both nodes and edges can contain several properties. Each ontology concept node corresponds to a node in the property graph, and the attribute of the node is the attribute edge of the concept. Each relationship edge corresponds to a directed edge between nodes in the property graph, and the attribute of the edge is the metadata of the relationship, such as relationship type, confidence, etc. The property graph is intuitive and flexible, and can express concepts, attributes, relationships and other elements at the same time. It is suitable for modeling complex semantic networks in the field of lead-acid batteries.

[0062] After clarifying the graph data model, we began to gradually build the lead-acid battery knowledge graph. First, add the ontology concept node to the graph. Traverse each ontology concept node and create a corresponding node in the graph. The label of the node is the name of the concept. For different instances of the same concept, such as different "lead-acid batteries", create multiple nodes and distinguish them by unique identifiers (such as URIs). Then, add attribute edges to the graph. Traverse the attribute edges of each ontology concept node and add them as attributes to the corresponding nodes. The key of the attribute is the name of the attribute, and the value of the attribute is the numerical value of the attribute. Next, add relationship edges to the graph. Traverse each relationship triple and create a directed edge between its subject node and object node. The label of the edge is the type of relationship. If necessary, add attributes to the relationship edge, such as metadata such as the confidence and source of the relationship. Through the above steps, all ontology concept nodes, attribute edges, and relationship edges are integrated into an attribute graph, and the preset fusion lead-acid battery knowledge graph is preliminarily constructed.

[0063] In order to optimize the quality and performance of the knowledge graph, it is also necessary to perform necessary processing on the initially constructed graph. On the one hand, it is necessary to eliminate noise and redundancy in the graph, such as removing duplicate nodes, deleting invalid edges, correcting incorrect attribute values, etc., to improve the accuracy and simplicity of the graph. On the other hand, it is necessary to enrich the semantic information of the graph, such as supplementing synonyms and abbreviations of concepts, mining the upper and lower relationships of concepts, and associating links to external knowledge bases, etc., to enhance the semantic expression ability of the graph. At the same time, it is necessary to optimize the storage and retrieval performance of the graph, and adopt graph databases, indexing mechanisms and other technologies to improve the query efficiency and scalability of the graph. After a series of optimization processes, a high-quality and high-performance lead-acid battery knowledge graph was finally obtained.

[0064] S402, determining a state node in a battery state causal chain where the current lead-acid battery is located according to the degree of deviation; Specifically, first, according to the degree of deviation of the key indicators of each state node, the abnormal degree of the battery at each state node is determined. The degree of deviation of each state node is compared with the preset abnormal level threshold to determine the abnormal level of the node. For example, multiple levels such as normal, mild abnormality, moderate abnormality, and severe abnormality can be set, and corresponding deviation thresholds can be set for different key indicators and state nodes. If the degree of deviation of a state node exceeds the corresponding abnormal level threshold, it is considered that the battery has an abnormal state of the corresponding level at the node. By determining the level of deviation of all state nodes, the abnormal distribution of the battery in the entire causal chain can be comprehensively evaluated. Then, the abnormal level of each state node is comprehensively analyzed to determine the comprehensive state stage in which the battery is currently located. Considering that the evolution of the battery state is a gradual process, the state nodes in the causal chain have a sequence and causal dependency. Usually, the battery state starts from a normal state, gradually goes through early decay, mid-term decay, late decay, and finally reaches a failure state. According to the abnormal level distribution of each state node, combined with the topological structure and logical relationship of the causal chain, the state stage that the battery is most likely to be in is inferred. For example, if the early decay nodes are abnormal and the mid-late decay nodes are normal, the battery may be in the early decay stage; if all decay nodes are abnormal, the battery may have entered the late decay stage. Through this comprehensive reasoning, a time and causal relationship can be established between the status nodes, forming a dynamic assessment of the overall battery status.

[0065] S403, calculating the state transition probability of each state node in the battery state causal chain; Specifically, first, feature data related to state transition are extracted from the battery operation history database. A batch of representative battery samples are selected to collect various monitoring data throughout their life cycle, such as voltage, current, temperature, capacity, internal resistance and other time series data. For each battery sample, the main state nodes it has experienced are identified, such as normal state, early decay state, mid-term decay state, late decay state, etc., and the start and end time and duration of each state node are recorded. Then, the key characteristic indicators on each state node, such as capacity decay rate, internal resistance growth rate, temperature mean, etc., are extracted to form a state node feature vector. By summarizing and analyzing the state node features of multiple battery samples, a state transition feature data set is formed. Then, the state transition feature data set is used to train the state transition probability model. Common modeling methods include Markov chain, Bayesian network, survival analysis, etc. Taking Markov chain as an example, the battery state causal chain is abstracted into a discrete time Markov chain, the state node corresponds to the state of the Markov chain, and the state transition corresponds to the state transfer of the Markov chain. By using methods such as maximum likelihood estimation, the parameters of the state transition probability matrix are learned from the state transition feature data to obtain a Markov chain model that describes the state transition law. Similarly, the Bayesian network can also establish a corresponding state transition probability model by constructing conditional dependencies between state nodes, and the survival analysis can also establish a distribution function of the state duration. Through model training and verification, the model with the best prediction performance is selected as the final state transition probability model.

[0066] After the state transition probability model is built, the model can be used to calculate the state transition probability of each state node. Given the state node that the battery is currently in, the probability of the battery transferring from the current state to other state nodes can be inferred using the trained state transition probability model based on the state transition feature vector.

[0067] S404, performing real-time state evaluation according to the state node and state transition probability of the current lead-acid battery state in the battery state causal chain, and obtaining a real-time state evaluation result.

[0068] Specifically, first, the state node where the current battery is located and the migration probability of each state node are obtained from steps S402 and S403. The real-time monitoring data of the current battery is input into the state node determination model to identify the position of the battery on the causal chain and determine the state node where it is located, such as normal state, early decay state, etc. At the same time, the feature vector of the current state node is input into the state migration probability model to calculate the probability distribution of transferring from the current state node to other state nodes. These two types of information describe the health state of the battery from static and dynamic perspectives respectively. Then, based on the state node and migration probability, a real-time state evaluation index system is constructed. For various key battery performances, such as capacity, life, efficiency, etc., a series of state evaluation indicators are designed to quantitatively describe the performance level of the battery in these performances. For example, the capacity state indicator can be measured by the ratio of the current capacity to the rated capacity, and the life state indicator can be measured by the ratio of the number of cycles used to the rated cycle life. Each evaluation indicator is associated with a specific state node and reflects the performance level of the battery in this state. By mapping the state node of the current battery to each evaluation indicator, a set of indicator values ​​reflecting the real-time state of the current battery can be obtained.

[0069] Next, the state transition probability is used to dynamically weight the real-time state evaluation index. Starting from the current state node, the probability distribution of each state node that the battery may reach in the future is calculated based on the state transition probability. The probabilities of these state nodes are used as weights and weighted averaged with the evaluation index values ​​under the corresponding states to obtain a set of expected state evaluation index values ​​that take into account the state transition risk. These expected values ​​reflect the evolution trend of the battery state over a period of time in the future and depict the expected path of changes in the battery health level.

[0070] Based on the above embodiment, as an optional embodiment, before the multimodal fusion feature vector is input into a preset battery performance degradation trend estimation model to obtain the battery performance degradation trend, the method further includes: S501, collecting lead-acid battery degradation history data, preprocessing the lead-acid battery degradation history data, and obtaining preprocessed lead-acid battery degradation history data; Specifically, first determine the collection indicators and data sources of the degradation history data based on the needs of estimating the battery performance degradation trend. Battery degradation history data generally includes performance parameters such as capacity, internal resistance, self-discharge rate, charge and discharge efficiency, and operating parameters such as voltage, current, and temperature. The data source can be the monitoring database of the battery management system, operation and maintenance record files, or third-party test reports. According to the different characteristics of the data source, design corresponding data collection interfaces and parsing rules, such as database query statements, text parsing scripts, etc., to extract structured degradation history data from heterogeneous data sources. For data from different battery samples and different time periods, data cleaning and alignment are required to ensure data integrity and consistency.

[0071] Then, the collected degradation history data is preprocessed to improve data quality and highlight the degradation trend characteristics. Common preprocessing operations include outlier removal, noise filtering, data normalization, feature construction, etc. Outlier removal identifies and removes abnormal data points that deviate significantly from the normal range by setting thresholds or statistical criteria. Noise filtering uses signal processing methods such as moving average and Kalman filtering to smooth short-term fluctuations and extract long-term trends. Data normalization maps parameters of different dimensions to the same scale to eliminate the impact of differences in value ranges. Feature construction is based on the original parameters and derives some composite features that reflect degradation rate and acceleration factor, such as capacity attenuation rate, internal resistance growth rate, etc. After a series of preprocessing operations, a high-quality degradation history data set is obtained.

[0072] After completing the preprocessing of the degradation history data, it is necessary to further divide the data set to construct the training set and test set of the degradation trend estimation model. The preprocessed degradation history data is divided into a training set and a test set according to a certain ratio, such as 8:2. The training set is used for parameter learning and optimization of the degradation trend estimation model, and the test set is used to evaluate the generalization performance and prediction effect of the model. In order to enhance the representativeness of the samples, the data division should try to ensure that the distribution of samples with different degradation modes and different working conditions in the training set and the test set is balanced. At the same time, in order to verify the stability of the model, methods such as cross-validation can be used to randomly divide the data set multiple times to evaluate the distribution of model performance.

[0073] S502, extracting features from the preprocessed lead-acid battery degradation history data to obtain degradation feature data; Specifically, firstly, according to the degradation mechanism and failure mode of lead-acid batteries, the degradation feature extraction method and indicator system are determined. Common degradation feature extraction methods include time domain analysis, frequency domain analysis, wavelet analysis, etc. Time domain analysis characterizes the overall trend and fluctuation level of the degradation process by calculating the statistics of time series data, such as mean, variance, peak, etc. Frequency domain analysis uses Fourier transform and other techniques to convert time series data into frequency domain to extract the periodicity and frequency characteristics of the degradation process. Wavelet analysis extracts degradation features at different time scales by multi-scale decomposition of time series, capturing local details and mutation points of the degradation process. For typical degradation modes such as capacity attenuation, internal resistance growth, and increased self-discharge, a series of characteristic indicators are designed to quantitatively evaluate the key attributes of the degradation process, such as rate, acceleration, and step change.

[0074] Then, the determined feature extraction method and indicator system are used to construct features for the preprocessed degradation history data. Taking time domain analysis as an example, the statistical characteristics of the time series data of key performance parameters such as capacity and internal resistance are calculated, such as the initial degradation value, the final degradation value, the degradation mean, the degradation standard deviation, etc., to characterize the starting point, end point, central trend and degree of dispersion of the entire degradation process. For derivative parameters such as degradation rate, the maximum value, minimum value, average value, etc. are calculated to reflect the extreme level and overall level of the degradation speed. Similarly, frequency domain analysis and wavelet analysis can extract the spectrum characteristics, energy distribution characteristics, multi-scale decomposition coefficients, etc. of the degradation process, and characterize the dynamic characteristics of the degradation trend from different angles.

[0075] S503, performing degradation pattern recognition on the pre-processed lead-acid battery degradation history data to obtain a degradation pattern; Specifically, firstly, based on the domain knowledge of lead-acid batteries, common degradation mode types and their discrimination rules are summarized. Common degradation modes include linear degradation mode, exponential acceleration mode, multi-stage degradation mode, etc. The linear degradation mode refers to the battery performance parameters showing an approximately linear downward trend, and the degradation rate is basically constant. The exponential acceleration mode refers to the exponential growth of the degradation rate, the degradation process is constantly accelerating, and the time to reach the scrap point is greatly shortened. The multi-stage degradation mode divides the entire degradation process into several stages, and the degradation rates of different stages are significantly different, showing the characteristics of piecewise linear or piecewise exponential. According to the typical characteristics of different degradation modes, a series of discrimination rules are designed, such as linear fit, exponential fit goodness, segmented inflection point position, etc., for qualitative discrimination and quantitative evaluation of degradation modes. Then, the degradation mode discrimination rules are used to identify the degradation mode of the preprocessed degradation history data. For each degradation history data sequence, its linear degradation trend line and exponential degradation trend line are fitted, and the corresponding goodness of fit indicators, such as determination coefficient, root mean square error, etc., are calculated. According to the comparison result between the goodness of fit index and the preset threshold, it is judged whether the data sequence belongs to the linear degradation mode or the exponential degradation mode. For data sequences with low goodness of fit, consider using a multi-stage degradation model for segmented fitting. Using optimization methods such as dynamic programming and greedy algorithm, the optimal segmentation point position is searched on the time axis, and the degradation process is divided into several linear or exponential degradation sub-stages. After degradation pattern recognition, the degradation mode type and its key parameters corresponding to each degradation history data are obtained, such as degradation rate, acceleration factor, segmentation node, etc.

[0076] S504, using battery domain knowledge, constructing a degradation mechanism battery state causal chain reflecting the degradation mechanism; Specifically, the key physical quantities involved in the degradation process of lead-acid batteries and their internal connections are sorted out. The battery degradation process is usually manifested as capacity decay and increased internal resistance, and these macroscopic performance degradation phenomena are an integrated reflection of changes in multiple microscopic states. At the positive electrode, the shedding of active substances, sulfate crystallization, plate corrosion, etc. will lead to a decrease in active substances and a decrease in active specific surface area. At the negative electrode, the shedding of active substances, sulfate crystallization, plate passivation, etc. will lead to a decrease in active specific surface area and deterioration of conductive properties. In the electrolyte, water evaporation and changes in acid composition will lead to a decrease in conductivity. Combining the relationship between these microscopic state quantities and macroscopic performance energies, a multi-layer state causal chain covering electrodes, electrolytes, and macroscopic performance is initially constructed, reflecting the corresponding relationship between performance degradation phenomena and internal state changes.

[0077] S505, constructing a battery performance degradation trend estimation model based on degradation feature data, degradation mechanism battery status causal chain and degradation mode, and constructing a battery performance degradation trend estimation model based on lead-acid battery degradation history data. When the prediction accuracy of the battery performance degradation trend estimation model is greater than a preset accuracy threshold, a preset battery performance degradation trend estimation model is obtained.

[0078] Specifically, according to the causal chain of degradation mechanism states, the basic framework of the battery performance degradation trend estimation model is built. The key state quantities in the causal chain are used as the state variables of the model. The state variables are causally transmitted and iteratively updated through the quantitative relationship in the causal chain, and finally the performance indicators of battery degradation are output. On the basis of the causal chain, the degradation pattern information obtained by data drive is introduced, and the quantitative relationship between the state quantities is segmented according to the characteristics of different degradation stages to achieve the staged characterization of the degradation trend. The operating conditions are used as the input variables of the model. By introducing the operating condition modulation function, the influence weights between the state quantities are dynamically adjusted to achieve the model's adaptive ability to the changes in the operating conditions. The battery performance degradation model constructed based on the mechanism framework, degradation mode and operating condition modulation has both the explanatory power of the mechanism and the flexibility of the data. Then, the degradation feature data is used to identify the parameters and optimize the structure of the model. Based on the degradation history data samples, the quantitative relationship parameters in the model, such as the influence coefficient between the state quantities and the weight coefficient of the operating condition modulation function, are fitted through parameter identification methods such as the least squares method and the maximum likelihood estimation. On the basis of parameter identification, model structure optimization algorithms such as regularization and feature selection are used to optimize the structural complexity of the model, reduce redundant state variables and causal relationships, and improve the generalization performance of the model. Through parameter identification and structural optimization, the degradation trend model can fit the degradation history data to the maximum extent while avoiding the risk of overfitting.

[0079] After completing the model construction and optimization, the prediction accuracy of the model is evaluated by cross-validation and other methods. The degradation history data is divided into a training set and a test set. The model is trained with the training set data and evaluated with the test set data to obtain the prediction accuracy indicators of the model, such as the root mean square error, mean absolute percentage error, etc. Through multiple cross-validations, the prediction performance and stability of the model are comprehensively evaluated. When the prediction accuracy of the model is greater than the preset accuracy threshold, it is considered that the model has met the expected prediction performance requirements and the preset battery performance degradation trend estimation model is obtained.

[0080] S104, inputting the multimodal fusion feature vector into a preset battery performance degradation trend estimation model to obtain the battery performance degradation trend; Specifically, we first build a battery performance degradation trend estimation model based on a large amount of historical operating data. This model is based on the long short-term memory network (LSTM), and describes the dynamic law of battery performance indicators changing over time by sequentially learning the feature vectors of multiple time steps. During the model training process, the degradation curves of key performance indicators such as capacity and internal resistance are used as the target, and end-to-end supervised learning is performed using historical data, so that the model can adaptively fit the degradation patterns of different batteries.

[0081] After obtaining the trained battery performance degradation trend estimation model, the multimodal fusion feature vector obtained in step S102 is used as the input of the model, and the performance degradation trend of the battery in the future period of time is obtained through forward propagation calculation. Specifically, the model gradually predicts the performance indicator values ​​of each future time step, such as capacity, internal resistance, etc., based on the input feature vector, and connects these predicted values ​​into a complete degradation curve. The degradation curve reflects the long-term degradation trend of the battery in the current state, which is of great significance for predicting the remaining service life of the battery and formulating maintenance strategies.

[0082] By using the preset battery performance degradation trend estimation model for long-term prediction, we can fully explore the degradation laws and trend information contained in the historical operation data and learn the general characteristics of the battery degradation process. This data-driven method can not only overcome the limitations of physical models in long-term prediction, but also adaptively fit the degradation patterns under different working conditions and environments, with stronger robustness and extrapolation capabilities.

[0083] S105, comprehensively evaluating the state of the lead-acid battery based on the real-time state evaluation result and the battery performance degradation trend to obtain a comprehensive evaluation result.

[0084] Specifically, the real-time status evaluation results and the battery performance degradation trend are first fused at the feature level. The key nodes in the battery status causal chain obtained in step S103 are used as real-time status features, and the key time points on the battery performance degradation trend obtained in step S104 are used as long-term trend features to form a comprehensive evaluation feature set. These features reflect the health status of the battery from different angles, including static and dynamic, local and global multi-granular information. Then, based on the comprehensive evaluation feature set, a multi-level evaluation index system is constructed. The index system takes the battery health status as the core, designs evaluation indicators from multiple dimensions such as capacity, internal resistance, self-discharge rate, cycle life, etc., and divides the indicators into different levels according to the dependency and importance between the indicators. For example, the top-level indicator is the battery health, the middle-level indicators include capacity retention rate, internal resistance growth rate, etc., and the bottom-level indicators include active material shedding degree, electrolyte concentration, etc. Through this hierarchical and graded indicator design, the battery health status can be evaluated from macro to micro, from overall to local. Next, the analytic hierarchy process (AHP) is used to quantitatively evaluate and weightedly fuse the indicators at each level. First, we invite experts in the battery field to compare the importance of each layer of indicators and obtain a judgment matrix. Then, we use the eigenvalue method to calculate the maximum eigenvalue and corresponding eigenvector of the judgment matrix, and normalize the eigenvector as the weight vector of the indicator. Finally, we use the weighted summation method to integrate the evaluation results of each layer of indicators layer by layer from bottom to top to obtain a comprehensive evaluation score of the battery health status.

[0085] On the other hand, the present application also provides a battery operation status monitoring and evaluation system, such as Figure 2 The system comprises: Data acquisition module 1, used to collect multi-modal monitoring data of the lead-acid battery during the charging and discharging process; Feature extraction module 2, used for preprocessing multimodal monitoring data based on deep learning multi-view encoder to obtain multimodal fusion feature vector; The real-time status evaluation module 3 is used to perform causal chain reasoning on the multimodal fusion feature vector based on the preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains, and to perform status evaluation on the lead-acid battery based on the battery status causal chain to obtain a real-time status evaluation result; The degradation trend prediction module 4 is used to input the multimodal fusion feature vector into a preset battery performance degradation trend estimation model to obtain the battery performance degradation trend; The comprehensive evaluation module 5 is used to comprehensively evaluate the state of the lead-acid battery based on the real-time state evaluation result and the battery performance degradation trend to obtain a comprehensive evaluation result.

[0086] Please refer to Figure 3 The application also discloses an electronic device. Figure 3 The electronic device 300 may include: at least one processor 301 , at least one network interface 304 , a user interface 303 , a memory 305 , and at least one communication bus 302 .

[0087] The communication bus 302 is used to realize the connection and communication between these components.

[0088] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0089] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0090] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field~Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processor (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented separately through a chip.

[0091] Among them, the memory 305 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read~Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may optionally be at least one storage system located away from the aforementioned processor 301. Refer to Figure 3 , the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of a battery operation status monitoring and evaluation method.

[0092] exist Figure 3 In the electronic device 300 shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call the application program storing the road assessment method in the memory 305. When executed by one or more processors 301, the electronic device 300 executes one or more methods in the above-mentioned embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited to the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application. In the above embodiments, the description of each embodiment has its own emphasis. For the parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0093] In the several implementations provided in this application, it should be understood that the disclosed system can be implemented in other ways. For example, the system embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of the system or unit can be electrical or other forms.

[0094] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0095] The present application also provides a computer storage medium that can store multiple instructions, which are suitable for being loaded and executed by a processor as described above. Figure 1 The road assessment method of the embodiment shown in the figure can be found in the specific implementation process. Figure 1 The specific description of the illustrated embodiment will not be repeated here.

[0096] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0097] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory, including a number of instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0098] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.

[0099] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A battery operation status monitoring and evaluation method, characterized in that: The method comprises: Collect multi-modal monitoring data of lead-acid batteries during the charging and discharging process; A multi-view encoder based on deep learning pre-processes the multi-modal monitoring data to obtain a multi-modal fusion feature vector; Performing causal chain reasoning on the multimodal fusion feature vector based on a preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains, and performing status evaluation on the lead-acid battery based on the battery status causal chains to obtain a real-time status evaluation result; Inputting the multimodal fusion feature vector into a preset battery performance degradation trend estimation model to obtain a battery performance degradation trend; Based on the real-time status evaluation result and the battery performance degradation trend, the status of the lead-acid battery is comprehensively evaluated to obtain a comprehensive evaluation result.

2. The method according to claim 1, characterized in that The multi-view encoder based on deep learning pre-processes the multi-modal monitoring data to obtain a multi-modal fusion feature vector, including: Unifying the multimodal monitoring data to the same scale and distribution to obtain preprocessed multimodal monitoring data; Multi-view feature extraction is performed on the preprocessed multi-modal monitoring data to obtain the multi-modal fusion feature vector.

3. The method according to claim 1, characterized in that The multimodal fusion feature vector is subjected to causal chain reasoning based on the preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains, including: Linking and matching the multimodal fusion feature vector with entities and attributes in the preset fusion lead-acid battery knowledge graph to obtain a mapping relationship network between the multimodal fusion feature vector and the preset fusion lead-acid battery knowledge graph; Performing a causal chain path search based on the mapping relationship network to obtain a candidate causal chain set, and performing causal strength calculation on the causal chains in the candidate causal chain set to obtain the causal strength of each causal chain in the candidate causal chain set; The causal chains are sorted based on the causal strengths, and the first N causal chains with the largest causal strengths are selected as a series of battery status causal chains.

4. The method according to claim 1, characterized in that The state evaluation of the lead-acid battery based on the battery state causal chain is performed to obtain a real-time state evaluation result, including: Matching the multimodal monitoring data with the battery status causal chain to obtain key indicator values ​​in the multimodal monitoring data, and comparing and calculating the key indicator values ​​with corresponding state nodes in the battery status causal chain to obtain a degree of deviation between the key indicator values ​​and a preset state threshold; Determine, according to the degree of deviation, a state node in the battery state causal chain where the current lead-acid battery is located; Calculating the state transition probability of each state node in the battery state causal chain; A real-time state evaluation is performed according to the state node of the current lead-acid battery state in the battery state causal chain and the state transition probability to obtain the real-time state evaluation result.

5. The method according to claim 1, characterized in that Before inputting the multimodal fusion feature vector into a preset battery performance degradation trend estimation model to obtain the battery performance degradation trend, the method further includes: Collecting lead-acid battery degradation history data, and preprocessing the lead-acid battery degradation history data to obtain preprocessed lead-acid battery degradation history data; Extracting features from the preprocessed lead-acid battery degradation history data to obtain degradation feature data; Performing degradation pattern recognition on the preprocessed lead-acid battery degradation history data to obtain a degradation pattern; Using battery domain knowledge, a causal chain of battery status reflecting degradation mechanisms is constructed; A battery performance degradation trend estimation model is constructed based on the degradation characteristic data, the degradation mechanism battery status causal chain and the degradation mode, and a battery performance degradation trend estimation model is constructed based on the lead-acid battery degradation history data. When the prediction accuracy of the battery performance degradation trend estimation model is greater than a preset accuracy threshold, the preset battery performance degradation trend estimation model is obtained.

6. The method according to claim 5, characterized in that Before performing causal chain reasoning on the multimodal fusion feature vector based on the preset fusion lead-acid battery knowledge graph to obtain a series of battery state causal chains, the method further includes: Acquire lead-acid battery knowledge text related to lead-acid batteries, and construct a lead-acid battery ontology knowledge base based on the lead-acid battery knowledge text; Obtain sample data of lead-acid battery monitoring materials, and map the sample data of lead-acid battery monitoring materials to an ontology framework based on the lead-acid battery ontology knowledge base to form a semantic network including concept nodes, attribute edges, and relationship edges, and obtain the preset fused lead-acid battery knowledge graph.

7. The method according to claim 6, characterized in that The lead-acid battery inspection material sample data is mapped to the ontology framework based on the lead-acid battery ontology knowledge base to form a semantic network including concept nodes, attribute edges, and relationship edges, and the preset fused lead-acid battery knowledge graph is obtained, including: Preprocessing the lead-acid battery monitoring data sample data to obtain lead-acid battery monitoring data; Based on the lead-acid battery ontology knowledge base, the lead-acid battery monitoring data is data mapped, entities in the lead-acid battery monitoring data are mapped to corresponding ontology concept nodes in the ontology knowledge base, and attributes in the structured data are mapped to attribute edges of the ontology concept nodes; Based on the lead-acid battery ontology knowledge base, relationship extraction is performed on the lead-acid battery monitoring data, the semantic relationship between the named entities in the lead-acid battery monitoring data is extracted, and the semantic relationship is mapped into a relationship edge between the ontology concept nodes; Based on the ontology concept nodes, the attribute edges and the relationship edges, the preset fused lead-acid battery knowledge graph is constructed.

8. A battery operation status monitoring and evaluation system, characterized in that: The system comprises: A data acquisition module is used to collect multi-modal monitoring data of lead-acid batteries during the charging and discharging process; A feature extraction module, used for a multi-view encoder based on deep learning, to pre-process the multi-modal monitoring data to obtain a multi-modal fusion feature vector; A real-time status assessment module, used to perform causal chain reasoning on the multimodal fusion feature vector based on a preset fusion lead-acid battery knowledge graph to obtain a series of battery status causal chains, and perform status assessment on the lead-acid battery based on the battery status causal chain to obtain a real-time status assessment result; A degradation trend prediction module, used for inputting the multimodal fusion feature vector into a preset battery performance degradation trend estimation model to obtain the battery performance degradation trend; The comprehensive evaluation module is used to comprehensively evaluate the state of the lead-acid battery based on the real-time state evaluation result and the battery performance degradation trend to obtain a comprehensive evaluation result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: It includes a processor, a memory and a transceiver, the memory is used to store instructions, the transceiver is used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 7.

Images