Battery system electric connection abnormal fault diagnosis method based on fault data features

Abstract

The present application relates to the field of vehicle battery abnormality diagnosis method, and particularly relates to a battery system electric connection abnormality fault diagnosis method based on fault data characteristics, comprising the following steps: step 1, collecting historical real vehicle fault data of multiple fault vehicles, and extracting fault characteristics of electric connection abnormality of the real vehicle fault data; step 2, combing the logical relationship between the fault characteristics through fault imaging and analytic hierarchy process, and building the fault characteristics and the logical relationship to form a diagnosis fault tree; step 3, obtaining real-time fault data of the vehicle, extracting fault performance information in the real-time fault data, and determining fault causes in the diagnosis fault tree according to the fault performance information. The present application accurately and timely locates the causes of the electric connection abnormality based on the diagnosis fault tree.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of vehicle battery abnormality diagnosis method, and particularly relates to a battery system electric connection abnormality fault diagnosis method based on fault data characteristics. BACKGROUND

[0002] The pure electric vehicle fire accident is caused by complex failure modes, among which the battery system electric connection abnormality is one of the failure modes with extremely high risk. Once it occurs, it is easy to cause arc fire accident. Especially for the burned vehicles, only data can be used to analyze the fire cause.

[0003] The power battery system electric connection abnormality is divided into two types according to the system characteristics, single parallel system and multi-parallel system. The single parallel system means that all battery monomers are connected in series to form a battery system, and the multi-parallel system means that the connection mode of the single battery cell in the same battery system is both series connection and parallel connection. Because the data performance of the multi-parallel system electric connection abnormality is completely different from that of the single parallel system, the current battery system electric connection abnormality fault tree has no fault data characteristics or only has single parallel system data characteristics for guidance, and it is difficult to accurately locate the fault cause based on the fault tree. SUMMARY

[0004] The present application aims to provide a battery system electric connection abnormality fault diagnosis method based on fault data characteristics, so as to accurately and timely locate the cause of the electric connection abnormality.

[0005] The battery system electric connection abnormality fault diagnosis method based on fault data characteristics in the present application comprises the following steps:

[0006] Step 1, collecting historical real vehicle fault data of multiple fault vehicles, and extracting fault characteristics of the real vehicle fault data electric connection abnormality;

[0007] Step 2, combing the logical relationship between the fault characteristics by fault portrait and analytic hierarchy process, and building the fault characteristics and the logical relationship to form a diagnosis fault tree;

[0008] Step 3, obtaining real-time fault data of the vehicle, extracting fault performance information in the real-time fault data, and determining the fault cause in the diagnosis fault tree according to the fault performance information.

[0009] The beneficial effects of the present application are as follows:

[0010] By extracting fault characteristics from the real vehicle data of multiple fault vehicles, the extracted fault characteristics are more comprehensive and accurate. The logical relationship of the extracted fault characteristics is extracted and a diagnosis fault tree is built. The fault cause is determined in the diagnosis fault tree according to the fault performance information of the vehicle. The cause of the electric connection abnormality can be accurately and timely located based on the diagnosis fault tree.

[0011] Further, in the step 1, the single-parallel system is determined according to that the ohmic resistance of the single battery is in the first preset range and the battery system capacity is not attenuated, and the multi-parallel system is determined according to that the ohmic resistance of the single battery is in the second preset range and the battery system capacity is in the third preset range.

[0012] The beneficial effect is that the parameters of the battery are judged to determine whether the fault is caused by the electrical connection, and the accuracy of subsequent electrical connection fault reason judgment is improved.

[0013] Further, the first preset range is R NGΩ >1.5*R OKΩ , R NGΩ is the ohmic resistance of the faulty single battery, R OKΩ is the ohmic resistance of the normal single battery, the second preset range is 1.0*R OKΩ <R NGΩ <1.2*R OKΩ , and the third preset range is C NG <C OK *0.8, C NG is the faulty battery system capacity, and C OK is the normal battery system capacity.

[0014] The beneficial effect is that by setting each range, the battery of the vehicle is distinguished on the fault tree as a single-parallel system or a multi-parallel system, and the judgment of the electrical connection fault in the power battery system is more accurate.

[0015] Further, in the step 2, the electrical connection abnormality of the power battery system is taken as a top event, the connection relationship and the component fault of each component under the top event are taken as intermediate events, and the reasons corresponding to the intermediate events are taken as bottom events.

[0016] The beneficial effect is that the possible reasons for the power battery fault are formed into multiple events, and the fault reasons are clearer.

[0017] Further, in the step 2, the connection relationship of the component is taken as a first layer intermediate event, the component fault is taken as a second layer intermediate event, and the reasons corresponding to the second layer intermediate event are taken as bottom events.

[0018] The beneficial effect is that the connection relationship of the component and the component fault are taken as two layers of intermediate events respectively, and the logical reasons for the power battery fault are clearer.

[0019] Further, in the step 2, the logical relationship includes an OR gate, and the logical relationship is added between the top event and the first layer intermediate event, the first layer intermediate event and the second layer intermediate event, and the second layer intermediate event and the bottom event to form a diagnostic fault tree.

[0020] The beneficial effect is that by adding logical relationships between the vertical layers, the hierarchical relationships and logic of the diagnostic fault tree become clearer.

[0021] Furthermore, in step 3, the power battery system is first determined to be a single-parallel system or a multi-parallel system based on real-time fault data. After determining whether the power battery system is a single-parallel system or a multi-parallel system, the fault is determined by judging from the top event layer by layer, and the bottom event is determined to be the cause of the fault.

[0022] The beneficial effect is that it first determines whether the power battery system is a single parallel system or a multi-parallel system, and then locates the cause of the fault, avoiding misjudgment of the cause of the fault due to errors in the identification of data features in the early stage.

[0023] Furthermore, in step 1, fault features are extracted using discrete convolutional wavelet transform signal processing and point neuron technology.

[0024] The beneficial effect is that by using two algorithms to extract fault features, the fault features can be accurately classified, thereby extracting complete fault features.

[0025] Furthermore, the first-level intermediate events include, by logical relation, internal cell connector failures and external cell connector failures. The second-level intermediate events under the internal cell connector failure include, by logical relation, adapter breakage and tab breakage. The bottom events under the adapter breakage failure include, by logical relation, adapter quality defects and adapter-to-terminal welding defects. The bottom events under the tab breakage failure include, by logical relation, core loosening and tab-to-adapter welding defects. The second-level intermediate events under the external cell connector failure include, by logical relation, cross-module connector detachment and cell aluminum bar detachment. The bottom events under the cross-module connector detachment failure include, by logical relation, stress fatigue fracture, welding defects, bolt loosening, and incoming material quality defects. The bottom events under the cell aluminum bar detachment failure include, by logical relation, stress fatigue fracture, aluminum bar incoming material quality defects, and aluminum bar-to-terminal welding quality defects.

[0026] The beneficial effects are: for events at each level of the fault tree for abnormal electrical connections, the overall logic is clear and the root causes of the fault are fully included. Attached Figure Description

[0027] Figure 1 This is a flowchart of an embodiment of the battery system electrical connection anomaly fault diagnosis method based on fault data characteristics according to the present invention;

[0028] Figure 2 This is a schematic diagram of an abnormal fault tree in an embodiment of the battery system electrical connection abnormality fault diagnosis method based on fault data characteristics of the present invention;

[0029] Figure 3This is a voltage characteristic spectrum of a single-cell battery fault in a single-parallel system, as shown in an embodiment of the battery system electrical connection abnormality fault diagnosis method based on fault data characteristics of the present invention.

[0030] Figure 4 This is a voltage characteristic map of a single cell fault in a multi-parallel system, as shown in an embodiment of the battery system electrical connection abnormality fault diagnosis method based on fault data characteristics of the present invention. Detailed Implementation

[0031] The following detailed description provides further details on specific implementation methods.

[0032] Example

[0033] like Figure 1 As shown, the battery system electrical connection anomaly fault diagnosis method based on fault data characteristics includes the following steps:

[0034] Step 1: Collect historical vehicle fault data from multiple faulty vehicles. The historical vehicle data is collected and stored through the vehicle's built-in battery management system (BMS) to ensure the accuracy and authenticity of the data.

[0035] Based on the condition that the ohmic resistance of a single cell is within a first preset range and the battery system capacity shows no degradation, it is determined to be a single-parallel system. The first preset range is R. NGΩ >1.5*R OKΩ R NGΩ R is the ohmic resistance of the faulty individual cell. OKΩ The resistance of a normal single cell is used as the reference value. Based on the fact that the resistance of a single cell is within a second preset range and the battery system capacity is within a third preset range, it is determined to be a multi-parallel system. The second preset range is 1.0*R. OKΩ <R NGΩ <1.2*R OKΩ The third preset range is C NG <C OK *0.8, C NG For the capacity of the faulty battery system, C OK This represents the normal battery system capacity. Fault characteristics of abnormal electrical connections in real-vehicle fault data were extracted using Discrete Convolutional Wavelet Transform (DCWT) and point neuron technology, respectively. These fault characteristics revealed significant differences in the data after the occurrence of abnormal electrical connections in single-parallel and multi-parallel systems.

[0036] Step 2: Take the abnormal electrical connection of the power battery system as the top event, take the connection relationship of multiple components under each top event and the component failure as intermediate events, take the connection relationship of the components as the first-level intermediate event, take the component failure as the second-level intermediate event, take the cause corresponding to the second-level intermediate event as the bottom event, and take the cause corresponding to the intermediate event as the bottom event.

[0037] By profiling faults and using the hierarchical analysis method, the logical relationships between fault characteristics are identified, and the causal events related to abnormal electrical connections are analyzed. This involves identifying all potential hazards and weak points in the system, including hardware faults such as equipment components, software faults, human errors, and environmental factors. All root causes related to the accident are identified and used as causal events in the event tree. The fault characteristics and logical relationships are then used to construct a diagnostic fault tree. The logical relationships include OR gates. These logical relationships are then added between the top event and the first-level intermediate events, the first-level intermediate events and the second-level intermediate events, and the second-level intermediate events and the bottom event to form the diagnostic fault tree.

[0038] like Figure 2 As shown, electrical connection abnormalities are purely physical failure modes. Based on the internal electrical connections of the power battery system, the module electrical connections, and the cell electrical connections, the top event is an electrical connection abnormality in the battery system. Electrical connection abnormalities include faults in the internal connections of the cell and faults in the external connections of the cell. Faults in the internal connections of the cell and faults in the external connections of the cell are the first-level intermediate events. Faults in the external connections of the cell include the cross-membrane copper busbar connection and the aluminum busbar connection between cells. The cross-membrane copper busbar connection and the aluminum busbar connection between cells are the second-level intermediate events under the faults in the external connections of the cell. Faults in the internal connections of the cell include the breakage of the adapter piece and the breakage of the tab inside the cell. The breakage of the adapter piece and the breakage of the tab are the second-level intermediate events under the faults in the internal connections of the cell.

[0039] The causes of each intermediate event in the second layer are identified as the base events. The base events for adapter breakage include two base events: adapter quality defects and welding defects between the adapter and the pole. The base events for tab breakage include two base events: core loosening and welding defects between the tab and the adapter. The base events for cross-membrane assembly connector detachment failure include four base events: stress fatigue fracture, welding defects, bolt loosening, and incoming material quality defects. The base events for cell aluminum bar detachment failure include three base events: stress fatigue fracture, aluminum bar incoming material quality defects, and welding quality defects between the aluminum bar and the pole.

[0040] Step 3: Obtain real-time vehicle fault data and extract fault performance information from it. First, determine whether the power battery system is a single-parallel or multi-parallel system based on the real-time fault data. For example, extract the voltage characteristic spectrum of the operating data before and after the fault. The voltage characteristic spectrum is as follows: Figure 3 and Figure 4 As shown, fault behavior information is identified, such as... Figure 3 The voltage characteristic spectrum shows that the fault manifestation information exhibits a unique form with a clear voltage and current response relationship during the discharge process. The unique form is that when the current is positive, the faulty cell has the lowest voltage in the system; when the current is negative, the faulty cell has the highest voltage in the system. At this time, the ohmic resistance R of the single cell connected to the faulty point... NGΩ=1.9*R OKΩ Satisfying R NGΩ >1.5*R OKΩ Furthermore, since the battery system capacity did not decrease and it was a single-parallel system, the fault was investigated layer by layer in the fault tree, ultimately pinpointing the cause of the fault as a specific bottom-level event in the fault tree. For example... Figure 4 The voltage characteristic spectrum shows the fault manifestation information of a single battery cell suddenly experiencing a significant decrease in capacity during normal operation. The rated capacity is 150Ah, and suddenly at a certain cycle fault location, the single battery cell capacity drops to 94.5Ah, i.e., C... NG <C OK *0.8, R NGΩ =1.12*R OKΩ Satisfies 1.0*R OKΩ <R NGΩ <1.2*R OKΩ And C NG <C OK The *0.8 condition, pointing to a fault in the aluminum bus or internal connector of a single cell in a multi-parallel system, will not be misdiagnosed as an abnormal cell capacity degradation fault tree. This avoids misjudgment of the fault cause due to errors in the initial data feature identification. In a multi-parallel system, analyzing in the direction of cell capacity degradation would lead to a misdiagnosis. That is, the cause of the fault is determined in the diagnostic fault tree based on the fault manifestation information. After determining whether the power battery system is a single-parallel or multi-parallel system, the fault is determined by judging layer by layer from the top event to identify the bottom event as the cause of the fault.

[0041] Although the fault tree forms of single-parallel and multi-parallel systems are roughly the same, their data manifestations are completely different. Relying solely on existing battery system fault trees with no fault data characteristics or only single-parallel system data characteristics for fault cause diagnosis will result in significant errors.

[0042] This embodiment obtains historical real-vehicle fault data from faulty pure electric vehicles and extracts fault features from this data. These features are then used to build a fault tree for cell electrical connection anomalies, improving the versatility of the established fault tree for root cause analysis of different pure electric vehicles. The fault tree established in this embodiment forms a complete failure system from the battery system to the internal structure of the cells, not limited to electrical connection anomalies between cells or modules. It comprehensively considers different electrical connection anomaly modes such as single-parallel, multi-parallel, and internal cell faults, improving the realism and reproducibility of pure electric vehicle battery faults. Compared to existing methods that only analyze electrical connection anomalies between cells or modules, this embodiment, by combining fault data features for different system structures, can effectively guide the identification of the causes of electrical connection anomalies in single-parallel and multi-parallel systems. It encompasses more than 90% of the faults that pure electric vehicle batteries can generate from multiple aspects of the cells, improving the accuracy of the root cause analysis by 80%.

[0043] The above descriptions are merely embodiments of the present invention, and common knowledge regarding specific structures and characteristics is not elaborated upon here. It should be noted that those skilled in the art can make various modifications and improvements without departing from the structure of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A method for diagnosing abnormal electrical connections in battery systems based on fault data characteristics, characterized in that, Includes the following steps: Step 1: Collect historical vehicle fault data from multiple faulty vehicles and extract fault characteristics of abnormal electrical connections from the vehicle fault data. Step 2: Use fault profiling and hierarchical analysis to sort out the logical relationships between fault features, and build a diagnostic fault tree by combining fault features and logical relationships; The abnormal electrical connection of the power battery system is taken as the top event, the connection relationship of multiple components under each top event and the component failure are taken as intermediate events, the cause corresponding to the intermediate event is taken as the bottom event, the connection relationship of the components is taken as the first-level intermediate event, the component failure is taken as the second-level intermediate event, and the cause corresponding to the second-level intermediate event is taken as the bottom event. The logical relationship includes an OR gate. This logical relationship is added to form a diagnostic fault tree between the top event and the first-level intermediate event, the first-level intermediate event and the second-level intermediate event, and the second-level intermediate event and the bottom event. The first-level intermediate events include faults in the internal and external connectors of the battery cell, both based on the OR logic. The second-level intermediate events under the internal connector fault include broken adapter pieces and broken tabs, both based on the OR logic. The bottom events under the broken adapter piece fault include quality defects in the adapter piece and defects between the adapter piece and the terminal post, both based on the OR logic. Welding defects, the bottom events under the tab breakage fault include, or logically related, core loosening, and tab-to-transfer piece welding defects; the second intermediate events under the cell external connection fault include, or logically related, cross-module connector loosening and cell aluminum bar loosening; the bottom events under the cross-module connector loosening fault include, or logically related, stress fatigue fracture, welding defects, bolt loosening, and incoming material quality defects; the bottom events under the cell aluminum bar loosening fault include, or logically related, stress fatigue fracture, aluminum bar incoming material quality defects, and aluminum bar-to-terminal welding quality defects. Step 3: Obtain real-time fault data of the vehicle, extract fault performance information from the real-time fault data, and determine the cause of the fault in the diagnostic fault tree based on the fault performance information.

2. The battery system electrical connection anomaly fault diagnosis method based on fault data characteristics according to claim 1, characterized in that: In step 1, a single-parallel system is determined based on the ohmic resistance of a single cell being within a first preset range and the battery system capacity not decreasing. A multi-parallel system is determined based on the ohmic resistance of a single cell being within a second preset range and the battery system capacity being within a third preset range.

3. The battery system electrical connection anomaly fault diagnosis method based on fault data characteristics according to claim 2, characterized in that: The first preset range is , The ohmic resistance of the faulty individual cell. The ohmic resistance of a normal single cell is given by the second preset range. The third preset range is , For the capacity of the faulty battery system, This is the normal battery system capacity.

4. The battery system electrical connection anomaly fault diagnosis method based on fault data characteristics according to claim 1, characterized in that: In step 3, the power battery system is first determined to be a single-parallel system or a multi-parallel system based on real-time fault data. After determining whether the power battery system is a single-parallel system or a multi-parallel system, the fault is determined by judging from the top event layer by layer and identifying the bottom event as the cause of the fault.

5. The battery system electrical connection anomaly fault diagnosis method based on fault data characteristics according to claim 1, characterized in that: In step 1, fault features are extracted using discrete convolutional wavelet transform signal processing and point neuron technology.

Images