Online prediction and evaluation method for residual life of power battery

Abstract

The invention relates to the technical field of power battery management, and discloses a power battery residual life online prediction and evaluation method comprising the following steps: S1, obtaining operation data of a power battery pack, and monitoring a local maintenance or local replacement event; s2, determining a replacement object and a non-replacement object; s3, constructing a historical inheritance mark; s4, constructing a mixed-age life state; s5, determining a constraint relation of each object to the life of the whole package; s6, executing residual life online prediction to obtain a residual life prediction result; and S7, evaluating the residual life prediction result to obtain a residual life evaluation result. When a battery is maintained and replaced, a replaced object and a non-replaced object are distinguished, a historical inheriting mark is constructed, life historical information is inherited or a life initial state is reconstructed according to the historical inheriting mark, and a mixed-age life state after local maintenance or local replacement is formed, so that subsequent residual life online prediction and evaluation have a clear object boundary and a state basis.

Description

Technical Field

[0001] This invention relates to the field of power battery management technology, specifically to a method for online prediction and evaluation of the remaining life of a power battery. Background Technology

[0002] Power batteries are core energy components in new energy vehicles, electric energy storage systems, and other fields. Their operating status and remaining life directly affect the safety and reliability of the system. Online prediction and assessment of the remaining life of power batteries involves analyzing battery operating data to predict the time or number of cycles the battery can continue to use in its current state, and evaluating the prediction results to assist in operation and maintenance decisions and safety management. In practical applications, power batteries are usually operated as battery packs composed of individual battery cells, modules, or battery units. Their life status has the characteristic of evolving over time. Existing online prediction methods for the remaining life of power batteries are usually based on historical operating data to model the battery as a whole or its components, and continuously update the prediction results during operation.

[0003] However, in current technology, when a power battery pack undergoes partial repair or replacement, there will be both replaced and unreplaced components within the battery pack. Existing methods usually treat them as a single object, uniformly inheriting historical lifespan information or resetting lifespan status for all objects. This makes it difficult to accurately distinguish the lifespan status boundaries of different objects, thus affecting the accuracy of online prediction and evaluation results of the remaining lifespan of the power battery. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an online prediction and evaluation method for the remaining lifespan of power batteries, solving the problem of difficulty in accurately distinguishing the lifespan status of power batteries after partial repair or replacement.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for online prediction and evaluation of the remaining life of a power battery, comprising: S1. Obtain the operating data of the power battery pack and monitor whether the power battery pack has undergone partial repair or replacement events; S2. Upon detecting the partial repair or partial replacement event, determine the replacement and non-replacement items in the power battery pack; The replacement objects are battery cells, modules, or battery units that have been replaced, while the non-replaced objects are battery cells, modules, or battery units that have not been replaced. S3. Construct a history inheritance marker based on the replaced object and the unreplaced object, wherein the history inheritance marker is used to characterize whether the corresponding object inherits the life history information before repair or replacement; S4. Based on the historical inheritance marker, inherit the life history information before repair or replacement for the unreplaced object, and reset or rebuild the initial life state for the replaced object to obtain the mixed-age life state after partial repair or partial replacement. S5. Based on the mixed-age life status and the operating data, determine the constraint relationship between each object in the power battery pack and the overall life of the pack. S6. Perform online prediction of the remaining life of the power battery pack according to the constraint relationship to obtain the prediction result of the remaining life of the power battery pack. S7. Based on the remaining life prediction results and the mixed-age life status, evaluate the remaining life prediction results of the power battery pack to obtain the remaining life evaluation results.

[0006] Preferably, step S1 specifically includes: The system acquires operational data for each individual battery cell, module, or battery unit in the power battery pack, including voltage data, current data, temperature data, and state of charge data. Obtain the maintenance records, configuration records, object coding information, or connection relationship information corresponding to the power battery pack; Based on the operational data and the maintenance records, configuration records, object coding information, or connection relationship information, monitor whether the power battery pack has undergone partial maintenance or partial replacement events.

[0007] Preferably, step S2 specifically includes: After detecting a partial repair or replacement event in the power battery pack, identify the individual battery cell, module or battery unit in the power battery pack that has experienced changes in object coding, connection relationship, repair marking, or sudden changes in operating parameters. The identified individual battery cells, modules, or battery units are determined as replacement targets; The battery cells, modules, or battery units in the power battery pack that are not replaced are identified as not replaced.

[0008] Preferably, step S3 specifically includes: Construct a first history inheritance tag for the replaced object; Construct a second history inheritance marker for the unreplaced object; The first historical inheritance marker is used to indicate that the corresponding object does not inherit the life history information before partial repair or partial replacement, and the second historical inheritance marker is used to indicate that the corresponding object inherits the life history information before partial repair or partial replacement. The historical inheritance marker includes an object identifier field and an inheritance status field. The object identifier field is used to characterize the corresponding battery cell, module, or battery unit, and the inheritance status field is used to characterize the inheritance method of the life history information of the corresponding object. The historical inheritance marker is generated based on the object encoding information, connection relationship information, or maintenance records, and the inheritance status field is updated based on the operation data during subsequent operation.

[0009] Preferably, step S4 specifically includes: Based on the second historical inheritance marker, the life history information before partial repair or partial replacement is inherited for the unreplaced object; Based on the first historical inheritance marker, the initial lifespan state of the replacement object is reset, or the initial lifespan state is reconstructed based on the initial operating data of the replacement object after partial repair or partial replacement. The lifetime history information inherited by the unreplaced object is combined with the initial lifetime state of the replaced object to construct a mixed-age lifetime state after partial repair or partial replacement. The historical inheritance marker serves as a control parameter for constructing the mixed-age lifespan state, and is used to control the way the lifespan history information of the corresponding object is introduced.

[0010] Preferably, the step of reconstructing the initial lifespan state based on the initial operating data of the replaced object after partial repair or replacement specifically includes: Obtain the first round of operational data of the replaced object after partial repair or partial replacement; Based on the first round of operational data, voltage change rate, current response characteristics, and temperature response characteristics were extracted. Based on the preset mapping relationship between life state parameters and operating characteristics, the voltage change rate, current response characteristics and temperature response characteristics are solved in reverse to determine the initial life state parameters of the replacement object. The initial lifespan state of the replacement object is constructed based on the initial lifespan state parameters.

[0011] Preferably, step S5 specifically includes: Based on the mixed-age lifespan status, the lifespan status parameters corresponding to each object in the power battery pack are determined. Constraint functions can be constructed based on the lifetime state parameters of each object, or fuzzy membership functions can be constructed based on the lifetime state parameters of each object. The lifespan state parameters corresponding to each object are compared with preset lifespan constraints, and the constraint relationship between each object in the power battery pack and the overall lifespan is determined according to the constraint function or the fuzzy membership function.

[0012] Preferably, step S6 specifically includes: Based on the aforementioned constraint relationship, the constraint object that constrains the remaining lifespan of the power battery pack is determined. Obtain the lifetime status parameters and operational data corresponding to the constraint object; The remaining lifespan of the constraint object is determined based on the lifespan status parameters and operating data corresponding to the constraint object. Based on the remaining lifespan corresponding to the constrained object, the remaining lifespan prediction result of the power battery pack is obtained; The historical inheritance marker and the mixed-age lifespan status serve as inputs for online prediction of remaining lifespan, and are used to control the participation of different objects' lifespan history information.

[0013] Preferably, step S7 specifically includes: Based on the mixed-age lifespan status, determine the evaluation parameters corresponding to the remaining lifespan prediction results; The evaluation parameters are compared with preset evaluation conditions; Based on the comparison results, the evaluation results corresponding to the remaining life prediction results are obtained; The evaluation parameters are calculated based on the lifespan parameters and their changing trends of each object in the mixed-age lifespan state.

[0014] Preferably, the evaluation results include availability markers, warning markers, or trust level markers.

[0015] This invention provides a method for online prediction and evaluation of the remaining life of a power battery. It has the following beneficial effects: 1. This invention distinguishes between replaced and unreplaced objects, constructs historical inheritance markers, and inherits lifespan history information or reconstructs the initial lifespan state accordingly, forming a mixed-age lifespan state after partial repair or partial replacement, so that subsequent online prediction and assessment of remaining lifespan has clear object boundaries and state basis.

[0016] 2. When the initial lifespan state of the replacement object cannot be directly obtained, the present invention extracts operating features based on the first round of operating data of the replacement object after partial repair or replacement, and determines the initial lifespan state of the replacement object according to the mapping relationship between lifespan state parameters and operating features. This allows the replacement object to establish a lifespan starting point that can be used for subsequent calculations even in the absence of direct initial state records, thereby ensuring that the power battery pack after partial repair or replacement can continue to undergo online prediction of remaining lifespan.

[0017] 3. This invention determines the constraint relationship between each object in the power battery pack and the overall lifespan based on the mixed-age lifespan status and operating data. On this basis, it performs online prediction and evaluation of the remaining lifespan, so that the remaining lifespan prediction result not only reflects the remaining lifespan of the power battery pack, but also outputs the corresponding evaluation result in combination with the current mixed-age lifespan status. Thus, the remaining lifespan prediction result in the scenario of partial repair or partial replacement has judgment information corresponding to the object status. Attached Figure Description

[0018] Figure 1 This is a flowchart of an online prediction and evaluation method for the remaining life of a power battery according to the present invention. Detailed Implementation

[0019] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see the appendix Figure 1 This invention provides a method for online prediction and evaluation of the remaining life of a power battery, comprising: S1. Obtain the operating data of the power battery pack and monitor whether the power battery pack has undergone partial repair or replacement events; Furthermore, step S1 specifically includes: The system acquires operational data for each individual battery cell, module, or battery unit in the power battery pack. This operational data includes voltage data, current data, temperature data, and state of charge data. Obtain the maintenance records, configuration records, object coding information, or connection relationship information corresponding to the power battery pack; Based on operational data, maintenance records, configuration records, object coding information, or connection relationship information, monitor whether the power battery pack has undergone partial maintenance or partial replacement events.

[0021] Specifically, the system first acquires the operating data of each battery cell, module, or cell in the power battery pack through the battery management system, sampling unit, or vehicle controller. The operating data may include voltage data, current data, temperature data, and state of charge data. The above operating data characterizes the current operating status of each object in the power battery pack and serves as the basis for subsequent judgment on whether the object status is continuous and whether there are abnormal changes. At the same time, the system further acquires the maintenance records, configuration records, object coding information, or connection relationship information corresponding to the power battery pack. The maintenance records can be used to characterize whether the power battery pack has undergone maintenance or replacement operations, the configuration records can be used to characterize the configuration changes before and after maintenance, the object coding information can be used to distinguish the identity of different battery cells, modules, or cells, and the connection relationship information can be used to characterize whether the connection structure between each object has been adjusted. After obtaining the above information, the operating data is correlated with maintenance records, configuration records, object coding information or connection relationship information to monitor whether the power battery pack has experienced a partial maintenance or partial replacement event. Specifically, when there is a partial maintenance or partial replacement identifier in the maintenance record, or the object coding information changes, or the connection relationship information changes, or some objects in the operating data show a significant change relative to the historical operating state, it can be determined that the power battery pack has experienced a partial maintenance or partial replacement event, and subsequent identification and processing of replaced and unreplaced objects will be triggered. If no partial maintenance or partial replacement event is detected, it can be considered that the objects in the power battery pack are still in a continuous operating state, and subsequent processing will not enter the life state reconstruction process based on the partial maintenance or partial replacement scenario. For example, during the operation of a power battery pack, the voltage, temperature and state of charge data of each module are continuously collected, and maintenance records and object code information are read simultaneously. When a module is reconnected to the power battery pack after maintenance, and its object code is inconsistent with the record before maintenance, and the voltage change and temperature response of the module in the initial stage of re-operation are significantly different from the historical record, it can be determined that a partial replacement event has occurred in the power battery pack, and then proceed to the subsequent steps to distinguish between the replaced object and the non-replaced object.

[0022] S2. In the event of a partial repair or replacement, determine the replacement and non-replacement items in the power battery pack. The replacement targets are the battery cells, modules, or battery units that have been replaced, while the non-replaced targets are the battery cells, modules, or battery units that have not been replaced. Furthermore, step S2 specifically includes: After detecting a partial repair or replacement event in the power battery pack, identify the individual battery cells, modules or battery units in the power battery pack that have experienced changes in object coding, connection relationships, repair markings or sudden changes in operating parameters. The identified individual battery cells, modules, or battery units are determined as replacement targets; The battery cells, modules, or battery units in the power battery pack that are not replaced are identified as not replaced.

[0023] Specifically, after detecting a partial repair or replacement event in the power battery pack, each battery cell, module, or unit within the power battery pack is identified one by one. During identification, the object coding information, connection relationship information, and repair marking information of each object before and after the repair are read, and the corresponding object's operating data before and after the repair is retrieved. The object coding information before and after the repair is compared to determine whether the object's identity has changed. The connection relationship information before and after the repair is compared to determine whether the object's connection location, connection order, or topology has changed. The repair marking information is matched with the object identifier to determine whether the corresponding object has a repair, replacement, or regrouping record. The operating data after the repair is put back into operation is compared with the historical operating data before the repair to determine whether the corresponding object has discontinuous changes in voltage, temperature response, or state of charge. Based on the above comparison and matching results, the battery cells, modules, or units that have experienced changes in object coding, connection relationships, repair markings, or sudden changes in operating parameters are identified. After identifying the above objects, the identified battery cells, modules, or battery units are determined as replacement objects, and a replacement object set is generated. The replacement object set is used to characterize the scope of objects that no longer directly use the pre-repair life history information after partial repair or partial replacement. Subsequently, the remaining battery cells, modules, or battery units in the power battery pack other than the replacement objects are determined as non-replacement objects, and a non-replacement object set is generated. The non-replacement object set is used to characterize the scope of objects that continue to retain the pre-repair life history information after partial repair or partial replacement. Through the above processing, all objects in the power battery pack are divided into two categories: replacement objects and non-replacement objects. The division results of the two categories of objects are used as the object-level input for constructing the history inheritance tag in S3. In one specific embodiment, a power battery pack contains 8 modules. After a partial replacement event is detected, the object code information, connection relationship information, maintenance records, and corresponding voltage and temperature data of the 8 modules before and after maintenance are read. After comparison, it is found that the object code of the 5th module is inconsistent with the record before maintenance. The maintenance record contains the replacement identifier of the module. In addition, in the initial stage after restarting, the voltage change trend and temperature response of the 5th module are not continuous with the historical record before maintenance. Therefore, the 5th module is identified as the replacement object. The other modules do not show the above changes and are identified as non-replacement objects. Thus, a set of replacement objects and a set of non-replacement objects with clear boundaries can be obtained. Subsequently, historical inheritance tags that do not inherit history and those that inherit history can be constructed based on the division results.

[0024] S3. Construct a history inheritance marker based on the replaced object and the non-replaced object. The history inheritance marker is used to characterize whether the corresponding object inherits the life history information before the repair or replacement. Furthermore, step S3 specifically includes: Construct a first history inheritance marker for the replaced object; Construct a second history inheritance marker for objects that have not been replaced; Among them, the first historical inheritance marker is used to indicate that the corresponding object does not inherit the life history information before partial repair or partial replacement, and the second historical inheritance marker is used to indicate that the corresponding object inherits the life history information before partial repair or partial replacement. The historical inheritance marker includes an object identifier field and an inheritance status field. The object identifier field is used to represent the corresponding battery cell, module or battery unit, and the inheritance status field is used to represent the inheritance method of the life history information of the corresponding object. Historical inheritance markers are generated based on object coding information, connection relationship information, or maintenance records, and the inheritance status field is updated based on the running data during subsequent operation.

[0025] Specifically, after obtaining the set of replaced objects and the set of unreplaced objects, historical inheritance tags are constructed for the two types of objects respectively. That is, each battery cell, module or battery unit in the set of replaced objects is written into the historical inheritance tag table one by one, and a first historical inheritance tag is configured to represent "not inheriting the life history information before partial repair or partial replacement". Each battery cell, module or battery unit in the set of unreplaced objects is written into the historical inheritance tag table one by one, and a second historical inheritance tag is configured to represent "inheriting the life history information before partial repair or partial replacement". In this way, a correspondence between objects and historical inheritance methods is formed at the object level, so that each object in the power battery pack has a clear historical information introduction boundary when constructing the life status in the future. To facilitate the subsequent construction of mixed-age life states, the historical inheritance marker is represented as an object-level control variable. ,in, Indicates the first in the power battery pack Individual battery cells, modules, or battery units. Indicates the current moment. Indicates the first The historical inheritance markers corresponding to each object at the current moment, satisfying: ; when When, it indicates the first The object is a replacement object, corresponding to the first history inheritance marker, meaning it does not inherit the life history information before partial repair or replacement. When, it indicates the first The object is an unreplaced object, corresponding to the second historical inheritance mark, that is, inheriting the life history information before partial repair or partial replacement. The above historical inheritance mark is not only used for object classification, but also as a control variable for subsequent mixed-age life status construction and online prediction of remaining life, used to control the way the historical life information of the corresponding object is introduced. The historical inheritance marker is stored in the form of structured data. The structured data includes at least an object identifier field and an inheritance status field. The object identifier field is used to record the object code, module number or single-unit location identifier of the corresponding battery cell, module or battery unit to achieve unique correspondence between objects. The inheritance status field is used to record the historical inheritance method corresponding to the object, that is, "inherited" or "not inherited". When generating historical inheritance tags, the determined set of replaced objects and the set of non-replaced objects are associated with object coding information, connection relationship information or maintenance records, respectively, so that the object identifier field and inheritance status field of each object are synchronously written into the tag table. After this process, a historical inheritance tag table covering all objects of the power battery pack can be obtained. The historical inheritance tag table serves as the direct input for constructing mixed-age life status in S4. In this embodiment, the generation of historical inheritance markers is not based solely on a one-time classification result, but rather retains an update interface during subsequent operation. Specifically, during continuous operation of the power battery pack, the operation data of each object is continuously acquired, and the newly added operation data is associated with the existing object identifier field and inheritance status field. When the operation data indicates that an object has undergone a change in connection relationship or object code, or when maintenance or replacement information for the corresponding object is added to the maintenance record, the inheritance status field corresponding to that object is updated. Thus, the historical inheritance markers can be kept consistent with the subsequent operation status of the power battery pack, avoiding the continued use of the original inheritance status in the case of secondary maintenance, secondary replacement, or regrouping, and ensuring that the object inheritance boundary used when constructing the mixed-age life status is consistent with the actual object status. In one specific embodiment, a power battery pack contains 8 modules. The output result of S2 indicates that the 5th module is to be replaced, while the other 7 modules are not to be replaced. At this time, a first historical inheritance mark is generated for the 5th module, and its inheritance status field is set to 0. A second historical inheritance mark is generated for the other 7 modules, and their inheritance status fields are set to 1. Correspondingly, in the historical inheritance mark table, the object identifier field corresponding to the 5th module is "Module 5", and the inheritance status field is 0. The object identifier fields corresponding to the 1st, 2nd, 3rd, 4th, 6th, 7th, and 8th modules are their respective module numbers, and their inheritance status fields are all 1. If the 7th module is repaired again during subsequent operation and replacement information is added to the repair record, the inheritance status field corresponding to the 7th module is updated from 1 to 0. In this way, in subsequent processing, the 5th and 7th modules no longer inherit the pre-repair life history information, while the other modules continue to inherit the pre-repair life history information, thereby providing an object-level control basis for the accurate construction of mixed-age life status.

[0026] S4. Based on the historical inheritance marker, inherit the life history information before repair or replacement for the unreplaced object, and reset or rebuild the initial life state for the replaced object to obtain the mixed-age life state after partial repair or partial replacement. Furthermore, step S4 specifically includes: Based on the second historical inheritance marker, the life history information before partial repair or partial replacement is inherited for objects that have not been replaced. Based on the first historical inheritance marker, reset the initial life state of the replaced object, or reconstruct the initial life state based on the initial operating data of the replaced object after partial repair or partial replacement. The lifetime history information inherited by the unreplaced object is combined with the initial lifetime state of the replaced object to construct the mixed-age lifetime state after partial repair or partial replacement. Among them, the historical inheritance marker serves as a control parameter for constructing mixed-age life states, and is used to control the way the life history information of the corresponding object is introduced.

[0027] Specifically, the lifetime status of unreplaced objects and replaced objects is processed according to the historical inheritance mark. For unreplaced objects corresponding to the second historical inheritance mark, the lifetime history information stored before partial repair or replacement is read and the lifetime history information is used as the historical state input of the object at the current moment. For replaced objects corresponding to the first historical inheritance mark, the lifetime history information before partial repair or replacement is no longer called, and the initial lifetime state processing is performed on them. For replacement objects, when there is readily available initial state data, the post-repair test data, factory calibration data, or state parameters recorded during initial connection are written into the object state table as the initial lifespan state. When data cannot be directly obtained, the initial operating data of the replacement object after partial repair or replacement is read, and its initial lifespan state is determined based on this initial operating data, so that the replacement object has a starting state that can participate in subsequent state calculations. Through the above processing, the non-replaced object maintains the continuity of its original lifespan evolution, while the replacement object participates in subsequent calculations with a new lifespan starting point. After processing the states of both unreplaced and replaced objects, the states of the two types of objects are combined to obtain the mixed-age lifespan state of the power battery pack after partial repair or replacement. During the combination process, the historical inheritance flag is used as an object-level control parameter to select the lifespan state for each object. Objects inheriting history adopt their historical lifespan state, while objects not inheriting history adopt their initial lifespan state. For the first... An object, whose mixed-age lifetime state is represented as follows: ; in, Indicates the first The mixed-age lifetime state of an object at the current moment. Indicates the corresponding historical inheritance marker, This indicates the product's lifespan history before repair or replacement. Indicates the initial lifespan state after repair or replacement, when When, the historical lifetime status is adopted, when At that time, the initial lifespan state is adopted, thereby forming a unified state expression that includes objects with different lifespan starting points in the same state space. In one specific embodiment, a power battery pack comprises eight modules, with the fifth module being the replacement target and the remaining modules not being replaced. For the non-replaced modules, their historical lifespan states stored before maintenance are directly used as the current state input. For the fifth module, its historical data is not retrieved; instead, the initial state parameters obtained after maintenance are written into the state table. Subsequently, based on the historical inheritance tags corresponding to each module, the historical states of the seven non-replaced modules are combined with the initial state of the fifth module to form the mixed-age lifespan state of the entire pack. This mixed-age lifespan state serves as the input for subsequent calculations of the overall pack lifespan constraint relationship and online prediction of remaining lifespan.

[0028] Furthermore, the initial lifespan state is reconstructed based on the initial operating data of the replaced object after partial repair or replacement, specifically including: Obtain the first round of operational data for the replaced object after partial repair or replacement; Voltage change rate, current response characteristics, and temperature response characteristics were extracted from the first round of operational data. Based on the mapping relationship between the preset life state parameters and operating characteristics, the voltage change rate, current response characteristics and temperature response characteristics are solved in reverse to determine the initial life state parameters of the replacement object. Construct the initial lifetime state of the replacement object based on the initial lifetime state parameters.

[0029] Specifically, when the replacement object cannot be directly obtained in its usable initial life state after partial repair or replacement, the first round of operating data after the replacement object is put back into operation is read. The first round of operating data can be selected from the operating data segments of the replacement object during the first charging process, the first discharging process, the first complete start-stop process, or the first time to reach the preset running time. The voltage data, current data, and temperature data during this process are recorded in chronological order. The selection of the first round of operating data enables the dynamic response of the replacement object after reconnecting to the power battery pack to be completely preserved, thereby providing an input basis for the subsequent reconstruction of the initial life state. After obtaining the initial operating data, voltage change rate, current response characteristics, and temperature response characteristics are extracted from the data. The voltage change rate can be calculated from the voltage difference between adjacent sampling moments and the sampling time interval during the initial operation. The current response characteristics can be extracted from the current change amplitude, response duration, or current fluctuation pattern during the initial operation. The temperature response characteristics can be extracted from the temperature change amplitude, temperature rise rate, or temperature stabilization time during the initial operation. To uniformly represent the characteristics of the replaced object during the initial operation, the voltage change rate, current response characteristics, and temperature response characteristics can be combined into an operating feature vector. ; in, Indicates the first The running feature vector corresponding to each replaced object. Indicates the first Voltage change rate characteristics of the replaced object. Indicates the first Current response characteristics of a replacement object Indicates the first The temperature response characteristics of the replaced object are used to extract features from the first round of running data, thereby converting the original running data into state representation quantities for inverse solution. After obtaining the operational feature vector, the operational feature vector is solved inversely according to the preset mapping relationship between lifetime state parameters and operational features to determine the initial lifetime state parameters of the replacement object. The mapping relationship is used to characterize the impact of changes in lifetime state parameters on changes in operational features, and can be given through a pre-established state-feature correspondence table, regression relationship, or parameterized mapping model. In this embodiment, the initial lifetime state parameters of the replacement object are represented as follows: Then its determination process can be expressed as: ; in, Indicates the first The initial state parameters of the lifetime of the replacement object. This represents the mapping relationship between lifetime state parameters and operational characteristics. This represents the operational feature vector extracted from the data from the first round of operation. This represents the lifetime state parameter that minimizes the target scalar. The error square norm is represented. When solving in reverse, the state parameter that minimizes the error between the mapped output and the actual running feature vector is selected as the initial state parameter of the lifetime of the replacement object. Thus, in the absence of direct initial state information, the initial state of the lifetime can be restored based on the first round of response after the replacement object is put back into operation. After obtaining the initial life state parameters, the initial life state of the replacement object is constructed based on the initial life state parameters and written into the object state table. The initial life state can be composed of one or more life state parameters, which are used to characterize the initial life level of the replacement object after partial maintenance or partial replacement. After the construction is completed, the replacement object has a state input that can participate in the combination of mixed-age life states. Subsequently, it can be combined with the life history state inherited by the non-replaced object to form the mixed-age life state of the power battery pack.

[0030] S5. Based on the mixed-age life status and operating data, determine the constraint relationship between each object in the power battery pack and the overall life of the pack. Furthermore, the S5 steps specifically include: Based on the mixed-age life state, determine the life state parameters corresponding to each object in the power battery pack. Constraint functions can be constructed based on the lifetime state parameters of each object, or fuzzy membership functions can be constructed based on the lifetime state parameters of each object. The life state parameters corresponding to each object are compared with the preset life constraints, and the constraint relationship of each object in the power battery pack on the life of the entire pack is determined according to the constraint function or fuzzy membership function.

[0031] Specifically, after obtaining the mixed-age life status of the power battery pack, the life status parameters corresponding to each battery cell, module or battery unit are extracted by combining the operating data. The life status parameters are used to characterize the current life status of each object and its changes, and are written into the corresponding object status table as input for subsequent constraint relationship calculation. After obtaining the lifetime state parameters corresponding to each object, the lifetime state parameters are mapped to the corresponding constraint quantities to characterize the degree of constraint that object places on the overall package lifetime. For the first... An object, whose constraints are represented as follows: ; in, Indicates the first The constraint quantity corresponding to each object Indicates the first Lifetime state parameters of an object This represents the mapping relationship between lifetime status parameters and constraint quantities. The mapping relationship can be set according to preset lifetime constraint conditions, so that objects with poor lifetime status or large changing trends correspond to larger constraint quantities. Subsequently, the constraint quantities corresponding to each object are compared with the preset lifespan constraint conditions, and the constraint relationship between each object in the power battery pack and the overall lifespan is determined based on the comparison results. Objects with larger constraint quantities correspond to objects that play a major role in constraining the overall lifespan, while other objects correspond to secondary or non-dominant constraint objects, thus forming a set of object-level constraint relationships, which are used to determine the constraint objects in subsequent online prediction of remaining lifespan. For example, in a power battery pack, after calculating the constraint quantities for each module, if the constraint quantity of the third module is the largest, then the third module is identified as the object that plays a major constraint role on the overall lifespan of the pack. The constraint degree of the remaining modules is determined according to the magnitude of their constraint quantities. The constraint relationship obtained in this way can be directly used for the selection of key objects in subsequent remaining life prediction.

[0032] S6. Perform online prediction of the remaining life of the power battery pack according to the constraint relationship, and obtain the prediction result of the remaining life of the power battery pack. Furthermore, step S6 specifically includes: Based on the constraint relationship, the constraint object that constrains the remaining life of the power battery pack is determined. Obtain the lifetime status parameters and operational data corresponding to the constraint object; The remaining lifetime of the constraint object is determined based on the lifetime status parameters and operating data of the constraint object. Based on the remaining lifespan of the constrained object, the remaining lifespan prediction result of the power battery pack is obtained; Among them, historical inheritance markers and mixed-age life status serve as inputs for online prediction of remaining lifespan, controlling how different objects' lifespan history information participates.

[0033] Specifically, the constraint objects that constrain the remaining life of the power battery pack are determined based on the constraint relationship. The constraint objects can be a single battery cell, a single module, a single battery unit, or a set of constraint objects composed of multiple objects. The constraint objects are extracted from all objects in the power battery pack, and the life status parameters and operating data corresponding to the constraint objects are read, so that the subsequent online prediction of the remaining life is focused on the scope of objects that have a real constraint effect on the life of the entire pack. After reading the lifespan state parameters and operational data corresponding to the constrained object, the historical inheritance flag and mixed-age lifespan state are used as inputs for online prediction of remaining lifespan. The historical inheritance flag distinguishes whether the lifespan state of the constrained object originates from historical lifespan information before repair or replacement, or from the initial lifespan state after repair or replacement. The mixed-age lifespan state characterizes the unified lifespan state of the constrained object at the current moment. Based on the lifespan state parameters, operational data, historical inheritance flag, and mixed-age lifespan state, the remaining lifespan corresponding to the constrained object is determined. To characterize the predicted remaining lifespan of the power battery pack, the remaining lifespan corresponding to the constrained object can be taken as the remaining lifespan of the entire pack, expressed as: ; in, This indicates the predicted remaining lifespan of the power battery pack. This indicates the remaining lifetime of the constrained object. This indicates the constraint object number that constrains the overall lifespan of the package; After determining the remaining lifespan of the constraint object, the remaining lifespan of the constraint object is output as the remaining lifespan prediction result of the power battery pack. If the constraint object is a set of multiple objects, the remaining lifespan prediction result of the power battery pack can be determined based on the remaining lifespan of each object in the constraint object set. Thus, the online prediction process of the remaining lifespan of the power battery pack after partial repair or replacement no longer directly processes all objects uniformly, but selects the objects that actually play a constraint role based on the constraint relationship, and combines the historical inheritance mark and mixed age lifespan status to control the participation of different objects' lifespan history information, thereby obtaining the remaining lifespan prediction result of the power battery pack. For example, in a power battery pack, after determining that the third module is the constraint object that constrains the life of the entire pack, the life status parameters and operating data corresponding to the third module are read, and combined with the historical inheritance mark and mixed-age life status corresponding to the module, the remaining life of the third module is determined. Then, the remaining life of the third module is output as the remaining life prediction result of the entire pack. If the constraint relationship indicates that the third module and the fifth module together constitute a set of constraint objects, then the remaining life prediction result of the power battery pack is further determined according to the remaining life of the third module and the fifth module respectively.

[0034] S7. Based on the remaining life prediction results and mixed-age life status, evaluate the remaining life prediction results of the power battery pack to obtain the remaining life assessment results.

[0035] Furthermore, the S7 steps specifically include: Based on the mixed-age life status, determine the evaluation parameters corresponding to the remaining life prediction results; Compare the evaluation parameters with the preset evaluation conditions; Based on the comparison results, the evaluation results corresponding to the remaining life prediction results are obtained; The evaluation parameters are calculated based on the life state parameters of each object in the mixed-age life state and their changing trends.

[0036] Furthermore, the assessment results include availability markers, warning markers, or trust level markers.

[0037] Specifically, after obtaining the predicted remaining life of the power battery pack, the predicted remaining life is evaluated in conjunction with the mixed-age life status. During the evaluation, the life status parameters and their changing trends corresponding to each object in the mixed-age life status are read, and the evaluation parameters corresponding to the predicted remaining life are calculated based on these life status parameters and their changing trends. The evaluation parameters are used to characterize the degree of matching and stability between the current predicted remaining life and the actual life status of the power battery pack. For ease of unified calculation, the evaluation parameters can be expressed as: ; in, Indicates the evaluation parameters, This represents the set of lifetime state parameters corresponding to each object in a mixed-age lifetime state. This represents the set of trends in the lifetime state parameters corresponding to each object. This represents the evaluation parameter calculation function, which allows the current state and evolution trend of each object in the mixed-age lifespan state to be uniformly mapped into quantitative parameters for evaluating the remaining lifespan prediction results. After obtaining the evaluation parameters, they are compared with preset evaluation conditions. Based on the comparison results, the evaluation result corresponding to the remaining lifetime prediction result is determined. The preset evaluation conditions can be single-threshold conditions or multi-threshold tiered conditions. When using single-threshold conditions, the usability of the remaining lifetime prediction result can be determined based on whether the evaluation parameters reach the preset threshold. When using multi-threshold tiered conditions, the confidence level corresponding to the remaining lifetime prediction result can be determined based on the interval in which the evaluation parameters are located. If usability is used as the evaluation result, it can be represented as: ; in, Indicates availability. Indicates the preset evaluation threshold. When When this condition is met, it indicates that the current remaining lifetime prediction results meet the preset evaluation conditions. When this occurs, it indicates that the current remaining life prediction results do not meet the preset evaluation conditions; When the evaluation results are output in a hierarchical manner, the corresponding confidence level marker can be determined based on the comparison relationship between the evaluation parameters and multiple preset thresholds. The evaluation results may include an availability marker, a warning marker, or a confidence level marker. The availability marker is used to indicate whether the current remaining life prediction result meets the usage conditions, the warning marker is used to indicate whether the current remaining life prediction result triggers the preset warning conditions, and the confidence level marker is used to indicate the confidence level of the current remaining life prediction result. After the evaluation results and the remaining life prediction results are associated and output, the online prediction results of the remaining life of the power battery pack after partial repair or partial replacement can simultaneously have result values ​​and status determination information. For example, in a power battery pack, after obtaining the remaining life prediction result of the entire pack, the life status parameters and their changing trends of each module in the mixed-age life state are further read, and the corresponding evaluation parameters are calculated. If the evaluation parameter is higher than the preset evaluation threshold, the available flag corresponding to the remaining life prediction result is set to valid. If the evaluation parameter is in the warning threshold range, the corresponding warning flag is output. If a graded method is adopted, the corresponding confidence level flag is output according to the range in which the evaluation parameter is located. Thus, the remaining life prediction result of the power battery pack can not only reflect the size of the remaining life, but also reflect the evaluation status of the prediction result in the current mixed-age life state.

[0038] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for online prediction and evaluation of the remaining life of a power battery, characterized in that, include: S1. Obtain the operating data of the power battery pack and monitor whether the power battery pack has undergone partial repair or replacement events; S2. Upon detecting the partial repair or partial replacement event, determine the replacement and non-replacement items in the power battery pack; The replacement objects are battery cells, modules, or battery units that have been replaced, while the non-replaced objects are battery cells, modules, or battery units that have not been replaced. S3. Construct a history inheritance marker based on the replaced object and the unreplaced object, wherein the history inheritance marker is used to characterize whether the corresponding object inherits the life history information before repair or replacement; S4. Based on the historical inheritance marker, inherit the life history information before repair or replacement for the unreplaced object, and reset or rebuild the initial life state for the replaced object to obtain the mixed-age life state after partial repair or partial replacement. S5. Based on the mixed-age life status and the operating data, determine the constraint relationship between each object in the power battery pack and the overall life of the pack. S6. Perform online prediction of the remaining life of the power battery pack according to the constraint relationship to obtain the prediction result of the remaining life of the power battery pack. S7. Based on the remaining life prediction results and the mixed-age life status, evaluate the remaining life prediction results of the power battery pack to obtain the remaining life evaluation results.

2. The method for online prediction and evaluation of the remaining life of a power battery according to claim 1, characterized in that, The S1 step specifically includes: The system acquires operational data for each individual battery cell, module, or battery unit in the power battery pack, including voltage data, current data, temperature data, and state of charge data. Obtain the maintenance records, configuration records, object coding information, or connection relationship information corresponding to the power battery pack; Based on the operational data and the maintenance records, configuration records, object coding information, or connection relationship information, monitor whether the power battery pack has undergone partial maintenance or partial replacement events.

3. The method for online prediction and evaluation of the remaining life of a power battery according to claim 1, characterized in that, The S2 step specifically includes: After detecting a partial repair or replacement event in the power battery pack, identify the individual battery cell, module or battery unit in the power battery pack that has experienced changes in object coding, connection relationship, repair marking, or sudden changes in operating parameters. The identified individual battery cells, modules, or battery units are determined as replacement targets; The battery cells, modules, or battery units in the power battery pack that are not replaced are identified as not replaced.

4. The method for online prediction and evaluation of the remaining life of a power battery according to claim 3, characterized in that, The S3 step specifically includes: Construct a first history inheritance tag for the replaced object; Construct a second history inheritance marker for the unreplaced object; The first historical inheritance marker is used to indicate that the corresponding object does not inherit the life history information before partial repair or partial replacement, and the second historical inheritance marker is used to indicate that the corresponding object inherits the life history information before partial repair or partial replacement. The historical inheritance marker includes an object identifier field and an inheritance status field. The object identifier field is used to characterize the corresponding battery cell, module, or battery unit, and the inheritance status field is used to characterize the inheritance method of the life history information of the corresponding object. The historical inheritance marker is generated based on the object encoding information, connection relationship information, or maintenance records, and the inheritance status field is updated based on the operation data during subsequent operation.

5. The method for online prediction and evaluation of the remaining life of a power battery according to claim 4, characterized in that, The S4 step specifically includes: Based on the second historical inheritance marker, the life history information before partial repair or partial replacement is inherited for the unreplaced object; Based on the first historical inheritance marker, the initial lifespan state of the replacement object is reset, or the initial lifespan state is reconstructed based on the initial operating data of the replacement object after partial repair or partial replacement. The lifetime history information inherited by the unreplaced object is combined with the initial lifetime state of the replaced object to construct a mixed-age lifetime state after partial repair or partial replacement. The historical inheritance marker serves as a control parameter for constructing the mixed-age lifespan state, and is used to control the way the lifespan history information of the corresponding object is introduced.

6. The method for online prediction and evaluation of the remaining life of a power battery according to claim 5, characterized in that, The step of reconstructing the initial lifespan state based on the initial operating data of the replaced object after partial repair or replacement specifically includes: Obtain the first round of operational data of the replaced object after partial repair or partial replacement; Based on the first round of operational data, voltage change rate, current response characteristics, and temperature response characteristics were extracted. Based on the preset mapping relationship between life state parameters and operating characteristics, the voltage change rate, current response characteristics and temperature response characteristics are solved in reverse to determine the initial life state parameters of the replacement object. The initial lifespan state of the replacement object is constructed based on the initial lifespan state parameters.

7. The method for online prediction and evaluation of the remaining life of a power battery according to claim 1, characterized in that, The S5 step specifically includes: Based on the mixed-age lifespan status, the lifespan status parameters corresponding to each object in the power battery pack are determined. Constraint functions can be constructed based on the lifetime state parameters of each object, or fuzzy membership functions can be constructed based on the lifetime state parameters of each object. The lifespan state parameters corresponding to each object are compared with preset lifespan constraints, and the constraint relationship between each object in the power battery pack and the overall lifespan is determined according to the constraint function or the fuzzy membership function.

8. The method for online prediction and evaluation of the remaining life of a power battery according to claim 1, characterized in that, Step S6 specifically includes: Based on the aforementioned constraint relationship, the constraint object that constrains the remaining lifespan of the power battery pack is determined. Obtain the lifetime status parameters and operational data corresponding to the constraint object; The remaining lifespan of the constraint object is determined based on the lifespan status parameters and operating data corresponding to the constraint object. Based on the remaining lifespan corresponding to the constrained object, the remaining lifespan prediction result of the power battery pack is obtained; The historical inheritance marker and the mixed-age lifespan status serve as inputs for online prediction of remaining lifespan, and are used to control the participation of different objects' lifespan history information.

9. The method for online prediction and evaluation of the remaining life of a power battery according to claim 1, characterized in that, The S7 step specifically includes: Based on the mixed-age lifespan status, determine the evaluation parameters corresponding to the remaining lifespan prediction results; The evaluation parameters are compared with preset evaluation conditions; Based on the comparison results, the evaluation results corresponding to the remaining life prediction results are obtained; The evaluation parameters are calculated based on the lifespan parameters and their changing trends of each object in the mixed-age lifespan state.

10. The method for online prediction and evaluation of the remaining life of a power battery according to claim 9, characterized in that, The assessment results include availability markers, warning markers, or trust level markers.

Images