Power battery pack monitoring system based on satellite industrial internet

Abstract

The invention relates to the technical field of battery monitoring, in particular to a power battery pack monitoring system based on the satellite industrial internet, which comprises a data acquisition terminal, a communication module, a data storage module and a diagnosis processing module, a communication module and a time or mileage mark are uploaded and cached through a satellite network, a storage module stores data and diagnosis results, a cloud end calculates the difference between the input voltage of an inverter and the voltage of a battery pack in a segmented mode and builds a set with current, linear fitting is carried out to obtain initial resistance, outliers are removed, and fitting is carried out to obtain filtering resistance after moving average noise reduction. And converting the average temperature and the material coefficient into a reference temperature, generating a threshold value based on normal sample distribution, outputting a health state or an alarm, and distinguishing anomalies of an external wire harness, a busbar and a battery. The method can solve the problem that in the actual operation process of the automobile power battery pack, when interconnected parts are abnormal, it is difficult to identify the abnormal parts timely and accurately.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery monitoring, and particularly relates to a power battery pack monitoring system based on satellite industrial internet. BACKGROUND

[0002] In applications such as electric vehicles, the power battery pack is a key energy unit of electric equipment, and its interconnection structure is subjected to comprehensive stresses such as vibration, thermal cycling and current impact for a long time. The changes in the contact state in the connection path, conductor aging, connection loosening and other phenomena can cause electrical performance degradation, and further may cause heating, efficiency decline and even safety risks. In the prior art, the evaluation of the interconnection health state is mostly dependent on offline measurement, disassembly inspection or special test means in the factory or maintenance scene, which can obtain relatively accurate results, but is difficult to meet the continuous monitoring needs of the vehicle during in-service operation, and is also not conducive to long-term trend tracking and preventive maintenance for large-scale vehicle fleets.

[0003] On the other hand, in the prior art, the commonly used remote monitoring and diagnosis is mostly based on ground cellular networks or local area networks. When the vehicle is in areas with insufficient communication conditions such as mountains, deserts, islands and cross-border trunk lines, problems such as discontinuous data return, difficulty in real-time updating of diagnosis models and lagging of alarm triggering are prone to occur. At the same time, the running signals under actual road conditions have characteristics such as strong noise, frequent working condition switching and significant temperature changes, which makes it easy to be disturbed by abnormal points and environmental factors when inferring the interconnection state based on running data, causing false positives or false negatives; and the interconnection abnormality and the single battery abnormality may be coupled with each other in performance, further increasing the positioning and decision difficulty. SUMMARY

[0004] In view of the above technical problems, the present application provides a power battery pack monitoring system based on satellite industrial internet, which solves the problem that the interconnection components of the power battery pack are difficult to be identified in time and accurately when they are aged or abnormal during actual operation, especially in the scene where the communication conditions of the vehicle running area are limited. The traditional remote monitoring based on ground network is prone to data discontinuity and diagnosis lag, thereby causing difficulty in early discovery and disposal of fault hazards.

[0005] Other characteristics and advantages of the present application will become apparent from the following detailed description, or will be learned by practice of the present application.

[0006] According to an aspect of the present application, a power battery pack monitoring system based on satellite industrial internet is provided, which comprises a data acquisition terminal, a communication module, a data storage module and a diagnosis processing module. The data acquisition terminal is configured to acquire monitoring data from a battery management system, an inverter, a low-voltage power supply device, and a temperature sensor, the monitoring data including at least battery pack voltage, battery pack current, inverter input voltage, inverter input current, and battery module temperature; The communication module is configured to time and / or mileage mark the monitoring data and send it to the diagnostic processing module through a satellite network; The data storage module is configured to store the monitoring data and diagnostic results generated by the diagnostic processing module; The diagnostic processing module is built in the cloud and includes a segmented construction unit, a voltage difference calculation unit, an interconnection resistance estimation unit, a multi-step filtering unit, a temperature compensation unit, and a fault determination unit. The segmented construction unit is configured to segment the monitoring data to obtain data segments. The voltage difference calculation unit is configured to calculate the voltage difference between the inverter input voltage and the battery pack voltage in each data segment and form a current-voltage difference data set with the corresponding battery pack current. The interconnection resistance estimation unit is configured to perform linear fitting on the current-voltage difference data set to obtain an initial trend line and use the slope of the initial trend line as an initial estimate of the interconnection resistance. The multi-step filtering unit is configured to remove abnormal data points deviating from the initial trend line based on statistical outlier criteria and perform adaptive moving average on the current-voltage difference data set to reduce noise and obtain a multi-step filtered interconnection resistance by fitting again. The temperature compensation unit is configured to obtain the average operating temperature corresponding to the data segment and convert the multi-step filtered interconnection resistance to a reference temperature according to the temperature correction coefficient of the conductor material to obtain a temperature-compensated interconnection resistance. The fault determination unit is configured to compare the temperature-compensated interconnection resistance with a reference model constructed from the statistical distribution of temperature-compensated interconnection resistances of normal samples to output the health status or fault warning of the battery pack and distinguish external wiring harness abnormalities, busbar abnormalities, or single battery abnormalities when preset diagnostic rules are met.

[0007] Further, the data acquisition terminal comprises a vehicle-mounted data recorder, an interface adapter and a time synchronization unit, the interface adapter is respectively connected with the battery management system, the inverter, the low-voltage power supply device and the temperature sensor to synchronously acquire the monitoring data at a preset sampling period, and add time stamp and / or running mileage information to the monitoring data under the unified time reference of the time synchronization unit; wherein the monitoring data further comprises at least one auxiliary diagnostic parameter for decomposing a fault component, the auxiliary diagnostic parameter at least comprises at least one of the following: total voltage of the battery pack, output voltage of the battery pack, battery voltage sum calculated from a set of single battery voltages, single battery voltage extreme difference, output voltage of the low-voltage power supply device, output current of the low-voltage power supply device, driving motor torque, driving motor speed, vehicle running mode, environmental temperature, and charge / discharge direction identifier; and the communication module is configured to perform integrity check, compression encoding and local caching on the monitoring data on the vehicle side, and sends the monitoring data to the diagnostic processing module by using the batch uploading and breakpoint resuming mechanism.

[0008] Further, the fault determination unit comprises a fault positioning subunit, the fault positioning subunit is configured to construct a plurality of current-voltage difference profiles based on different voltage information in the auxiliary diagnostic parameter and obtain a resistance characteristic set respectively; the different voltage information at least comprises: total voltage between the battery pack, output voltage of the battery pack, battery voltage information provided by the battery management system, battery voltage sum calculated from a set of single battery voltages, and battery voltage extreme difference calculated from the set of single battery voltages; the resistance characteristic set at least comprises: a first resistance characteristic representing a busbar and battery connection path, a second resistance characteristic representing a busbar and battery connection path and superimposing a wire harness and connector path, a third resistance characteristic representing an equivalent path between batteries, and a fourth resistance characteristic representing consistency change of the batteries; wherein the fault positioning subunit determines a wire harness and connector resistance component according to the difference between the first resistance characteristic and the second resistance characteristic, determines a busbar resistance component according to the difference between the first resistance characteristic and the third resistance characteristic, and determines a single battery abnormality indication according to the fourth resistance characteristic; the fault determination unit outputs a component-level diagnostic conclusion based on the combination rule of the resistance component and the temperature-compensated interconnection resistance to realize the differentiation of external wire harness abnormality, busbar abnormality and single battery abnormality.

[0009] Further, the segment construction unit is configured to divide the monitoring data into a plurality of data segments according to running mileage accumulation, running time accumulation or vehicle running mode, and perform data quality control on each data segment, the data quality control comprises: The system determines whether the battery pack current within the data segment covers a preset current variation range to meet the excitation conditions for resistance identification, whether the voltage difference has a monotonically approximately linear relationship with the battery pack current, eliminates sample points containing sensor saturation, communication packet loss, or state switching transients, and eliminates charging and non-traction conditions based on the vehicle operation mode. After passing the data quality control, the voltage difference calculation unit calculates the difference between the inverter input voltage and the battery pack voltage to obtain the original voltage difference sequence, and aligns the original voltage difference sequence with the battery pack current in time to form a scatter pair; The voltage difference calculation unit is also configured to perform current binning or sliding window aggregation on the scatter pairs, and calculate the mean, weighted mean or median of the original voltage difference sequence in each current interval or window as a representative voltage difference to form a profile point set that is deredundant and easy to fit.

[0010] Furthermore, the multi-step filtering unit includes an outlier removal subunit, a redundancy suppression subunit, a window smoothing subunit, and a refitting subunit, wherein: The outlier removal subunit is configured to use the initial trend line as a reference to calculate the residual of each profile point in the profile point set relative to the initial trend line, and determine the outlier criterion based on the statistical dispersion of the residuals, so as to remove abnormal data points that deviate from the initial trend line. The redundancy suppression subunit is configured to perform aggregation on profile points with repeated or nearly repeated current values ​​to reduce fitting bias caused by repeated sampling. The window smoothing subunit is configured to perform moving average, moving median, or exponential smoothing on the profile point set after outlier removal and redundancy suppression, and adaptively select a smoothing window between the first window width and the second window width according to the degree of fluctuation of the battery pack current in different current ranges, so that a larger window is used in the range with large current fluctuations and a smaller window is used in the range with small current fluctuations. The refitting subunit is configured to perform a second linear fit on the smoothed profile point set, using the slope of the second linear fit as the multi-step filter interconnect resistance, and writing the difference between the multi-step filter interconnect resistance and the initial estimate as a noise level index into the data storage module.

[0011] Furthermore, the temperature compensation unit is configured to obtain the temperature sequence corresponding to the data segment from the battery module temperature, and obtain the average operating temperature by time averaging, steady-state interval filtering, or abnormal temperature removal. The temperature compensation unit is further configured to select a conductor material temperature correction coefficient based on the material type of the conductive components in the battery pack, and convert the multi-step filtering interconnect resistance to the reference temperature based on the linear approximation relationship of conductor resistance changing with temperature, so as to eliminate the influence of temperature fluctuation on resistance judgment; wherein, the material type includes at least busbar material, connector material and wire harness conductor material, and the temperature compensation unit calculates the resistance compensation amount for different material types and performs weighted synthesis to obtain the temperature-compensated interconnect resistance that is closer to the actual structure; and when the average operating temperature is unavailable, the temperature compensation unit generates an alternative temperature input based on the ambient temperature, vehicle thermal management status or historical temperature model.

[0012] Furthermore, the baseline model includes at least one of a threshold model based on statistical distribution and an aging model based on trend. The threshold model obtains a center position parameter and a dispersion parameter by fitting the temperature-compensated interconnect resistance of normal samples according to the distribution, and generates an adaptive alarm threshold based on the dispersion parameter. The aging model obtains a trend term representing progressive aging by smoothing and regressing the change sequence of the temperature-compensated interconnect resistance with operating mileage or operating time. The fault determination unit is configured to calculate the deviation of the temperature-compensated interconnect resistance under test relative to the center position parameter within each data segment, and determine, in conjunction with the trend term, whether the deviation is caused by short-term noise, environmental changes, or the battery pack interconnect system.

[0013] Furthermore, the diagnostic processing module also includes an efficiency evaluation unit, which is configured to calculate power and energy consumption within a preset driving range and associate it with the temperature compensation interconnection resistor. The efficiency evaluation unit includes: a low-voltage power calculation subunit, a battery output power calculation subunit, an inverter input power calculation subunit, a motor output power calculation subunit, and an efficiency and energy consumption calculation subunit. The low-voltage power calculation subunit is used to multiply the output voltage of the low-voltage power supply device by the output current of the low-voltage power supply device to obtain the low-voltage output power; The battery output power calculation subunit is used to multiply the battery pack voltage and the battery pack current to obtain the battery terminal power, and subtract the low-voltage output power to obtain the battery output power. The inverter input power calculation subunit is used to multiply the inverter input voltage and the inverter input current to obtain the inverter terminal power, and subtract the low-voltage output power to obtain the inverter input power. The motor output power calculation subunit is used to convert the drive motor speed into angular velocity and multiply it by the drive motor torque to obtain the motor output power. The efficiency and energy consumption calculation subunit is used to accumulate the output power of the motor and the output power of the battery within the preset driving range and take the ratio to obtain the battery efficiency, accumulate the output power of the motor and the input power of the inverter within the preset driving range and take the ratio to obtain the motor efficiency, and perform numerical integration on the time series of the battery output power and perform energy unit conversion to obtain the energy consumption index. The efficiency evaluation unit is further configured to update the motor efficiency only when the drive motor is in a preset high-efficiency operating range, and to use the motor efficiency to verify the comparability of the battery efficiency, so as to distinguish the efficiency change caused by motor aging from the efficiency change caused by battery pack aging.

[0014] The technical solution disclosed herein has the following beneficial effects: This disclosure utilizes monitoring data obtainable during vehicle operation to establish a continuous assessment capability of interconnected status in the cloud. Through satellite industrial internet, it achieves wider coverage and more stable remote transmission, enabling vehicles to maintain the necessary data links for diagnosis even in areas with weak or unavailable terrestrial networks, thereby improving the continuity of monitoring and the timeliness of alarms. By rationally organizing and quality-controlling operational data, and employing more robust mechanisms for handling anomalies and noise, the impact of random road disturbances on diagnostic conclusions can be reduced, enhancing the repeatability and comparability of results across different vehicles, environments, and operating conditions.

[0015] Furthermore, this disclosure introduces a normalization process for the impact of temperature, ensuring that interconnect status indicators have consistent discriminative significance under different thermal conditions, which is beneficial for distinguishing between gradual aging and sudden anomalies. Simultaneously, by comprehensively utilizing multi-source voltage information, the ability to differentiate fault types can be improved without adding complex hardware, providing more targeted prompts for maintenance decisions. Combined with an alarm and maintenance closed-loop mechanism, actual handling information can inversely promote model and rule updates, thereby improving diagnostic accuracy during long-term operation and supporting collaborative assessment of preventative maintenance, remote operation and maintenance, and efficiency degradation. Attached Figure Description

[0016] Figure 1 This is a structural block diagram of a power battery pack monitoring system based on satellite industrial internet, as described in the embodiments of this specification. Figure 2 This is a structural block diagram of the battery interconnection system in the embodiments of this specification. Detailed Implementation

[0017] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this disclosure more comprehensive and complete, and to fully convey the concept of the example embodiments to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more of the specific details omitted, or other methods, components, apparatus, steps, etc., can be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0018] Furthermore, the accompanying drawings are merely illustrative of this disclosure. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0019] This disclosure provides a power battery pack monitoring system based on satellite industrial internet. (Refer to...) Figure 1 The diagram shown is a structural block diagram of a power battery pack monitoring system based on satellite industrial internet, according to one embodiment. The monitoring system includes a data acquisition terminal 101, a communication module 102, a data storage module 103, and a diagnostic processing module 104.

[0020] The data acquisition terminal 101 is configured to acquire monitoring data from the battery management system, inverter, low-voltage power supply device and temperature sensor. The monitoring data includes at least the battery pack voltage, battery pack current, inverter input voltage, inverter input current and battery module temperature. The data acquisition terminal 101 includes an on-board data recorder, an interface adapter, and a time synchronization unit. The interface adapter establishes communication connections with the battery management system, inverter, low-voltage power supply device, and temperature sensor, respectively, to synchronously collect monitoring data according to a preset sampling period, and adds timestamps and / or mileage information to the monitoring data under the unified time reference of the time synchronization unit. The monitoring data also includes at least one auxiliary diagnostic parameter for decomposing faulty components. The auxiliary diagnostic parameter includes at least one of the following: total battery pack voltage Vc1, battery pack output voltage Vc2, battery voltage and Vc3 calculated from the set of individual battery voltages, extreme difference of individual battery voltages, low-voltage power supply device output voltage, low-voltage power supply device output current, drive motor torque, drive motor speed, vehicle operating mode, ambient temperature, and charging / discharging direction indication. Furthermore, the communication module is configured to perform integrity verification, compression encoding, and local caching on the on-board side of the monitoring data, and to send the monitoring data to the diagnostic processing module using a batch upload and breakpoint resume mechanism.

[0021] The data acquisition terminal's role is to collect key quantities related to the high-voltage battery pack and its connection system during vehicle operation in a faithful, synchronous, and traceable manner, ensuring that the voltage, current, temperature, and operating condition information required for subsequent calculations are aligned under the same time reference. In real-world vehicle data acquisition scenarios, the monitoring terminal can record the aforementioned key diagnostic indicators such as voltage, current, and temperature at a frequency of 1Hz and send the data to the server at minute intervals to reduce the impact of communication interruptions on data loss and improve transmission reliability. This monitoring data is not limited to a small number of signals; under real-world operating conditions, it can cover multiple parameters (e.g., approximately 40 parameters), including battery pack voltage, current, and inverter input voltage, to support the differentiation of multiple fault modes and health assessment. The time synchronization unit provides a unified time reference for the collected data, allowing data from different controller / sensor links to be timestamped consistently upon writing. In automotive applications, mileage information can also be added under the same time reference, forming time-mileage dual-indexed sequence data, facilitating playback, statistics, and comparison by distance window or event window (e.g., extracting voltage-current points by fixed mileage segments to form I-V curves).

[0022] like Figure 2 As shown, Figure 2This is a structural block diagram of a battery interconnection system, showing the structure of one battery pack. The system comprises multiple such battery packs. The battery pack voltages are specifically Vc1, Vc2, and Vc3, where Vc1 is the total battery pack voltage, Vc2 is the battery pack output voltage, and Vc3 is the sum of the individual cell voltages. The battery pack includes a busbar 201, an external wiring harness 202, and individual cells 203. The busbar 201 connects multiple batteries and efficiently conducts current. The external wiring harness 202 transmits power and provides electrical connection between the battery system and external power / drive systems (it may also include communication wiring harnesses). Since the busbar and wiring harness do not participate in electrochemical reactions, their failure / aging is more often manifested as increased connection resistance, leading to systemic effects such as decreased energy efficiency and increased energy consumption. Therefore, the data acquisition terminal needs to collect not only the voltage and current (Vc1, Vc2, and battery pack current) exhibited by the battery pack, but also quantities that can reflect consistency and connection anomalies at the individual cell level (e.g., the set of individual cell voltages used to obtain Vc3, and the extreme value difference calculated from the set of individual cell voltages; by definition, the maximum / minimum voltage difference of an individual cell can be expressed as Vc4). By simultaneously retaining Vc1, Vc2, Vc3, and the individual cell extreme value difference, a more sufficient observational basis can be provided for subsequent diagnostic inferences such as decomposing faulty components into busbars / wiring harnesses / individual cell anomalies, without changing the boundary responsibilities of the acquisition terminal.

[0023] When the monitoring data also includes at least one auxiliary diagnostic parameter for decomposing faulty components, the data acquisition terminal can synchronously collect information such as the output voltage / current of the low-voltage power supply device, drive motor torque, drive motor speed, vehicle operating mode, ambient temperature, and charging / discharging direction indication at the same sampling cycle, and output it after aligning the timestamps with the high-voltage side voltage and current. The reason why the output voltage VL and output current IL of the low-voltage power supply device are needed is that its output power can be directly obtained from the formula: This power quantity is used in energy flow analysis to subtract low-voltage side consumption from the battery output power, thereby avoiding miscounting low-voltage load power consumption as drive-side energy consumption. Correspondingly, the battery output power can be expressed as: ,in and These correspond to the battery pack voltage and battery pack current, respectively; the mechanical output power of the drive motor can be expressed as: ,in For motor angular velocity, The power calculation is based on motor torque (the data acquisition terminal only needs to collect speed / torque data). Furthermore, to support the evaluation of the battery pack connection status, the data acquisition terminal must simultaneously collect the inverter input voltage and the total battery pack voltage, forming a data item related to the difference between the two: the resistance of the connection system composed of multiple battery packs can be characterized by the voltage difference between the inverter input and the total battery pack voltage, combined with the battery pack current. This voltage difference provides a key clue to changes in the connection system resistance; when a voltage difference-current I-V relationship is subsequently established, according to Ohm's law V=IR, the slope of the trend line corresponds to the resistance.

[0024] In terms of communication and data management, the communication module is configured to perform integrity verification, compression encoding, and local caching of monitoring data on the vehicle side, and send it to the diagnostic processing module using a batch upload and breakpoint resume mechanism. Consistent with the minute-level segmented upload approach, continuous sampling data can be cached on the vehicle side into fixed time slices (e.g., 1 minute) before transmission. This reduces the probability of packet loss during satellite network fluctuations through local caching and batch upload, and fills in missing segments after connection recovery by relying on breakpoint resume, ensuring the continuity and availability of long-term real road data. Meanwhile, the interface adapter, as a convergence and protocol adaptation node for multi-source data, can uniformly parse, align, and package different messages from the BMS, inverter, low-voltage power supply device, and temperature sensor. Under the unified time base provided by the time synchronization unit, it adds timestamps and / or mileage information to each record, enabling subsequent extraction of voltage difference-current points by mileage segment or time period, and comparison of data from different vehicles / different stages, to directly use the same dimension index for positioning and association, thus meeting the overall requirements of synchronous acquisition, unified time base, and traceable upload in the embodiment.

[0025] The communication module 102 is configured to time and / or mileage mark the monitoring data and send it to the diagnostic processing module via a satellite network.

[0026] The data storage module 103 is configured to store monitoring data and diagnostic results generated by the diagnostic processing module.

[0027] The diagnostic processing module 104 is built in the cloud and includes: a segmentation construction unit, a voltage difference calculation unit, an interconnect resistance estimation unit, a multi-step filtering unit, a temperature compensation unit, and a fault determination unit. The segmentation construction unit is configured to segment the monitoring data to obtain data segments; the voltage difference calculation unit is configured to calculate the voltage difference between the inverter input voltage and the battery pack voltage within each data segment, and combine the voltage difference with the corresponding battery pack current to form a current-voltage difference dataset; the interconnect resistance estimation unit is configured to perform linear fitting on the current-voltage difference dataset to obtain an initial trend line, and use the slope of the initial trend line as the initial estimate of the interconnect resistance; the multi-step filtering unit is configured to... The statistical outlier criterion removes outlier data points that deviate from the initial trend line, and an adaptive moving average is performed on the current-voltage difference dataset to reduce noise. The multi-step filtered interconnect resistance is then fitted again. The temperature compensation unit is configured to obtain the average operating temperature corresponding to the data segment and convert the multi-step filtered interconnect resistance to the reference temperature based on the temperature correction coefficient of the conductor material to obtain the temperature-compensated interconnect resistance. The fault determination unit is configured to compare the temperature-compensated interconnect resistance with the benchmark model, which is constructed from the statistical distribution of the temperature-compensated interconnect resistance of normal samples, to output the health status or fault alarm of the battery pack, and to distinguish external wiring harness abnormalities, busbar abnormalities, or individual cell abnormalities when the preset diagnostic rules are met.

[0028] In cloud-based implementation, the diagnostic processing module 104 can coordinate with the data access service of the central server: after the monitoring records generated on the vehicle side are transmitted to the server, the access service first performs sequential splicing and consistency verification according to vehicle identifier, timestamp, and mileage index, and then writes them into the data storage module to form a traceable time-series data stream. To reduce data gaps caused by link fluctuations, the monitoring data can be organized by continuous acquisition on the terminal side and time-slice reception on the server side. That is, key diagnostic quantities are recorded at a fixed sampling frequency on the vehicle monitoring terminal and sent to the central server at fixed time intervals. This segmented transmission reduces the impact of connection interruptions on continuity, ensuring that the cloud can stably obtain sufficiently fine-grained operational data for diagnostic analysis. In deployments targeting fleets or long-term operations, the cloud can uniformly archive monitoring data from different vehicles and different acquisition cycles, and form a historical sample set for benchmark statistics and alarm threshold construction (e.g., forming a stable distribution characterization of the connection system status through long-term monitoring data), providing a data foundation for subsequent health assessments.

[0029] The cloud-side calculation of interconnection status can adopt an I-V profile-based processing paradigm: the difference between the inverter input voltage and the total battery pack voltage is used as an observation reflecting the voltage drop of the interconnection system. The relationship between this difference and the battery pack current can provide key clues for changes in the interconnection system resistance. In specific implementation, the data can be organized into several segments according to preset mileage intervals or driving segments, and the voltage difference-current scatter points in each segment can be plotted as an I-V relationship. The slope of the trend line is used to obtain the initial resistance result. Since there is noise and duplicate points in the real vehicle data, the cloud can first generate a baseline trend line from the original scatter points, and then remove outliers that deviate from the trend line according to the standard deviation of the residuals (for example, using sigma exceeding a preset multiple as a criterion). Subsequently, the data is smoothed by moving average, and the window size is adaptively adjusted according to the degree of current fluctuation. Finally, a refined trend line is generated on the smoothed data to obtain a more stable resistance estimate Rm, thereby improving the consistency and noise resistance of the resistance calculation. The key purpose of the above processing link is to converge the resistance estimates obtained from different mileage segments from "dispersed and abrupt" to a comparable sequence, so that the cloud can identify abnormal offsets and gradual degradation based on historical distribution.

[0030] The fault determination unit includes a fault location subunit, which is configured to construct multiple sets of current-voltage difference profiles based on different voltage information in the auxiliary diagnostic parameters and obtain resistance feature sets for each. The different voltage information includes at least: the total voltage across the battery pack, the battery pack's output voltage, the battery pack voltage information provided by the battery management system, the sum of battery voltages obtained by summing the individual battery voltage sets, and the extreme difference in battery voltage calculated from the individual battery voltage sets. The resistance feature sets include at least: a first resistance feature representing the connection path between the busbar and the battery, a second resistance feature representing the connection path between the busbar and the battery and superimposed with the path between the harness and the connector, a third resistance feature representing the equivalent path between batteries, and a fourth resistance feature representing changes in battery consistency. The fault location subunit determines the harness and connector resistance components based on the difference between the first and second resistance features, determines the busbar resistance components based on the difference between the first and third resistance features, and determines the individual battery abnormality indication based on the fourth resistance feature. The fault determination unit outputs component-level diagnostic conclusions based on the combination rules of the resistance components and the temperature-compensated interconnect resistance to distinguish between external harness abnormalities, busbar abnormalities, and individual battery abnormalities.

[0031] The fault location subunit can utilize multi-source voltage information available to the vehicle to form multiple sets of current-voltage difference profiles, thereby separating the equivalent resistance characteristics reflected at different measurement points / different voltage information in the same segment of operating data. Specifically, in addition to using the battery pack current as the horizontal axis, the vertical axis can select voltage information or voltage differences related to the interconnect structure, including the total battery pack voltage, the external output terminal voltage, the battery pack voltage information given by the battery management system, the sum of battery voltages obtained by summing the individual cell voltage sets, and the extreme differences of individual cell voltages, etc., to obtain the corresponding profiles and their trend line slopes respectively; among them, the extreme differences of individual cell voltages can be used as a direct observation to characterize the consistency changes of individual cells, and are used to form a complementary relationship with interconnect resistance characteristics.

[0032] Specifically, when resistance estimation is needed for alarm determination, temperature sensitivity can be normalized in the cloud to reduce the interference of temperature fluctuations on diagnostic results. A temperature correction formula can be used: Make corrections, among which Reference temperature Temperature correction resistor (e.g., at 25℃) This is the processed resistance estimate. This is the temperature coefficient of the conductor (e.g., 0.00403 / ℃ for copper conductors). The average operating temperature is used; by converting the resistance to a unified reference temperature, resistance fluctuations caused by environmental and thermal conditions can be more clearly distinguished from resistance changes caused by aging or mechanical abnormalities. At the fault location level, the cloud can further combine different voltage measurement points to form multiple sets of resistance characteristics, and interpret them as characteristic components related to different connection paths or consistency, such as: resistance components related to busbar and individual cell connections, resistance components after superimposing external wiring harnesses, resistance components related to equivalent paths between individual cells, and consistency components characterizing voltage dispersion between individual cells; this component concept can correspond to the definition of resistance types (such as resistance components used to characterize different meanings of busbar-individual cell, busbar-individual cell-wiring harness, paths between individual cells, and voltage variations between individual cells), thereby outputting more targeted component-level conclusions in the cloud based on the differential relationship of different resistance components.

[0033] In one embodiment, the segmentation unit is configured to divide the monitoring data into multiple data segments based on the cumulative mileage, cumulative operating time, or vehicle operating mode, and to perform data quality control on each data segment.

[0034] The segmentation construction unit, upon receiving time-series data arranged by timestamp and mileage index in the cloud, can prioritize organizing data segments using continuous driving intervals at fixed mileage intervals as the basic segmentation granularity. This ensures that each data segment corresponds to a relatively continuous operating process suitable for constructing current-voltage difference relationships. This approach aligns with the concept of building current-voltage difference maps based on real-vehicle monitoring data, which plots the relationship between inverter input and battery pack voltage difference relative to battery pack current within fixed mileage intervals for comparative analysis across different intervals. When the monitoring data includes vehicle operating modes or charging / discharging direction indicators, the segmentation construction unit can also combine these state variables before and after segmentation to determine whether the data segment is in a stable operating segment related to traction drive. This avoids including abrupt changes in data during charging, non-traction, or mode switching into the same segment, thus ensuring consistency in subsequent profile construction. Furthermore, segments within a short distance interval can be considered representative intervals to generate a representative point or a set of representative points, enabling the formation of comparable profile sequences over long distances (e.g., demonstrating the process of calculating representative points within a certain distance range). Data quality control includes: determining whether the battery pack current in the data segment covers the preset current change range to meet the excitation conditions for resistance identification; determining whether the voltage difference has a monotonically approximately linear relationship with the battery pack current; removing sample points containing sensor saturation, communication packet loss, or state switching transients; and removing charging conditions and non-traction conditions based on the vehicle operating mode.

[0035] Here, in terms of data quality control, the segmented construction unit can perform identifiability-oriented filtering on each data segment: on the one hand, it checks whether the battery pack current within the segment has sufficient variation range and fluctuation coverage to ensure that the voltage difference-current scatter plot has expansion on the horizontal axis, thus reflecting the law of voltage drop variation with current in the connection system; on the other hand, it checks whether the relationship between voltage difference and current shows an approximately linear trend to eliminate nonlinear dispersion caused by extreme operating conditions, transient impacts, or measurement anomalies. In real vehicle monitoring data processing, noise and duplicate data points can cause deviations in trend judgment based on scatter plots, which is one of the reasons why segment quality needs to be controlled at the segmentation stage. In addition, if there are jump points caused by sensor saturation, communication loss, or transient segments caused by control mode switching within a segment, the segmented construction unit can remove sample points or truncate the segment accordingly, so that the data entering the subsequent current-voltage difference profile construction comes as much as possible from a continuous and physically consistent operating range, thereby reducing the dispersion introduced by non-steady-state factors.

[0036] After passing data quality control, the voltage difference calculation unit aligns the inverter input voltage and battery pack voltage on the same time base, forming a time-by-time original voltage difference sequence, and pairs it with the battery pack current at the same time. The voltage difference between the inverter input and the total battery pack voltage provides a direct observation entry point for the equivalent voltage drop of the connected system and is used to construct a current-voltage difference graph to extract connected system state information. To reduce the redundancy caused by high-frequency noise and repeated sampling in the scatter pairs, the voltage difference calculation unit can aggregate the scatter pairs along the current dimension without changing the physical meaning of the data, generating representative profile points: for example, binning by current intervals or summarizing by sliding windows on the current axis. Within each interval / window, a representative value (such as mean, weighted mean, or median) is calculated for the original voltage difference sequence, and the sample size, dispersion, or time coverage corresponding to the representative value is recorded to characterize the stability of the representative point. The resulting profile point set is more uniform and less redundant on the current axis, which can serve as the input basis for establishing trend relationships and calculating slopes. Furthermore, the way this profile point set is constructed is consistent with the overall processing path of plotting voltage differences against current within a fixed mileage interval and conducting analysis accordingly.

[0037] In one embodiment, the multi-step filtering unit includes an outlier removal subunit, a redundancy suppression subunit, a window smoothing subunit, and a refitting subunit, wherein: the outlier removal subunit is configured to calculate the residual of each profile point in the profile point set relative to the initial trend line based on the initial trend line, and determine the outlier criterion based on the statistical dispersion of the residuals, so as to remove abnormal data points that deviate from the initial trend line.

[0038] As an explanation, to ensure the stability of the interconnection resistance estimates obtained from the current-voltage difference profile under real road conditions, the multi-step filtering unit can first use the initial trend line as a baseline to statistically determine the deviation of the profile point set. Specifically, the observed voltage difference at each profile point can be subtracted from the predicted voltage difference corresponding to the initial trend line under the same current to obtain the residual; then, the dispersion of this residual set (e.g., standard deviation) can be used as a sigma index to set outlier criteria and remove points with excessive deviation. These outliers usually originate from noise disturbances or transient anomalies under extreme operating conditions. Removing them can significantly reduce the bias of a small number of outliers on the slope of the trend line, thereby improving the reliability and consistency of subsequent resistance calculations.

[0039] The redundancy suppression subunit is configured to aggregate profile points with repeated or nearly repeated current values ​​to reduce fitting bias caused by repeated sampling. The window smoothing subunit is configured to perform moving average, moving median, or exponential smoothing on the profile point set after outlier removal and redundancy suppression, and adaptively select a smoothing window between a first window width and a second window width based on the degree of fluctuation of the battery pack current in different current ranges, so that a larger window is used in the range with large current fluctuations and a smaller window is used in the range with small current fluctuations.

[0040] As a supplement, after outlier removal, the redundancy suppression subunit and window smoothing subunit can further address profile redundancy caused by repeated sampling and random noise. In actual vehicle monitoring data, the battery pack current repeatedly exhibits the same or similar values ​​in certain intervals, leading to dense stacking of profile points near the same current, making the fitting results more susceptible to being dominated by local point clusters. Simultaneously, voltage difference measurements may also be superimposed with high-frequency disturbances, causing unnecessary dispersion of scatter points near the trend line. To address this, profile points corresponding to repeated or nearly repeated currents can be aggregated on the current axis, allowing the same current neighborhood to participate in subsequent calculations with representative values; further, smoothing methods such as moving averages can be used to reduce noise in the profile. The implementation of moving averages can be understood as: setting a sliding window on the current axis, averaging the voltage differences of the profile points covered by the window, and redrawing the smoothed profile curve; its effect is to simultaneously reduce noise and redundancy caused by repeated points, making the profile closer to its true linear relationship. Furthermore, the window smoothing subunit can employ an adaptive window strategy to balance robustness in strong fluctuation ranges with precision in weak fluctuation ranges. In regions where battery pack current fluctuations are more pronounced, using a small window will result in insufficient smoothing, and the profile curve will still retain significant random jitter. Conversely, in regions with more stable current, an excessively large window may lead to over-smoothing, resulting in a loss of profile detail. Therefore, the window size can be selected based on the degree of variation in battery pack current across different regions: a larger moving average window should be used in high-fluctuation regions, while a smaller window should be used in other regions to maintain a balance between overall accuracy and local stability.

[0041] The refit subunit is configured to perform a second linear fit on the smoothed profile point set, using the slope of the second linear fit as the multi-step filter interconnect resistance, and writing the difference between the multi-step filter interconnect resistance and the initial estimate as a noise level index into the data storage module.

[0042] After completing the outlier removal, redundancy suppression, and window smoothing processes described above, the refitted sub-unit can regenerate a refined trend line based on the processed profile point set. The slope of this refined trend line is then used to obtain the multi-step filtered interconnect resistance. This process can be understood as performing a linear trend extraction again using a denoised and deredundant profile, so that the obtained resistance estimate converges from the scattered state of the original points into a more stable sequence. In actual long-term data processing, the resistance estimate obtained after multi-step filtering shows a stronger convergence trend compared to the initial estimate, and can continuously generate a large number of resistance estimates across multiple driving intervals to characterize the trajectory of interconnect state changes. Simultaneously, the difference between the multi-step filtered interconnect resistance and the initial estimate can be written into the data storage module as a quantitative representation of noise level or data quality, enabling the identification of intervals with large profile dispersion in subsequent analysis and allowing for backtracking checks or weight reduction.

[0043] In one embodiment, the temperature compensation unit is configured to obtain the temperature sequence corresponding to the data segment from the battery module temperature, and obtain the average operating temperature through time averaging, steady-state interval filtering, or abnormal temperature elimination; the temperature compensation unit is also configured to select a conductor material temperature correction coefficient according to the material type of the conductive components in the battery pack, and convert the multi-step filtered interconnect resistance to a reference temperature based on the linear approximation relationship of conductor resistance changing with temperature, so as to eliminate the influence of temperature fluctuation on resistance judgment; wherein, the material type includes at least busbar material, connector material and wire harness conductor material, the temperature compensation unit calculates the resistance compensation amount for different material types and performs weighted synthesis to obtain a temperature-compensated interconnect resistance that is closer to the actual structure; and when the average operating temperature is unavailable, the temperature compensation unit generates an alternative temperature input based on the ambient temperature, vehicle thermal management status, or historical temperature model.

[0044] In one specific implementation, after receiving the battery module temperature corresponding to the data segment, the temperature compensation unit can first form a time-stamp-aligned temperature sequence and then perform consistency matching between this temperature sequence and the boundary of the data segment to ensure that the average operating temperature reflects the true thermal state within the same data segment. The average operating temperature can be obtained using a combination of time averaging, steady-state filtering, and anomaly removal strategies: for example, missing values ​​and jump points can be removed from the temperature sequence first, and then the average can be taken within a range where temperature changes are gradual and without significant abrupt changes, thereby reducing temperature distortion caused by operating condition switching or communication discontinuity; when multiple temperature sampling points exist, consistency checks can be performed on the multiple temperatures before summarizing them to avoid deviations caused by individual measurement point failures. The purpose of this processing method is to use the average operating temperature as input for subsequent temperature correction, forming a temperature baseline corresponding to the trend of vehicle travel distance changes. This facilitates the comparability and normalization of interconnection resistance estimates from different data segments under the same temperature baseline (e.g., visualizing the change of average operating temperature with travel distance, and then performing temperature compensation on the resistance values ​​obtained from multi-step filtering based on this).

[0045] After obtaining the average operating temperature, the temperature compensation unit can normalize the temperature of the multi-step filtering interconnect resistors based on the temperature sensitivity of conductor resistance. This makes resistance changes more reflective of aging or mechanical abnormalities, rather than temperature fluctuations themselves. The temperature compensation unit can use the temperature correction formula mentioned above. The temperature correction coefficient of the conductor material is used to quantify the sensitivity of conductor resistance to temperature changes. The temperature compensation unit can select the corresponding temperature correction coefficient according to the material type of different conductive components in the battery pack (for example, establishing coefficient mapping tables for busbar materials, connector materials, and wire harness conductor materials), and calculate the corresponding compensation resistance for each material before weighted synthesis to obtain a temperature-compensated interconnect resistor that more closely resembles the actual interconnect structure. By unifying the resistance to a standard reference temperature, the consistency of diagnostic results can be maintained under different operating temperature conditions, and resistance fluctuations caused by temperature are relatively isolated from resistance changes caused by aging or mechanical faults, thereby improving the reliability of interconnect fault identification.

[0046] When the average operating temperature is unavailable, the temperature compensation unit can generate alternative temperature inputs to ensure the temperature compensation link remains uninterrupted. For example, it can prioritize using ambient temperature as an approximate input, or, when vehicle thermal management status information is available, correct the ambient temperature based on the thermal management status to obtain an estimated temperature. Alternatively, it can establish a distance / time-temperature mapping based on a historical temperature model, and regress the predicted temperature value from historical data of the same vehicle model or type to use as an alternative. The temperature correction formula mentioned above is used as a reference. The core constraint of using an alternative temperature input is to maintain consistency in the reference temperature conversion, ensuring that subsequent results are accurate. It remains under a comparable and unified benchmark; at the implementation level, it can simultaneously record the temperature source identifier (measured / estimated) and the estimated confidence level, so as to interpret and trace the credibility of the temperature compensation results during subsequent data playback or diagnostic verification.

[0047] In one embodiment, the baseline model includes at least one of a threshold model based on statistical distribution and an aging model based on trend. The threshold model obtains a center position parameter and a dispersion parameter by fitting the distribution of the temperature compensation interconnect resistance of normal samples, and generates an adaptive alarm threshold based on the dispersion parameter. The aging model obtains a trend term representing progressive aging by smoothing and regressing the change sequence of the temperature compensation interconnect resistance with operating mileage or operating time. The fault determination unit is configured to calculate the deviation of the temperature compensation interconnect resistance under test relative to the center position parameter in each data segment, and determine whether the deviation is caused by short-term noise, environmental changes or the battery pack interconnect system in combination with the trend term.

[0048] To ensure the interpretability and transferability of the threshold model, data segments from vehicles in a healthy state can be used as a normal sample set. The temperature-compensated interconnect resistance corresponding to each data segment is then used as the statistical object, forming a resistance sample sequence that accumulates over mileage or time. When characterizing the probability distribution of this sample sequence, the samples are first normalized and their probability density or histogram distribution is plotted to observe whether they exhibit statistical characteristics of clustering around a certain central location. A suitable distribution family can then be selected for fitting. Actual monitoring and analysis show that the temperature-compensated interconnect resistance in healthy vehicles typically exhibits a certain fluctuation bandwidth. When the value exceeds this fluctuation bandwidth, it is more likely to correspond to degradation or anomalies in the interconnect system. Therefore, the threshold model can be set as an adaptive interval jointly defined by the "central location parameter + dispersion parameter" to achieve the effect of self-updating the threshold based on data for different vehicles and under different environments. Furthermore, the normalized distribution curves of multi-vehicle comparisons show a clear central trend and are generally approximately normally distributed, providing a data basis for constructing alarm thresholds using the central location parameter and dispersion parameter.

[0049] To avoid false alarms or missed alarms caused by relying solely on instantaneous thresholds, aging models can introduce long-term trend terms to model the temperature-compensated interconnect resistance along the sequence of operating mileage or operating time. Firstly, the temperature-compensated interconnect resistance can be concatenated into a time series or mileage series according to data segments, and smoothing can be used to suppress short-cycle operating condition disturbances, making the series present a more stable evolution trajectory on a macroscopic scale. Secondly, after obtaining the smoothed resistance series, regression can be used to extract trend terms representing gradual changes, thereby characterizing the slow degradation process of the interconnect system with use. Based on this, the calculation of deviation is not limited to "whether the sample point exceeds the threshold," but can be further combined with the trend term to determine whether the deviation is due to short-term random disturbances, overall drift caused by environmental changes, or aging accumulation corresponding to a continuous increase (or continuous change) in the trend term. For example, when the temperature-compensated interconnect resistance fluctuates but still generally follows the central trend distribution, it is likely to be judged as short-term noise or operating condition changes. Conversely, when the temperature-compensated interconnect resistance exhibits a continuous deviation exceeding the healthy fluctuation bandwidth over a longer mileage range, it is more consistent with the characteristics of interconnect system degradation or anomalies.

[0050] Furthermore, the threshold model and aging model can employ a parallel fusion decision strategy: the threshold model provides a quick conclusion on whether the current segment is abnormal; the aging model provides explanations for whether the abnormality is persistent and cumulative. Deviation can be expressed as the difference between the measured temperature compensation interconnect resistance and the center position parameter, and then normalized by combining the dispersion parameter to make the deviations of different vehicles comparable. When the deviation increases instantaneously but the trend term remains stable, it can be attributed to short-term noise or environmental changes. When both the deviation and the trend term show persistent changes, a higher confidence level judgment of interconnect system degradation can be output. In addition, the baseline model can continuously update the center position parameter and dispersion parameter as normal samples accumulate, thereby maintaining the adaptive capability of the alarm threshold under long-term vehicle operation, seasonal changes, and changes in operating conditions, and improving the stability and consistency of fault alarms.

[0051] In one embodiment, the diagnostic processing module further includes an efficiency evaluation unit, which is configured to calculate power and energy consumption within a preset driving range and associate it with a temperature compensation interconnection resistor. The efficiency evaluation unit includes: a low-voltage power calculation subunit, a battery output power calculation subunit, an inverter input power calculation subunit, a motor output power calculation subunit, and an efficiency and energy consumption calculation subunit.

[0052] Specifically, when the efficiency evaluation unit performs a unified calculation of the power links within a preset driving segment in the cloud, it can treat low-voltage power as parallel load power outside the drive link. This allows it to be separated when calculating traction-related power, thus avoiding the inclusion of power consumed by the vehicle's low-voltage system (such as onboard electronics and thermal management loads) in the traction energy calculation. Accordingly, low-voltage power can be obtained by multiplying the output voltage and output current of the low-voltage power supply device using the following formula: The battery output power can be obtained by subtracting the low-voltage power from the instantaneous power at the battery terminal using the following formula: The motor output power can be obtained by multiplying the angular velocity and torque using the following formula: .

[0053] The low-voltage power calculation subunit multiplies the output voltage and output current of the low-voltage power supply device to obtain the low-voltage output power; the battery output power calculation subunit multiplies the battery pack voltage and current to obtain the battery terminal power, and subtracts the low-voltage output power to obtain the battery output power; the inverter input power calculation subunit multiplies the inverter input voltage and current to obtain the inverter terminal power, and subtracts the low-voltage output power to obtain the inverter input power; the motor output power calculation subunit converts the drive motor speed into angular velocity and multiplies it with the drive motor torque to obtain the motor output power; the efficiency and energy consumption calculation subunit accumulates the motor output power and battery output power within a preset driving range and compares them to obtain the battery efficiency, accumulates the motor output power and inverter input power within a preset driving range and compares them to obtain the motor efficiency, and performs numerical integration on the time series of battery output power and performs energy unit conversion to obtain the energy consumption index.

[0054] Furthermore, to obtain battery efficiency and reduce the impact of instantaneous fluctuations on the results, the efficiency and energy consumption calculation subunit can accumulate the motor output power and battery output power separately within a preset driving range, and then calculate the battery efficiency by comparing the "accumulated output / accumulated input", which can be specifically expressed as: ,in and In engineering implementation, this can be understood as the summation of the corresponding power sequence within a preset driving segment. However, due to the degradation of drive motor performance, only the above formula may be used. If efficiency comparison discrepancies occur, the efficiency evaluation unit can further incorporate inverter input power and motor efficiency for cross-verification. The inverter input power can be obtained by multiplying the inverter input voltage and inverter input current and subtracting the low-voltage power using the following formula: ; Motor efficiency can be obtained by comparing the cumulative value of motor output power to inverter input power over a preset driving range using the following formula: Based on this, the efficiency evaluation unit can update the motor efficiency using only data from the high-efficiency region in the motor torque-speed efficiency map. This enhances the representativeness of motor efficiency to changes in drive link status and verifies the comparability of motor efficiency with battery efficiency, thereby distinguishing as much as possible the efficiency changes caused by motor-side degradation from those caused by changes in the battery side (including the interconnect system).

[0055] The efficiency evaluation unit is also configured to update the motor efficiency only when the drive motor is in a preset high-efficiency operating range, and to use the motor efficiency to verify the comparability of the battery efficiency, so as to distinguish the efficiency change caused by motor aging from the efficiency change caused by battery pack aging.

[0056] In addition, the efficiency and energy consumption calculation subunit can also calculate energy consumption indicators based on the battery output power sequence, which can be used to characterize the economic changes of the whole vehicle in a preset driving range from the perspective of "energy consumption". Energy consumption can be obtained by performing numerical integration on the time series of battery output power according to equation (8), and the power integration result is converted into electrical energy consumption through unit conversion: ; Where T is the time to complete the preset travel segment, and when When recording at a fixed sampling frequency, numerical integration can be used to approximate the discrete points by accumulating them. By correlating energy consumption indicators with temperature-compensated interconnect resistance, the efficiency assessment unit can be used to reveal the system-level indirect effects that may result from an increase in interconnect resistance: increased interconnect resistance may cause higher power transmission losses, reduce the effective power transmitted to the drive link, and disturb the vehicle's power and thermal balance, thus manifesting as decreased efficiency and increased energy consumption during operation. This correlation does not require the temperature-compensated interconnect resistance to be directly included in the power calculation formula, but rather achieves performance-side corroboration and quantitative evaluation through synchronous statistics of efficiency / energy consumption and temperature-compensated interconnect resistance within the same driving segment.

[0057] This invention utilizes monitoring data obtainable during vehicle operation to create a continuous assessment capability of interconnected status in the cloud. Through satellite industrial internet, it achieves wider coverage and more stable remote transmission, enabling vehicles to maintain the necessary data links for diagnosis even in areas with weak or unavailable terrestrial networks, thereby improving the continuity of monitoring and the timeliness of alarms. By rationally organizing and controlling the operational data according to operating conditions, and employing a more robust processing mechanism for anomalies and noise, the impact of random road disturbances on diagnostic conclusions can be reduced, enhancing the repeatability and comparability of results across different vehicles, environments, and operating conditions.

[0058] Furthermore, this invention introduces a normalization process for the influence of temperature, ensuring that interconnection status indicators have consistent discriminative significance under different thermal states, which is beneficial for distinguishing between gradual aging and sudden anomalies. Simultaneously, by comprehensively utilizing multi-source voltage information, the ability to differentiate fault types can be improved without adding complex hardware, providing more targeted prompts for maintenance decisions. Combined with an alarm and maintenance closed-loop mechanism, actual handling information can inversely promote model and rule updates, thereby improving diagnostic accuracy during long-term operation and supporting collaborative assessment of preventative maintenance, remote operation and maintenance, and efficiency degradation.

[0059] It should be noted that although several modules or units of the system have been mentioned in the detailed description above, this division is not mandatory. In fact, according to exemplary embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0060] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0061] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A power battery pack monitoring system based on satellite industrial internet, characterized in that, The monitoring system includes a data acquisition terminal, a communication module, a data storage module, and a diagnostic processing module. The data acquisition terminal is configured to acquire monitoring data from the battery management system, inverter, low-voltage power supply device and temperature sensor. The monitoring data includes at least the battery pack voltage, battery pack current, inverter input voltage, inverter input current and battery module temperature. The communication module is configured to time- and / or mileage-mark the monitoring data and transmit it to the diagnostic processing module via a satellite network; The data storage module is configured to store the monitoring data and the diagnostic results generated by the diagnostic processing module. The diagnostic processing module is built in the cloud and includes: a segmentation unit, a voltage difference calculation unit, an interconnect resistance estimation unit, a multi-step filtering unit, a temperature compensation unit, and a fault determination unit. The segmentation unit is configured to segment the monitoring data to obtain data segments. The voltage difference calculation unit is configured to calculate the voltage difference between the inverter input voltage and the battery pack voltage within each data segment, and combine the voltage difference with the corresponding battery pack current to form a current-voltage difference dataset. The interconnect resistance estimation unit is configured to perform linear fitting on the current-voltage difference dataset to obtain an initial trend line, and use the slope of the initial trend line as an initial estimate of the interconnect resistance. The multi-step filtering unit is configured to... To remove outlier data points deviating from the initial trend line based on statistical outlier criteria, and to perform adaptive moving average on the current-voltage difference dataset for noise reduction, a multi-step filtered interconnect resistance is obtained by fitting again. The temperature compensation unit is configured to acquire the average operating temperature corresponding to the data segment, and convert the multi-step filtered interconnect resistance to a reference temperature based on the temperature correction coefficient of the conductor material to obtain the temperature-compensated interconnect resistance. The fault determination unit is configured to compare the temperature-compensated interconnect resistance with a benchmark model, which is constructed from the statistical distribution of the temperature-compensated interconnect resistance of normal samples, to output the health status or fault alarm of the battery pack, and to distinguish external wiring harness abnormalities, busbar abnormalities, or individual cell abnormalities when preset diagnostic rules are met.

2. The power battery pack monitoring system based on satellite industrial internet according to claim 1, characterized in that, The data acquisition terminal includes an on-board data recorder, an interface adapter, and a time synchronization unit. The interface adapter establishes communication connections with the battery management system, the inverter, the low-voltage power supply device, and the temperature sensor, respectively, for synchronously acquiring the monitoring data according to a preset sampling period, and adding timestamps and / or mileage information to the monitoring data under the unified time reference of the time synchronization unit. The monitoring data also includes at least one auxiliary diagnostic parameter for decomposing faulty components. The auxiliary diagnostic parameter includes at least one of the following: total battery pack voltage, battery pack output voltage, sum of battery voltages calculated from the set of individual battery voltages, extreme difference of individual battery voltages, output voltage of the low-voltage power supply device, output current of the low-voltage power supply device, drive motor torque, drive motor speed, vehicle operating mode, ambient temperature, and charging / discharging direction indicator. Furthermore, the communication module is configured to perform integrity verification, compression encoding, and local caching on the monitoring data on the on-board side, and to send the monitoring data to the diagnostic processing module using a batch upload and breakpoint resume mechanism.

3. The power battery pack monitoring system based on satellite industrial internet according to claim 2, characterized in that, The fault determination unit includes a fault location subunit, which is configured to construct multiple sets of current-voltage difference profiles based on different voltage information in the auxiliary diagnostic parameters and obtain resistance feature sets respectively. The different voltage information includes at least: the total voltage across the battery pack, the voltage at the battery pack's external output terminals, the battery pack voltage information provided by the battery management system, the sum of battery voltages obtained by summing the individual battery voltage sets, and the extreme difference in battery voltage calculated from the individual battery voltage sets; the resistance feature set includes at least: a first resistance feature characterizing the connection path between the busbar and the battery, a second resistance feature characterizing the connection path between the busbar and the battery and superimposed on the path between the harness and the connector, a third resistance feature characterizing the equivalent path between batteries, and a fourth resistance feature characterizing changes in battery consistency; wherein, the fault location subunit determines the harness and connector resistance component based on the difference between the first resistance feature and the second resistance feature, determines the busbar resistance component based on the difference between the first resistance feature and the third resistance feature, and determines the individual battery abnormality indication based on the fourth resistance feature; the fault determination unit outputs a component-level diagnostic conclusion based on the combination rules of the resistance components and the temperature compensation interconnection resistance, so as to distinguish between external harness abnormalities, busbar abnormalities, and individual battery abnormalities.

4. The power battery pack monitoring system based on satellite industrial internet according to claim 1, characterized in that, The segmentation unit is configured to divide the monitoring data into multiple data segments based on the cumulative mileage, cumulative operating time, or vehicle operating mode, and to perform data quality control on each data segment, the data quality control including: The system determines whether the battery pack current within the data segment covers a preset current variation range to meet the excitation conditions for resistance identification, whether the voltage difference has a monotonically approximately linear relationship with the battery pack current, eliminates sample points containing sensor saturation, communication packet loss, or state switching transients, and eliminates charging and non-traction conditions based on the vehicle operation mode. After passing the data quality control, the voltage difference calculation unit calculates the difference between the inverter input voltage and the battery pack voltage to obtain the original voltage difference sequence, and aligns the original voltage difference sequence with the battery pack current in time to form a scatter pair; The voltage difference calculation unit is also configured to perform current binning or sliding window aggregation on the scatter pairs, and calculate the mean, weighted mean or median of the original voltage difference sequence in each current interval or window as a representative voltage difference to form a profile point set that is deredundant and easy to fit.

5. The power battery pack monitoring system based on satellite industrial internet according to claim 4, characterized in that, The multi-step filtering unit includes an outlier removal subunit, a redundancy suppression subunit, a window smoothing subunit, and a refitting subunit, wherein: The outlier removal subunit is configured to use the initial trend line as a reference to calculate the residual of each profile point in the profile point set relative to the initial trend line, and determine the outlier criterion based on the statistical dispersion of the residuals, so as to remove abnormal data points that deviate from the initial trend line. The redundancy suppression subunit is configured to perform aggregation on profile points with repeated or nearly repeated current values ​​to reduce fitting bias caused by repeated sampling. The window smoothing subunit is configured to perform moving average, moving median, or exponential smoothing on the profile point set after outlier removal and redundancy suppression, and adaptively select a smoothing window between the first window width and the second window width according to the degree of fluctuation of the battery pack current in different current ranges, so that a larger window is used in the range with large current fluctuations and a smaller window is used in the range with small current fluctuations. The refitting subunit is configured to perform a second linear fit on the smoothed profile point set, using the slope of the second linear fit as the multi-step filter interconnect resistance, and writing the difference between the multi-step filter interconnect resistance and the initial estimate as a noise level index into the data storage module.

6. The power battery pack monitoring system based on satellite industrial internet according to claim 1, characterized in that, The temperature compensation unit is configured to obtain the temperature sequence corresponding to the data segment from the battery module temperature, and obtain the average operating temperature by time averaging, steady-state interval filtering or abnormal temperature removal. The temperature compensation unit is further configured to select a conductor material temperature correction coefficient based on the material type of the conductive components in the battery pack, and convert the multi-step filtering interconnect resistance to the reference temperature based on the linear approximation relationship of conductor resistance changing with temperature, so as to eliminate the influence of temperature fluctuation on resistance judgment; wherein, the material type includes at least busbar material, connector material and wire harness conductor material, and the temperature compensation unit calculates the resistance compensation amount for different material types and performs weighted synthesis to obtain the temperature-compensated interconnect resistance that is closer to the actual structure; and when the average operating temperature is unavailable, the temperature compensation unit generates an alternative temperature input based on the ambient temperature, vehicle thermal management status or historical temperature model.

7. The power battery pack monitoring system based on satellite industrial internet according to claim 1, characterized in that, The baseline model includes at least one of a threshold model based on statistical distribution and an aging model based on trend. The threshold model obtains a center position parameter and a dispersion parameter by fitting the distribution of the temperature compensation interconnect resistance of normal samples, and generates an adaptive alarm threshold based on the dispersion parameter. The aging model obtains a trend term representing gradual aging by smoothing and regressing the change sequence of the temperature compensation interconnect resistance with operating mileage or operating time. The fault determination unit is configured to calculate the deviation of the temperature compensation interconnect resistance under test relative to the center position parameter in each data segment, and determine whether the deviation is caused by short-term noise, environmental changes, or the battery pack interconnect system in combination with the trend term.

8. The power battery pack monitoring system based on satellite industrial internet according to claim 1, characterized in that, The diagnostic processing module also includes an efficiency evaluation unit, which is configured to calculate power and energy consumption within a preset driving range and associate it with the temperature compensation interconnection resistor. The efficiency evaluation unit includes: a low-voltage power calculation subunit, a battery output power calculation subunit, an inverter input power calculation subunit, a motor output power calculation subunit, and an efficiency and energy consumption calculation subunit. The low-voltage power calculation subunit is used to multiply the output voltage of the low-voltage power supply device by the output current of the low-voltage power supply device to obtain the low-voltage output power; The battery output power calculation subunit is used to multiply the battery pack voltage and the battery pack current to obtain the battery terminal power, and subtract the low-voltage output power to obtain the battery output power. The inverter input power calculation subunit is used to multiply the inverter input voltage and the inverter input current to obtain the inverter terminal power, and subtract the low-voltage output power to obtain the inverter input power. The motor output power calculation subunit is used to convert the drive motor speed into angular velocity and multiply it by the drive motor torque to obtain the motor output power. The efficiency and energy consumption calculation subunit is used to accumulate the output power of the motor and the output power of the battery within the preset driving range and take the ratio to obtain the battery efficiency, accumulate the output power of the motor and the input power of the inverter within the preset driving range and take the ratio to obtain the motor efficiency, and perform numerical integration on the time series of the battery output power and perform energy unit conversion to obtain the energy consumption index. The efficiency evaluation unit is further configured to update the motor efficiency only when the drive motor is in a preset high-efficiency operating range, and to use the motor efficiency to verify the comparability of the battery efficiency, so as to distinguish the efficiency change caused by motor aging from the efficiency change caused by battery pack aging.

Images