A lithium iron phosphate battery soc cloud collaborative estimation method for a hybrid vehicle

Abstract

The application belongs to the technical field of new energy automobile power battery state estimation, and discloses a hybrid power vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method, which comprises the following steps: a vehicle end identifies comprehensive operating conditions containing a working mode and a driving response state, carries out power consistency verification of a power side and a battery side to obtain a power coupling error, determines a correction weight of an equivalent current of the vehicle power side according to the power coupling error, and fuses the correction weight with a measured current to obtain a recursive current; a dynamic interference index is used to screen credible candidate segments, and effective capacity and current zero deviation are updated based on the credible calibration segments; a cloud end obtains a geographic environment state to generate a correction parameter, which is returned to the vehicle end for smoothing absorption; and the vehicle end outputs a final SOC estimation value based on the fused recursive current and the corrected parameter. The application improves the accuracy, stability and environmental adaptability of the lithium iron phosphate battery SOC estimation under complex conditions.

Description

Technical Field

[0001] This invention belongs to the field of new energy vehicle power battery state estimation technology, and particularly relates to a vehicle-cloud collaborative estimation method for the state of charge (SOC) of lithium iron phosphate batteries in hybrid vehicles. Background Technology

[0002] The State of Charge (SOC) estimation results of the power battery in hybrid vehicles have a significant impact on vehicle energy management, power distribution, driving range prediction, and battery safety protection. For hybrid vehicles equipped with lithium iron phosphate batteries, there are multiple operating states during operation, including electric drive-dominated, engine-mechanical drive, electromechanical coupled drive, engine-generated power supplementation, and regenerative braking. Under different operating states, the power flow directions between the engine mechanical transmission path, the electric motor electric drive path, the engine power generation path, the regenerative braking path, and the battery differ, making the sources of power variation on the battery side more complex, thus affecting the accuracy and stability of the SOC estimation results.

[0003] Currently, the main methods for estimating the State of Charge (SOC) of power batteries include the ampere-hour integration method, the open-circuit voltage method, the equivalent circuit model method, the Kalman filter method, and data-driven estimation methods. These methods can achieve a certain level of accuracy in SOC estimation under normal operating conditions. However, for lithium iron phosphate batteries, the open-circuit voltage-SOC curve is relatively flat over a wide SOC range, and the change in terminal voltage caused by SOC changes is not significant. Under dynamic vehicle operating conditions, the battery terminal voltage is also affected by factors such as current-ohmic voltage drop, polarization, charge-discharge hysteresis, temperature changes, and transient load fluctuations, making it difficult to obtain sufficiently discriminative SOC reference information based solely on the terminal voltage or open-circuit voltage-SOC calibration relationship. Furthermore, the ampere-hour integration method is susceptible to cumulative errors due to current zero bias and effective capacity decay. During the complex operation of hybrid vehicles, power source switching, engine power generation for supplemental energy, regenerative braking, and strong transient power fluctuations further increase the complexity of battery-side power changes, leading to accumulated SOC recursive errors or decreased model parameter adaptability.

[0004] Existing SOC estimation methods typically fail to adequately consider the powertrain operating modes, driving behavior, and power response states of hybrid vehicles, making it difficult to accurately distinguish the impacts of engine mechanical paths, electric motor electric drive paths, engine power generation paths, and regenerative braking paths on battery-side power changes. Furthermore, existing methods rarely incorporate external environmental information such as vehicle location, altitude, ambient temperature, weather conditions, road gradient, and road surface conditions to adaptively correct effective capacity, charge / discharge efficiency, power coupling error thresholds, and dynamic disturbance thresholds. In addition, existing methods often lack power coupling verification mechanisms between the vehicle's powertrain and battery sides for different operating modes, as well as reliable screening mechanisms for highly disruptive segments such as rapid acceleration, low-speed start-stop, and power mode switching. This allows abnormal segments to participate in effective capacity updates and current zero-bias estimation, thus affecting the long-term accuracy and stability of SOC estimation. Summary of the Invention

[0005] This invention addresses the issues of insufficient accuracy, stability, and environmental adaptability in estimating the state of charge (SOC) of lithium iron phosphate batteries under complex operating conditions in hybrid vehicles, and proposes a vehicle-cloud collaborative estimation method for the SOC of lithium iron phosphate batteries in hybrid vehicles.

[0006] The main technical idea of ​​this invention is as follows: The vehicle-side identifies the hybrid power system's operating mode, driving behavior, and power response status based on multi-source real-time data such as battery voltage, battery current, battery temperature, engine power, motor mechanical output power, accessory load power, regenerative braking power, and vehicle operating status, forming a comprehensive operating condition. Based on this comprehensive operating condition, the corresponding power coupling relationship between the vehicle's power side and the battery side is selected, and the corresponding battery-side power is estimated. This is then compared with the actual battery power to obtain the power coupling error. Furthermore, based on the comprehensive operating condition and the power coupling error, the correction weight for the vehicle's power-side equivalent current participating in the SOC recursion is determined, forming a fused SOC recursive current. This suppresses battery current integral drift under stable and reliable operating conditions and reduces or disables the correction effect of the vehicle's power-side equivalent current under strong disturbance conditions such as rapid acceleration, low-speed start-stop, regenerative braking, or power mode switching. Meanwhile, the vehicle-side screens reliable calibration segments based on dynamic interference indicators and segment reliability, and uses these reliable calibration segments to update the battery's effective capacity and current zero bias; the cloud-side combines vehicle location, altitude, ambient temperature, weather conditions, road gradient, road surface conditions, and data up to the specified point... Historical operational data at each sampling time is used to perform geographic environment perception corrections on effective capacity, current zero bias, charging and discharging efficiency, thresholds, and weight parameters, and the cloud-based corrected parameters are then transmitted back to the vehicle. The vehicle smoothly fuses the local parameters with the cloud-based corrected parameters to obtain... , and The system then calculates the final SOC estimate based on the fused SOC recursive current output. This forms a vehicle-cloud collaborative estimation closed loop consisting of "multi-source data input, power coupling verification, reliable segment calibration, cloud environment correction, and final SOC recursion."

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A vehicle-cloud collaborative estimation method for the state of charge (SOC) of lithium iron phosphate batteries in hybrid vehicles includes the following steps:

[0009] Step 1: The vehicle acquires vehicle operation data, and the cloud acquires external environment data corresponding to the vehicle's current location;

[0010] Step 2: The vehicle-side identifies the comprehensive operating conditions, including the hybrid power system's operating mode, driving behavior, and power response status, based on the vehicle's operating data.

[0011] Step 3: The vehicle selects the corresponding power coupling relationship according to the hybrid system's operating mode, estimates the battery-side power corresponding to the vehicle's power side, and performs consistency verification with the actual power at the battery side to obtain the power coupling error.

[0012] Step 4: The vehicle side determines the correction weight of the equivalent current on the vehicle power side based on the comprehensive operating conditions and power coupling error, and merges the measured current of the battery with the equivalent current on the vehicle power side to obtain the SOC recursive current.

[0013] Step 5: The vehicle-side continuously running data is divided into multiple segments, a dynamic interference index is constructed, and a reliable candidate segment is selected based on this index. The segment that meets the reliability requirements and the SOC reference value reliability requirements is determined as a reliable calibration segment. The battery effective capacity and current zero bias are updated based on the reliable calibration segment.

[0014] Step 6: The cloud obtains the geographical environment status based on the vehicle's current location, combines the comprehensive operating conditions and historical operating data to generate cloud correction parameters and send them back to the vehicle. The vehicle smoothly absorbs the cloud correction parameters.

[0015] Step 7: Based on the SOC recursive current, the corrected effective capacity, current zero bias, and charge / discharge efficiency, the vehicle outputs the final SOC estimate.

[0016] Furthermore, in step two, the hybrid power system operating modes include at least an electric drive-dominated mode, an engine mechanical drive mode, an electromechanical coupling drive mode, an engine power generation supplementation mode, and a braking energy recovery mode; the driving behavior and power response states include at least a rapid acceleration response state, a stable cruise state, a low-speed start-stop state, a power mode switching state, and a braking energy recovery strong disturbance state; each operating mode is determined based on engine output power, motor mechanical output power, braking signal, battery current, and related thresholds; each driving behavior and power response state is determined based on vehicle acceleration, accelerator pedal opening, vehicle speed, battery current fluctuation, power change, and related thresholds; the above threshold parameters are updated based on the vehicle's powertrain structure, vehicle control strategy, battery aging status, ambient temperature, road slope, and road surface conditions.

[0017] Furthermore, in step three, segmented power coupling relationships are established according to different hybrid power system operating modes; the estimated battery-side power on the vehicle's power side is determined by the power balance relationship of the motor's mechanical output power, accessory load power, engine power generation entering the battery side, and braking energy recovery entering the battery side under different operating modes; the power coupling error is the absolute value of the difference between the actual power at the battery end and the estimated battery-side power on the vehicle's power side; when the power coupling error is less than or equal to the power coupling error threshold related to the comprehensive operating conditions, the consistency is high; when the power coupling error is greater than the power coupling error threshold, a large deviation is judged, and the participation weight of the equivalent current on the vehicle's power side is reduced in the subsequent SOC recursion process; the power coupling error threshold is determined based on the hybrid power system operating mode, driving behavior and power response status, battery temperature, road slope, road surface condition, vehicle historical operating data, and cloud-based geographic environment perception correction parameters.

[0018] Furthermore, in step four, the vehicle power-side equivalent current is obtained by dividing the estimated battery power by the battery terminal voltage; the comprehensive operating condition correction coefficient is jointly determined by the basic correction weight corresponding to the hybrid system operating mode, the basic correction weight corresponding to driving behavior and power response state, and the power coupling error factor, with a value range of 0 to 1; when the vehicle is in a stable cruising state and the power coupling error is less than or close to the power coupling error threshold, the participation weight of the vehicle power-side equivalent current is increased; when the vehicle is in a rapid acceleration response state, low-speed start-stop state, strong disturbance state of braking regeneration, or power mode switching state, the participation weight of the vehicle power-side equivalent current is decreased; when the power coupling error is significantly greater than the threshold or the vehicle is in a strong disturbance state, the comprehensive operating condition correction coefficient is set to zero; the fused SOC recursive current is determined by the weighted sum of the measured battery current and the vehicle power-side equivalent current.

[0019] Furthermore, in step five, the continuous vehicle operation data is divided into several operation segments according to time windows, SOC change intervals, operating mode change times, or driving behavior change times. For each operation segment, a dynamic interference index is constructed, which is obtained by weighted fusion of one or more of the following: hysteresis influence degree, polarization influence degree, current fluctuation degree, temperature fluctuation degree, insufficient voltage recovery degree, power coupling error fluctuation degree, road slope change influence degree, and road surface slipperiness influence degree. The weights of various features are adjusted according to the comprehensive operating conditions and geographical environment. When the dynamic interference index is less than or equal to the dynamic interference threshold, the corresponding operation segment is marked as a reliable candidate segment. When the dynamic interference index is greater than the dynamic interference threshold, it is marked as a strong interference segment. The strong interference segment does not participate in effective capacity update and current zero bias estimation. The reliability of the segment is determined by the dynamic interference index. The larger the dynamic interference index, the lower the reliability of the segment.

[0020] Furthermore, in step five, the SOC reference value is obtained from a reference source that meets preset reliability conditions. The reference source includes at least one of the following: the SOC reference value obtained from the OCV-SOC calibration curve after the vehicle is stationary; the SOC reference value determined by recursively combining the historical SOC of the vehicle with historical high-reliability SOC reference information in the cloud during a stable operation segment; the SOC anchor point obtained during charging cutoff, full charge calibration, or low SOC boundary calibration states; the SOC reference value obtained by backtracking and fusing historical operation data in the cloud; when the start and end points of the segment both meet the requirements of low dynamic interference, small power coupling error, low voltage change rate, and SOC span not less than a preset threshold, the SOC reference value is determined to be reliable.

[0021] Furthermore, in step five, the candidate effective capacity is determined based on the ratio of the current integral within a trusted calibration segment to the change in the SOC reference value; the online effective capacity is obtained by weighted fusion based on the segment credibility of multiple trusted calibration segments, and then fused with the basic effective capacity according to the capacity update coefficient to obtain the updated effective capacity; the current zero bias is estimated by minimizing the charge closure error objective function based on multiple trusted calibration segments; the effective capacity and current zero bias are updated using an alternating estimation method, first estimating the effective capacity based on the current current zero bias, and then estimating the current zero bias based on the updated effective capacity, until convergence.

[0022] Furthermore, in step six, the cloud-generated correction parameters include at least the corrected effective capacity, current zero bias, and charge / discharge efficiency, as well as one or more of the following: power coupling error threshold, dynamic interference threshold, dynamic interference feature weight, credibility penalty coefficient, and capacity update coefficient. The cloud corrects the battery's effective capacity and charge / discharge efficiency based on ambient temperature, corrects the power coupling error threshold based on altitude, corrects the power coupling error threshold and dynamic interference threshold based on road slope changes, and reduces the credibility of the regenerative braking segment based on rain, snow, or slippery road conditions. The vehicle-side smoothly absorbs the cloud-corrected parameters based on the cloud parameter update coefficients, which are determined based on positioning accuracy, environmental data timeliness, and data source reliability.

[0023] Furthermore, in step seven, the final SOC recursion adopts the ampere-hour integral form, the recursive current adopts the fused SOC recursive current, and the effective capacity, current zero bias, and charging / discharging efficiency adopt the values ​​updated by the vehicle-side trusted calibration segment and corrected by the cloud-based geographical environment perception. In the strong interference segment, the correction weight of the vehicle power-side equivalent current is reduced or turned off and the segment is prohibited from participating in parameter updates. In the trusted calibration segment, the segment is allowed to participate in effective capacity updates, current zero bias estimation, and model parameter identification.

[0024] Furthermore, the cloud platform combines long-term vehicle operation data from multiple days or cycles to retrospectively analyze the SOC estimation residuals, power coupling errors, dynamic interference indicators, segment reliability, effective capacity estimation results, and current zero bias estimation results under different geographical regions, altitudes, weather conditions, and road slopes, continuously optimizing individual vehicle parameters and environmental correction rules.

[0025] The vehicle-cloud collaborative estimation method for the state of charge (SOC) of lithium iron phosphate batteries in hybrid vehicles, as proposed in this invention, has the following advantages:

[0026] 1. This invention constructs a comprehensive operating condition identification method for hybrid vehicles. By identifying various hybrid system operating modes, including electric drive-dominated, engine-mechanical drive, electromechanical coupling drive, engine-generated power supplementation, and regenerative braking, and combining these with driving behaviors and power response states such as rapid acceleration response, stable cruise, low-speed start-stop, power mode switching, and strong disturbances caused by regenerative braking, a comprehensive operating condition is formed. This allows for the differentiation of the impact of engine mechanical path, electric motor electric drive path, engine-generated power path, and regenerative braking path on battery-side power changes, providing a basis for subsequent power coupling error judgment, SOC recursive current fusion, reliable candidate segment selection, and reliable calibration segment parameter updates.

[0027] 2. This invention establishes a power coupling verification and SOC recursive current fusion mechanism between the vehicle power side and the battery side. By establishing segmented power coupling relationships based on different hybrid system operating modes, the power of the vehicle power side is calculated to estimate the battery side power, and consistency verification is performed with the actual power at the battery end to obtain the power coupling error. Furthermore, based on the comprehensive operating conditions and the power coupling error, the correction weight for the equivalent current of the vehicle power side participating in SOC recursion is determined. This enables the use of vehicle power side information to suppress zero bias and long-term integral drift of the battery current under stable and reliable operating conditions. Under strong disturbance conditions such as rapid acceleration, low-speed start-stop, strong disturbances from regenerative braking, or power mode switching, the correction effect of the vehicle power side is reduced or turned off, thereby improving the accuracy and stability of real-time SOC recursion under complex operating conditions.

[0028] 3. This invention constructs a parameter update mechanism that combines trusted segment calibration with cloud-based geographic environment perception parameter correction. It evaluates dynamic interference from factors such as hysteresis, polarization, current fluctuations, temperature fluctuations, insufficient voltage recovery, power coupling error fluctuations, road slope changes, and road surface slipperiness. Trustworthy candidate segments are selected based on dynamic interference indicators and segment trustworthiness. Trustworthy calibration segments are then determined by combining the reliability of the SOC reference value. Only operating segments with low dynamic interference, high trustworthiness, and reliable SOC reference values ​​are used for effective capacity updates and current zero-bias estimation, avoiding contamination of parameter update results by strong disturbance segments. Simultaneously, the cloud corrects battery effective capacity, charge / discharge efficiency, power coupling error threshold, dynamic interference threshold, dynamic interference weight, trustworthiness penalty coefficient, and capacity update coefficient based on the vehicle's current location, altitude, ambient temperature, weather conditions, road slope, and road surface conditions. Furthermore, it continuously optimizes individual vehicle parameters and environmental correction rules based on long-term operating data, thereby improving the adaptability of the SOC estimation method to regional environmental changes, road slope changes, weather changes, battery aging, and individual vehicle differences. Attached Figure Description

[0029] Figure 1 This is a flowchart of the overall SOC vehicle-cloud collaborative estimation method of the present invention;

[0030] Figure 2 A flowchart for identifying comprehensive operating conditions;

[0031] Figure 3 A flowchart illustrating the power coupling between the vehicle's powertrain side and battery side, as well as the SOC recursive current fusion process.

[0032] Figure 4 Flowchart for dynamic interference evaluation, reliable candidate fragment selection, and reliable calibration fragment parameter update;

[0033] Figure 5 Flowchart for correcting parameters for cloud-based geographic environment perception. Detailed Implementation

[0034] To better understand the purpose, structure, and function of this invention, the following detailed description, in conjunction with the accompanying drawings, provides a method for vehicle-cloud collaborative estimation of SOC of lithium iron phosphate batteries in hybrid vehicles.

[0035] This invention provides a vehicle-cloud collaborative estimation method for the state of charge (SOC) of lithium iron phosphate batteries in hybrid vehicles, as detailed below:

[0036] Step 1: Acquire multi-source operational data from the vehicle and the cloud. Specifically, during vehicle operation, vehicle operational data is acquired through the vehicle-side data processing module or the vehicle-cloud platform. This vehicle operational data originates from the battery management system, vehicle controller, motor controller, engine controller, braking control system, vehicle networking terminal, or a historical operational database in the cloud.

[0037] The vehicle operating data includes battery pack voltage, battery pack current, battery temperature, vehicle speed, vehicle acceleration, accelerator pedal opening, braking signal, motor operating status, engine operating status, engine output power, motor mechanical output power, accessory load power, regenerative braking power, and the vehicle's current location data. The vehicle's current location data includes longitude, latitude, and altitude, which can be obtained from an onboard positioning module, a vehicle-to-everything (V2X) terminal, or a cloud-based map platform.

[0038] Furthermore, the cloud acquires external environmental data corresponding to the vehicle's current location. This external environmental data includes one or more of the following: ambient temperature, weather conditions, road slope, road elevation changes, rainfall, snowfall, and road surface slipperiness. This external environmental data can be obtained from cloud-based meteorological services, map services, road environment databases, vehicle-to-everything (V2X) platforms, or vehicle sensors.

[0039] Let the first The set of vehicle operation data at each sampling time point is as follows:

[0040] ,

[0041] In the formula, This refers to the battery terminal voltage. Battery current, For battery temperature, For vehicle speed, To accelerate the vehicle, This refers to the accelerator pedal opening. This is a braking signal. This refers to the mechanical output power of the electric motor. This refers to the engine's output power. This represents the actual power that enters the battery side in the engine's power generation path. This represents the actual power entering the battery side in the regenerative braking path. For the load power of the accessory, The vehicle's latitude and longitude location, The cloud displays the vehicle's altitude based on its location. and altitude Obtain the corresponding set of geographic environment states .

[0042] Vehicle acceleration can be obtained from the change in vehicle speed between adjacent sampling times:

[0043]

[0044] In the formula, This represents the sampling time interval.

[0045] Based on the aforementioned vehicle-side operation data, powertrain power data, and cloud environment data, the vehicle-side utilizes the vehicle operation data set. Identify the current hybrid system operating mode, driving behavior, and power response status, and complete subsequent calculations of actual battery power, vehicle power-side estimated battery power, and SOC recursive current fusion; the cloud utilizes geographic environment status sets. and up to the Historical operating data at each sampling time is used to correct effective capacity, current bias, charging and discharging efficiency, as well as related threshold and weight parameters, and the corrected parameters from the cloud are transmitted back to the vehicle. The vehicle then smoothly fuses the corrected parameters from the cloud and participates in the process. The final SOC is calculated at each sampling time.

[0046] Step 2: Based on the vehicle operation data obtained in Step 1, identify the comprehensive operating conditions of the hybrid vehicle. Specifically, the vehicle-side system uses the vehicle operation data set... Identify the comprehensive operating conditions of hybrid vehicles. The comprehensive operating conditions include two parts: the hybrid system operating mode and the driving behavior and power response status.

[0047] The hybrid system operating mode is used to represent the power flow relationship between the engine, motor, power generation path, mechanical transmission path, regenerative braking path, and battery in a hybrid vehicle. Driving behavior and power response status represent the vehicle's operating response status caused by driver actions such as acceleration, braking, and low-speed start-stop, and is reflected by power distribution in the power system, battery current fluctuations, and operating mode switching.

[0048] Let the first The hybrid power system operating mode at each sampling time is as follows: Driving behavior and power response status are The comprehensive operating condition is Then we have:

[0049]

[0050]

[0051]

[0052] In the formula, This is a function for identifying the operating mode of a hybrid power system. A function for identifying driving behavior and power response states. For the first A set of vehicle operation data at each sampling time. Hybrid system operating mode. Used to characterize the power flow relationship between the engine, motor, power generation path, mechanical transmission path, regenerative braking path, and battery; driving behavior and power response status. Used to characterize the dynamic response features of a vehicle during acceleration, cruising, low-speed start-stop, power response switching, and regenerative braking. (Comprehensive operating conditions) Depend on and Together they constitute.

[0053] In one implementation, the battery discharge direction is set to positive and the charging direction to negative, and the braking signal is valid when... When the braking signal is invalid In one implementation, the hybrid power system operates in a certain mode. It can include electric drive-dominated mode, engine mechanical drive mode, electromechanical coupling drive mode, engine power generation and regenerative braking mode, and can be judged according to the rules shown in Table 1.

[0054] Table 1 Rules for Determining the Operating Mode of Hybrid Power Systems

[0055]

[0056] in, This refers to the engine's output power. This refers to the mechanical output power of the electric motor. This represents the actual power that enters the battery side in the engine's power generation path. This represents the actual power entering the battery side in the regenerative braking path. This represents the battery current. This is the engine intervention threshold. The mechanical drive threshold of the engine. This is the threshold value for motor drive. For motor auxiliary drive threshold, The threshold for engine power generation. The threshold for regenerative braking energy. This is the battery discharge current threshold. The battery charging current threshold. It is only used to distinguish between the mechanical drive mode and the electromechanical coupling drive mode of the engine, and does not represent the actual power variable.

[0057] Driving behavior and power response status It includes at least the rapid acceleration response state, stable cruise state, low-speed start-stop state, power mode switching state, and strong disturbance state of regenerative braking, which can be judged according to the rules shown in Table 2.

[0058] Table 2 Rules for Judging Driving Behavior and Power Response Status

[0059]

[0060] in, To accelerate the vehicle, This refers to the accelerator pedal opening. For vehicle speed, This refers to the mechanical output power of the electric motor. This is a braking signal. For the first The road surface wetness status indicator at each sampling time. When When, it indicates that the vehicle is in a state of rain, snow, or slippery road conditions; when This indicates that the vehicle was not in a state of rain, snow, or on a slippery road. It should be noted that... This is used to indicate the road surface wetness / slippery condition at the sampling time, distinguishing it from the later indication used to indicate the first... Credibility of each running segment .

[0061] Indicates the first The sampling time is the endpoint, and the length is... The standard deviation of battery current within a time window of each sampling point is used to characterize the degree of battery current fluctuation within a preset time window. This indicates the number of times the accelerator pedal state, braking signal, or current direction changes within that time window. Indicates the first The sampling time is the endpoint, and the length is... The actual power entering the battery side in the regenerative braking path within the time window of each sampling point. The standard deviation of the braking energy recovery power is used to characterize the degree of fluctuation in braking energy recovery power; , , , and These represent the changes in battery current, motor mechanical output power, engine output power, engine power generated into the battery, and braking energy recovery power into the battery at adjacent sampling times.

[0062] The acceleration threshold for rapid acceleration, To rapidly accelerate the accelerator pedal opening threshold, To accelerate the motor power threshold, To stabilize the cruise acceleration threshold, The current fluctuation threshold, The threshold for low-speed judgment is set. This is the threshold for the number of low-speed start-stop switching operations. , , , and These are the current surge threshold, motor power surge threshold, engine power surge threshold, engine power surge threshold for power surge entering the battery side, and regenerative braking power surge threshold, respectively. The threshold for power fluctuations in the regenerative braking energy input to the battery side.

[0063] To avoid misjudgments caused by single-point noise, instantaneous load fluctuations, or sensor jitter, rapid acceleration response, low-speed start-stop, power mode switching, and strong disturbances during regenerative braking can be continuously monitored within a preset time window. Let the preset time window length be... The sampling time interval is Then the number of sampling points corresponding to this preset time window is:

[0064]

[0065] In the formula, Indicates length is The number of sampling points corresponding to the preset time window. This indicates rounding up. If the corresponding judgment condition is continuously satisfied for at least [time value] within the preset time window... Each sampling point, and This confirms that the vehicle is in the corresponding driving behavior and power response state. The minimum number of sampling points required for continuous confirmation can be determined based on the vehicle controller sampling cycle, powertrain response time, typical driving condition experiments, or historical operating data from the cloud.

[0066] The aforementioned threshold parameters are not limited to fixed constants. They can be determined by existing vehicle calibration parameters, typical driving condition experiments, historical operating data from the cloud, or cloud-based geographic environment perception correction parameters. They can also be updated based on the vehicle's powertrain structure, vehicle control strategy, battery aging status, ambient temperature, road slope, and road surface condition.

[0067] The comprehensive operating conditions obtained through the above methods It is used not only to characterize the current power flow direction and driving response state of the vehicle's power system, but also to select the corresponding power coupling relationship between the vehicle's power side and battery side, determine the power coupling error threshold, calculate the SOC recursive current fusion weight, and screen reliable candidate segments.

[0068] Step 3: Obtain comprehensive operating conditions Then, the power coupling relationship between the vehicle's powertrain and battery sides is established. Specifically, this is based on the hybrid system's operating mode. Select the corresponding power coupling relationship between the vehicle's power side and battery side to estimate the first... Estimated battery-side power at each sampling time point The power coupling relationship between the vehicle's power side and the battery side is used to characterize the impact of motor drive power, engine power generation entering the battery side, braking energy recovery power, and accessory load power on the battery side power under different hybrid power system operating modes.

[0069] In one implementation, the battery discharge direction is set to positive, and the battery charging direction to negative. The actual battery power is calculated based on the battery terminal voltage and battery current.

[0070]

[0071] In the formula, For the first The actual power at the battery terminal at each sampling time. For the first Battery terminal voltage at each sampling time, For the first The battery current at each sampling time. If the current direction is defined differently in the actual vehicle data, the current and power signs will be adjusted accordingly.

[0072] The estimated battery-side power of the vehicle's powertrain is expressed as follows:

[0073]

[0074] In the formula, The battery-side power estimated for the vehicle's powertrain; This is a power coupling calculation function related to the operating mode of the hybrid power system; This refers to the actual mechanical output power of the motor. For the load power of the accessory; This represents the actual power entering the battery side in the engine's power generation path; This represents the actual power entering the battery side in the regenerative braking path.

[0075] The estimated battery power on the vehicle's power side can be determined according to the relationship shown in Table 3 under different hybrid system operating modes.

[0076] Table 3 Estimated battery power on the vehicle's power side under different hybrid system operating modes

[0077]

[0078] in, This is the conversion factor between the motor's mechanical output power, determined by the motor efficiency calibration results, and the battery-side power. According to the aforementioned power definition, the actual power entering the battery side in the engine power generation path and the braking energy recovery path is directly deducted as the battery-side charging power in the power balance calculation in Table 3, and is not repeatedly multiplied by the power generation efficiency or braking energy recovery efficiency.

[0079] In both engine-mechanical drive mode and electromechanical coupling drive mode, the power of the engine directly participating in vehicle drive via the mechanical path is not directly included in the battery-side power balance; only the actual mechanical output power of the motor, the power of the accessory load, and the power actually entering the battery side via the power generation path are included in the battery-side power estimation.

[0080] When the motor is not involved in driving or auxiliary driving When the motor has a small auxiliary output but it has not yet reached the motor auxiliary drive threshold, it will still be based on the actual measured output. Participate in battery-side power estimation.

[0081] Furthermore, the power coupling error between the actual power at the battery terminal and the estimated battery-side power on the vehicle's powertrain side is calculated:

[0082]

[0083] In the formula, For the first The battery-side power coupling error at each sampling time is used to characterize the consistency between the actual power at the battery end and the battery-side power estimated based on the hybrid power system operating mode.

[0084] When the following conditions are met:

[0085]

[0086] When the estimated battery power on the vehicle's power side is consistent with the actual measured power on the battery side, it is determined that the two values ​​are highly consistent; when the following conditions are met:

[0087]

[0088] When a significant discrepancy is found between the estimated battery power on the vehicle's power side and the actual measured power at the battery terminal, the weighting of the equivalent current on the vehicle's power side is reduced in subsequent SOC calculations. In the formula, The power coupling error threshold is related to the overall operating conditions.

[0089] The power coupling error threshold It is not limited to a fixed constant and can be determined based on the hybrid system's operating mode, driving behavior and power response status, battery temperature, road slope, road surface condition, vehicle historical operating data, and cloud-based geographic environment perception correction parameters.

[0090] Using the above method, the vehicle's powertrain side estimates the battery-side power. It does not directly replace the actual power of the battery. Instead, it serves as redundant verification and auxiliary correction information for the actual power at the battery end. When the power coupling error is small and the vehicle is under stable and reliable operating conditions, the equivalent current on the vehicle's power side is allowed to participate in the SOC recursion; when the power coupling error is large or the vehicle is under strong disturbance operating conditions, the correction effect of the equivalent current on the vehicle's power side is reduced or turned off, thereby avoiding SOC recursion deviation caused by power side estimation errors.

[0091] Step 4: Estimate the battery-side power based on the calculated vehicle powertrain side power. and power coupling error Then, SOC recursive current fusion is performed. Specifically, the battery-side power is estimated from the vehicle's powertrain side. Converted to equivalent battery current on the vehicle's power side :

[0092]

[0093] In the formula, For the first The equivalent battery current at each sampling time is obtained by estimating the battery-side power from the vehicle's powertrain side. For the first Battery terminal voltage at each sampling time, This is used to prevent small positive numbers with a denominator of zero.

[0094] Based on comprehensive operating conditions and power coupling error Determine the comprehensive operating condition correction factor for the vehicle's power-side equivalent current to participate in the SOC recursion:

[0095]

[0096] In the formula, For the first The comprehensive operating condition correction coefficient at each sampling time. These are the basic correction weights corresponding to the operating modes of the hybrid power system. The basic correction weights corresponding to driving behavior and power response status. The power coupling error threshold is related to the overall operating conditions.

[0097] in, and All values ​​are within the range of The weighting function within. The aforementioned The specific modes can be determined by referring to tables for electric drive-dominated mode, engine-mechanical drive mode, electromechanical coupling drive mode, engine-generated energy replenishment mode, and regenerative braking mode; the aforementioned... The values ​​can be determined by looking up tables for rapid acceleration response, stable cruise, low-speed start-stop, power mode switching, and strong disturbance during regenerative braking. The initial value of the weighting function can be determined by existing vehicle calibration parameters or typical driving condition experiments, and can be updated based on historical operating data from the cloud and geographic environment perception correction parameters.

[0098] Obtain the basic correction weights corresponding to the operating modes of the hybrid power system. Basic correction weights corresponding to driving behavior and power response state and power coupling error Afterwards, the first was determined comprehensively. The comprehensive operating condition correction coefficient for the vehicle power-side equivalent current at each sampling time participating in the SOC recursion. .in, satisfy:

[0099]

[0100] When the vehicle is in a stable cruise state, and the power coupling error Less than or close to the power coupling error threshold When the vehicle is in a state of rapid acceleration, low-speed start-stop, strong disturbance state due to regenerative braking, or power mode switching, increase the weight of the equivalent current on the vehicle's power side in the SOC recursion; when the vehicle is in a state of rapid acceleration, low-speed start-stop, strong disturbance state due to regenerative braking, or power mode switching, decrease the weight of the equivalent current on the vehicle's power side in the SOC recursion; when the power coupling error is significantly greater than the threshold, or when the vehicle is in a strong disturbance state such as power mode switching, the weight can be increased. Approaching zero or equal to zero.

[0101] Measured battery current Equivalent current of the vehicle's power side By merging, the SOC recursive current is obtained:

[0102]

[0103] In the formula, For the first The SOC recursive current after fusion at each sampling time point For the first The measured current of the battery at each sampling time. For the first The equivalent battery current on the vehicle's power side at each sampling time. For the first The comprehensive working condition correction coefficient at each sampling time.

[0104] The merged SOC recursive current The current input serves as the final SOC recursive calculation. The effective capacity, zero-bias current, and charge / discharge efficiency are determined by the reliable calibration segment parameter update in step six and the cloud-based geographic environment perception parameter correction in step seven, respectively. Step eight is based on the fused SOC recursive current. Corrected effective capacity Corrected current zero bias and corrected charge / discharge efficiency Output the final SOC estimate.

[0105] In this way, the present invention does not directly replace the measured current at the battery terminal with the estimated current at the vehicle power side. Instead, when the estimated battery power at the vehicle power side is highly consistent with the actual power at the battery terminal and the overall operating conditions are relatively stable, the equivalent current at the vehicle power side is used as the redundant correction input for the final SOC recursion. Under strong disturbance conditions such as rapid acceleration, low-speed start-stop, braking regeneration, or power mode switching, the correction effect of the equivalent current at the vehicle power side is reduced or turned off, thereby avoiding deviation in the final SOC recursion due to transient power fluctuations or power side estimation errors.

[0106] Step 5: After completing the SOC recursive current fusion, construct a dynamic disturbance index and screen reliable candidate segments. Because rapid acceleration, low-speed start-stop, power mode switching, and regenerative braking can cause significant fluctuations in battery current, voltage, power, and temperature, and may amplify the effects of hysteresis, polarization, and insufficient voltage recovery in lithium iron phosphate batteries, not all operating data is suitable for effective capacity updates and zero-bias current estimation. To avoid strong disturbance segments contaminating subsequent parameter update results, this invention further divides continuous operating data into segments and constructs a dynamic disturbance index related to comprehensive operating conditions.

[0107] Specifically, the continuous vehicle operation data is divided into several operation segments according to time windows, SOC change intervals, changes in powertrain operating modes, or changes in driving behavior and power response states. Let the first segment be... The sampling intervals corresponding to each running segment are:

[0108]

[0109] In the formula, For the first The starting sampling time of each running segment, For the first The final sampling time of each running segment. The comprehensive operating condition corresponding to each operating segment is denoted as: Geographical environment status is denoted as .

[0110] Regarding the first Each runtime segment is used to construct a dynamic interference index:

[0111]

[0112] In the formula, For the first Dynamic interference indicators for each running segment For the first The first running segment The normalized values ​​of the dynamic disturbance features. Based on comprehensive operating conditions and geographical environment The jointly determined first Dynamic interference feature weights, The number of features participating in dynamic interference evaluation.

[0113] The dynamic interference characteristics include one or more of the following: hysteresis effect, polarization effect, current fluctuation effect, temperature fluctuation effect, insufficient voltage recovery effect, power coupling error fluctuation effect, road slope change effect, and road surface slipperiness effect.

[0114] Among these, the characteristics of various dynamic interferences can be determined by measurable statistics within the operating segment. The degree of hysteresis can be determined based on the number of battery current direction switching times, the number of battery charge / discharge state switching times, or the charge / discharge switching frequency per unit time within the operating segment; the degree of polarization can be determined based on the mean absolute value of battery current, the standard deviation of battery current, the mean residual of estimated battery terminal voltage, or the voltage recovery slope after a sudden current change within the operating segment; the degree of current fluctuation can be determined based on the standard deviation of battery current, the maximum current change, or the peak-to-peak current within the operating segment; the degree of temperature fluctuation can be determined based on the difference between the highest and lowest battery temperatures, the temperature change rate, or the temperature standard deviation within the operating segment; the degree of insufficient voltage recovery can be determined based on the battery terminal voltage change rate, the resting recovery time, the voltage change after resting, or the voltage recovery slope within the operating segment; and the degree of power coupling error fluctuation can be determined based on the power coupling error within the operating segment. The mean, maximum value, standard deviation, or peak-to-peak value are used to determine the impact of road slope changes; the impact of road slope changes can be determined based on the road slope, slope change rate, altitude change rate, or altitude change per unit travel distance corresponding to the vehicle's current location; the impact of road surface slipperiness can be determined based on rainfall, snowfall, road surface slipperiness indicators, or fluctuations in regenerative braking power.

[0115] Various dynamic interference features are normalized before participating in weighted fusion. In one implementation, the first... Dynamic interference features can be normalized as follows:

[0116]

[0117] In the formula, For the first The first running segment The original numerical values ​​of the dynamic interference characteristics. and These are the minimum and maximum reference values ​​for this type of dynamic interference characteristic, respectively. This is used to prevent small positive numbers with a denominator of zero. Normalized Used to characterize the relative intensity of the corresponding dynamic disturbance feature.

[0118] Dynamic interference feature weights It is not limited to a fixed constant, but is adjusted according to the comprehensive operating conditions and geographical environment. Specifically, in the rapid acceleration response state, the weights corresponding to the current fluctuation degree, polarization influence degree, and power coupling error fluctuation degree are increased; in the low-speed start-stop state, the weights corresponding to the number of current direction switching times, hysteresis influence degree, and insufficient voltage recovery degree are increased; in the power mode switching state, the weights corresponding to the current mutation, power mutation, and power coupling error fluctuation degree are increased; in the strong disturbance state of regenerative braking, the weights corresponding to the regenerative braking power fluctuation degree and the influence of road surface slipperiness are increased; in the stable cruise state, the weights related to dynamic interference are reduced or the segment reliability is increased.

[0119] In obtaining dynamic interference indicators Then, based on the comprehensive operating conditions and geographical environment Set dynamic interference threshold:

[0120]

[0121] When the following conditions are met:

[0122]

[0123] At that time, the first A running segment is marked as a reliable candidate segment; when the following condition is met:

[0124]

[0125] At that time, the first Each running segment is marked as a strong interference segment. The strong interference segment can participate in the final SOC recursion, but it does not participate in subsequent effective capacity updates, current zero-bias estimation, and model parameter identification.

[0126] Furthermore, to differentiate the contribution of different credible candidate segments to subsequent parameter updates, segment credibility is calculated based on the dynamic interference index:

[0127]

[0128] In the formula, For the first The reliability of each running segment. This is a reliability penalty coefficient determined jointly by comprehensive operating conditions and geographical environment status. Dynamic interference index. The larger the value, the higher the credibility of the fragment. The lower the value; the lower the dynamic interference index The smaller the value, the higher the credibility of the fragment. The higher.

[0129] Among these, stable cruise control corresponds to a smaller reliability penalty coefficient; rapid acceleration response, low-speed start-stop, power mode switching, and strong disturbance during regenerative braking correspond to a larger reliability penalty coefficient. When the vehicle is in environmental conditions such as low temperature, high slope, rain, snow, or slippery road surfaces, the cloud can adjust the dynamic disturbance threshold based on geographic environment perception parameters. Dynamic interference feature weights And credibility penalty coefficient Make corrections.

[0130] Through the above methods, the present invention can evaluate the dynamic disturbance degree of the operating segment based on the comprehensive operating conditions and geographical environment, and select operating segments with low dynamic disturbance and high reliability as reliable candidate segments, providing a reliable data foundation for subsequent effective capacity updates, current zero bias estimation and model parameter identification.

[0131] Step Six: After completing the screening of credible candidate fragments and obtaining the fragment credibility. Then, the effective capacity and current zero bias are updated based on the trusted calibration segments. Specifically, after completing the screening of trusted candidate segments and obtaining the segment trustworthiness... Next, the reliability of the SOC reference value of the credible candidate segment is further assessed. Only when the segment's credibility meets the requirements and the SOC reference value is reliable, is the corresponding running segment determined as a credible calibration segment and used for battery effective capacity update, current zero bias estimation, and model parameter identification.

[0132] For the If there are 1 credible candidate segments, and the following conditions are met:

[0133]

[0134] If the SOC reference values ​​at the start and end points of the running segment are reliable, then the running segment is determined to be a reliable calibration segment. Where, This is the threshold for fragment credibility.

[0135] The SOC reference value is not directly derived from the current recursive SOC estimate, but is obtained from a reference source that meets preset reliability conditions. The reference source includes at least one of the following: When the vehicle's idle time exceeds a preset idle time threshold, and the local slope of the lithium iron phosphate battery OCV-SOC calibration curve corresponding to the stable battery voltage meets preset reliability conditions, the SOC reference value is obtained based on the stable battery voltage, battery temperature, and the lithium iron phosphate battery OCV-SOC calibration curve; when the vehicle is in a stable operating segment with low dynamic interference, power coupling error below a threshold, current fluctuation below a threshold, and voltage change rate below a threshold, the SOC reference value is determined by the vehicle's historical SOC recursive result up to the end of the operating segment, combined with historical high-reliability SOC reference information from the cloud; the SOC anchor point obtained when the battery is in a charging cutoff, full-charge calibration, low SOC boundary calibration, or when the vehicle controller and battery management system output a reliable calibration state; or the SOC reference value obtained by backtracking and fusing high-reliability segments of the same vehicle under similar temperatures, similar operating conditions, and similar environments based on multi-day or multi-cycle historical operating data from the cloud. When the OCV-SOC curve corresponding to the current stable voltage is in the voltage plateau region and the local slope does not meet the preset confidence conditions, the SOC is not inferred solely from the stable voltage. Instead, the SOC reference value is determined by combining the historical high confidence SOC results in the cloud, the charging cutoff state, the full charge calibration state, the low SOC boundary calibration state, or the confidence calibration state output by the vehicle control system.

[0136] When the start and end points of a segment both satisfy low dynamic interference, small power coupling error, low voltage change rate, and a SOC span not less than a preset threshold, the SOC reference values ​​for the start and end points of the segment are considered reliable. The SOC span condition can be expressed as:

[0137]

[0138] In the formula, and The first SOC reference values ​​for the start and end points of a reliable calibration segment. This is the minimum SOC span threshold.

[0139] For a reliable calibration segment, candidate effective capacity is estimated based on the intra-segment current integral and the change in the SOC reference value:

[0140]

[0141] In the formula, For the first The candidate effective capacity is calculated from a number of reliable calibration segments. and These are the start and end times of the segment, respectively. For a moment The corresponding coulomb efficiency or charge / discharge efficiency, For a moment The corresponding battery current, This is the current zero-bias correction amount used within the segment. and These are the SOC reference values ​​for the start and end points of the segment, respectively.

[0142] The online effective capacity is obtained by weighted fusion based on the segment credibility of multiple trusted calibration segments:

[0143]

[0144] In the formula, The online effective capacity is obtained by weighted fusion of multiple trusted calibration segments. The number of reliable calibration segments that participate in the effective capacity estimation.

[0145] The online effective capacity is combined with the basic effective capacity to obtain the updated effective capacity:

[0146]

[0147] In the formula, For the updated effective battery capacity, Based on effective capacity, Let be the capacity update factor, and satisfy:

[0148]

[0149] Capacity update factor The capacity update reliability can be determined based on the reliability of the capacity update. This reliability can be determined by one or more of the following: the number of reliable calibration segments, the average reliability of the segments, the SOC span, the dispersion of candidate effective capacity, temperature stability, and power coupling error stability. When the number of reliable calibration segments is large, the segment reliability is high, the SOC span is large, and the dispersion of candidate capacity is small, the capacity update coefficient is increased; conversely, the capacity update coefficient is decreased.

[0150] Since the SOC reference values ​​at the start and end of the reliable calibration segment meet the reliability conditions, the change in SOC obtained based on the current integral within this segment should be consistent with the change in the SOC reference value. If the battery current has zero bias, it will cause a charge closure error between the current integral and the change in the SOC reference value. Therefore, the current zero bias can be used as the variable to be estimated, and the current zero bias can be estimated based on the charge closure errors of multiple reliable calibration segments.

[0151] Specifically, let the zero-bias variable of the current to be estimated be... The zero-bias estimate of the current is obtained by minimizing the following objective function:

[0152]

[0153] In the formula, For the first The zero bias of the vehicle-end local current at each sampling time is estimated by a reliable calibration segment and used as the local input for cloud correction; For the zero-bias variable of the current to be estimated, The number of reliable calibration segments participating in the current zero bias estimation, For the first The reliability of a reliable calibration segment. and The first SOC reference values ​​for the start and end points of a reliable calibration segment. and The first The start and end times of a reliable calibration segment. For a moment The corresponding coulomb efficiency or charge / discharge efficiency, For a moment The corresponding battery current, This represents the effective capacity of the currently used battery.

[0154] In one implementation, the effective capacity and current zero bias are updated using an alternating estimation method. First, the effective capacity is estimated based on the current current zero bias, then the current zero bias is estimated based on the updated effective capacity, and the iteration stops when the changes in effective capacity and current zero bias obtained from two consecutive updates are both less than a preset threshold.

[0155] In this way, the present invention uses only the operating segments with low dynamic interference, high reliability and reliable SOC reference values ​​to update the effective capacity and current zero bias, avoiding the pollution of capacity estimation and current zero bias estimation results by strong interference segments such as rapid acceleration, low-speed start-stop, strong disturbances of braking recovery and power mode switching.

[0156] Step 7: Based on the SOC recursive current fusion and trusted segment parameter update completed on the vehicle side, perform cloud-based geographic environment perception parameter correction. Specifically, the vehicle side obtains the vehicle's latitude, longitude, and altitude through the onboard positioning module or vehicle-to-everything (V2X) terminal, and updates the vehicle's current location, battery status, and overall operating conditions. Power coupling error Dynamic interference index Credibility of fragments The effective capacity estimation results and the current zero bias estimation results are uploaded to the cloud.

[0157] The cloud obtains a set of geographic environment conditions based on the vehicle's current location:

[0158]

[0159] In the formula, For ambient temperature, For weather conditions, This refers to road slope or terrain variation characteristics. This refers to conditions such as rain, snow, or slippery roads. This indicates the altitude of the vehicle's location.

[0160] Before cloud-based correction, the vehicle-side effective local capacity, zero-bias current, and charging / discharging efficiency were as follows: , and .

[0161] in, This refers to the vehicle-side local effective capacity updated based on trusted calibration fragments in step six. This is the zero bias current estimated based on the reliable calibration segment in step six, and used as the cloud-corrected zero bias current adopted locally at the front end; To correct the estimated or calibrated charging and discharging efficiency values ​​used locally on the vehicle side in the cloud, the values ​​can be determined based on the vehicle's local efficiency calibration parameters, battery temperature, current direction, current rate, and overall operating conditions.

[0162] Based on comprehensive operating conditions Geographical environment status and up to the Historical vehicle operation data at each sampling time point is used to determine the effective capacity environmental correction coefficient. Current zero bias correction offset and environmental correction factor for charge and discharge efficiency And obtain the corrected parameters from the cloud:

[0163]

[0164]

[0165]

[0166] In the formula, , and These are the cloud-corrected effective capacity, zero-bias current, and charge / discharge efficiency, respectively. This is the effective capacity environmental correction factor. This is the offset for zero bias correction of cloud current. This is an environmental correction factor for charge / discharge efficiency. , and Based on comprehensive operating conditions Geographical environment status The determination is made from one or more of the following factors: battery temperature, altitude, road slope, weather conditions, road surface slipperiness, SOC estimation residual, power coupling error, and vehicle historical operating data.

[0167] Furthermore, a set of correction parameters is formed in the cloud:

[0168]

[0169] In the formula, Based on comprehensive operational conditions in the cloud and geographical environment Corrected power coupling error threshold The cloud provides a comprehensive overview of the operational status based on the running segments. and geographical environment Corrected dynamic interference threshold The corrected version in the cloud Dynamic interference feature weights, This is the credibility penalty coefficient after cloud-based correction. The cloud-corrected capacity update coefficients are used to update the power coupling error threshold, dynamic interference threshold, dynamic interference feature weight, credibility penalty coefficient, and capacity update coefficients respectively after the cloud-corrected thresholds and weight parameters are sent back to the vehicle.

[0170] When a vehicle is in a low-temperature environment, the cloud reduces the effective battery capacity and charge / discharge efficiency, and increases the weights related to polarization effects, insufficient voltage recovery, and temperature fluctuations. When a vehicle is in a high-temperature environment, the cloud corrects the charge / discharge efficiency and increases the weight of temperature fluctuation penalties. When a vehicle is in a high-altitude area, the cloud adjusts the thresholds related to engine output capacity, engine power generation judgment, and power coupling error based on altitude. When a vehicle is in mountainous areas or on steep roads, the cloud adjusts the power coupling error threshold and dynamic interference threshold based on changes in road slope to prevent normal load changes on sloped roads from being misjudged as abnormal power disturbances. When a vehicle is in rain, snow, or slippery road conditions, the cloud reduces the reliability of regenerative braking segments, increases the weights related to regenerative braking power fluctuations and slippery road conditions, and reduces the weight of such segments in effective capacity updates and current zero-bias estimation.

[0171] The cloud-based parameter correction does not directly replace the vehicle's local parameters; instead, the vehicle updates the coefficients based on the cloud-based parameters. Smooth absorption is performed. For the effective battery capacity, the vehicle first performs a smooth update based on the difference between the cloud-corrected capacity and the vehicle's local capacity:

[0172]

[0173] Because the cloud-based correction capacity meets the requirements:

[0174]

[0175] Therefore, we can conclude that:

[0176]

[0177] For zero current bias, since cloud-based correction mainly manifests as an offset correction of the local zero bias at the vehicle end, the vehicle end updates the coefficients based on cloud parameters. Smoothing the corrected offset yields:

[0178]

[0179] Regarding charging and discharging efficiency, the vehicle-side system smoothly updates the efficiency based on the difference between the cloud-corrected efficiency and the vehicle-side local efficiency.

[0180]

[0181] Because cloud-based correction efficiency meets:

[0182]

[0183] Therefore, we can conclude that:

[0184]

[0185] In the formula, , and These are the effective capacity, zero-bias current, and charge / discharge efficiency used for the final SOC calculation after smooth fusion at the vehicle end; Update the coefficients for parameters in the cloud, and satisfy the following:

[0186]

[0187] when At that time, the vehicle-side does not use cloud-based correction values; when At that time, the vehicle-side parameters are corrected entirely via the cloud; when At that time, the vehicle end smoothly absorbs the cloud correction amount proportionally.

[0188] Cloud parameter update coefficient Determined based on the reliability of cloud-based geographic environment information. When positioning accuracy is high, weather data update time is short, road slope data is complete, environmental data sources are reliable, and historical environmental correction effects are good, the accuracy is improved. When positioning accuracy is low, weather data is lagging, road slope data is missing, or the reliability of environmental data sources is insufficient, the accuracy will decrease. This is to avoid corrupted vehicle parameters due to erroneous environmental data.

[0189] Furthermore, the cloud can also combine long-term vehicle operation data from multiple days or periods to retrospectively optimize the rules for correcting individual vehicle parameters and geographic environment perception parameters. The long-term operation data includes one or more of the following under different geographical regions, altitudes, weather conditions, and road slopes: SOC estimation residuals, power coupling errors, dynamic interference indicators, segment reliability, effective capacity estimation results, and current zero bias estimation results.

[0190] In this way, the cloud can not only make real-time or near-real-time corrections to the vehicle's SOC estimation parameters based on the vehicle's current geographical environment, but also continuously optimize individual vehicle parameters and environmental correction rules using long-term vehicle operation data, thereby improving the SOC estimation method's adaptability to regional environmental changes, weather changes, road gradient changes, battery aging, and individual vehicle differences.

[0191] Step 8: After completing the SOC recursive current fusion in Step 4, the reliable calibration segment parameter update in Step 6, and the cloud-based geographic environment perception parameter correction in Step 7, the vehicle-side uses the fused SOC recursive current... Effective capacity after vehicle-side local parameter updates and cloud-based geographic environment perception corrections Zero bias current and charge / discharge efficiency Output the first The unique final SOC estimate at each sampling time. .

[0192] The final SOC recursive form is:

[0193]

[0194] In the formula, For the first A unique final SOC estimate for each sampling time; For the first The final SOC estimate at each sampling time point; The charging and discharging efficiency is the result of updated parameters of the trusted calibration segment and correction based on cloud-based geographic environment perception. The fused SOC recursive current obtained in step four; The current is zero bias after being estimated by a reliable calibration segment and smoothed in the cloud; The sampling time interval; This represents the effective battery capacity after updating trusted calibration fragment parameters and correcting for cloud-based geographic environment awareness.

[0195] In the high-interference segment, the vehicle-side reduction or shutdown of the correction weight of the vehicle power-side equivalent current in the SOC recursion is implemented, and this segment is prohibited from participating in effective capacity updates and current zero-bias estimation. In the reliable calibration segment, this segment is allowed to participate in effective capacity updates, current zero-bias estimation, and model parameter identification. Thus, vehicle power-side power coupling verification, reliable segment calibration, effective capacity updates, current zero-bias estimation, and cloud-based environmental perception parameter correction all work together in the final SOC recursion process by fusing current, effective capacity, charge / discharge efficiency, current zero-bias, and related threshold parameters.

[0196] Through the above methods, the present invention can determine the source of battery power change based on the working mode of the hybrid system, determine the intensity of data disturbance based on driving behavior and power response status, and optimize the effective capacity, current zero bias, charge and discharge efficiency, power coupling error threshold, dynamic interference threshold, dynamic interference weight and credibility penalty coefficient through cloud-based geographic environment perception parameter correction and long-term operation data backtracking optimization. This improves the accuracy, stability and environmental adaptability of lithium iron phosphate battery SOC estimation under complex operating conditions of hybrid vehicles.

[0197] Reference Figure 1The SOC vehicle-cloud collaborative estimation method of this invention consists of real-time vehicle-side estimation, reliable segment parameter updates, and cloud-based geographic environment perception correction. The vehicle-side acquires operational data such as battery, motor, engine, braking system, and vehicle location, identifies comprehensive operating conditions, and completes power coupling verification between the vehicle's power side and battery side, SOC recursive current fusion, dynamic interference evaluation, and reliable calibration segment selection. The cloud obtains geographic environment information such as ambient temperature, weather conditions, road slope, road surface condition, and altitude based on the vehicle's location, corrects the vehicle-side parameters, and transmits the corrections back. The vehicle-side smoothly absorbs the cloud-based corrections and outputs the final SOC estimation result.

[0198] Reference Figure 2 The vehicle-side system identifies the hybrid power system's operating mode, driving behavior, and power response status based on data such as engine output power, electric motor mechanical output power, power generated by the engine and sent to the battery, regenerative braking power, battery current, vehicle speed, vehicle acceleration, accelerator pedal opening, and braking signal, and generates comprehensive operating conditions.

[0199] Reference Figure 3 The vehicle-side establishes the power coupling relationship between the vehicle power side and the battery side based on the comprehensive operating conditions, calculates the power coupling error between the actual power of the battery side and the estimated power of the battery side on the vehicle power side, and determines the correction weight of the equivalent current of the vehicle power side participating in the SOC recursion based on the power coupling error and the comprehensive operating conditions, thus obtaining the fused SOC recursive current.

[0200] Reference Figure 4 The vehicle-side divides continuous operating data into segments and selects reliable calibration segments based on dynamic interference indicators, segment reliability, SOC reference value reliability, and SOC reference value change range. Only reliable calibration segments are used to update the battery effective capacity and current zero bias, avoiding the impact of strong disturbance segments such as rapid acceleration, low-speed start-stop, power mode switching, and regenerative braking on the parameter update results.

[0201] Reference Figure 5 The cloud generates geographic environment perception correction parameters based on information such as the vehicle's current location, ambient temperature, weather conditions, road slope, road surface condition, and altitude. The vehicle smoothly absorbs these correction parameters according to the cloud's parameter update coefficients to obtain the final effective capacity used for SOC recursion. Zero bias current and charge / discharge efficiency .

[0202] To further illustrate the technical solution of the present invention, a specific embodiment is given below.

[0203] Example 1

[0204] (1) Acquisition of multi-source operation data from vehicle and cloud

[0205] This embodiment focuses on hybrid vehicles equipped with lithium iron phosphate batteries. The vehicle-side sampling period is:

[0206]

[0207] The battery discharge direction is defined as the positive direction, and the charging direction is defined as the negative direction. The basic effective capacity is determined based on the vehicle's factory capacity calibration results and the effective capacity saved from the previous parameter update cycle. In this embodiment:

[0208]

[0209] The vehicle is currently traveling on a gentle slope in the urban area. The vehicle's location, obtained by the onboard positioning module and verified through cloud-based map and weather services, indicates an ambient temperature of 5°C and a road gradient of approximately [missing information]. The vehicle's location is at an altitude of approximately The weather is sunny and the road surface is dry.

[0210] In the At each sampling time, the vehicle operation data collected by the vehicle terminal is as follows: Battery terminal voltage is

[0211]

[0212] Battery current is

[0213]

[0214] The battery temperature is 8℃, and the vehicle speed is... The vehicle acceleration is The accelerator pedal opening is The braking signal is invalid. Engine output power is zero, and the electric motor's mechanical output power is...

[0215]

[0216] The power of the accessory load is

[0217]

[0218] The actual power entering the battery side in both the engine power generation path and the regenerative braking path is zero, that is:

[0219]

[0220]

[0221] In this embodiment, the battery terminal voltage Battery current Battery temperature is collected by the battery management system; engine power, motor mechanical output power, accessory load power and braking energy recovery power are collected by the vehicle controller, motor controller, engine controller and braking control system; vehicle location, ambient temperature, weather conditions, road slope, altitude and road surface conditions are obtained by the on-board positioning module, vehicle networking terminal and cloud services.

[0222] Among them, vehicle-side battery data and powertrain system data are used to calculate the actual battery power, estimate the battery power on the vehicle's power side, calculate power coupling error, and calculate the fused SOC recursive current; cloud-based geographic environment data and data up to the [number missing] [item missing] Historical operating data at each sampling time point is used to correct effective capacity, current zero bias, charge / discharge efficiency, and related threshold and weight parameters. The following values ​​are only used to illustrate the implementation process of this invention and do not constitute a limitation on the scope of protection of this invention.

[0223] (2) Comprehensive operating condition identification

[0224] Based on the fact that the engine is not currently engaged, the mechanical output power of the electric motor is greater than the electric motor drive threshold, and the battery current is in the discharge direction, the vehicle is determined to be in electric drive dominant mode.

[0225] Meanwhile, the vehicle acceleration is small, the accelerator pedal opening changes smoothly, the braking signal is invalid, the battery current fluctuation is small, and no significant sudden changes in battery current, motor mechanical output power, engine output power, engine power generation entering the battery side, or braking energy recovery power are detected. Therefore, the current driving behavior and power response state are determined to be a stable cruise state.

[0226] Therefore, we obtain the first... Overall operating conditions at each sampling time:

[0227]

[0228] in, It is an electric drive-dominated mode. To maintain a stable cruise state.

[0229] (3) Verification of power coupling between vehicle power side and battery side

[0230] In this step, the battery management system provides the battery terminal voltage. and battery current The vehicle controller and related controllers provide the mechanical output power of the motor. Attachment load power Power generated by the engine and fed into the battery and the power of regenerative braking entering the battery side .

[0231] To maintain consistency in calculation units, power will be uniformly calculated using watts, i.e.:

[0232]

[0233]

[0234] Calculate the actual power at the battery terminals based on the battery terminal voltage and battery current:

[0235]

[0236] Substituting the current data, we get:

[0237]

[0238] Right now:

[0239]

[0240] According to the vehicle motor efficiency calibration table, at the operating point corresponding to the current motor speed and output torque, the motor drive efficiency is: Based on the conversion relationship between motor mechanical output power and battery-side power in Table 3, the motor power conversion factor is taken as follows:

[0241]

[0242] The vehicle is currently in electric drive-dominated mode, with zero power output from the engine to the battery and zero power output from regenerative braking. Therefore, the estimated battery-side power output from the vehicle's powertrain side is:

[0243]

[0244] Substituting the current data, we get:

[0245]

[0246] Right now:

[0247]

[0248] Further calculations were performed to determine the power coupling error between the actual power at the battery end and the estimated battery-side power on the vehicle's powertrain side:

[0249]

[0250] Substituting the current data, we get:

[0251]

[0252] Right now:

[0253]

[0254] Based on the power coupling error threshold calibration parameters corresponding to the electric drive-dominated mode and stable cruise state, determine:

[0255]

[0256] because:

[0257]

[0258] This indicates that the estimated battery power on the vehicle's power side is highly consistent with the actual battery power, allowing the equivalent current on the vehicle's power side to participate in the subsequent SOC recursive current fusion.

[0259] In this calculation process, the battery terminal voltage and battery current are used to calculate the actual power at the battery terminal. The mechanical output power of the electric motor and the power of the accessory load are used to calculate the estimated battery-side power of the vehicle's powertrain. The power coupling error calculated by the two methods This is used to determine the correction weights for the equivalent current on the vehicle's power side in the next step. The State of Charge (SOC) is not directly output in this step.

[0260] (4) SOC recursive current fusion

[0261] In this step, the overall operating conditions are considered. The fundamental correction weights used to determine the equivalent current on the vehicle's power side, and the power coupling error. It is used to reflect the reliability of the vehicle's power estimation results.

[0262] First, the estimated battery power on the vehicle's power side is converted into the vehicle's equivalent battery current on the power side:

[0263]

[0264] in, To prevent the division by tiny positive numbers with a denominator of zero, in this embodiment, Compared to This can be ignored; substituting the current data, we get:

[0265]

[0266] Based on the basic correction weight calibration table corresponding to the electric drive-dominated mode and stable cruise state, the following is determined:

[0267]

[0268]

[0269] The comprehensive working condition correction factor is:

[0270]

[0271] Substituting the current data, we get:

[0272]

[0273]

[0274] The measured battery current is combined with the equivalent battery current on the vehicle's power side to obtain the SOC recursive current:

[0275]

[0276] Substituting the current data, we get:

[0277]

[0278]

[0279] Therefore, the SOC recursive current is integrated. It simultaneously includes measured current information from the battery and vehicle powertrain information verified by power coupling, representing the SOC recursive current derived from the fusion of multi-source data from the vehicle. It does not represent the SOC estimation result alone, but serves as the current input for subsequent final SOC recursion.

[0280] (5) Dynamic interference evaluation and screening of credible candidate fragments

[0281] In the Before the final SOC calculation at the sampling time, the vehicle-side data is based on the data up to the sampling time. The historical continuous data collected at each sampling time point is used to divide the data into segments and filter reliable candidate segments. The following uses one of the durations as... The process of evaluating dynamic interference is illustrated using historical operational segments as an example.

[0282] The representative comprehensive operating condition for this segment is electric drive-dominated—stable cruise mode, corresponding to geographical environmental conditions including low temperature, gentle slope, sunny weather, dry road surface, and approximately... Altitude. The average battery current within this segment is:

[0283]

[0284] The vehicle-side local coulomb efficiency is:

[0285]

[0286] Based on the dynamic interference characteristics such as the degree of hysteresis, polarization, current fluctuation, temperature fluctuation, and power coupling error fluctuation, and combined with the dynamic interference characteristic weights under the corresponding comprehensive operating conditions and geographical environment of this segment, the dynamic interference index is calculated as follows:

[0287]

[0288] Based on the dynamic interference threshold calibration parameters corresponding to the current comprehensive operating conditions and geographical environment, the following is determined:

[0289]

[0290] because:

[0291]

[0292] Therefore, this running segment was identified as a reliable candidate segment.

[0293] Furthermore, let the credibility penalty coefficient be:

[0294]

[0295] The credibility of the fragment is:

[0296]

[0297] Substituting the current data, we get:

[0298]

[0299] Let the credibility threshold of the segment be:

[0300]

[0301] because:

[0302]

[0303] Therefore, this running segment meets the credibility requirements of a credible candidate segment.

[0304] (6) Parameter update of trusted calibration fragment

[0305] For the aforementioned reliable candidate segments, the reliability of the SOC reference values ​​at the segment start and end points is further determined.

[0306] In this embodiment, the stable voltage recovery information of the lithium iron phosphate battery is not directly used as the basis for the current SOC output, but is used to determine the reliability of the SOC reference value. When the local slope of the lithium iron phosphate battery OCV-SOC calibration curve corresponding to the stable voltage meets the preset confidence condition, the stable voltage can be used as one of the SOC reference information; when it is in the voltage plateau region and the local slope is insufficient, the SOC is not inferred solely from the stable voltage, but the SOC reference value is determined by combining the historical high-confidence SOC results in the cloud, the full charge calibration state, the low SOC boundary calibration state, the charging cut-off state, or the confidence calibration state output by the battery management system.

[0307] In this embodiment, both the start and end points of the historical operation segment meet the SOC reference value reliability conditions of low dynamic interference, small power coupling error, low current fluctuation, and low voltage change rate. After fusing the vehicle-side stable operation information with the cloud-based historical high-reliability SOC results, the SOC reference values ​​for the start and end points of this segment are obtained as follows:

[0308]

[0309]

[0310] The change in the SOC reference value for this segment is:

[0311]

[0312] Let the minimum SOC span threshold be:

[0313]

[0314] because:

[0315]

[0316] Therefore, this segment was identified as a reliable calibration segment.

[0317] In this embodiment, the effective capacity and current zero bias are updated using an alternating estimation method. First, the effective capacity is estimated based on the current current zero bias, and then the current zero bias is estimated based on the updated effective capacity. The iteration stops when the changes in effective capacity and current zero bias obtained from two consecutive updates are both less than a preset convergence threshold.

[0318] After iterative convergence, the zero-bias estimation result of the vehicle-end local current is obtained:

[0319]

[0320] The candidate effective capacity of the first reliable calibration segment is calculated using the converged vehicle-end local current with zero bias:

[0321]

[0322] Substituting into the calculation, we get:

[0323]

[0324] Two other reliable calibration fragments were obtained using the same method, with corresponding candidate effective capacity and fragment reliability as follows:

[0325]

[0326]

[0327] The online effective capacity is obtained by weighted fusion based on the segment credibility of multiple trusted calibration segments:

[0328]

[0329] Substituting the current data, we get:

[0330]

[0331]

[0332] The capacity update coefficients are determined based on the number of reliable calibration fragments, average reliability, SOC span, and the dispersion of candidate effective capacity:

[0333]

[0334] The updated local effective capacity on the vehicle side is:

[0335]

[0336] Substituting the current data, we get:

[0337]

[0338]

[0339] Therefore, the cloud-based correction uses the local parameters adopted by the front-end vehicle:

[0340]

[0341]

[0342] The vehicle-side local charging and discharging efficiency is determined based on the battery temperature, current direction, current rate, and efficiency calibration parameters corresponding to the current comprehensive operating conditions.

[0343]

[0344] The vehicle-side local effective capacity, current zero bias, and charging / discharging efficiency are used for subsequent cloud-based geographic environment perception parameter correction.

[0345] (7) Correction of cloud-based geographic environment perception parameters

[0346] The cloud does not directly calculate or output the first... Instead of SOC at each sampling time, it is a set of geographic environment states. and up to the Using historical operational data from each sampling moment as input, the effective local capacity of the vehicle terminal is assessed. Zero bias of local current at vehicle end Vehicle-side local charging and discharging efficiency Perform environmental perception residual correction.

[0347] The vehicle uploads its current location, overall operating conditions, power coupling error, dynamic interference index, segment reliability, effective capacity estimation results, and local current zero-bias estimation results to the cloud. The cloud then obtains ambient temperature, weather conditions, road gradient, road surface condition, and altitude based on the current location.

[0348] In this embodiment, an ambient temperature of 5°C is used to select the capacity and efficiency residual correction parameters under low-temperature conditions; road slope and altitude approximately Used to select power coupling error threshold correction partitions related to slope and elevation changes; sunny and dry road conditions indicate that the reliability of brake recovery segments is downweighted under rain, snow or slippery road conditions.

[0349] Since the vehicle is currently in a low-temperature, gentle-slope environment, the cloud-based system uses historical SOC estimation residuals from similar temperatures, slopes, and overall operating conditions to correct the vehicle's local parameters for environmental awareness. It's important to note that the cloud-based correction of the vehicle's effective local capacity is not a direct addition of the full capacity reduction ratio corresponding to the ambient temperature. Instead, it uses a correction coefficient determined based on historical estimation residuals from similar operating conditions and environments, relative to the vehicle's local estimation results, to avoid redundant corrections for environmental impacts such as low temperatures.

[0350] In this embodiment, the effective capacity environmental correction coefficient, the charge / discharge efficiency environmental correction coefficient, and the current zero bias correction offset are determined by the cloud as follows:

[0351]

[0352]

[0353]

[0354] This yields the cloud-based correction parameters:

[0355]

[0356]

[0357]

[0358]

[0359]

[0360]

[0361] The vehicle positioning accuracy is high and the road slope data is complete, but the weather data update time is relatively long. Therefore, the reliability of the cloud-based geographic environment information is rated as medium, and the cloud parameter update coefficient is determined as follows:

[0362]

[0363] The vehicle-side smoothly absorbs the cloud-based correction parameters to obtain the final result used for the first... Effective capacity of SOC recursion at each sampling time:

[0364]

[0365] Substituting the current data, we get:

[0366]

[0367]

[0368] The final current with zero bias is:

[0369]

[0370]

[0371]

[0372] The final charge / discharge efficiency is:

[0373]

[0374]

[0375]

[0376] This yields the result used for the first The vehicle-side smooth fusion parameters derived from the final SOC at each sampling time:

[0377]

[0378]

[0379]

[0380] (8) Final SOC recursion and strong disturbance protection

[0381] The endpoint of the aforementioned trusted calibration segment corresponds to the first At the sampling time, the SOC reference value obtained through reliability determination is used as the first sampling time. The initial value for the final SOC recursion at each sampling time:

[0382]

[0383] Therefore, the parameters required for the final SOC recursion have a clear source: the fused SOC recursion current. Effective capacity is obtained by comprehensively identifying operating conditions, verifying power coupling, and fusing current data from vehicle-side battery voltage, battery current, and power system power. Zero bias current and charge / discharge efficiency It is obtained by combining the vehicle-side trusted calibration results with cloud-based geographic environment data and historical operation data for correction.

[0384] In this embodiment, the current sign convention defines the battery discharge direction as positive. Because:

[0385]

[0386] This indicates that the measured battery current has a positive zero bias, meaning the measured discharge current is larger than the actual discharge current. Therefore, the following approach is used in the final SOC calculation:

[0387]

[0388] Zero bias correction is applied to the recursive current.

[0389] According to the final SOC recursive formula:

[0390]

[0391] Substituting the current data, we get:

[0392]

[0393]

[0394] Therefore, the first The final SOC estimate for each sampling time point is approximately:

[0395]

[0396] During subsequent operation, if the vehicle enters a state of rapid acceleration response, low-speed start-stop, power mode switching, or strong disturbance state due to regenerative braking, the comprehensive operating condition correction coefficient will be reduced. When the power coupling error increases significantly or the power mode changes rapidly, the following can be set:

[0397]

[0398] At this point, the final SOC calculation mainly relies on the measured battery current after current zero-bias correction. Simultaneously, the corresponding running segments are marked as strong interference segments. These strong interference segments can participate in the final SOC calculation, but they do not participate in subsequent effective capacity updates, current zero-bias estimation, and model parameter identification.

[0399] Through the above implementation process, the vehicle-side utilizes battery data, power system power data, and vehicle operating status data to form a fused SOC recursive current. The cloud utilizes vehicle location, geographic environment data, and historical operational data to generate a corrected effective capacity. Zero bias current and charge / discharge efficiency The vehicle-side system substitutes all the above parameters into the final SOC recursive formula and outputs a unique final SOC estimate. This enables vehicle-cloud collaborative estimation of the state of charge (SOC) of lithium iron phosphate power batteries under complex operating conditions and in different geographical environments for hybrid vehicles.

[0400] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A vehicle-cloud collaborative estimation method for the state of charge (SOC) of lithium iron phosphate batteries in hybrid vehicles, characterized in that, Includes the following steps: Step 1: The vehicle acquires vehicle operation data, and the cloud acquires external environment data corresponding to the vehicle's current location; Step 2: The vehicle-side identifies the comprehensive operating conditions, including the hybrid power system's operating mode, driving behavior, and power response status, based on the vehicle's operating data. Step 3: The vehicle selects the corresponding power coupling relationship according to the hybrid system's operating mode, estimates the battery-side power corresponding to the vehicle's power side, and performs consistency verification with the actual power at the battery side to obtain the power coupling error. Step 4: The vehicle side determines the correction weight of the equivalent current on the vehicle power side based on the comprehensive operating conditions and power coupling error, and merges the measured current of the battery with the equivalent current on the vehicle power side to obtain the SOC recursive current. Step 5: The vehicle-side continuously running data is divided into multiple segments, a dynamic interference index is constructed, and a reliable candidate segment is selected based on this index. The segment that meets the reliability requirements and the SOC reference value reliability requirements is determined as a reliable calibration segment. The battery effective capacity and current zero bias are updated based on the reliable calibration segment. The vehicle's continuous operation data is divided into several operation segments according to time windows, SOC change intervals, operating mode changes, or driving behavior changes. For each operation segment, a dynamic interference index is constructed by weighting and fusing one or more of the following factors: hysteresis effect, polarization effect, current fluctuation, temperature fluctuation, insufficient voltage recovery, power coupling error fluctuation, road slope change effect, and road surface slipperiness. The weights of each feature are adjusted according to the overall operating conditions and geographical environment. When the dynamic interference index is less than or equal to the dynamic interference threshold, the corresponding running segment is marked as a reliable candidate segment; when the dynamic interference index is greater than the dynamic interference threshold, it is marked as a strong interference segment, and the strong interference segment does not participate in the effective capacity update and current zero bias estimation; the segment reliability is determined by the dynamic interference index, and the larger the dynamic interference index, the lower the segment reliability. The SOC reference value is obtained from a reference source that meets preset reliability conditions. The reference source includes at least one of the following: the SOC reference value obtained from the OCV-SOC calibration curve after the vehicle is stationary; or the SOC reference value determined by combining the vehicle's historical SOC with the cloud's historical high-reliability SOC reference information during a stable operation segment. SOC anchor point obtained under charging cutoff, full charge calibration, or low SOC boundary calibration state; The SOC reference value is obtained by backtracking and fusing historical operation data in the cloud. When the start and end points of the segment both meet the requirements of low dynamic interference, small power coupling error, low voltage change rate, and SOC span not less than the preset threshold, the SOC reference value is judged to be reliable. Step 6: The cloud obtains the geographical environment status based on the vehicle's current location, combines the comprehensive operating conditions and historical operating data to generate cloud correction parameters and send them back to the vehicle. The vehicle smoothly absorbs the cloud correction parameters. Step 7: Based on the SOC recursive current, the corrected effective capacity, current zero bias, and charge / discharge efficiency, the vehicle outputs the final SOC estimate.

2. The hybrid vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method according to claim 1, characterized in that, In step two, the hybrid power system operating modes include at least electric drive-dominated mode, engine mechanical drive mode, electromechanical coupling drive mode, engine power generation supplementation mode, and regenerative braking mode; the driving behavior and power response states include at least rapid acceleration response state, stable cruise state, low-speed start-stop state, power mode switching state, and regenerative braking strong disturbance state; each operating mode is determined based on engine output power, motor mechanical output power, braking signal, battery current, and related thresholds; each driving behavior and power response state is determined based on vehicle acceleration, accelerator pedal opening, vehicle speed, battery current fluctuation, power change, and related thresholds; the above threshold parameters are updated based on the vehicle's powertrain structure, vehicle control strategy, battery aging status, ambient temperature, road slope, and road surface conditions.

3. The hybrid vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method according to claim 1, characterized in that, In step three, segmented power coupling relationships are established according to different hybrid power system operating modes. The estimated battery power on the vehicle's power side is determined by the power balance relationship of the motor mechanical output power, accessory load power, engine power generation entering the battery side, and braking energy recovery entering the battery side under different operating modes. The power coupling error is the absolute value of the difference between the actual power at the battery end and the estimated battery power on the vehicle's power side. When the power coupling error is less than or equal to the power coupling error threshold related to the comprehensive operating conditions, the consistency is high. When the power coupling error is greater than the power coupling error threshold, a large deviation is judged, and the participation weight of the equivalent current on the vehicle's power side is reduced in the subsequent SOC recursion process. The power coupling error threshold is determined based on the hybrid power system operating mode, driving behavior and power response status, battery temperature, road slope, road surface condition, vehicle historical operating data, and cloud-based geographic environment perception correction parameters.

4. The hybrid vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method according to claim 1, characterized in that, In step four, the equivalent current on the vehicle's power side is obtained by dividing the estimated battery power on the vehicle's power side by the battery terminal voltage. The comprehensive operating condition correction coefficient is jointly determined by the basic correction weight corresponding to the hybrid system's operating mode, the basic correction weight corresponding to driving behavior and power response state, and the power coupling error factor, with a value range of 0 to 1. When the vehicle is in a stable cruising state and the power coupling error is less than or close to the power coupling error threshold, the participation weight of the equivalent current on the vehicle's power side is increased. When the vehicle is in a rapid acceleration response state, a low-speed start-stop state, a strong disturbance state due to braking regeneration, or a power mode switching state, the participation weight of the equivalent current on the vehicle's power side is decreased. When the power coupling error is significantly greater than the threshold or the vehicle is in a state of strong disturbance, the comprehensive operating condition correction coefficient is set to zero. The combined SOC recursive current is determined by the weighted sum of the measured battery current and the equivalent current on the vehicle's power side.

5. The hybrid vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method according to claim 1, characterized in that, In step five, the candidate effective capacity is determined based on the ratio of the current integral within the trusted calibration segment to the change in the SOC reference value; the online effective capacity is obtained by weighted fusion based on the segment credibility of multiple trusted calibration segments, and then fused with the basic effective capacity according to the capacity update coefficient to obtain the updated effective capacity; Zero current bias is estimated by minimizing a charge closure error objective function based on multiple reliable calibration segments; The effective capacity and current zero bias are updated using an alternating estimation method. First, the effective capacity is estimated based on the current current zero bias, and then the current zero bias is estimated based on the updated effective capacity, until convergence.

6. The hybrid vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method according to claim 1, characterized in that, In step six, the cloud-generated correction parameters include at least the corrected effective capacity, current zero bias, and charge / discharge efficiency, as well as one or more of the following: power coupling error threshold, dynamic interference threshold, dynamic interference feature weight, credibility penalty coefficient, and capacity update coefficient. The cloud corrects the battery's effective capacity and charge / discharge efficiency based on ambient temperature, corrects the power coupling error threshold based on altitude, corrects the power coupling error threshold and dynamic interference threshold based on road slope changes, and reduces the credibility of the regenerative braking segment based on rain, snow, or slippery road conditions. The vehicle-side smoothly absorbs the cloud-corrected parameters based on the cloud parameter update coefficients, which are determined based on positioning accuracy, environmental data timeliness, and data source reliability.

7. The hybrid vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method according to claim 1, characterized in that, In step seven, the final SOC recursion adopts the ampere-hour integral form, the recursive current adopts the fused SOC recursive current, and the effective capacity, current zero bias and charging and discharging efficiency adopt the values ​​updated by the vehicle-side trusted calibration segment and corrected by cloud-based geographical environment perception. In the strong interference segment, the correction weight of the vehicle power side equivalent current is reduced or turned off and the segment is prohibited from participating in parameter updates. In the trusted calibration segment, the segment is allowed to participate in effective capacity updates, current zero bias estimation and model parameter identification.

8. The hybrid vehicle lithium iron phosphate battery SOC vehicle-cloud collaborative estimation method according to claim 1, characterized in that, The cloud platform also combines long-term vehicle operation data from multiple days or cycles to retrospectively analyze the SOC estimation residuals, power coupling errors, dynamic interference indicators, segment reliability, effective capacity estimation results, and current zero bias estimation results under different geographical regions, altitudes, weather conditions, and road slopes, continuously optimizing individual vehicle parameters and environmental correction rules.

Images