Battery soc estimation method, device, and engineering machine, medium, and product

Abstract

The present disclosure provides a battery SOC estimation method, device, engineering machinery, medium and product. The battery SOC estimation method comprises: obtaining the working parameters of the battery in the engineering machinery, wherein the working parameters comprise the current value, voltage value and temperature value of the battery; determining the current working condition mode of the engineering machinery according to the working parameters; processing the working parameters by using multiple estimators respectively to obtain multiple candidate SOC values, multiple initial reliabilities corresponding to the multiple candidate SOC values, and an SOC reference value of the battery; determining a global reference reliability according to the current working condition mode and; calibrating the initial reliability of each candidate SOC value by using the global reference reliability to obtain the comprehensive reliability of each candidate SOC value; and determining a target SOC value according to the multiple candidate SOC values and the multiple comprehensive reliabilities corresponding to the multiple candidate SOC values.

Description

Technical Field

[0001] This disclosure relates to the fields of battery management and engineering machinery technology, and in particular to a battery SOC estimation method, apparatus, engineering machinery, medium, and product. Background Technology

[0002] Currently, the construction machinery industry is accelerating its transformation from traditional fuel-powered systems to electrification and intelligentization. This transformation makes the power battery system the core energy carrier for hybrid and pure electric construction machinery. The Battery Management System (BMS) is the core control unit for the safe and efficient operation of the power battery system. Among the many state parameters managed by the BMS, the State of Charge (SOC) directly represents the remaining usable charge of the battery and is the primary indicator for measuring the battery's current energy level. Therefore, accurate SOC estimation is the foundation for the battery management system to perform power limiting, energy optimization, and lifespan management. Summary of the Invention

[0003] The inventors noted that in related technologies, battery SOC estimation is mainly based on the ampere-hour integration method, open circuit voltage (OCV) calibration method, extended Kalman filtering method, and equivalent circuit model method. However, due to the significant unique characteristics of the operating environment and conditions of hybrid engineering machinery, calibration opportunities are scarce, current fluctuations are severe, and operating modes are frequently switched. Under these circumstances, traditional SOC estimation methods are prone to problems such as large cumulative errors, high risk of miscalibration, and model divergence, making it difficult to meet the needs of practical applications.

[0004] In view of this, this disclosure proposes a battery SOC estimation method, device, engineering machinery, medium, and product, which can estimate the battery SOC using a multimodal estimator based on high-precision data acquisition and calibrate it by combining multidimensional reliability, thereby effectively improving the accuracy and reliability of battery SOC estimation for engineering machinery under various operating conditions.

[0005] According to a first aspect of this disclosure, a battery SOC estimation method is proposed, comprising: acquiring the operating parameters of a battery in construction machinery, wherein the operating parameters include the battery's current value, voltage value, and temperature value; determining the current operating mode of the construction machinery based on the operating parameters; and processing the operating parameters using multiple estimators to obtain multiple candidate SOC values ​​and multiple initial confidence levels corresponding one-to-one with the multiple candidate SOC values, as well as a battery SOC reference value. Based on the current operating conditions and The global baseline credibility is determined; the initial credibility of each candidate SOC value is calibrated using the global baseline credibility to obtain the comprehensive credibility of each candidate SOC value; and the target SOC value is determined based on the multiple candidate SOC values ​​and the multiple comprehensive credibility corresponding to the multiple candidate SOC values.

[0006] In some embodiments, processing the operating parameters using multiple estimators includes: using a first estimator based on an adaptive filtering algorithm to process the operating parameters to obtain a first candidate SOC value and a corresponding first initial confidence level; and using a second estimator based on an ampere-hour integration algorithm to process the operating parameters to obtain... The second estimator, based on the dynamic fragment correction algorithm, processes the working parameters to obtain the second candidate SOC value and corresponding second confidence level associated with the low-end region, and the third candidate SOC value and corresponding third confidence level associated with the high-end region. The third estimator, based on the feature state observation algorithm, processes the working parameters to obtain the fourth candidate SOC value and corresponding fourth confidence level associated with the boundary region, and the fifth candidate SOC value and corresponding fifth confidence level associated with the platform region.

[0007] In some embodiments, obtaining the second candidate SOC value associated with the low-end region and the corresponding second confidence level includes: determining that the battery is currently in a quasi-steady-state segment when the duration of the current value being less than a preset current threshold is greater than a preset duration threshold and the voltage change rate is less than a preset change rate threshold; and within the quasi-steady-state segment, determining the second candidate SOC value and the corresponding second confidence level based on the average current value, the average voltage value, the current temperature value, and the first candidate SOC value.

[0008] In some embodiments, obtaining a third candidate SOC value and a corresponding third confidence level associated with the high-end region includes: determining a third candidate SOC value and a corresponding third confidence level based on the charging rate, the current temperature value, and the current voltage value when the battery is in a charging state and the current current value and the current temperature value are within a predetermined current range and a predetermined temperature range, respectively.

[0009] In some embodiments, obtaining the fourth candidate SOC value and the corresponding fourth confidence level associated with the boundary region includes: When the engineering machinery is in a stable mode and is located in the boundary region, the fourth candidate SOC value and the corresponding fourth confidence level are determined based on the current voltage value.

[0010] In some embodiments, obtaining the fifth candidate SOC value and the corresponding fifth confidence level associated with the platform region includes: If the battery is in a platform region and the duration of the engineering machinery in stable mode exceeds a predetermined duration threshold, it is determined that the battery is currently in an ultra-long stable segment. Within the ultra-long stable segment, the fifth candidate SOC value and the corresponding fifth confidence level are determined based on the average current value, the average voltage value, and the current temperature value.

[0011] In some embodiments, determining the target SOC value based on multiple candidate SOC values ​​and multiple comprehensive confidence levels corresponding to the multiple candidate SOC values ​​includes: determining the weight of each candidate SOC value based on the multiple comprehensive confidence levels; and calculating the weighted sum of the multiple candidate SOC values ​​based on the weight of each candidate SOC value to obtain the target SOC value.

[0012] In some embodiments, determining the weight of each candidate SOC value based on multiple comprehensive confidence levels includes: detecting whether the second to fifth candidate SOC values ​​and their corresponding comprehensive confidence levels meet predetermined conditions; and, if the second to fifth candidate SOC values ​​and their corresponding comprehensive confidence levels meet predetermined conditions, determining the weight of each candidate SOC value based on multiple comprehensive confidence levels.

[0013] In some embodiments, detecting whether the second candidate SOC values ​​to the fifth candidate SOC values ​​and their corresponding overall credibility meet predetermined conditions includes: detecting whether the overall credibility corresponding to the second candidate SOC values ​​to the fifth candidate SOC values ​​is lower than the global benchmark credibility; if the overall credibility corresponding to the second candidate SOC values ​​to the fifth candidate SOC values ​​is lower than the global benchmark credibility, then according to The target SOC value is determined by the first candidate SOC value; if the overall credibility of any candidate SOC value from the second to the fifth candidate SOC value is not lower than the global benchmark credibility, then it is checked whether the second to the fifth candidate SOC values ​​and the corresponding overall credibility meet the predetermined conditions.

[0014] In some embodiments, calibrating the initial confidence level of each candidate SOC value includes: determining the support level corresponding to each candidate SOC value based on the current operating condition mode; determining the risk value corresponding to each candidate SOC value based on the stability of the current operating condition mode and operating parameters; and determining the risk value corresponding to each candidate SOC value based on the relationship between the candidate SOC value and the operating parameters. The absolute difference and The voltage sensitivity of the region is used to determine the potential benefit value corresponding to each candidate SOC value; based on the global baseline confidence, initial confidence, support, risk value and potential benefit value, the comprehensive confidence value corresponding to each candidate SOC value is determined.

[0015] In some embodiments, determining the current operating mode of the construction machinery based on operating parameters includes: determining that the construction machinery is currently in a stable mode when the current change rate is less than a first current change rate threshold and the current fluctuation intensity is less than a first current fluctuation intensity threshold; determining that the construction machinery is currently in a fluctuating mode when the current change rate is not less than the first current change rate threshold and less than a second current change rate threshold, and the current fluctuation intensity is not less than the first current fluctuation intensity threshold and less than the second current fluctuation intensity threshold, wherein the first current change rate threshold is less than the second current change rate threshold and the first current fluctuation intensity threshold is less than the second current fluctuation intensity threshold; and determining that the construction machinery is currently in a severe mode when the current change rate is not less than the second current change rate threshold and the current fluctuation intensity is not less than the second current fluctuation intensity threshold.

[0016] In some embodiments, when the current operating mode is a stable mode or a fluctuating mode, the current sampling value provided by the first sensor is used as the battery current value; and when the current operating mode is a severe mode, the battery current value is determined based on the current sampling value provided by the first sensor and the current sampling value provided by the second sensor.

[0017] In some embodiments, when the current operating mode is severe mode, the battery current value is a weighted sum of the current sample value provided by the first sensor and the current sample value provided by the second sensor.

[0018] In some embodiments, a dynamic current error compensation factor is determined based on the current current change rate and the current absolute value of the battery; the product of the dynamic current error compensation factor and the weighted sum is calculated to obtain the compensated current value.

[0019] In some embodiments, when the global benchmark confidence level is consistently below a preset lower limit threshold and it is detected that the construction machinery is about to enter a low-load phase, the battery's usage range is expanded so that the battery can enter the boundary region. During the process of expanding the usage range, if the risk value corresponding to the candidate SOC value increases, the expansion of the usage range is stopped or the original usage range is restored. If the battery completes calibration in the boundary region, the usage range is shrunk to a safe range.

[0020] According to a second aspect of this disclosure, a battery SOC estimation apparatus is provided, comprising: a memory configured to store instructions; and a processor configured to execute instructions, causing the battery SOC estimation apparatus to perform the battery SOC estimation method as described above.

[0021] According to a third aspect of this disclosure, an engineering machine is proposed, including the battery SOC estimation device as described above.

[0022] According to a fourth aspect of this disclosure, a computer-readable storage medium is proposed that stores computer program instructions thereon, which, when executed by a processor, implement the battery SOC estimation method as described above.

[0023] According to a fifth aspect of this disclosure, a computer program product is proposed, comprising a computer program that, when executed by a processor, implements the battery SOC estimation method as described above.

[0024] Other features and advantages of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic flowchart of a battery SOC estimation method according to some embodiments of the present disclosure; Figure 2 This is a schematic flowchart of dynamic current error compensation according to some embodiments of the present disclosure; Figure 3 This is a schematic diagram of the structure of several estimators according to some embodiments of the present disclosure; Figure 4 This is a schematic flowchart illustrating the dynamic correction of the low-end region according to some embodiments of this disclosure; Figure 5 This is a schematic diagram of the initial confidence calibration process according to some embodiments of this disclosure; Figure 6 This is a schematic diagram of the process for calibrating candidate SOC values ​​according to some embodiments of this disclosure; Figure 7 This is a flowchart illustrating the expansion of battery usage range according to some embodiments of this disclosure; Figure 8 This is a schematic diagram of the structure of a battery SOC estimation device according to some embodiments of the present disclosure. Detailed Implementation

[0027] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this disclosure or its application or use. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0028] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of this disclosure.

[0029] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0030] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0031] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0032] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0033] In related technologies, there are three main methods for estimating battery SOC. Method 1: Limiting the battery SOC to an intermediate plateau region during normal use (e.g., arrive When the cumulative throughput reaches a threshold (e.g., 10C-15C), the SOC usage range is opened to the high-end and low-end regions for OCV calibration. After calibration, it is restored to the original usage range. Method 2: During the plateau period of lithium iron phosphate batteries, the first and second correction values ​​are obtained by looking up OCV and dynamic internal resistance tables respectively. These are then weighted and fused with the ampere-hour integral estimate. The weights are dynamically adjusted according to whether the battery SOC is in the plateau period. Method 3: In the low-end region, the DC internal resistance (DCR) compensation is calculated using short-time steady-state data to obtain OCV for table lookup calibration. In the high-end charging region, calibration is performed under full-charge conditions using a mapping table constructed from charging rate, temperature, and terminal voltage.

[0034] The inventors noted the following shortcomings in the related technologies: In Method 1, the cumulative throughput is used as the sole calibration trigger condition, resulting in a single decision dimension and failing to comprehensively consider information such as the dynamics of the working conditions of the engineering machinery, the reliability of current measurement, and model adaptability, leading to inaccurate and untimely SOC estimation; In Method 2, the weight adjustment relies on the battery plateau period and fails to consider the problem of decreased current sampling accuracy under extreme conditions, resulting in low accuracy of subsequent fusion and correction algorithms; In Method 3, the calibration actions in the high-end and low-end regions lack assessment and protection against the risks of the current working conditions, and direct table lookup under severe conditions is prone to miscalibration.

[0035] In summary, current related technologies generally suffer from problems such as a single calibration trigger dimension, uncontrolled data sampling accuracy, and a lack of systematic error control in calibration actions, resulting in low accuracy and poor reliability of battery SOC estimation.

[0036] In view of this, this disclosure proposes a battery SOC estimation method, device, engineering machinery, medium, and product, which can estimate the battery SOC using a multimodal estimator based on high-precision data acquisition and calibrate it by combining multidimensional reliability, thereby effectively improving the accuracy and reliability of battery SOC estimation for engineering machinery under various operating conditions.

[0037] Figure 1 This is a schematic flowchart of a battery SOC method according to some embodiments of the present disclosure.

[0038] like Figure 1 As shown, the battery SOC method includes step S101. Step S106.

[0039] In step S101, the operating parameters of the battery in the engineering machinery are obtained, including the current value, voltage value and temperature value of the battery.

[0040] In some embodiments, the battery's operating parameters can be obtained through a BMS that includes multiple sensors, an electronic control unit, and a communication network.

[0041] For example, multiple sensors may include current sensors, voltage sensors, and temperature sensors.

[0042] In some embodiments, the current sensor may include a Hall effect current sensor and a shunt current sensor.

[0043] In some embodiments, a 16-bit successive approximation analog-to-digital converter (ADC) can be used for isolated sampling of the voltage. For example, the sampling rate per channel is greater than 1kSPS, and the accuracy is... .

[0044] In some embodiments, the temperature sensor may include negative temperature coefficient thermistors and digital temperature sensors distributed within the battery module. For example, an accuracy of... .

[0045] In some embodiments, the battery's operating parameters can be obtained by reading data stored in the memory.

[0046] For example, memory may include non-volatile memory (NVM), electrically erasable programmable read-only memory (EEPROM), and embedded flash memory.

[0047] In some embodiments, the SOC baseline value, the initial value of the global baseline confidence, the cumulative absolute capacity, historical operating condition records, the state of health (SOH), the model parameter lookup table, and the SOC-OCV curve can also be obtained through the memory.

[0048] In step S102, the current operating mode of the construction machinery is determined based on the working parameters.

[0049] In some embodiments, if the rate of change of current is less than a first rate of change of current threshold and the intensity of current fluctuation is less than a first intensity of current fluctuation threshold, it is determined that the construction machinery is currently in a stable mode.

[0050] In some embodiments, if the rate of change of current is not less than a first rate of change of current threshold and is less than a second rate of change of current threshold, and the intensity of current fluctuation is not less than a first intensity of current fluctuation threshold and is less than a second intensity of current fluctuation threshold, it is determined that the construction machinery is currently in a fluctuation mode, wherein the first rate of change of current threshold is less than the second rate of change of current threshold, and the first intensity of current fluctuation threshold is less than the second intensity of current fluctuation threshold.

[0051] In some embodiments, if the rate of change of current is not less than a second rate of change of current threshold and the intensity of current fluctuation is not less than a second intensity of current fluctuation threshold, it is determined that the construction machinery is currently in a severe mode.

[0052] In some embodiments, the rate of change of current can be calculated using the following formula (1).

[0053] (1)

[0054] In formula (1), The rate of change of current, This is the current sampled current value. This is the previous sampled current value. This represents the sampling time interval.

[0055] In some embodiments, the intensity of current fluctuations can be determined by the standard deviation of the current values ​​within a sliding window. This can be represented by a variable. For example, a sliding window can represent the 100 most recent sampling points.

[0056] In some embodiments, based on a current change rate threshold and current fluctuation intensity threshold The current operating conditions of construction machinery are divided into three modes.

[0057] For example, The first current change rate threshold, The second current change rate threshold, and ; The first current fluctuation intensity threshold, This is the second current fluctuation intensity threshold, and .

[0058] In some embodiments, when and At that time, it is determined that the construction machinery is currently in L1 stable mode. For example, the construction machinery is moving at a constant speed or stopping briefly.

[0059] In some embodiments, when and At that time, it is determined that the construction machinery is currently in L2 fluctuation mode. For example, the construction machinery is performing a general compound action.

[0060] In some embodiments, when and At that time, it is determined that the construction machinery is currently in L3 severe mode. For example, the construction machinery is performing a heavy-load impact or bucket cutting.

[0061] In some embodiments, when the current operating mode is L1 stable mode or L2 fluctuating mode, the current sampling value provided by the first sensor is used as the battery current value.

[0062] In some embodiments, the first sensor may be a main current sensor.

[0063] For example, the first sensor could be a Hall effect current sensor with a range of [missing information]. The accuracy is Response time is lower than The sampling period is 10ms.

[0064] In some embodiments, when the current operating mode is L3 severe mode, the battery current value is determined based on the current sampling value provided by the first sensor and the current sampling value provided by the second sensor.

[0065] In some embodiments, the second sensor may be an auxiliary high-precision current sensor.

[0066] For example, the second sensor could be a shunt current sensor with a range of [missing information]. The accuracy is Response time is lower than The sampling period is 5ms.

[0067] In some embodiments, the fusion strategy of the operating mode of the construction machinery and sensor data is shown in Table 1.

[0068] Table 1. Working conditions and sensor data fusion strategies for construction machinery

[0069] In some embodiments, when the current operating mode is L3 severe mode, the battery current value is a weighted sum of the current sample value provided by the first sensor and the current sample value provided by the second sensor.

[0070] In some embodiments, the instantaneous difference between the current sampling values ​​provided by the first sensor and the second sensor is calculated.

[0071] For example, the instantaneous difference can be calculated using the following formula (2).

[0072] (2)

[0073] In formula (2), The current sampling value provided by the first sensor, The current sampling value provided by the second sensor This is the instantaneous difference between the current sample values.

[0074] In some embodiments, based on the instantaneous difference of current sample values and current current amplitude The current sampling values ​​provided by the first sensor are respectively The current sampling value provided by the second sensor Assign instantaneous weights and .

[0075] For example, instantaneous weights can be determined by looking up a weighting table based on current amplitude and instantaneous difference.

[0076] In some embodiments, the current current amplitude The current sampling value provided by the first sensor can be obtained. or the current sampling value provided by the second sensor. ,or and The average value.

[0077] In some embodiments, the weighted sum of the battery sample values ​​is the fusion current, which can be calculated by the following formula (3).

[0078] (3)

[0079] In formula (3), To merge current, Current sampling value provided for the first sensor Instantaneous weights, Current sampling value provided for the second sensor Instantaneous weights.

[0080] In some embodiments, when the instantaneous difference of the current sample values If the threshold is exceeded continuously, a health check of the sensor can be triggered. For example, the sensor may proactively report a fault or cut off the data stream.

[0081] In some embodiments, the fused current can be subjected to adaptive filtering based on operating conditions to improve the stability and reliability of the output current data.

[0082] For example, in L3 severe mode, a stronger filter can be used to suppress high-frequency noise while ensuring the dynamic response of the sensor.

[0083] Under harsh operating conditions, the polarization effect caused by rapid current changes may lead to "hidden" charge transfer. Since this charge is not fully accounted for by the integrator at the moment of the current change, it will affect the ampere-hour integral output. Significant errors can occur, thus affecting the accuracy of the overall SOC estimation.

[0084] In some embodiments, a dynamic current error compensation factor can be determined based on the current current change rate and the current absolute value of the battery to compensate for the instantaneous error in the subsequent calculation of the ampere-hour integral under harsh operating conditions.

[0085] Figure 2 This is a schematic flowchart of dynamic current error compensation according to some embodiments of the present disclosure.

[0086] like Figure 2 As shown, the input region 201 includes the current change rate. and absolute value of current The lookup area 202 includes a two-dimensional lookup table. The output region 203 includes a dynamic current error compensation factor. .

[0087] In some embodiments, The value of is usually greater than 1.

[0088] For example, The value can be 1.0. 1.2.

[0089] In some embodiments, dynamic current error compensation factor Two-dimensional lookup table It can be obtained through calibration using a large amount of dynamic operating condition test data.

[0090] According to the rate of change of current and absolute value of current Determine the dynamic current error compensation factor Two-dimensional lookup table Examples are shown in Table 2.

[0091] Table 2 Dynamic Current Error Compensation Factor Two-dimensional lookup table Example

[0092] In some embodiments, the product of the dynamic current error compensation factor and the weighted sum is calculated to obtain the compensated current value.

[0093] For example, in L3 severe mode, the current value used for ampere-hour integration is instantaneously compensated, and the specific calculation formula is shown in formula (4).

[0094] (4)

[0095] In formula (4), For the compensated current, This is a weighted sum of the current sample values, which is the fused current subsequently used for ampere-hour integration. This is the dynamic current error compensation factor.

[0096] In this way, the current value used for ampere-hour integration can be equivalently compensated, improving the reliability of the input data and thus further improving the accuracy of subsequent SOC estimation.

[0097] In the above embodiments, the current working condition of the construction machinery is divided into three modes according to the battery's operating parameters. In the severe mode, the sampling data of dual sensors is used as the current value, which can improve the anti-interference ability and stability of current sampling under harsh working conditions, and provide a reliable data foundation for subsequent battery SOC estimation, thereby further reducing the risk of miscalibration.

[0098] In step S103, multiple estimators are used to process the working parameters to obtain multiple candidate SOC values ​​and multiple initial confidence levels corresponding one-to-one with the multiple candidate SOC values, as well as the SOC reference value of the battery. .

[0099] Figure 3 This is a schematic diagram of the structure of several estimators according to some embodiments of this disclosure.

[0100] like Figure 3 As shown, the input area 301 includes the battery's operating parameters and the engineering machinery's operating mode; the estimation area 302 includes estimators M1, M2, and M3; and the output area 303 includes multiple candidate SOC values ​​and multiple initial confidence levels corresponding one-to-one with the candidate SOC values, as well as the battery's SOC reference value. .

[0101] In some embodiments, multiple estimators take the aforementioned processed data as input, run in parallel, and each outputs a candidate SOC value and a corresponding initial confidence level.

[0102] In some embodiments, a first estimator is used to process the working parameters based on an adaptive filtering algorithm to obtain a first candidate SOC value and a corresponding first initial confidence level.

[0103] In some embodiments, an equivalent circuit model can be constructed using a first estimator to describe the dynamic behavior of the battery.

[0104] For example, the equivalent circuit model can be a first-order or second-order RC equivalent circuit model.

[0105] In some embodiments, the model parameters of the equivalent circuit model can be determined based on a pre-calibrated lookup table.

[0106] For example, for a second-order RC equivalent circuit model, a pre-calibrated multidimensional lookup table can be used ( T To determine the model parameters (SOC, SOH), R 0, R 1, C 1, R 2, C 2), of which TThis indicates the current temperature value, SOC indicates the current state of charge, and SOH indicates the current health status. R 0 indicates transient response resistance. R 1 represents the resistance parameter of the first RC network. C 1 represents the capacitance parameter of the first RC network. R 2 represents the resistance parameters of the second RC network. C 2 represents the capacitance parameter of the second RC network.

[0107] For example, the state vector of the second-order RC equivalent circuit model is ,in, Indicates the current state of charge. Indicates the first polarization voltage. This represents the second polarization voltage.

[0108] In some embodiments, nonlinear state estimation algorithms can be used to predict, solve, and correct the equivalent circuit model.

[0109] In some embodiments, the nonlinear state estimation algorithm may be an Extended Kalman Filter (EKF) algorithm or an Unscented Kalman Filter (UKF) algorithm.

[0110] In some embodiments, an adaptive strategy can be used to adjust the parameters of the nonlinear state estimation algorithm to adapt to different operating conditions.

[0111] In some embodiments, the process noise covariance matrix Q and the measurement noise covariance matrix R of the nonlinear estimation algorithm can be dynamically adjusted according to the current operating condition.

[0112] For example, in L3 severe mode, Q is increased to actively amplify the uncertainty of model predictions, while R is decreased to improve the confidence in the input high-precision fused current data.

[0113] In some embodiments, the first candidate SOC value is Its corresponding first level of credibility is .

[0114] In some embodiments, the determination can be based on the stability of the filtered innovation, the trace of the estimated covariance matrix, and the current operating mode. .

[0115] For example, in L3 intense mode, the intensity can be appropriately reduced. The baseline value is used to limit the interference of oscillations caused by high model mismatch under severe operating conditions on the SOC estimation.

[0116] In some embodiments, the operating parameters are processed using a second estimator based on an ampere-hour integral algorithm to obtain... .

[0117] In some embodiments, the following formula (5) can be used to calculate the result. .

[0118] (5)

[0119] In formula (5), This is the current SOC reference value for the battery. This is the integral value of ampere-hours from the previous moment. For Coulomb efficiency, For the fusion current value, The sampling period is This refers to the battery's capacity value. This is the battery's state of health coefficient.

[0120] In some embodiments, dynamic current error compensation can be performed on the fused current to correct the impact of rapid current changes caused by sensor oscillation on the ampere-hour integral under harsh operating conditions, thereby improving the accuracy of subsequent SOC estimation.

[0121] For example, the compensated current in formula (4) can be used. Replace fusion current Perform ampere-hour integration.

[0122] In some embodiments, the battery capacity value can be determined based on a pre-calibrated lookup table.

[0123] For example, it can be based on the current temperature value of the battery. T And the health status coefficient (SOH), through a two-dimensional lookup table To obtain the battery capacity value in real time. .

[0124] In some embodiments, the second estimator processes the working parameters based on the dynamic fragment correction algorithm to obtain a second candidate SOC value and a corresponding second confidence level associated with the low-end region, and a third candidate SOC value and a corresponding third confidence level associated with the high-end region.

[0125] In some embodiments, if the duration of the current value being less than a preset current threshold is greater than a preset duration threshold and the voltage change rate is less than a preset change rate threshold, it is determined that the battery is currently in a quasi-steady-state segment.

[0126] In some embodiments, within a quasi-steady-state segment, a second candidate SOC value and a corresponding second confidence level are determined based on the average current value, the average voltage value, the current temperature value, and the first candidate SOC value.

[0127] Figure 4 This is a schematic diagram of the process for dynamic correction of the low-end region according to some embodiments of this disclosure.

[0128] like Figure 4 As shown, the low-end region dynamic correction of the second estimator includes step S401. Step S407.

[0129] In step S401, the real-time current of the battery is continuously monitored. I ,Voltage V and temperature T .

[0130] In step S402, the current is determined. I Is the mutation greater than the threshold? a If the real-time current I The mutation is greater than the threshold a If the real-time current is... I The mutation is not greater than the threshold. a Then return to step S401.

[0131] In step S403, the absolute value of the current is determined. Is it less than the threshold? b Duration t 1 second, and voltage change rate Less than the threshold c If the above conditions are met, proceed to step S404; if the above conditions are not met, return to step S402.

[0132] In step S404, it is determined that the battery is currently in a quasi-steady-state segment, and the average current value is recorded. Average voltage and initial sudden change current value .

[0133] For example, when and At that time, it is determined that the battery is currently in a quasi-steady-state segment.

[0134] In some embodiments, The average value of the current after it has stabilized within the quasi-steady-state segment can be taken. The average value of the voltage after it has stabilized within the quasi-steady-state segment can be taken.

[0135] In some embodiments, You can take the difference between the current value before the transition and the current value after the transition.

[0136] In some embodiments, when a quasi-steady-state segment begins with a sudden change from a larger current to a smaller current, the quasi-steady-state segment is marked as a valid correction window.

[0137] In step S405, according to , , T and the first candidate SOC value The dynamic internal resistance compensation coefficient is obtained by looking up the table. .

[0138] In some embodiments, it can be As an index reference value ,according to , , T and Through a pre-calibrated four-dimensional lookup table To get .

[0139] In step S406, according to , and Calculate the open-circuit voltage .

[0140] For example, open circuit voltage It can be calculated using the following formula (6).

[0141] (6)

[0142] In formula (6), Open circuit voltage, The average voltage. This is the average current. To be based on temperature T The health status (SOH) is determined by a two-dimensional lookup table. The reference ohmic internal resistance was found in the middle. This is the dynamic internal resistance compensation coefficient.

[0143] In step S407, according to The second candidate SOC value is obtained by looking up the table. .

[0144] In some embodiments, it may be based on This is obtained by looking up the pre-calibrated low-end OCV-SOC table. .

[0145] In some embodiments, the duration of the quasi-steady-state segment can be used as a reference. t 1 and the amplitude of the initial sudden change current Determine the second credibility .

[0146] For example, Duration of the quasi-steady-state segment t 1. Positively correlated with the amplitude of the initial sudden change current. Negative correlation.

[0147] In this way, by monitoring the rate of change of current and voltage in real time, the quasi-steady-state segment of small current after a large current impact can be accurately identified as an effective correction window, avoiding erroneous calibration under harsh operating conditions. Furthermore, the influence of residual polarization voltage on open-circuit voltage under steady state can be effectively compensated by the dynamic internal resistance compensation coefficient, thereby achieving short-time non-static calibration. This significantly improves the long-term stability and reliability of SOC estimation in the low-charge area under the frequent impact operating conditions of engineering machinery.

[0148] In some embodiments, when the battery is in a charging state and the current current value and the current temperature value are within a predetermined current range and a predetermined temperature range, respectively, a third candidate SOC value and a corresponding third confidence level are determined based on the charging rate, the current temperature value, and the current voltage value.

[0149] In some embodiments, the third candidate SOC value can be determined based on a pre-calibrated lookup table.

[0150] For example, it can be based on the charging rate. CR Current temperature value T and current voltage value U By searching the pre-calibrated 3D table To obtain the third candidate SOC value .

[0151] In some embodiments, third credibility It can be determined based on the stability of the current and temperature values.

[0152] For example, Can be compared with current value I and temperature value T The stability is positively correlated.

[0153] In this way, through a dynamic high-end correction mechanism, it is not necessary to wait for the battery to be fully charged to the cutoff voltage during charging. The SOC candidate value can be obtained based on the charging rate, temperature and voltage within the operating window where the current and temperature are stable. This effectively decouples the interference of polarization voltage on open circuit voltage during charging. By correlating the reliability with the stability of current and temperature, the reliability of calibration is quantified, thereby achieving incomplete charge calibration and significantly improving the stability and accuracy of SOC estimation for construction machinery under various operating conditions.

[0154] In some embodiments, the second estimator may output , , Three candidate SOC values ​​and , Two levels of credibility.

[0155] In some embodiments, a third estimator is used to process the working parameters based on a feature state observation algorithm to obtain a fourth candidate SOC value and a corresponding fourth confidence level associated with the boundary region, and a fifth candidate SOC value and a corresponding fifth confidence level associated with the platform region.

[0156] In some embodiments, the battery is continuously monitored to see if it enters a preset high-slope calibration boundary region.

[0157] In some embodiments, the extent of the boundary region is the SOC value. or SOC value .

[0158] For example, it can be done through The range in the OCV-SOC curve determines whether the battery has entered the high-slope region of the curve.

[0159] In some embodiments, When the engineering machinery is in a stable mode and is located in the boundary region, the fourth candidate SOC value and the corresponding fourth confidence level are determined based on the current voltage value.

[0160] In some embodiments, the fourth candidate SOC value can be determined based on a pre-calibrated lookup table.

[0161] For example, the fourth candidate SOC value can be obtained from a pre-calibrated OCV-SOC table based on the OCV or the current voltage value. .

[0162] In some embodiments, a fourth confidence level may be applied when the boundary region condition is met. It was determined to be an extremely high value.

[0163] For example, when the battery enters the boundary region and the construction machinery is in L1 stable mode, it can... Set as .

[0164] In some embodiments, If the construction machinery is in a stable mode for a period of time that exceeds a predetermined time threshold while it is in a platform area, it is determined that the battery is currently in an ultra-long stable segment.

[0165] In some embodiments, the duration of the construction machinery being in L1 stable mode exceeds a predetermined duration threshold. Under these circumstances, it can be determined that the construction machinery is in a stable operating condition for an extended period of time.

[0166] In some embodiments, when the construction machinery is in a stable operating condition for an extended period of time, the battery is continuously monitored to see if it enters a preset calibration platform area.

[0167] In some embodiments, the platform area is defined by a SOC value within... arrive between.

[0168] For example, it can be done through The range in the OCV-SOC curve determines whether the battery has entered the plateau region of the curve.

[0169] In some embodiments, within an ultra-long stationary segment, a fifth candidate SOC value and a corresponding fifth confidence level are determined based on the average current value, the average voltage value, and the current temperature value.

[0170] In some embodiments, the average voltage within the ultra-long stationary segment can be recorded. and average current .

[0171] For example, average current It can approach 0.

[0172] In some embodiments, the fifth candidate SOC value can be determined based on a pre-calibrated lookup table.

[0173] For example, it can be based on OCV and the current temperature value. T The fourth candidate SOC value is obtained by using a pre-calibrated "plateau voltage-temperature-aging mapping table". .

[0174] In some embodiments, a pre-calibrated "plateau voltage-temperature-aging mapping table" can be established by statistically analyzing the steady-state voltage of batteries with different aging levels at different temperatures in the plateau region using big data.

[0175] In some embodiments, the fifth level of credibility It can be determined based on the duration of the ultra-long stationary segment.

[0176] For example, It can be positively correlated with the duration of ultra-long stationary segments, but is lower than the fourth confidence level of boundary region calibration. .

[0177] For example, The highest setting can be set to .

[0178] In this way, a calibration system covering the entire SOC range is constructed through dual-modal collaboration between the boundary region and the platform region. This effectively suppresses the long-term cumulative effects of sensor drift, current error, and model mismatch, thereby significantly improving the stability and accuracy of SOC estimation throughout the entire life cycle of construction machinery.

[0179] In some embodiments, the third estimator may output and Two candidate SOC values ​​and their corresponding and Two levels of credibility.

[0180] In step S104, based on the current operating mode and Determine the global baseline credibility.

[0181] In some embodiments, global benchmark confidence The initial value is 100.

[0182] In some embodiments, It decreases as the absolute capacity of the ampere-hour integral increases.

[0183] For example, the cumulative absolute capacity of the ampere-hour integral can be calculated using the following formula (7).

[0184] (7)

[0185] In formula (7), Accumulated absolute capacity for ampere-hour integration. I The current value used for ampere-hour integration.

[0186] In some embodiments, along with The decay rate, which decreases with the increase of SOC, is dynamically related to the local slope of the OCV-SOC curve where the historical operating conditions and the current SOC value are located.

[0187] For example, the higher the proportion of L3 severe mode in the historical operating condition file, the greater the attenuation acceleration factor and the greater the attenuation rate.

[0188] In some embodiments, in the current When the value is in the high-slope region of the OCV-SOC curve, the decay slows down; in the current When the value is in the plateau region of the OCV-SOC curve, the decay accelerates.

[0189] For example, It can be calculated using the following formula (8).

[0190] (8)

[0191] In formula (8), For global benchmark credibility, Accumulated absolute capacity for ampere-hour integration. This represents the percentage of each operating condition mode in the historical operating condition archive. For the present The slope value in the OCV-SOC curve.

[0192] In this way, by dynamically coupling the decay rate of global benchmark confidence with the cumulative ampere-hour throughput, historical operating conditions, and the local slope of the OCV-SOC curve where the current SOC is located, the dynamic optimization and control of the calibration strategy under all operating conditions is realized, making the global benchmark confidence a precise quantitative indicator of the overall SOC estimation confidence, thereby effectively improving the safety and reliability of construction machinery in extreme operating scenarios.

[0193] In step S105, the initial confidence of each candidate SOC value is calibrated using the global benchmark confidence level to obtain the comprehensive confidence level of each candidate SOC value.

[0194] Figure 5 This is a schematic diagram of the initial confidence calibration process according to some embodiments of this disclosure.

[0195] like Figure 5 As shown, the initial confidence calibration includes step S501. S504.

[0196] In step S501, the support level corresponding to each candidate SOC value is determined according to the current operating mode.

[0197] In some embodiments, for the second candidate SOC value If the construction machinery is currently in L3 operating mode, then The corresponding support level can be set to extremely low.

[0198] In some embodiments, for the fourth candidate SOC value If the construction machinery is currently in L1 operating mode, then The corresponding support level can be set to extremely high.

[0199] In step S502, the risk value corresponding to each candidate SOC value is determined based on the current operating mode and the stability of the operating parameters.

[0200] In some embodiments, the risk value corresponding to each candidate SOC value may be proportional to the current operating mode, current change trend, and voltage fluctuation intensity.

[0201] In some embodiments, the risk value corresponding to the candidate SOC value obtained by looking up a table is usually higher than the risk value corresponding to the candidate SOC value obtained by an integration algorithm or a filtering algorithm.

[0202] For example, the third candidate SOC value The corresponding risk value can be higher than the first candidate SOC value. The corresponding risk value.

[0203] In step S503, based on each candidate SOC value and The absolute difference and The voltage sensitivity of the region is used to determine the potential benefit value corresponding to each candidate SOC value.

[0204] In some embodiments, the potential return value corresponding to each candidate SOC value is related to the sum of the candidate SOC value and the return value. The absolute difference is positively correlated with, and The voltage sensitivity in the region of the OCV-SOC curve is positively correlated.

[0205] For example, candidate SOC values and The absolute difference is , The voltage sensitivity in the OCV-SOC curve region is the reciprocal of the curve slope.

[0206] In some embodiments, when When the value is in the plateau region of the OCV-SOC curve, even if the candidate SOC value is similar to... The absolute difference is relatively large, and since the voltage sensitivity is low at this time, the potential benefit value will also be low.

[0207] In some embodiments, when When the value is in the high slope region of the OCV-SOC curve, even if the candidate SOC value is similar to... The absolute difference is small, and because the voltage sensitivity is high at this time, the potential benefit value may also be high.

[0208] In step S504, the overall credibility of each candidate SOC value is determined based on the global baseline credibility, initial credibility, support, risk value, and potential benefit value.

[0209] For example, the overall credibility of each candidate SOC value can be calculated using the following formula (9).

[0210] (9) In formula (9), Used to represent the i-th candidate SOC value The overall credibility corresponding to the i-th candidate SOC value is... For global benchmark credibility, Let be the initial confidence level corresponding to the i-th candidate SOC value. Let SOC be the support value corresponding to the i-th candidate SOC value. The risk value corresponding to the i-th candidate SOC value. Let be the potential return value corresponding to the i-th candidate SOC value.

[0211] In this way, by constructing a multi-dimensional credibility assessment and calibration mechanism, the rationality and fault tolerance of fusion decision-making are greatly improved, ensuring intervention and correction at critical calibration windows, and achieving the unity of safety and effectiveness of calibration strategies. This can significantly improve the accuracy and reliability of SOC estimation for construction machinery under extreme operating scenarios.

[0212] In step S106, the target SOC value is determined based on multiple candidate SOC values ​​and multiple comprehensive confidence levels corresponding to the multiple candidate SOC values.

[0213] Figure 6 This is a schematic diagram of the process for calibrating candidate SOC values ​​according to some embodiments of this disclosure.

[0214] like Figure 6 As shown, candidate SOC value calibration includes step S601. Step S607.

[0215] In step S601, the second to fifth candidate SOC values ​​and their corresponding overall credibility are detected.

[0216] In step S602, it is detected whether the overall credibility corresponding to the second to fifth candidate SOC values ​​is lower than the global benchmark credibility. If the overall credibility corresponding to the second to fifth candidate SOC values ​​is lower than the global benchmark credibility, proceed to step S603; if the overall credibility corresponding to any one of the second to fifth candidate SOC values ​​is not lower than the global benchmark credibility, proceed to step S604.

[0217] In step S603, according to The target SOC value is determined by combining the first candidate SOC value with the target SOC value.

[0218] For example, when , , and All below At that time, the target SOC value by Mainly, but retain Corrections.

[0219] In step S604, it is detected whether the second to fifth candidate SOC values ​​and their corresponding overall credibility meet a predetermined condition. If the second to fifth candidate SOC values ​​and their corresponding overall credibility do not meet the predetermined condition, the process proceeds to step S605; if the second to fifth candidate SOC values ​​and their corresponding overall credibility meet the predetermined condition, the process proceeds to step S606.

[0220] In some embodiments, a predetermined condition can be determined to be met only if the second to fifth candidate SOC values ​​and their corresponding overall credibility simultaneously satisfy the first condition of credibility and benefit, the second condition of low-risk window, and the third condition of smooth transition.

[0221] In some embodiments, the first condition for being credible and beneficial is that the overall credibility corresponding to the candidate SOC value is greater than the sum of the global baseline credibility and the predetermined fault tolerance value, and the potential benefit value corresponding to the candidate SOC value is greater than the predetermined potential benefit threshold.

[0222] For example, the first condition of being credible and beneficial can be represented by the following formula (10).

[0223] (10) In formula (10), Used to represent the i-th candidate SOC value The overall credibility corresponding to the i-th candidate SOC value is... For global benchmark credibility, For the predetermined fault tolerance value, Let $\frac{i}{i}$ be the potential return value corresponding to the $i$-th candidate SOC value. This is a predetermined potential revenue threshold.

[0224] In some embodiments, the second condition for a low-risk window is that the risk value corresponding to the candidate SOC value is less than a predetermined risk threshold, and the current construction machinery is at the end of an L1 stable mode or a stable L2 volatile mode.

[0225] For example, in the second condition of the low-risk window ,in The risk value corresponding to the i-th candidate SOC value. This is a predetermined risk threshold.

[0226] In some embodiments, the third condition for a smooth transition is that the candidate SOC value is equal to... The absolute difference is less than the predetermined mutation threshold of SOC.

[0227] For example, the third condition for a smooth transition is ,in Let i be the candidate SOC value. This is the SOC reference value. A predetermined mutation threshold is set for SOC.

[0228] For example, It can be set to .

[0229] In some embodiments, if the candidate SOC value is the same as If the absolute difference exceeds the predetermined mutation threshold of SOC, then progressive calibration is initiated.

[0230] For example, comparing candidate SOC values ​​with The absolute difference is divided into N parts (e.g., N=5), and one part is corrected every time interval T (e.g., T is 3 minutes) until the correction is complete.

[0231] In step S605, calibration is skipped.

[0232] For example, when the second candidate SOC value and the corresponding overall credibility If the above-mentioned predetermined conditions are not met, the value will not be used to calibrate the SOC reference value.

[0233] In step S606, the weight of each candidate SOC value is determined based on multiple comprehensive confidence levels.

[0234] For example, the weight of each candidate SOC value can be calculated using the following formula (11).

[0235] (11)

[0236] In formula (11), Used to represent the confidence level corresponding to the i-th candidate SOC value The weight of the i-th candidate SOC value, and This represents the overall credibility corresponding to the i-th or j-th candidate SOC value.

[0237] In step S607, the weighted sum of multiple candidate SOC values ​​is calculated based on the weight of each candidate SOC value to obtain the target SOC value.

[0238] For example, the target SOC value can be calculated using the following formula (12).

[0239] (12)

[0240] In formula (12), , For the target SOC value, Let i be the candidate SOC value. The weight is the value of the i-th candidate SOC.

[0241] In this way, by real-time detection of multiple candidate SOC values ​​and their corresponding comprehensive credibility to ensure they meet multiple predetermined conditions, it is possible to effectively prevent the use of unreliable candidate SOC values ​​as calibration data to update the target SOC value. Furthermore, by weighting, dynamic fusion and safety calibration of multi-dimensional SOC estimation are achieved, thereby significantly improving the accuracy and reliability of SOC estimation for construction machinery under extreme working conditions.

[0242] In some embodiments, when the global benchmark confidence level remains below a preset lower limit threshold and it is detected that the construction machinery is about to enter a low-load phase, the battery usage range is expanded so that the battery enters the boundary region.

[0243] Figure 7 This is a flowchart illustrating the process of expanding the battery usage range according to some embodiments of this disclosure.

[0244] like Figure 7 As shown, expanding the battery usage range includes step S701. Step S706.

[0245] In step S701, the global benchmark confidence level is detected. And the low-load phase of construction machinery.

[0246] In step S702, determine Whether it remains below the threshold and whether the construction machinery is about to enter a low-load phase. If so... If the load remains below the threshold and the construction machinery is about to enter a low-load phase, proceed to step S703; if If the load does not remain below the threshold, or if the construction machinery is not about to enter a low-load phase, then return to step S701.

[0247] For example, when When the load remains below the threshold, historical operating cycles can be used to predict whether the construction machinery is about to enter a low-load phase.

[0248] In step S703, the battery's usage range is expanded so that the battery can enter the boundary area.

[0249] For example, "range relaxation" suggestions can be proactively generated to guide the battery into a calibrable boundary region.

[0250] In step S704, it is determined whether the risk corresponding to the candidate SOC value has increased. If the risk corresponding to the candidate SOC value has not increased, proceed to step S705; if the risk corresponding to the candidate SOC value has increased, proceed to step S706.

[0251] For example, during the process of widening the execution range, the risk value corresponding to the candidate SOC value can be monitored in real time. This is to determine whether the risk corresponding to the candidate SOC value has increased.

[0252] In step S705, if the battery has completed calibration in the boundary region, the range will be narrowed to a safe range.

[0253] For example, once the battery has completed low-risk calibration in the boundary region, the SOC usage range should be immediately reduced to a safe range.

[0254] In step S706, the expansion of the usage range is stopped or the original usage range is restored.

[0255] In some embodiments, if the risk value corresponding to the candidate SOC value increases during the process of expanding the usage range, the expansion of the usage range is stopped or the original usage range is restored.

[0256] For example, during the process of widening the execution range, the risk value corresponding to the candidate SOC value can be monitored in real time. If the risk value increases, immediately suspend the widening or restore the original range.

[0257] In this way, through dynamic management of the SOC usage range, intelligent collaboration between proactively creating calibration opportunities and ensuring safety is achieved without sacrificing battery health, thereby significantly improving the timeliness and safety of SOC estimation throughout the battery's entire life cycle.

[0258] The implementation process of the battery SOC estimation method will be described in detail below with reference to a specific embodiment of this disclosure.

[0259] It should be understood that this embodiment is only used to explain this disclosure and is not intended to limit this disclosure.

[0260] For construction machinery that requires SOC estimation, a power-on initialization is performed after the battery is powered on.

[0261] After the state is restored, read the key data saved at the time of the last power-down from the NVM. The key data may include: SOC baseline value. If there is no historical data, initialize to Global benchmark credibility Initialized to 100; cumulative absolute capacity Units are Ah, initialized to 0; historical archives of the percentage of time spent on the 10 most recent operating conditions. This refers to the respective percentages of L1 / L2 / L3 operating modes; the battery health status (SOH) is initialized to... ; R 0( T (, SOH) R 1(C 1, T Two-dimensional / three-dimensional lookup tables such as (SOC); OCV-SOC curve tables or temperature-voltage-SOC three-dimensional arrays, for example, temperature points... SOC step size is .

[0262] Next, set the initialization variables. These include the accumulated value of the ampere-hour integral. ; Current queue of sliding window buffer The length is 100, and the step size is 5ms, used to calculate the standard deviation. Voltage change rate queue in sliding window buffer Quasi-steady-state segment timer Boundary area timer .

[0263] Next, a battery self-test is performed, including current sensor deviation calibration (i.e., zero drift detection) and voltage channel open circuit detection.

[0264] In the main loop of SOC estimation, the following steps are executed sequentially, wherein the main loop of SOC estimation is periodically triggered. variable.

[0265] Step S101: Obtain the operating parameters of the battery in the construction machinery, including the battery's current value, voltage value, and temperature value.

[0266] Acquire main sensor current The period is 10ms. If the auxiliary sensor is enabled, it will collect data simultaneously. The period is 5ms.

[0267] Collect total battery voltage V and temperature T Total battery voltage V That is, the sum of the voltages of all series-connected cells or the highest / lowest voltage, temperature T Take the average or highest value of all temperature sensors.

[0268] Step S102: Determine the current working mode of the construction machinery based on the working parameters.

[0269] Calculate the rate of change of current And update the current standard deviation in the sliding window of 100 sampling points. .

[0270] like and If so, the current operating mode of the construction machinery is L1 stable mode; like and If so, the current operating mode of the construction machinery is L2 fluctuation mode; if and If so, the current operating mode of the construction machinery is L3 severe mode.

[0271] The data transmission period is adjusted according to the operating mode: 20ms for L1 stable mode, 10ms for L2 fluctuating mode, and 5ms for L3 severe mode.

[0272] If the current operating mode is L3 severe mode and the auxiliary sensor is enabled, then dual-sensor dynamic fusion will be performed.

[0273] Step S103: The working parameters are processed using multiple estimators to obtain multiple candidate SOC values ​​and multiple initial confidence levels corresponding one-to-one with the candidate SOC values, as well as the battery's SOC reference value. .

[0274] For the estimator M1, a first-order RC equivalent circuit model is adopted, and the state vector... Observation vector .

[0275] It should be noted that a first-order RC equivalent circuit model is used here for the sake of simplifying the calculation, but a second-order RC equivalent circuit model can also be used.

[0276] The state equation of the first-order RC equivalent circuit model is shown in formula (13).

[0277] , (13) In formula (13), For Coulomb efficiency, the value is 0.99; Battery capacity, corrected from SOH; This is the polarization time constant, obtained from the lookup table.

[0278] The observation equations of the first-order RC equivalent circuit model are shown in Equation (14).

[0279] (14)

[0280] The above equations are iterated using an extended Kalman filter to calculate the state prediction values. Sum of error covariance prediction And calculate the Kalman gain. K Using voltage measurements Update state estimation Covariance .

[0281] Based on the current operating mode, the process noise covariance matrix Q and the measurement noise covariance matrix R are dynamically set to adaptively adjust the filtering algorithm.

[0282] For estimator M1, an ampere-hour integral and dynamic segment correction algorithm is used.

[0283] The specific formula for calculating the ampere-hour integral is shown in formula (15).

[0284] (15)

[0285] In formula (15), SOH is read from NVM with an initial value of 100%, and its capacity decreases with each cycle.

[0286] The condition detection for dynamic low-end correction is as follows And the rate of change of voltage And duration .

[0287] For example If it is a 500Ah battery, then A.

[0288] When the conditions are met, record the average voltage of that segment. and average current And record the amplitude of the current surge at the start of the segment. , This is the current difference between the sampling point before the start of the segment and the first sampling point at the start of the segment.

[0289] The specific formula for calculating the OCV estimate is shown in formula (16).

[0290] (16)

[0291] In formula (16), the dynamic internal resistance compensation coefficient The value is from the pre-calibrated three-dimensional table. It was found that, among them Take the current .

[0292] Find the low-end OCV-SOC table to obtain SOC range ,interval .

[0293] Set initial credibility ,For example Lasting 3 seconds , lasting 5 seconds .

[0294] The condition for dynamic high-end correction is the absolute value of the current during charging. exist Between, temperature T exist Between, and the rate of change of voltage Lasts for 2 seconds.

[0295] Based on charging rate CR ,temperature T Terminal voltage V Look up the three-dimensional table ( CR , T , V )get .

[0296] Set initial credibility Its value is fixed or equal to CR Stability-related.

[0297] Estimator M2 output , and their respective credibility, and SOC reference values .

[0298] For estimator M3, a boundary and plateau region feature observation algorithm is used.

[0299] The conditions for boundary region observation are as follows: or ,and It lasts for 5 seconds, and the operating mode is L1 stable mode.

[0300] Directly from the terminal voltage V The standard full-range OCV-SOC table is consulted to obtain the results. The initial credibility level is set to .

[0301] The conditions for long-term steady-state observation in the platform region are: And the operating mode is in L1 stable mode for more than 10 minutes, and .

[0302] Statistical analysis of the average voltage during this period According to temperature T With the current SOH, look up the pre-calibrated platform area mapping table. get The initial credibility level is set to .

[0303] Estimator M3 output and And its corresponding credibility.

[0304] Step S104, based on the current operating mode and Determine the global benchmark credibility .

[0305] Step S105: Using the global baseline confidence level, the initial confidence level of each candidate SOC value is calibrated to obtain the comprehensive confidence level of each candidate SOC value.

[0306] Step S106: Determine the target SOC value based on multiple candidate SOC values ​​and multiple comprehensive confidence levels corresponding to the multiple candidate SOC values.

[0307] The weighted fusion output decision of the target SOC value is calculated by the following formula (17).

[0308] , (17) For table lookup-type candidate SOC values , , and If both conditions are met ; That is, the risk is lower than ; That is, the jump is less than Then perform calibration.

[0309] The calibration content allows for the selection of the SOC candidate value and its corresponding value. Participate in the aforementioned weighted fusion.

[0310] like Not satisfied but and If the conditions are met, then progressive calibration is initiated.

[0311] The content of progressive calibration is to adjust the difference The calibration is performed in 5 steps, with a 3-minute interval between each step, and the difference is corrected each time. .

[0312] In the above SOC estimation process, dynamic management decisions for SOC intervals can also be performed.

[0313] when If there have been no calibration actions in the past hour, and a low load is predicted within the next 10 minutes (e.g., currently in L1 mode and current remains below 0.1C), then the SOC usage range will be temporarily extended. (Original interval is) ).

[0314] Once boundary calibration is complete, immediately restore the original range. If the operating mode is detected as L3 severe mode during the relaxation period, immediately stop the relaxation.

[0315] In the above SOC estimation process, dynamic error compensation can also be selected.

[0316] Real-time calculation of current change rate .like and Then through a two-dimensional table Find and determine the dynamic error compensation factor Its value range is and use Used for ampere-hour integration; otherwise .

[0317] In the idle task, you can also choose to perform parameter learning and updating.

[0318] Whenever an L1 steady mode is detected to last for more than 60 seconds, the voltage, current, and temperature data of that segment are stored in a buffer, up to a maximum of 1000 points.

[0319] Once the buffer is full (100 points), the recursive least squares method is used to identify the model parameters of the first-order RC equivalent circuit. R 0、 R 1 and C 1. The result is used to update the corresponding table in the lookup table. T The value of the SOC point is used, for example, by employing an exponentially weighted moving average with a smoothing factor of 0.2.

[0320] In addition, the historical operating condition file is updated every 24 hours based on the cumulative proportion of L1 / L2 / L3 mode duration, so as to be used for global reliability decay calculation.

[0321] Upon receiving a power-down interrupt or a timed save task, the following critical states are written to NVM.

[0322] Key conditions include: current ampere-hour integral benchmark. Global benchmark credibility Cumulative absolute capacity Historical working conditions archives The latest increments of the battery health status (SOH) and the model parameter lookup table.

[0323] Finally, dual backup storage and CRC checksums are used to prevent data corruption.

[0324] Figure 8 This is a schematic diagram of the structure of a battery SOC estimation device according to some embodiments of the present disclosure.

[0325] like Figure 8 As shown, the battery SOC estimation device 800 includes a memory 801, a processor 802, a bus 803, an input / output interface 804, a network interface 805, and a storage interface 806.

[0326] In some embodiments, the battery SOC estimation device 800 may include a memory 801 configured to store instructions; and a processor 802 configured to execute instructions, causing the battery SOC estimation device 800 to perform the battery SOC estimation method in any of the embodiments of this disclosure.

[0327] The memory 801 may include, for example, system memory, non-volatile storage media, etc. The system memory may store, for example, an operating system, application programs, a boot loader, and other programs. The system memory may include volatile storage media, such as random access memory (RAM) and / or cache memory. The non-volatile storage media may store, for example, instructions for executing corresponding embodiments of at least one battery SOC estimation method disclosed herein. Non-volatile storage media include, but are not limited to, disk storage, optical storage, flash memory, etc.

[0328] The processor 802 can be implemented using a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistors, or other discrete hardware components. Correspondingly, it can be implemented by the central processing unit (CPU) running instructions in the memory 401 to execute the corresponding steps, or it can be implemented using dedicated circuitry to execute the corresponding steps.

[0329] Bus 803 can use any of the various bus architectures. For example, bus architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, and Peripheral Component Interconnect (PCI) bus.

[0330] The interfaces 804, 805, and 806 of the battery SOC estimation device 800, as well as the memory 801 and processor 802, can be connected via bus 803. Input / output interface 804 provides a connection interface for input / output devices such as monitors, mice, and keyboards. Network interface 805 provides a connection interface for various networked devices. Storage interface 806 provides a connection interface for external storage devices such as floppy disks, USB flash drives, and SD cards.

[0331] This disclosure also provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the battery SOC estimation method of any of the above embodiments.

[0332] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the battery SOC estimation method of any of the above embodiments.

[0333] In this embodiment, by employing condition-adaptive multi-rate sampling and dynamic fusion of dual sensors, current integration errors under extreme conditions are effectively suppressed, providing a reliable data foundation for solving the SOC estimation problem of construction machinery. A model-based filtering algorithm, a data-based lookup table correction algorithm, and a feature-state-based observation algorithm are organically combined. The filtering algorithm dominates under dynamic conditions, the lookup table algorithm plays a calibration role in steady-state segments, and high-reliability anchor points are provided in boundary region feature observations, achieving a balance between accuracy and robustness across all operating conditions. Through a five-dimensional reliability assessment model and a triple calibration threshold mechanism, the system can comprehensively assess risks and benefits, minimizing miscalibration and improving system safety and user experience. A lookup table-based dynamic current compensation factor is also introduced. This method effectively simulates dynamic polarization effects with extremely low computational cost, improving the accuracy of ampere-hour integration during transient processes and providing more accurate input for filtering algorithms. By exploring various opportunities such as boundary calibration and long-term steady-state reference in the platform region, and combining proactive interval predictive management, it can maintain accuracy throughout the entire lifecycle even without full charging / resting conditions. The solution provided in this disclosure integrates source accuracy improvement, multi-modal collaborative advantages, intelligent decision-making to eliminate miscalibration, dynamic error compensation, and lifecycle adaptation, thereby significantly improving the accuracy and reliability of SOC estimation for construction machinery under extreme operating conditions.

[0334] In some embodiments, the functional units described above may be implemented as general-purpose processors, programmable logic controllers (PLCs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or any suitable combination thereof for performing the functions described herein.

[0335] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0336] The description in this disclosure is provided for illustrative and descriptive purposes only and is not intended to be exhaustive or to limit the disclosure to its forms. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of this disclosure and to enable those skilled in the art to understand this disclosure and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A method for estimating battery SOC, comprising: Obtain the operating parameters of the battery in the construction machinery, wherein the operating parameters include the current value, voltage value and temperature value of the battery; Based on the operating parameters, determine the current operating mode of the construction machinery; The operating parameters are processed using multiple estimators to obtain multiple candidate SOC values ​​and multiple initial confidence levels corresponding one-to-one with the candidate SOC values, as well as the SOC reference value of the battery. ; Based on the current operating mode and the Determine the global baseline credibility; Using the global baseline confidence level, the initial confidence level of each candidate SOC value is calibrated to obtain the comprehensive confidence level of each candidate SOC value; The target SOC value is determined based on the multiple candidate SOC values ​​and the multiple comprehensive confidence levels corresponding to the multiple candidate SOC values.

2. The battery SOC estimation method according to claim 1, wherein, The process of processing the operating parameters using multiple estimators includes: The first estimator, based on an adaptive filtering algorithm, processes the working parameters to obtain a first candidate SOC value and a corresponding first initial confidence level. The second estimator, based on the ampere-hour integration algorithm, processes the operating parameters to obtain the... ; The second estimator, based on the dynamic fragment correction algorithm, processes the working parameters to obtain a second candidate SOC value and a corresponding second confidence level associated with the low-end region, and a third candidate SOC value and a corresponding third confidence level associated with the high-end region. The working parameters are processed using a third estimator based on a feature state observation algorithm to obtain a fourth candidate SOC value and a corresponding fourth confidence level associated with the boundary region, and a fifth candidate SOC value and a corresponding fifth confidence level associated with the platform region.

3. The battery SOC estimation method according to claim 2, wherein, The process of obtaining the second candidate SOC value associated with the low-end region and the corresponding second confidence level includes: If the duration of the current value being less than a preset current threshold is greater than a preset duration threshold, and the voltage change rate is less than a preset change rate threshold, it is determined that the battery is currently in a quasi-steady-state segment. Within the quasi-steady-state segment, the second candidate SOC value and the corresponding second confidence level are determined based on the average current value, the average voltage value, the current temperature value, and the first candidate SOC value.

4. The battery SOC estimation method according to claim 2, wherein, The process of obtaining the third candidate SOC value and the corresponding third confidence level associated with the high-end region includes: When the battery is in a charging state and the current current value and the current temperature value are within a predetermined current range and a predetermined temperature range, respectively, the third candidate SOC value and the corresponding third confidence level are determined based on the charging rate, the current temperature value, and the current voltage value.

5. The battery SOC estimation method according to claim 2, wherein, The process of obtaining the fourth candidate SOC value and the corresponding fourth confidence level associated with the boundary region includes: In the When the engineering machinery is in a stable mode and is located in the boundary region, the fourth candidate SOC value and the corresponding fourth confidence level are determined based on the current voltage value.

6. The battery SOC estimation method according to claim 2, wherein, The process of obtaining the fifth candidate SOC value and the corresponding fifth confidence level associated with the platform region includes: In the If the engineering machinery is in a stable mode for a period of time that exceeds a predetermined time threshold while it is in a platform area, it is determined that the battery is currently in an ultra-long stable segment. Within the ultra-long stable segment, the fifth candidate SOC value and the corresponding fifth confidence level are determined based on the average current value, the average voltage value, and the current temperature value.

7. The battery SOC estimation method according to claim 1, wherein, The step of determining the target SOC value based on the plurality of candidate SOC values ​​and the plurality of comprehensive confidence levels corresponding to the plurality of candidate SOC values ​​includes: The weight of each candidate SOC value is determined based on the multiple comprehensive credibility scores. Based on the weight of each candidate SOC value, a weighted sum of the multiple candidate SOC values ​​is calculated to obtain the target SOC value.

8. The battery SOC estimation method according to claim 7, wherein, The step of determining the weight of each candidate SOC value based on the multiple comprehensive credibility scores includes: Detect whether the second to fifth candidate SOC values ​​and their corresponding overall credibility meet predetermined conditions; If the second to fifth candidate SOC values ​​and their corresponding comprehensive credibility satisfy the predetermined conditions, the weight of each candidate SOC value is determined according to the multiple comprehensive credibility values.

9. The battery SOC estimation method according to claim 8, wherein, The step of detecting whether the second candidate SOC value to the fifth candidate SOC value and the corresponding overall confidence level meet the predetermined conditions includes: Detect whether the overall confidence levels corresponding to the second candidate SOC value to the fifth candidate SOC value are all lower than the global benchmark confidence level; If the overall confidence level corresponding to the second candidate SOC value to the fifth candidate SOC value is lower than the global benchmark confidence level, then according to the... The target SOC value is determined by combining the first candidate SOC value; If the overall credibility of any candidate SOC value from the second candidate SOC value to the fifth candidate SOC value is not lower than the global benchmark credibility, then it is checked whether the second candidate SOC value to the fifth candidate SOC value and the corresponding overall credibility meet a predetermined condition.

10. The battery SOC estimation method according to claim 1, wherein, The initial confidence level calibration for each candidate SOC value includes: Based on the current operating mode, determine the support level corresponding to each candidate SOC value; Based on the current operating mode and the stability of the operating parameters, determine the risk value corresponding to each candidate SOC value; Based on each candidate SOC value and the The absolute difference and the stated The voltage sensitivity of the region is used to determine the potential benefit value corresponding to each candidate SOC value; The overall credibility of each candidate SOC value is determined based on the global baseline credibility, the initial credibility, the support, the risk value, and the potential benefit value.

11. The battery SOC estimation method according to claim 1, wherein, Determining the current operating mode of the construction machinery based on the operating parameters includes: If the rate of change of current is less than the first threshold for the rate of change of current and the intensity of current fluctuation is less than the first threshold for the intensity of current fluctuation, it is determined that the engineering machinery is currently in a stable mode. If the current change rate is not less than the first current change rate threshold and less than the second current change rate threshold, and the current fluctuation intensity is not less than the first current fluctuation intensity threshold and less than the second current fluctuation intensity threshold, it is determined that the construction machinery is currently in a fluctuation mode, wherein the first current change rate threshold is less than the second current change rate threshold, and the first current fluctuation intensity threshold is less than the second current fluctuation intensity threshold. If the current change rate is not less than the second current change rate threshold and the current fluctuation intensity is not less than the second current fluctuation intensity threshold, it is determined that the engineering machinery is currently in a severe mode.

12. The battery SOC estimation method according to claim 11, wherein, When the current operating mode is the stable mode or the fluctuating mode, the current sampling value provided by the first sensor is used as the current value of the battery. as well as When the current operating mode is the severe mode, the current value of the battery is determined based on the current sampling value provided by the first sensor and the current sampling value provided by the second sensor.

13. The battery SOC estimation method according to claim 12, wherein, When the current operating mode is the severe mode, the current value of the battery is the weighted sum of the current sample value provided by the first sensor and the current sample value provided by the second sensor.

14. The battery SOC estimation method according to claim 13 further includes: The dynamic current error compensation factor is determined based on the current current change rate and the current absolute value of the battery. The product of the dynamic current error compensation factor and the weighted sum is calculated to obtain the compensated current value.

15. The battery SOC estimation method according to any one of claims 1-14, further comprising: When the global benchmark confidence level remains below a preset lower threshold and it is detected that the construction machinery is about to enter a low-load phase, the usage range of the battery is expanded so that the battery can enter the boundary zone. If the risk value corresponding to the candidate SOC value increases during the process of expanding the usage range, the expansion of the usage range shall be stopped or the original usage range shall be restored. Once the battery has been calibrated in the boundary region, the usage range will be narrowed to a safe range.

16. A battery SOC estimation device, comprising: The memory is configured to store instructions; The processor is configured to execute the instructions, causing the battery SOC estimation device to perform the battery SOC estimation method as described in any one of claims 1-15.

17. An engineering machine, comprising: The battery SOC estimation device as described in claim 16.

18. A computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the battery SOC estimation method as described in any one of claims 1-15.

19. A computer program product comprising a computer program, wherein, When the computer program is executed by the processor, it implements the battery SOC estimation method as described in any one of claims 1-15.

Images