Technical algorithm for estimating SOC of lithium battery in real time

Abstract

The invention discloses a technical algorithm for estimating the SOC of a lithium battery in real time, which relates to the technical field of lithium batteries and comprises a hardware acquisition and control module, a cloud service platform module and a core algorithm module. According to the technical algorithm for estimating the SOC of the lithium battery in real time, the MIAUKF of the core algorithm module is responsible for accurate estimation of the SOC, and the I-EKF is responsible for dynamic updating of the SOH and closed-loop linkage of data, so that the problem of estimation deviation of a single algorithm is solved; through self-adaptive correction of Q and R, battery characteristic changes under different working conditions are adapted; the local algorithm ensures real-time performance, the cloud platform provides computing power support and historical data optimization, and both speed and precision are considered; therefore, the terminal voltage and the cell temperature of the single lithium battery in the lithium battery pack can be monitored in real time with high precision, the common hidden dangers of the lithium battery such as overcharge, overdischarge and overhigh temperature rise of each single lithium battery are effectively prevented, the service life of the battery is prolonged, the use economy of the battery is improved, and the energy utilization rate of the battery pack is improved.

Description

Technical Field

[0001] This invention relates to the field of lithium battery technology, specifically to a real-time lithium battery SOC estimation algorithm. Background Technology

[0002] Lithium-ion batteries are a type of battery that uses lithium metal or lithium alloy as the positive / negative electrode material and a non-aqueous electrolyte solution. They are used in many fields such as power and manufacturing. In the railway locomotive industry, the performance advantages of lithium-ion batteries make them highly promising for use in the auxiliary power systems of high-speed trains. Lithium-ion batteries have high energy density and lifespan, which can better meet the increasing power consumption and long-term operation requirements of locomotives. At the same time, lithium-ion batteries charge quickly and efficiently, and have good charge and discharge performance, which can better meet the energy-saving and consumption-reducing requirements of locomotives. Lithium-ion batteries in railway locomotive energy storage are replacing lead-acid batteries. However, due to differences in the production and use processes, individual lithium-ion batteries can have different capacities and self-discharge rates. To ensure the safety of railway locomotives, it is necessary to monitor and provide early warnings for relevant parameters such as battery pack temperature, voltage, and charge / discharge current in real time, and to detect the health status of each individual battery in the energy storage system.

[0003] Existing lithium battery SOC technology algorithms typically have low accuracy in measuring parameters such as battery voltage, state of charge, and state of health, resulting in inadequate fault monitoring and safety protection. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a real-time lithium battery SOC estimation algorithm, which solves the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a real-time lithium battery SOC estimation algorithm, comprising a hardware acquisition and control module, a cloud service platform module, and a core algorithm module; The hardware acquisition and control module includes a central processing unit and an acquisition module; The cloud service platform module includes a data transmission module, a computing power support module, and a feedback control module. The core algorithm module includes a multi-inspiration adaptive unscented Kalman filter (MIAUKF) module, an extended Kalman filter (I-EKF) module, and an algorithm coordination module. The Multi-Innovation Adaptive Unscented Kalman Filter (MIAUKF) module introduces adaptive filtering (AF) and multi-innovation theory, enabling the process noise covariance Q and measurement noise covariance R of the algorithm to be adaptively corrected, thereby improving robustness and stability. This allows for real-time estimation of the lithium battery's State of Charge (SOC). Through iterative calculation using a mathematical model, the SOC estimation error is controlled within 0.01. The MIAUKF function for calculating the SOC of a lithium battery is as follows: ; ; The input values ​​Q and R are dynamically adjusted through an adaptive mechanism; while H is the observation matrix. For sigma point weights, For observed and predicted values; The extended Kalman filter I-EKF module is used to update the SOH of the lithium battery in real time, providing a dynamic correction basis for SOC estimation; The algorithm collaboration module's MIAUKF module and I-EKF module form a cyclic iteration mechanism. MIAUKF is responsible for real-time SOC estimation, while I-EKF is responsible for dynamic SOH updates. The data from both modules are linked to achieve online joint estimation of lithium battery SOC and SOH.

[0006] Furthermore, the central processing unit adopts ARM V9.2 platform technology and STM32L151RDT (64-bit) microcontroller with Cortex-X4 core to provide high-speed computing support.

[0007] Furthermore, the acquisition module is equipped with an AVR ATMEGA64 processor and a high-speed 16-bit AD converter to achieve high-precision acquisition of voltage, current, and temperature data, providing reliable input data for the algorithm.

[0008] Furthermore, the data transmission module uploads the battery data collected by the BMS device to the cloud platform in real time via a 4G / 5G communication module, with a latency as low as 50ms.

[0009] Furthermore, the computing power support module is based on a cloud platform and carries SOC and SOH evaluation models, adopts a multi-threaded computing mode, and matches the best prediction results.

[0010] Furthermore, the feedback control module is used to transmit the calculation results to the user terminal and simultaneously feed them back to the BMS main control module to issue battery control commands, ensuring a closed loop between algorithm estimation and actual application.

[0011] Furthermore, the extended Kalman filter I-EKF module includes the following specific procedures: (1) Input: Adaptive filter signal, initial SOH value of battery, real-time current data; (2) Calculate: the estimated SOH and voltage values ​​of the battery at time K+1; (3) Correction: Adjust the deviation by estimating the error at time K+1, and output the corrected SOH value; (4) Feedback: Real-time SOH data is sent back to the MIAUKF module to support accurate SOC estimation.

[0012] Furthermore, the algorithm collaboration module compensates for the static defects of the model parameters in the MIAUKF algorithm by dynamically representing time-varying parameters such as battery capacity and internal resistance through SOH, so that the SOC estimation always adapts to the characteristic changes of the battery throughout its entire life cycle.

[0013] Furthermore, the algorithm coordination module is designed with strict timing scheduling logic to meet the real-time requirements of algorithm execution: When the battery is under complex operating conditions, the data interaction frequency is increased to 100ms / time to prioritize the real-time correction of the SOC estimate; when the battery is stationary and the SOH is stable, the interaction frequency is reduced to 500ms / time to reduce hardware computing power consumption. During each data interaction, the module adds a timestamp and checksum to the transmitted data. If a data delay or checksum failure is detected, the cached valid data from the previous cycle is immediately used to ensure that the algorithm runs continuously without interruption.

[0014] This invention provides a real-time lithium battery SOC estimation algorithm, which has the following advantages: 1. This real-time lithium battery SOC estimation algorithm enables high-precision monitoring of lithium battery capacity, temperature, and consistency. It is primarily applied to the high-precision intelligent management of lithium batteries in railway locomotives and solar photovoltaic energy storage. The core algorithm module, MIAUKF, is responsible for accurate SOC estimation, while I-EKF handles dynamic SOH updates. This closed-loop data linkage solves the estimation bias problem of a single algorithm. Adaptive correction of Q and R values ​​accommodates changes in battery characteristics under different operating conditions. The local algorithm ensures real-time performance, while the cloud platform provides computing power support and historical data optimization, balancing speed and accuracy. Therefore, it can accurately monitor the terminal voltage and cell temperature of individual lithium batteries within the battery pack in real time, effectively preventing common lithium battery problems such as overcharging, over-discharging, and excessive temperature rise in each cell, extending battery life, improving battery economy, and increasing the energy utilization rate of the battery pack. Attached Figure Description

[0015] Figure 1 This is a schematic diagram illustrating the overall steps of the real-time lithium battery SOC estimation algorithm of the present invention.

[0016] Figure 2 This is a flowchart of the SOH estimation process based on the I-EKF algorithm, which is a real-time estimation algorithm for lithium battery SOC according to the present invention. Detailed Implementation

[0017] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0018] like Figures 1-2 As shown, the present invention provides a technical solution: a real-time lithium battery SOC estimation algorithm, including a hardware acquisition and control module, a cloud service platform module, and a core algorithm module; The hardware acquisition and control module includes a central processing unit and an acquisition module. The central processing unit adopts ARM V9.2 platform technology and an STM32L151RDT (64-bit) microcontroller with a Cortex-X4 core, providing high-speed computing support. The acquisition module is equipped with an AVR ATMEGA64 processor and a high-speed 16-bit AD converter to achieve high-precision acquisition of voltage, current and temperature data, providing reliable input data for the algorithm. The cloud service platform module includes a data transmission module, a computing power support module, and a feedback control module. The data transmission module uploads battery data collected by the BMS device to the cloud platform in real time via a 4G / 5G communication module, with a latency as low as 50ms. The computing power support module is based on the cloud platform and uses a multi-threaded computing mode to match the best prediction results. The feedback control module is used to transmit the calculation results to the user terminal and simultaneously feed back to the BMS main control module to issue battery control commands, ensuring a closed loop between algorithm estimation and actual application. The core algorithm modules include the Multi-Inspiration Adaptive Unscented Kalman Filter (MIAUKF) module, the Extended Kalman Filter (I-EKF) module, and the algorithm coordination module; The MIAUKF module of multi-innovation adaptive unscented Kalman filter introduces adaptive filtering AF and multi-innovation theory, which enables the process noise covariance Q and measurement noise covariance R of the algorithm to be adaptively corrected, thereby improving robustness and stability. This allows for real-time estimation of the state of charge (SOC) of lithium batteries. Through iterative calculation using a mathematical model, the SOC estimation error is controlled within 0.01. The MIAUKF function for calculating the SOC of a lithium battery is as follows: ; ; The input values ​​Q and R are dynamically adjusted through an adaptive mechanism; while H is the observation matrix. For sigma point weights, For observed and predicted values; The extended Kalman filter (I-EKF) module is used to update the lithium battery's state of harm in real time, providing a dynamic correction basis for SOC estimation. The MIAUKF module and I-EKF module of the algorithm collaboration module form a cyclic iteration mechanism. MIAUKF is responsible for real-time SOC estimation, and I-EKF is responsible for dynamic SOH updating. The data of the two are linked together to realize online joint estimation of lithium battery SOC and SOH.

[0019] The Extended Kalman Filter (I-EKF) module includes the following specific procedures: (1) Input: Adaptive filter signal, initial SOH value of battery, real-time current data; (2) Calculate: the estimated SOH and voltage values ​​of the battery at time K+1; (3) Correction: Adjust the deviation by estimating the error at time K+1, and output the corrected SOH value; (4) Feedback: Real-time SOH data is sent back to the MIAUKF module to support accurate SOC estimation.

[0020] The algorithm coordination module compensates for the static defects of the model parameters in the MIAUKF algorithm by dynamically representing time-varying parameters such as battery capacity and internal resistance through SOH, ensuring that the SOC estimation always adapts to the characteristic changes throughout the battery's entire life cycle. The algorithm coordination module also incorporates strict timing scheduling logic to meet the real-time requirements of the algorithm's operation. When the battery is under complex operating conditions, the data interaction frequency is increased to 100ms / time to prioritize the real-time correction of the SOC estimate; when the battery is stationary and the SOH is stable, the interaction frequency is reduced to 500ms / time to reduce hardware computing power consumption. During each data interaction, the module adds a timestamp and checksum to the transmitted data. If a data delay or checksum failure is detected, the cached valid data from the previous cycle is immediately activated to ensure that the algorithm runs continuously without interruption.

[0021] The real-time lithium battery SOC estimation algorithm includes the following specific steps: S1. Start the BMS hardware acquisition module to collect key parameters of the lithium battery pack in real time, including individual cell voltage, charging and discharging current, cell temperature and other auxiliary parameters; use the Cortex-X4 core microcontroller of the ARM V9.2 platform to preprocess the collected raw data to provide clean input for the algorithm. S2. Set the initial parameters of the MIAUKF module: initial value of process noise covariance Q0, initial value of measurement noise covariance R0, and initial estimates of state variables. Set the initial parameters of the I-EKF module: initial battery SOH value, initial filter gain value, and error threshold. S3. Based on the equivalent circuit model of a lithium battery, using the preprocessed current and voltage data, and through the unscented transformation of the UKF algorithm, generate sigma points and predict the state variables at time k+1. ; Adaptive filtering (AF) and multiple innovation theory are introduced to dynamically adjust the process noise covariance Q and measurement noise covariance R; the filter gain at time k+1 is calculated. and prediction error Update Q and R according to the formula: ; ; By combining the corrected Q, R, and observed values, the SOC estimate at time k+1 is calculated iteratively. Ensure that the SOC estimation error is ≤0.01; S4. Using the SOC estimate output by the MIAUKF module and the current data collected by the hardware as input, and combining the lithium battery aging model, calculate the predicted SOH value at time k+1. By comparing the predicted SOH value with the actual battery health status feedback, the estimation error at time k+1 is obtained. If the error exceeds the threshold, the filter gain is dynamically corrected through the I-EKF algorithm to adjust the predicted SOH value, and the corrected final SOH value at time k+1 is output. The real-time SOH value is then sent back to the MIAUKF module as the basis for dynamic correction of SOC estimation. The S5 and BMS main control modules upload the SOC estimate of MIAUKF, the SOH update value of I-EKF, and the raw collected data to the cloud service platform. The computing power support module of the cloud service platform loads the SOC / SOH evaluation model and performs secondary optimization on the local algorithm results through multi-threaded calculation to match the best prediction result. Verify the reasonableness of the results: If the SOC error is greater than ±2% or the SOH error is greater than ±2%, the cloud platform sends a parameter adjustment instruction to the local algorithm; S6. Display the optimized real-time SOC value (accuracy ≤ ±2%) and SOH value (accuracy ≤ ±2%) on the device screen and push them to the user terminal.

[0022] Based on the above, the algorithm of this invention achieves accurate dual calculation of SOC and SOH in terms of estimation accuracy. In the core algorithm module, the MIAUKF module introduces adaptive filtering and multiple innovation theory to dynamically correct the process noise covariance Q and measurement noise covariance R, controlling the SOC estimation error within 0.01. Ultimately, in conjunction with the I-EKF module, it achieves a high-precision standard of maximum SOC error ≤ ±2% and maximum SOH error ≤ ±2%, far exceeding the level of similar products in the industry. The hardware acquisition module is equipped with a high-precision processor and AD converter, with voltage detection accuracy ≤ ±0.3% and temperature detection accuracy ≤ ±1%, providing reliable data input for the algorithm and ensuring measurement accuracy from the source. Furthermore, the adaptive mechanism of the MIAUKF module can flexibly cope with different operating conditions such as battery charging and discharging and static storage, effectively resisting noise interference. The iterative mechanism constructed by the algorithm collaboration module dynamically updates and compensates for model parameter changes caused by battery aging through SOH updates, ensuring stable estimation accuracy throughout the entire lifespan. Simultaneously, the timing scheduling logic and data verification mechanism can increase the interaction frequency under complex operating conditions to ensure real-time correction. When data is abnormal, caching is activated to maintain operation, avoiding overall failure due to a single module failure, significantly enhancing the reliability of the algorithm.

[0023] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. 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 the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A real-time estimation of lithium battery SOC technique algorithm, characterized in that: The application relates to a lithium battery SOC and SOH estimation system. The hardware acquisition and control module comprises a central processing unit and an acquisition module. The cloud service platform module comprises a data transmission module, a computing power support module and a feedback regulation module. The core algorithm module comprises a multiple innovation adaptive unscented Kalman filter (MIAUKF) module, an extended Kalman filter (I-EKF) module and an algorithm coordination module. The MIAUKF module introduces adaptive filtering (AF) and multiple innovation theory, so that the process noise covariance Q and the measurement noise covariance R of the algorithm can be adaptively corrected, the robustness and stability are improved, the SOC (state of charge) of a lithium battery is estimated in real time, the SOC estimation error is controlled to be less than 0.01 through iterative calculation of a mathematical model, and the SOC estimation error is controlled to be less than 0.

01. The SOC estimation function of the MIAUKF module is as follows: ； ； wherein the input Q and R values are dynamically adjusted by an adaptive mechanism; and H is an observation matrix, is a sigma point weight, is an observation prediction value; The I-EKF module is used for updating the SOH (state of health) of the lithium battery in real time, and provides a dynamic correction basis for SOC estimation. The MIAUKF module and the I-EKF module of the algorithm coordination module form a cyclic iteration mechanism, the MIAUKF module is responsible for real-time SOC estimation, the I-EKF module is responsible for dynamic SOH updating, the data of the two modules are interactively linked, and online joint estimation of the SOC and the SOH of the lithium battery is realized.

2. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein: The central processing unit adopts an ARM V9.2 platform technology, an STM32L151RDT (64-bit) microcontroller with a Cortex-X4 core, and provides high-speed operation support.

3. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein: The acquisition module is equipped with an AVR ATMEGA64 processor and a high-speed 16-bit AD converter, realizes high-precision acquisition of voltage, current and temperature data, and provides reliable input data for the algorithm.

4. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein: The data transmission module uploads the battery data collected by a BMS device end to a cloud platform in real time through a 4G / 5G communication module, and the delay is as low as 50ms.

5. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein: The computing power support module is equipped with an SOC and SOH evaluation model based on the cloud platform, adopts a multi-threaded calculation mode, and matches the best prediction result.

6. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein: The feedback regulation module is used for transmitting the calculation result to a user end and feeding back the calculation result to a BMS master control module, and issues a battery regulation command, so that the algorithm estimation and the actual application are closed.

7. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein, The I-EKF module comprises the following specific processes: (1) input: adaptive filter signal, initial value of battery SOH, real-time current data; (2) calculation: battery SOH estimation value and voltage estimation value at K+1 moment; (3) correction: adjusting the deviation through the estimation error at K+1 moment, and outputting the corrected SOH value; (4) feedback: feeding back the real-time SOH data to the MIAUKF module, and supporting accurate estimation of the SOC.

8. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein: The algorithm coordination module dynamically characterizes time-varying parameters such as battery capacity and internal resistance through SOH, compensates for the static defects of model parameters in the MIAUKF algorithm, and makes the SOC estimation always adapt to the characteristic changes in the whole life cycle of the battery.

9. The real-time lithium battery SOC estimation technique algorithm of claim 1, wherein: The algorithm coordination module designs strict timing scheduling logic according to the real-time requirement of algorithm operation. When the battery is in complex working conditions, the data interaction frequency is increased to 100 ms / time to prioritize the real-time correction of SOC estimation; when the battery is stationary and SOH is stable, the interaction frequency is reduced to 500 ms / time to reduce hardware computing power consumption; During each data interaction, the module adds a timestamp and a check code to the transmitted data. If data delay or check failure is detected, the last period of valid data in the cache is immediately enabled to ensure continuous operation of the algorithm without interruption.

Images