Platform and method for estimating electric quantity state of energy storage system based on improved unscented Kalman filter

Abstract

The invention provides an energy storage system electric quantity state estimation platform and method based on improved unscented Kalman filtering, and relates to the technical field of electric quantity state estimation, and the method comprises the steps: carrying out the reference calibration through recognizing a standing condition, and matching a standing terminal voltage with an OCV-SOC curve to generate a three-dimensional calibration vector; performing space-time correlation on the real-time current, the voltage and the battery temperature with the reference points, predicting an SOC value by using unscented Kalman filtering, calculating a time attenuation weight and a space distance weight of each candidate reference point, and constructing a composite weight coefficient to form an effective reference set; and dynamically adjusting the covariance matrix of the process noise and the observation noise of the UKF based on the composite weight coefficient of the effective reference set, and updating the SOC estimated value. And establishing a dynamic safety evaluation domain on the SOC-temperature plane, and triggering early warning when an estimated value continuously exceeds a confidence interval for three times. According to the invention, the accuracy of electric quantity state estimation and the system security are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of state of charge estimation, and specifically provides a state of charge estimation platform and method for an energy storage system based on improved unscented Kalman filtering. Background Art

[0002] In modern energy systems, the application of energy storage technology is becoming increasingly widespread. Especially in the integration of renewable energy and the popularization of electric vehicles, the battery, as a core component, accurate estimation of its state of charge is crucial. Traditional state of charge estimation methods, such as the open-circuit voltage method and the ampere-hour metering method, often rely on static models and simple algorithms. Since these methods cannot dynamically adapt to the battery's operating state and environmental changes, the accuracy of SOC estimation is affected. At the same time, under different charge and discharge conditions of the battery, the internal resistance and polarization voltage will change accordingly. Traditional algorithms fail to effectively consider these dynamic factors, resulting in inaccurate assessment of the battery's health state, which in turn affects the safety and service life of the battery.

[0003] In addition, many existing technologies also have deficiencies in noise processing and anomaly detection. During the battery state estimation process, sensor noise and external interference often lead to data instability, making it difficult for the battery management system to make accurate decisions. Existing methods usually pay less attention to how to adjust the covariance of process noise and observation noise in real time to adapt to environmental changes. In addition, the anomaly detection mechanism is often relatively simple and cannot identify in a timely manner the situation where the SOC exceeds the safe range, increasing the risk of battery overheating and over-discharging. In summary, the deficiencies of traditional state of charge estimation technologies in dealing with dynamic characteristics, noise management, and anomaly detection need to be improved by more advanced methods to ensure the safe and reliable operation of the energy storage system.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The purpose of the present invention is to provide a state of charge estimation platform and method for an energy storage system based on improved unscented Kalman filtering to solve the problems raised in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solutions: A state of charge estimation platform for an energy storage system based on improved unscented Kalman filtering specifically includes: A reference point calibration module, which is used to identify the static condition and perform reference calibration during the operation of the energy storage system. By matching the static terminal voltage with the OCV-SOC curve, the reference SOC value is determined, and a three-dimensional calibration vector including the reference SOC value, battery temperature, and polarization voltage is generated as the reference reference point for state estimation; A spatio-temporal correlation analysis module, which is used to perform spatio-temporal correlation on the real-time collected current, voltage, and battery temperature with the reference reference point. The SOC value is predicted through the UKF unscented Kalman filter state equation, the time decay weight and spatial distance weight of each candidate reference point are calculated, a composite weight coefficient for SOC correction is constructed, and candidate reference points that meet the conditions of the composite weight coefficient are selected to form an effective reference set; A dynamic noise adjustment module, which is used to dynamically adjust the process noise covariance matrix and observation noise covariance matrix of the UKF based on the composite weight coefficient in the effective reference set. The state vector is updated through the observation equation, and the updated state vector includes the SOC estimated value, battery temperature, and polarization voltage; A safety assessment and warning module, which is used to establish a dynamic safety assessment domain with the SOC estimated value as the core in the SOC-battery temperature plane. When the SOC estimated value exceeds the confidence interval constructed by the reference reference point three times in a row, a warning is triggered, and the abnormal deviation vector is automatically recorded for model parameter correction. The corrected model parameters are applied to the UKF state estimation in the next working cycle.

[0007] Further, the specific logic for executing the reference point calibration module is as follows: When it is detected that the absolute value of the current is continuously less than and the voltage volatility is less than , it is determined as the static condition. The average terminal voltage in the last 30 seconds before the end of the static state is collected, and the closest SOC value is matched in the OCV-SOC curve database through the binary search method as the reference SOC value. At the same time, the current battery temperature is measured, and the polarization voltage at this time is identified using the least squares method and steady state values. Finally, a three-dimensional calibration vector including the reference SOC value, polarization voltage, and battery temperature is generated as the reference reference point for state estimation, and this three-dimensional calibration vector together with its calibration timestamp is stored in the reference reference point database. Among them, the three-dimensional calibration vector is expressed as follows: ; In the formula, is the reference SOC value, and are the polarization voltages, represents the first polarization voltage generated by the R1-C1 parallel circuit, Represents the second polarization voltage generated by the R2-C2 parallel circuit. R1 and R2 represent the resistors in the first polarization voltage generation circuit and the second polarization voltage generation circuit respectively, and C1 and C2 represent the capacitors in the first polarization voltage generation circuit and the second polarization voltage generation circuit respectively. is the battery temperature.

[0008] Furthermore, the current, voltage, and battery temperature collected in real time are spatially and temporally correlated with the reference reference point. The specific logic is as follows: Establish a spatio-temporal correlation model for the current at the current moment 、voltage 、and battery temperature ,retrieve the reference reference points within the time window in the reference reference point database and with a battery temperature difference not exceeding 5 °C as candidate reference points, where hours, represents the current moment; Substitute the current, voltage, and battery temperature at the current moment into the UKF unscented Kalman filter state equation to obtain the predicted SOC value at the current moment ; For each candidate reference point, calculate its corresponding time decay weight and spatial distance weight according to the following formula: ; In the formula, is the time decay weight of the th candidate reference point, is the calibrated time stamp of the th candidate reference point, is the index of the candidate reference point, is the time decay constant, used to control the decay rate of the weight over time, is the natural constant; is the spatial distance weight of the th candidate reference point, represents the predicted SOC value at the current moment , represents the reference SOC value calibrated by the th candidate reference point, represents the battery temperature measured at the current moment , represents the battery temperature at the time of calibration of the th candidate reference point.

[0009] Furthermore, construct a composite weight coefficient for SOC correction according to the time decay weight and spatial distance weight according to the following formula: ; In the formula, represents the composite weight coefficient of the th candidate reference point, and respectively represent the time decay weight and the spatial distance weight of the th candidate reference point, is the total number of candidate reference points; Calculate the composite weight coefficient for all candidate reference points, and select the candidate reference point with the composite weight coefficient to form an effective reference set.

[0010] Furthermore, based on the composite weight coefficients in the effective reference set, dynamically adjust the process noise covariance matrix and the observation noise covariance matrix of the UKF, and update the SOC estimated value through the observation equation. The specific logic is as follows: Based on the composite weight coefficients of each candidate reference point in the effective reference set, calculate the adjustment factor of the process noise covariance matrix. The formula is: ; In the formula, is the adjustment factor of the process noise covariance matrix, represents the composite weight coefficient of the th candidate reference point in the effective reference set, is the index of the candidate reference point in the effective reference set, represents the number of candidate reference points in the effective reference set, refers to the standard deviation of the composite weight coefficients in the effective reference set; According to the adjustment factor of the process noise covariance matrix, update the process noise covariance matrix : ; In the formula, is the initial covariance matrix, is the preset lower limit matrix; According to the distance between the current moment and the moment when the candidate reference point in the effective reference set is located, calculate the adjustment coefficient of the observation noise covariance matrix. The formula is as follows: ; In the formula, is the adjustment coefficient of the observation noise covariance matrix, represents the average value of the Euclidean distances of all candidate reference points; represents the current moment predicted SOC value, represents the The reference SOC value calibrated by the candidate reference points represents the current moment the measured battery temperature represents the battery temperature at the time of calibration of the candidate reference point; According to the adjustment coefficient of the observation noise covariance matrix update the observation noise covariance matrix : ; In the formula, is the initial observation noise covariance.

[0011] Furthermore, the UKF observation equation is used to update the SOC estimate value, and the formula is as follows: ; In the formula, is the observed voltage value at the current moment , represents the predicted open-circuit voltage, and the function adopts the piecewise function corresponding to the reference point with the highest weight in the effective reference set, represents the current moment predicted SOC value; is the current current at the moment, represents the ohmic internal resistance value corresponding to the reference point with the highest weight, and respectively represent the first polarization voltage and the second polarization voltage at the current moment , is the observation noise at the current moment , which follows the normal distribution , is the covariance matrix of the observation noise; Calculate the Kalman gain through the UKF unscented transform, and update the state vector. The mathematical expression is as follows: ; In the formula, represents the updated state vector at the moment, including the SOC estimate value, battery temperature and polarization voltage, is the predicted state vector at the moment, that is, the state prediction of the previous step, is the Kalman gain calculated at the moment, represents the actually measured voltage value at the moment; According to the Kalman gain Update the covariance matrix of the state estimate : ; In the formula, represents the updated state estimate covariance matrix at time , represents the predicted state covariance matrix at time , is the predicted residual covariance matrix, represents the Kalman gain transpose of.

[0012] Furthermore, the specific logic for executing the safety assessment and warning module is as follows: Establish a dynamic safety assessment domain with the SOC estimate as the core in the SOC-battery temperature plane. When it is detected that the SOC estimate exceeds the confidence interval constructed by the reference reference point in three consecutive samplings, it is marked as abnormal and a warning is triggered, and the abnormal deviation vector is automatically recorded. The abnormal deviation vector includes the SOC deviation, the battery temperature deviation, and the polarization voltage change amount for subsequent model parameter correction. The corrected model parameters will be applied to the UKF state estimate in the next working cycle; visualize the safety assessment domain with a heat map, mark the abnormal points with red pulses and display the deviation data; the SOC-battery temperature plane is a two-dimensional coordinate system, and its axis is the SOC estimate value, axis is the battery temperature; Among them, the SOC deviation is the difference between the current SOC estimate value and the SOC mean value in the reference reference point. The battery temperature deviation refers to the difference between the currently measured battery temperature and the battery temperature mean value in the reference reference point. The polarization voltage change amount refers to the difference between the currently measured polarization voltage and the polarization voltage mean value in the reference reference point.

[0013] The present invention also provides a method for estimating the state of charge of an energy storage system based on improved unscented Kalman filtering. The method for estimating the state of charge of an energy storage system based on improved unscented Kalman filtering is obtained by executing the above-mentioned platform for estimating the state of charge of an energy storage system based on improved unscented Kalman filtering, and includes: Step 1: During the operation of the energy storage system, identify the static condition and perform reference calibration. Determine the reference SOC value by matching the static terminal voltage with the OCV-SOC curve, and generate a three-dimensional calibration vector including the reference SOC value, the polarization voltage, and the battery temperature as the reference reference point for state estimation; Step 2: Spatially and temporally correlate the real-time collected current, voltage, and battery temperature with the reference points. Predict the SOC value through the UKF unscented Kalman filter state equation, calculate the time decay weight and spatial distance weight of each candidate reference point, construct a composite weight coefficient for SOC correction, and select the candidate reference points whose composite weight coefficients meet the conditions to form an effective reference set; Step 3: Dynamically adjust the process noise covariance matrix and observation noise covariance matrix of the UKF based on the composite weight coefficients in the effective reference set. Update the state vector through the observation equation. The updated state vector includes the SOC estimated value, battery temperature, and internal resistance; Step 4: Establish a dynamic safety assessment domain centered on the SOC estimated value in the SOC-temperature plane. Trigger an alarm when the SOC estimated value exceeds the confidence interval constructed by the reference points three times in a row, and automatically record the abnormal deviation vector for model parameter correction. The corrected model parameters are applied to the UKF state estimation in the next working cycle. Compared with the prior art, the beneficial effects of the present invention are: Through the dynamic calibration and spatio-temporal correlation analysis of the reference points, the SOC estimation of the present invention no longer depends on a fixed model, but can timely adapt to environmental changes and fluctuations in battery state, reducing the estimation error. Secondly, the introduction of the safety assessment and warning mechanism can issue an alarm in time when the SOC exceeds the safe range, reducing risks such as battery overheating and over-discharge, and improving the safety of the system. In addition, by recording and analyzing the abnormal deviation vector, it provides a reliable data basis for the precise correction of the subsequent model, further enhancing the intelligent level and operation efficiency of the entire energy storage system. These beneficial effects not only improve the safety and economy of battery use, but also provide new ideas and methods for the optimization of the battery management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic diagram of the platform module of the present invention; Figure 2 It is a schematic diagram of the overall method flow of the present invention; Figure 3 and Figure 4 They are respectively the change curves of the time decay weight and the spatial distance weight. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the following further details the present invention in conjunction with specific embodiments.

[0016] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0017] Embodiment: Please refer to Figure 1 , the present invention provides a technical solution: An energy storage system state of charge estimation platform based on improved unscented Kalman filter, specifically including: A reference point calibration module, which is used to identify the static condition and perform reference calibration during the operation of the energy storage system, determine the reference SOC value by matching the static terminal voltage with the OCV-SOC curve, and generate a three-dimensional calibration vector including the reference SOC value, battery temperature and polarization voltage as the reference point for state estimation; In this embodiment, the specific logic followed by the reference point calibration module is: when it is detected that the absolute value of the current is continuously less than and the voltage volatility is less than , it is determined as the static condition, the average terminal voltage in the last 30 seconds before the end of the static state is collected, the closest SOC value is matched in the OCV-SOC curve database by the binary search method as the reference SOC value, and at the same time the current battery temperature is measured, and the polarization voltage and steady-state values are identified using the least squares method. Finally, a three-dimensional calibration vector including the reference SOC value, polarization voltage and battery temperature is generated as the reference point for state estimation, and the three-dimensional calibration vector together with its calibration timestamp is stored in the reference point database, where the three-dimensional calibration vector has the following expression: ; In the formula, is the reference SOC value, and are the polarization voltages, represents the first polarization voltage generated by the R1-C1 parallel circuit, Represents the second polarization voltage generated by the R2-C2 parallel circuit. R1 and R2 respectively represent the resistors in the first polarization voltage generation circuit and the second polarization voltage generation circuit, and C1 and C2 respectively represent the capacitors in the first polarization voltage generation circuit and the second polarization voltage generation circuit. is the battery temperature.

[0018] The reference reference point calibration module realizes the accurate confirmation of the SOC value based on the OCV-SOC curve by identifying the static condition. The generated three-dimensional calibration vector includes the reference SOC value, battery temperature, and polarization voltage, providing a dynamic and reliable reference. Compared with the static calibration method of the prior art, it can better adapt to the changes of the battery under different working conditions, improving the accuracy and real-time performance of SOC estimation. In addition, through reasonable logical judgment and accurate data acquisition, this module significantly reduces the errors caused by environmental factors and equipment aging, ensuring the safety and reliability of the battery management system.

[0019] After adopting the reference reference point calibration module, the accuracy and stability of the entire energy storage system state of charge estimation platform are improved. This module provides a solid foundation for subsequent spatio-temporal correlation analysis, dynamic noise adjustment, and safety assessment and early warning, ensuring the effective matching of real-time data with the reference, thereby optimizing the effect of the UKF unscented Kalman filter. Through accurate reference SOC values and temperature data, the system can more effectively identify abnormal situations and trigger early warnings, enhancing the battery safety management ability and providing a reliable basis for the correction of model parameters, thereby improving the overall operation efficiency and lifespan of the energy storage system.

[0020] The spatio-temporal correlation analysis module is used to perform spatio-temporal correlation on the real-time collected current, voltage, and battery temperature with the reference reference point, predict the SOC value through the UKF unscented Kalman filter state equation, calculate the time decay weight and spatial distance weight of each candidate reference point, construct a composite weight coefficient for SOC correction, and select the candidate reference points that meet the conditions of the composite weight coefficient to form an effective reference set; In this embodiment, the specific logic for performing spatio-temporal correlation on the real-time collected current, voltage, and battery temperature with the nearest reference reference point is as follows: Establish a spatio-temporal correlation model for the current at the current time, voltage and battery temperature . Retrieve the reference reference points within the time window in the reference reference point database and with a battery temperature difference not exceeding 5°C as candidate reference points, where hours, represents the current time; Substitute the current, voltage, and battery temperature at the current moment into the UKF unscented Kalman filter state equation to obtain the predicted SOC value at the current moment; For each candidate reference point, calculate its corresponding time decay weight and spatial distance weight according to the following formulas: ; In the formula, is the time decay weight of the th candidate reference point, is the calibrated timestamp of the th candidate reference point, is the index of the candidate reference point, is the time decay constant, seconds, used to control the decay rate of the weight over time, is the natural constant; is the spatial distance weight of the th candidate reference point, represents the predicted SOC value at the current moment , represents the reference SOC value calibrated for the th candidate reference point, represents the battery temperature measured at the current moment , represents the battery temperature at the time of calibration of the th candidate reference point. The setting of the temperature deviation coefficient of 1 / 25 is designed based on the equivalent relationship of the battery electrochemical characteristics, reflecting the equivalence of the influence of temperature change and SOC change on the spatial distance weight under typical working conditions, and ensuring a reasonable match between the contribution degrees of temperature parameters and SOC parameters in the calculation of spatial distance weight.

[0021] Table 1: Statistical Table of Time Decay Weight and Spatial Distance Weight

[0022] It should be noted that in the above table, the time deviation corresponds to the item, the SOC deviation corresponds to the item, and the temperature deviation corresponds to the item.

[0023] Please refer to Figures 3 - 4 , in this data analysis, the time decay weight shows a typical exponential decay law. As the absolute value of the time deviation increases, the weight value gradually decays from 1 (perfect match) to nearly 0. When the time deviation reaches -4.75 hours, the weight has decayed to 0.0149, indicating that the system assigns a very low weight to historical reference points exceeding 4 hours. This decay characteristic ensures that the algorithm preferentially uses recent data.

[0024] From the tabular data, as the SOC deviation and temperature deviation increase, the changing trend of the spatial distance weight is observed: Initially, when the SOC and temperature deviations are small, the spatial distance weight remains at a high level. For example, when both the SOC and temperature are zero, the spatial distance weight reaches 1, indicating that the system is in an ideal state. However, as the SOC and temperature deviations increase, especially when the SOC reaches 0.5 or above, the spatial distance weight begins to decrease significantly and even becomes negative, indicating that the robustness of the system is severely affected. This trend shows that when the SOC and temperature deviations reach a certain threshold, the performance of the battery management system will be significantly worse than expected, which may lead to overheating, shortened lifespan, or safety hazards of the battery.

[0025] In summary, the increase in SOC and temperature deviations significantly affects the spatial distance weight, indicating the importance of accurately monitoring and regulating SOC and temperature during battery management. This analysis provides theoretical support for further battery management strategies and optimization, helping to improve the overall performance and safety of the battery system.

[0026] Based on the above analysis, the calculation of the spatio-temporal weight provides an assessment of the relative importance of each candidate reference point for the present invention. The combination of the time decay weight, SOC deviation, and temperature deviation provides a scientific basis for optimizing the battery state estimation. Through this analysis, the influence of different candidate reference points on the current state can be better understood, thus providing support for subsequent decision-making. This multi-dimensional evaluation method ensures the accuracy and reliability of the battery management system in complex environments.

[0027] A composite weight coefficient for SOC correction is constructed based on the time decay weight and spatial distance weight, and the formula is as follows: ; In the formula, represents the composite weight coefficient of the th candidate reference point, and respectively represent the time decay weight and spatial distance weight of the th candidate reference point, is the total number of candidate reference points; Calculate the composite weight coefficient for all candidate reference points, and select the candidate reference points with the composite weight coefficient to form an effective reference set.

[0028] The spatio-temporal correlation analysis module realizes the accurate prediction of the change of the SOC value by effectively matching the real-time collected current, voltage and battery temperature with the nearest reference point in space and time. Compared with the prior art, this module can dynamically consider the factors of time decay and spatial distance, ensure that the selected candidate reference points are closer to the current state in terms of time and temperature, and thus significantly improve the accuracy and timeliness of SOC estimation. This spatio-temporal correlation method based on UKF can better adapt to the rapid change of the battery state, reduce the estimation error caused by environmental factors, and comprehensively improve the reliability of the battery management system.

[0029] By adopting the spatio-temporal correlation analysis module, a more efficient SOC estimation process is realized for the overall energy storage system state of charge estimation platform. This module provides an accurate data basis for subsequent dynamic noise adjustment and safety assessment and early warning, enabling the system to correct and optimize the state estimation in a timely manner in real-time changes. The construction of the effective reference set and the generation of the composite weight coefficient significantly improve the effect of the UKF unscented Kalman filter, thus enhancing the intelligence and automation level of battery management. At the same time, the ability of this module in anomaly detection and early warning provides an important guarantee for the safe operation of the battery system and promotes the overall performance improvement of the energy storage system.

[0030] The dynamic noise adjustment module is used to dynamically adjust the process noise covariance matrix and the measurement noise covariance matrix of UKF based on the composite weight coefficient in the effective reference set, update the state vector through the observation equation, and the updated state vector includes the SOC estimated value, the battery temperature and the polarization voltage; In this embodiment, the process noise covariance matrix and the measurement noise covariance matrix of UKF are dynamically adjusted based on the composite weight coefficient in the effective reference set, and the SOC estimated value is updated through the observation equation. The specific logic is as follows: Based on the composite weight coefficients of each candidate reference point in the effective reference set, calculate the adjustment factor of the process noise covariance matrix, and the formula is: ; In the formula, is the adjustment factor of the process noise covariance matrix, represents the composite weight coefficient of the th candidate reference point in the effective reference set, represents the maximum value of the composite weight coefficients in the effective reference set, represents the number of candidate reference points in the effective reference set, is the index of the candidate reference point in the effective reference set, refers to the standard deviation of the composite weight coefficients in the effective reference set.

[0031] The dependent variable in the formula is the adjustment factor of the process noise covariance matrix, which is used to dynamically adjust the process noise covariance matrix in the UKF unscented Kalman filter algorithm to adapt to different working states and environmental conditions. The technical effect is reflected in its ability to dynamically optimize the process noise covariance matrix according to the composite weight coefficients of the candidate reference points in the effective reference set, reduce the state estimation error caused by noise, and improve the accuracy and robustness of SOC estimation. This adjustment mechanism enables the system to more flexibly cope with complex working conditions and improves the state estimation stability of the energy storage system. The two independent variables in the formula are respectively and , both of which are derived from the composite weight coefficients of the candidate reference points and reflect the matching degree between the reference points and the current state. represents the weight value of the reference point closest to the current state in the effective reference set, indicating whether there is a highly matching reference point in the reference set. The higher its value, the lower the noise adjustment requirement and the smaller the influence of the process noise. reflects the dispersion degree of the reference point weights in the effective reference set. The higher the dispersion degree, the greater the difference in weights within the reference set, and the state estimation may be affected by unstable reference points, requiring a higher adjustment factor to compensate for the noise. When has a large value, it indicates that there are highly matching reference points in the effective reference set, and the system has a lower adjustment requirement for the process noise. Therefore, will decrease, reducing the adjustment amplitude of the process noise covariance matrix. When has a large value, it indicates that the difference in the weight distribution of the reference points in the reference set is more obvious, and the system has a higher adjustment requirement for the process noise. Therefore, will increase to enhance the compensation effect for the process noise.

[0032] Update the process noise covariance matrix according to the adjustment factor of the process noise covariance matrix: ; In the formula, is the initial covariance matrix, and is the preset lower limit matrix; Calculate the adjustment coefficient of the observation noise covariance matrix according to the distance between the current moment and the moment when the candidate reference points in the effective reference set are located. The formula is as follows: ; In the formula, is the adjustment coefficient of the observation noise covariance matrix, represents taking the average value of the Euclidean distances of all candidate reference points; represents the current moment predicted SOC value, represents the The reference SOC value calibrated by a candidate reference point represents the current time The measured battery temperature represents the battery temperature at the time of calibrating the th candidate reference point; The dependent variable in the formula is the adjustment coefficient of the observation noise covariance matrix , which is used to dynamically adjust the observation noise covariance matrix in the UKF unscented Kalman filter algorithm to adapt to different environmental and state conditions. The technical effect is reflected in its ability to dynamically optimize the observation noise covariance matrix according to the average distance between the current time and the candidate reference points, thereby improving the accuracy and robustness of SOC estimation. By adjusting the noise covariance, the system can better reflect the uncertainty of the current state and ensure the stability and reliability of SOC estimation under variable operating conditions.

[0033] The independent variables in the formula are mainly , that is, the average Euclidean distance of the candidate reference points. This independent variable reflects the degree of proximity between the current state and each reference point in the reference set. When the distance between the current state and the reference point is small, it indicates a high degree of matching between the current state and the reference set. At this time, the system has a lower requirement for adjusting the observation noise. Therefore will decrease accordingly; conversely, when the distance is large, the matching degree is low, and the system will require a higher adjustment of the observation noise covariance to enhance the compensation for uncertainty. Therefore, the independent variable directly affects the magnitude of the dependent variable by reflecting the matching degree of the current state, thereby realizing the dynamic adjustment of the observation noise.

[0034] When the average distance of the candidate reference points increases, it indicates that the matching degree between the current state and the reference set decreases, which means that the system needs a higher adjustment of the observation noise covariance to cope with possible state estimation errors. Therefore will increase with the increase of the distance, thereby increasing the value of the observation noise covariance matrix. The design of this positive correlation relationship ensures that the system can dynamically adjust the observation noise according to the distance between the current state and the reference points, thereby maintaining the accuracy and stability of SOC estimation under different working conditions.

[0035] Update the observation noise covariance matrix according to the adjustment coefficient of the observation noise covariance matrix : ; In the formula, is the initial observation noise covariance.

[0036] Update the SOC estimated value using the UKF observation equation. The formula is as follows: ; Wherein, is the observed voltage value at the current moment , represents the open circuit voltage predicted based on , and the function adopts the piecewise function corresponding to the benchmark point with the highest weight in the effective reference set, represents the SOC value predicted at the current moment ; is the current at the current moment , represents the ohmic internal resistance value corresponding to the benchmark point with the highest weight, and respectively represent the first polarization voltage and the second polarization voltage at the current moment , is the observation noise at the current moment , which follows the normal distribution , is the covariance matrix of the observation noise; The Kalman gain is calculated through the unscented transform (UKF) of the UKF, and the state vector is updated. The mathematical expression is as follows: ; Wherein, represents the updated state vector at time , including the SOC estimated value, the battery temperature, and the polarization voltage, is the predicted state vector at time , that is, the state prediction of the previous step, is the Kalman gain calculated at time , represents the actually measured voltage value at time ; According to the Kalman gain , the covariance matrix of the state estimate is updated: ; Wherein, represents the updated covariance matrix of the state estimate at time , represents the predicted state covariance matrix at time , is the predicted residual covariance matrix, represents the transpose of the Kalman gain .

[0037] Reflects the uncertainty of state estimation. In Kalman filtering, the update of the covariance matrix is a crucial step to ensure that the system can adjust its confidence in the state according to new observation information. By updating the covariance matrix, the credibility of the current state can be evaluated more accurately. Represents the uncertainty assessment of the state without new observation information. This matrix is predicted based on the previous state estimate and the dynamic model, providing a preliminary covariance estimate. Determines the degree of uncertainty correction of the new observation information to the state estimate. The calculation of the Kalman gain takes into account the influence of the predicted state estimate covariance and the observation noise to ensure the optimal combination of prior information and measurement information during the update process. Represents the uncertainty introduced by the observation noise during the update process. By combining the Kalman gain with the residual covariance matrix, the updated covariance matrix Can effectively reflect the impact of the latest observation results on the uncertainty of the state estimate. In the update formula, the product term of the Kalman gain Is subtracted from the predicted covariance matrix, meaning that with the introduction of new observation information, the uncertainty of the state estimate usually decreases. This process makes the state estimate more accurate and improves the system's prediction ability for future states.

[0038] The dynamic noise adjustment module can dynamically adjust the process noise covariance matrix and the observation noise covariance matrix of the UKF by analyzing the composite weight coefficients in the effective reference set in real time. Compared with the existing technology, this module improves the adaptability of the system under different working conditions and environments, thereby reducing the impact of noise on the SOC estimation. This flexible noise processing mechanism enables the system to better cope with the fluctuations caused by battery state and environmental changes, ensuring the accuracy and reliability of the SOC estimation, and ultimately enhancing the operating efficiency and safety of the energy storage system.

[0039] After adopting the dynamic noise adjustment module, the performance of the overall energy storage system state of charge estimation platform has been significantly improved. By real-time adjusting the noise covariance matrix, the system can effectively optimize the state estimation process of the UKF, making the SOC estimation value more accurate and stable. This not only enhances the detection ability of abnormal situations but also provides a more solid foundation for subsequent safety assessment and early warning, reducing potential risks, and thus promoting the global safety and efficiency improvement of the energy storage system. Overall, the introduction of this module ensures the continuous and efficient operation of the system under complex working conditions and improves the intelligent level.

[0040] The safety assessment and early warning module is used to establish a dynamic safety assessment domain with the SOC estimated value as the core in the SOC-battery temperature plane. When the SOC estimated value exceeds the confidence interval constructed by the reference points three times in a row, an early warning is triggered, and the abnormal deviation vector is automatically recorded for model parameter correction. The corrected model parameters are applied to the UKF state estimation in the next working cycle; In this embodiment, the specific logic followed by the safety assessment and early warning module is as follows: A dynamic safety assessment domain with the SOC estimated value as the core is established in the SOC-battery temperature plane. When it is detected that the SOC estimated value exceeds the confidence interval constructed by the reference points in three consecutive samplings, it is marked as abnormal and an early warning is triggered, and the abnormal deviation vector is automatically recorded. The abnormal deviation vector includes the SOC deviation, the battery temperature deviation, and the polarization voltage change for subsequent model parameter correction. The corrected model parameters will be applied to the UKF state estimation in the next working cycle; the safety assessment domain is visualized as a heat map, and the abnormal points are marked with red pulses and the deviation data is displayed; the SOC-battery temperature plane is a two-dimensional coordinate system, and its x-axis is the SOC estimated value, y-axis is the battery temperature; Among them, the SOC deviation is the difference between the current SOC estimated value and the SOC mean value in the reference points, the battery temperature deviation refers to the difference between the currently measured battery temperature and the battery temperature mean value in the reference points, and the polarization voltage change refers to the difference between the currently measured polarization voltage and the polarization voltage mean value in the reference points.

[0041] By establishing a dynamic safety assessment domain with the SOC estimated value as the core, the safety assessment and early warning module can monitor the working state of the battery in real time and automatically trigger an early warning when the SOC value continuously exceeds the confidence interval. This mechanism has significant advantages compared with traditional static monitoring methods because it can identify potential safety hazards in a timely manner and avoid system failures or safety accidents caused by delayed responses. In addition, the module can automatically record the abnormal deviation vector, providing data support for subsequent model parameter correction, thereby continuously optimizing the performance of the battery management system and improving the overall safety and reliability.

[0042] After the introduction of the safety assessment and early warning module, the intelligence and safety of the entire energy storage system power state estimation platform have been significantly enhanced. The real-time monitoring and early warning capabilities of this module not only improve the rapid response ability to abnormal situations but also provide a basis for dynamic noise adjustment to ensure the accuracy of SOC estimation. When an abnormality occurs, the module can automatically adjust the model parameters, optimize the state estimation, and thus improve the stability and reliability of the system. This closed-loop feedback mechanism enables the battery management system to have stronger adaptability in a dynamic environment, ensuring the safe and efficient operation of the energy storage system and promoting the improvement of the overall technical level.

[0043] Please refer to Figure 2 , the method for estimating the state of charge of an energy storage system based on improved unscented Kalman filter, and the specific steps include: Step 1: During the operation of the energy storage system, identify the static condition and perform reference calibration. Determine the reference SOC value by matching the static terminal voltage with the OCV-SOC curve, and generate a three-dimensional calibration vector including the reference SOC value, polarization voltage, and battery temperature as the reference reference point for state estimation; Step 2: Spatially and temporally correlate the real-time collected current, voltage, and battery temperature with the reference reference point. Predict the SOC value through the UKF unscented Kalman filter state equation, calculate the time decay weight and spatial distance weight of each candidate reference point, construct a composite weight coefficient for SOC correction, and select the candidate reference points that meet the conditions of the composite weight coefficient to form an effective reference set; Step 3: Dynamically adjust the process noise covariance matrix and observation noise covariance matrix of the UKF based on the composite weight coefficient in the effective reference set, update the state vector through the observation equation, and the updated state vector includes the SOC estimated value, battery temperature, and internal resistance; Step 4: Establish a dynamic safety assessment domain centered on the SOC estimated value in the SOC-temperature plane. Trigger an alarm when the SOC estimated value continuously exceeds the confidence interval constructed by the reference reference point three times, and automatically record the abnormal deviation vector for model parameter correction. The corrected model parameters are applied to the UKF state estimation of the next working cycle.

[0044] The above formulas are all dimensionless and take their numerical calculations. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0045] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0046] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0047] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. An energy storage system state of charge estimation platform based on improved unscented Kalman filter, characterized in that Specifically, it includes: A reference point calibration module, which is used to identify the stationary condition and perform reference calibration during the operation of the energy storage system. By matching the stationary terminal voltage with the OCV-SOC curve, the reference SOC value is determined, and a three-dimensional calibration vector including the reference SOC value, battery temperature, and polarization voltage is generated as the reference reference point for state estimation; A spatio-temporal correlation analysis module, which is used to perform spatio-temporal correlation on the real-time collected current, voltage, and battery temperature with the reference reference point. The SOC value is predicted through the UKF unscented Kalman filter state equation, the time decay weight and spatial distance weight of each candidate reference point are calculated, a composite weight coefficient for SOC correction is constructed, and the candidate reference points that meet the conditions of the composite weight coefficient are selected to form an effective reference set; A dynamic noise adjustment module, which is used to dynamically adjust the process noise covariance matrix and observation noise covariance matrix of the UKF based on the composite weight coefficient in the effective reference set, and update the state vector through the observation equation. The updated state vector includes the SOC estimated value, battery temperature, and polarization voltage; A safety assessment and warning module, which is used to establish a dynamic safety assessment domain with the SOC estimated value as the core in the SOC-battery temperature plane. When the SOC estimated value exceeds the confidence interval constructed by the reference reference point three times in a row, a warning is triggered, and the abnormal deviation vector is automatically recorded for model parameter correction. The corrected model parameters are applied to the UKF state estimation in the next working cycle.

2. The state of charge estimation platform for an energy storage system based on improved unscented Kalman filter according to claim 1, characterized in that: The specific logic followed by the execution of the reference point calibration module is as follows: When it is detected that the absolute value of the current is continuously less than and the voltage volatility is less than for 5 minutes, it is determined as the static condition. The average terminal voltage in the last 30 seconds before the end of the static state is collected, and the closest SOC value is matched in the OCV-SOC curve database through the binary search method as the reference SOC value. At the same time, the current battery temperature is measured, and the polarization voltage and steady-state values are identified using the least squares method. Finally, a three-dimensional calibration vector containing the reference SOC value, polarization voltage, and battery temperature is generated as the reference point for state estimation, and this three-dimensional calibration vector along with its calibration timestamp is stored in the reference point database. Among them, the expression of the three-dimensional calibration vector is as follows: ； Wherein, is the reference SOC value, and are polarization voltages, represents the first polarization voltage generated by the R1-C1 parallel circuit, represents the second polarization voltage generated by the R2-C2 parallel circuit. R1 and R2 respectively represent the resistances in the first polarization voltage generation circuit and the second polarization voltage generation circuit, and C1 and C2 respectively represent the capacitances in the first polarization voltage generation circuit and the second polarization voltage generation circuit, is the battery temperature.

3. An estimation platform for the state of charge of an energy storage system based on an improved unscented Kalman filter according to claim 1, characterized in that: The specific logic for performing spatio-temporal correlation on the real-time collected current, voltage, and battery temperature with the reference reference point is as follows: Establish a spatio-temporal correlation model for the current moment of the current current , voltage and battery temperature , retrieve the reference points within the time window in the reference point database with a battery temperature difference not exceeding 5°C as candidate reference points, where hours, represents the current moment; Substitute the current, voltage, and battery temperature at the current moment into the UKF unscented Kalman filter state equation to obtain the predicted SOC value; For each candidate reference point, the corresponding time decay weight and spatial distance weight are calculated according to the following formulas: ； Wherein, is the time decay weight of the th candidate reference point, is the calibration timestamp of the th candidate reference point, is the index of the candidate reference point, is the time decay constant, which is used to control the decay rate of the weight over time, is the natural constant; is the spatial distance weight of the th candidate reference point, represents the SOC value predicted at the current time , represents the reference SOC value calibrated for the th candidate reference point, represents the battery temperature measured at the current time , represents the battery temperature at the time of calibration of the th candidate reference point.

4. A state of charge estimation platform for an energy storage system based on improved unscented Kalman filtering according to claim 3, characterized in that: The composite weight coefficient for SOC correction is constructed based on the time decay weight and spatial distance weight according to the following formula: ； In the formula, represents the composite weight coefficient of the th candidate reference point, and represent the time decay weight and the spatial distance weight of the th candidate reference point respectively, is the total number of candidate reference points; Calculate the composite weight coefficient for all candidate reference points, and select the candidate reference points with the composite weight coefficient to form an effective reference set.

5. A state of charge estimation platform for an energy storage system based on an improved unscented Kalman filter according to claim 1, characterized in that: The specific logic for dynamically adjusting the process noise covariance matrix and observation noise covariance matrix of the UKF based on the composite weight coefficient in the effective reference set and updating the SOC estimated value through the observation equation is as follows: Based on the composite weight coefficients of each candidate reference point in the effective reference set, the adjustment factor of the process noise covariance matrix is calculated according to the following formula: ； In the formula, is the adjustment factor of the process noise covariance matrix, represents the composite weight coefficient of the th candidate reference point in the effective reference set, is the index of the candidate reference point in the effective reference set, represents the number of candidate reference points in the effective reference set, refers to the standard deviation of the composite weight coefficients in the effective reference set; Adjustment factor for the process noise covariance matrix Update the process noise covariance matrix : ； In the formula, is the initial covariance matrix, is the preset lower limit matrix; According to the distance between the current moment and the moment when the candidate reference point in the effective reference set is located, the adjustment coefficient of the observation noise covariance matrix is calculated according to the following formula: ； In the formula, is the adjustment coefficient of the observation noise covariance matrix, represents taking the average of the Euclidean distances of all candidate reference points; represents the current time predicted SOC value, represents the reference SOC value calibrated by the th candidate reference point, represents the battery temperature measured at the current time represents the th battery temperature during the calibration of the candidate reference point; Adjustment coefficient according to the observation noise covariance matrix Update the observation noise covariance matrix : ； where is the initial observation noise covariance.

6. A state of charge estimation platform for an energy storage system based on improved unscented Kalman filtering according to claim 1, characterized in that: The SOC estimated value is updated using the UKF observation equation according to the following formula: ； Wherein, is the observed voltage value at the current moment, represents the open circuit voltage predicted based on , and the function adopts the piecewise function corresponding to the benchmark point with the highest weight in the effective reference set, represents the predicted SOC value at the current moment; is the current at the current moment, represents the ohmic internal resistance value corresponding to the benchmark point with the highest weight, and respectively represent the first polarization voltage and the second polarization voltage at the current moment, is the observation noise at the current moment, which follows a normal distribution , and is the covariance matrix of the observation noise; Calculate the Kalman gain through the UKF unscented transformation , and update the state vector. The mathematical expression is as follows: ; Wherein, represents the updated state vector at time , including the SOC estimation value, battery temperature and polarization voltage, is the predicted state vector at time , that is, the state prediction of the previous step, is the Kalman gain calculated at time , represents the actually measured voltage value at time ; According to the Kalman gain Update the covariance matrix of the state estimate : ； In the formula, represents the updated state estimation covariance matrix at time , represents the predicted state covariance matrix at time , is the predicted residual covariance matrix, represents the transpose of the Kalman gain .

7. A state of charge estimation platform for an energy storage system based on improved unscented Kalman filter according to claim 1, characterized in that: The specific logic for executing the safety assessment and warning module is as follows: A dynamic safety assessment domain with the SOC estimated value as the core is established in the SOC-battery temperature plane. When it is detected that the SOC estimated value exceeds the confidence interval constructed by the reference reference point in three consecutive samples, it is marked as abnormal and a warning is triggered, and the abnormal deviation vector is automatically recorded. The abnormal deviation vector includes the SOC deviation amount, battery temperature deviation amount, and polarization voltage change amount for subsequent model parameter correction. The corrected model parameters will be applied to the UKF state estimation in the next working cycle; Visualize the safety assessment domain with a heat map, mark abnormal points with red pulses and display deviation data; the SOC-battery temperature plane is a two-dimensional coordinate system, and its axis is the SOC estimated value, axis is the battery temperature; Among them, the SOC deviation is the difference between the current SOC estimated value and the SOC mean value in the reference point. The battery temperature deviation refers to the difference between the currently measured battery temperature and the battery temperature mean value in the reference point. The polarization voltage change refers to the difference between the currently measured polarization voltage and the polarization voltage mean value in the reference point.

8. A method for estimating the state of charge of an energy storage system based on an improved unscented Kalman filter, characterized in that: The method for estimating the state of charge of an energy storage system based on improved unscented Kalman filter is obtained by executing the platform for estimating the state of charge of an energy storage system based on improved unscented Kalman filter described in any one of claims 1-7, and includes: Step 1: During the operation of the energy storage system, identify the static condition and perform reference calibration. Determine the reference SOC value by matching the static terminal voltage with the OCV-SOC curve, and generate a three-dimensional calibration vector including the reference SOC value, polarization voltage, and battery temperature as the reference point for state estimation. Step 2: Spatially and temporally correlate the real-time collected current, voltage, and battery temperature with the reference point. Predict the SOC value through the UKF unscented Kalman filter state equation, calculate the time decay weight and spatial distance weight of each candidate reference point, construct a composite weight coefficient for SOC correction, and select the candidate reference points that meet the conditions of the composite weight coefficient to form an effective reference set. Step 3: Dynamically adjust the process noise covariance matrix and observation noise covariance matrix of the UKF based on the composite weight coefficient in the effective reference set, update the state vector through the observation equation, and the updated state vector includes the SOC estimated value, battery temperature, and internal resistance. Step 4: Establish a dynamic safety assessment domain with the SOC estimated value as the core in the SOC-temperature plane. When the SOC estimated value exceeds the confidence interval constructed by the reference point three times continuously, trigger an alarm, and automatically record the abnormal deviation vector for model parameter correction. The corrected model parameters are applied to the UKF state estimation in the next working cycle.

Images