Battery module state estimation method considering functional state dynamic characteristics

Abstract

The invention provides a battery module state estimation method considering functional state dynamic characteristics, and relates to the technical field of battery management. The method comprises the following steps: collecting characteristic parameters of a single battery, and calculating polarization voltage in the battery; according to the factors affecting the state estimation of the battery module, a state-of-charge model representing the short-time scale characteristic of the battery and a state-of-health model representing the long-time scale characteristic of the battery are constructed respectively; establishing a functional state dynamic model of the single battery and a state estimation model under a dual-time scale; and expanding the functional state dynamic model of the single battery and the state estimation model under the dual-time scale to a battery module level, and establishing the functional state dynamic model of the battery module and the state estimation model under the dual-time scale so as to calculate the state estimation value and the functional state of the battery module. According to the invention, the accuracy and dynamic response capability of battery state estimation are effectively improved, and a basis is provided for long-term health management of the battery.

Description

Technical Field

[0001] This invention relates to the field of battery management technology, and in particular to a method for estimating the state of a battery module that takes into account the dynamic characteristics of its functional state. Background Technology

[0002] With the accelerated scaling up of the global new energy vehicle industry, the accuracy of power battery system state estimation has become a core technological bottleneck restricting the improvement of vehicle energy efficiency, safety, and lifespan. High-precision state estimation is crucial for optimizing battery energy management and preventing overcharging and over-discharging, and is also an important foundation for achieving early fault warning and extending battery life. Studies have shown that accurate state estimation can improve the accuracy of electric vehicle range prediction by approximately 15%-20%, significantly reducing the battery system failure rate. Against this backdrop, the intelligent development of battery management systems places higher demands on the accuracy and dynamic response speed of state estimation technology throughout its entire lifecycle.

[0003] However, existing battery module state estimation methods still face significant challenges. Traditional methods often focus on individual cells, failing to fully account for the dynamic characteristics of battery functional state under conditions such as temperature, aging, and rate, resulting in peak power prediction errors generally exceeding 10%. At the module level, existing estimation strategies often ignore the nonlinear impact of inconsistencies between individual cells on the overall functional state under dynamic operating conditions, making it difficult to achieve coordinated tracking of state of charge and state of health. Furthermore, most models use a single time scale, unable to simultaneously adapt to the second-level changes in state of charge and the monthly / yearly evolution of state of health. In complex real-world operating conditions, estimation errors tend to accumulate and amplify, hindering the full realization of battery performance and the improvement of safety management levels. Summary of the Invention

[0004] To address the shortcomings of the existing technologies, this invention proposes a battery module state estimation method that takes into account the dynamic characteristics of functional states. This method is developed by establishing a dynamic functional state model and combining it with a dual-time-scale estimation strategy, extending the estimation object from individual battery cells to the module level. The aim is to effectively improve the accuracy and dynamic response capability of battery state estimation and provide a data foundation for the long-term health management of batteries.

[0005] This invention proposes a battery module state estimation method that takes into account the dynamic characteristics of functional states. The method includes the following steps:

[0006] Collect characteristic parameters of individual cells and calculate the electrochemical polarization voltage and concentration polarization voltage inside the cell;

[0007] Based on the factors affecting the state estimation of battery modules, a state of charge model characterizing the short-term characteristics of the battery and a health state model characterizing the long-term characteristics of the battery are constructed respectively.

[0008] Based on the electrochemical polarization voltage, concentration polarization voltage, state of charge model, and health state model, a dynamic model of the functional state of a single cell is established.

[0009] Based on the state of charge model and the state of health model, a state estimation model for a single cell under dual time scales is constructed.

[0010] Based on the functional state dynamic model of a single battery cell, a functional state dynamic model of the battery module is constructed.

[0011] Based on the state estimation model of a single cell under dual time scales and the functional state dynamic model of the battery module, a state estimation model of the battery module under dual time scales is constructed.

[0012] Based on the state estimation model of the battery module under dual time scales, the estimated value of the state of charge and the estimated value of the health state of the battery module at the current moment are calculated. Then, based on the functional state dynamic model of the battery module, the functional state of the battery module at the current moment is calculated.

[0013] Furthermore, the characteristic parameters of the single battery cell include: voltage, current, internal resistance, and capacitance.

[0014] Furthermore, the specific method for constructing a state-of-charge model characterizing the short-term characteristics of the battery and a health state model characterizing the long-term characteristics of the battery based on the factors affecting the state estimation of the battery module is as follows:

[0015] The factors affecting the state estimation of the battery module are identified, including: voltage, current, internal resistance, capacitance, state of charge, and state of health.

[0016] A dynamic model of the state of charge (SOC) characterizing the short-timescale properties of a battery is established based on the ampere-hour integral method.

[0017]

[0018] Among them, S OC (t) represents the current time. The state of charge; For the current moment Operating current; This refers to the rated capacity of a single battery cell; For Coulomb efficiency; For the current moment ; health status;

[0019] Establish a health state evolution model that characterizes the long-term characteristics of the battery, including: a health state evolution model that considers battery capacity decay and a health state evolution model that considers battery internal resistance growth.

[0020] The health state evolution model considering battery capacity degradation is as follows:

[0021]

[0022] Among them, S OH,Q For capacity-based health status; The number of complete cycles a single cell has undergone; This represents the average operating temperature of a single battery cell. This is the capacity decay coefficient; This is an empirical parameter for how capacity decay varies with the number of cycles; The activation energy for the aging reaction; This is the universal gas constant;

[0023] The health state evolution model considering the increase in battery internal resistance is as follows:

[0024]

[0025] in, The internal resistance in ohms under the current aging condition; The initial ohmic internal resistance of a single cell; This is the internal resistance growth coefficient; This is an empirical parameter showing how the internal resistance increases with the number of cycles. This is the activation energy during the internal resistance growth process; This represents the current state of health based on internal resistance under aging conditions.

[0026] Furthermore, the specific method for establishing a dynamic model of the functional state of a single cell based on the electrochemical polarization voltage, concentration polarization voltage, state of charge model, and health state model is as follows:

[0027] Based on the electrochemical polarization voltage and concentration polarization voltage, a dynamic response equation for the terminal voltage is established according to the polarization effect inside the battery.

[0028] Based on the dynamic response equation of the terminal voltage, a prediction model for the terminal voltage is established by iteratively calculating the terminal voltage at the current moment.

[0029] A maximum current constraint model is established based on the battery's voltage and temperature limitations.

[0030] Define the maximum power of the battery under both discharge and charging conditions;

[0031] Based on the predicted terminal voltage model, the maximum current constraint model, and the defined limiting power, a functional state dynamic model of a single cell is established, expressed as:

[0032]

[0033] in, This represents the maximum discharge power in the functional state. The maximum charging power for the functional state; Duration; The minimum permissible voltage for battery operation; The maximum permissible voltage for battery operation; The maximum allowable discharge current; This represents the maximum allowable charging current.

[0034] Furthermore, the specific method for constructing a state estimation model for a single cell under dual time scales based on the state of charge model and the health model is as follows:

[0035] Based on the state of charge model and the state of health model, the state vectors of a single cell are defined at short-term and long-term time scales, respectively.

[0036] Based on the definition of the state vector of a single cell in short and long time scales, a time-scale state transition function is established.

[0037] Based on the state transition function of the time scale, the observation residual correction mechanism is used to recursively estimate the state vector of a single cell at a short time scale and at a long time scale, so as to obtain the optimal estimate of the state vector of the single cell at both the short and long time scales, thus completing the dynamic modeling of the functional state of the single cell.

[0038] Furthermore, the method for recursively estimating the state vector of a single cell at a short time scale and at a long time scale using the observation residual correction mechanism based on the state transition function of the time scale is as follows:

[0039] Based on the state transition function of the time scale, state estimation models for individual cells under short and long time scales are established.

[0040] For short time scales, the sampling time is based on the previous short time scale. The optimal estimate of the state vector and the current long-scale sampling time The state vector estimate is obtained from the state prediction equation of a single cell over a short timescale, thus yielding the current short-timescale sampling time. State vector prediction value;

[0041] Obtain the current short-timescale sampling time of a single cell. The actual observation vector is obtained, and the observation residual is calculated. Then, based on the state update equation of a single cell at a short time scale, the current short time scale sampling time is updated. The predicted state vector value is updated to obtain the current short-timescale sampling moment. The optimal estimate of the state vector;

[0042] For long time scales, based on the sampling time of the previous long time scale The optimal estimate of the state vector and the current long-scale sampling time The average input vector is used to obtain the current long-term sampling time based on the state prediction equation for a single cell over a long time scale. State vector prediction value;

[0043] Using the sampling time from the previous long time scale up to the current long-term sampling time The accumulated observation residual vector is used to construct virtual observation increments over a long time scale;

[0044] Based on the state update equation of a single cell over a long time scale, the virtual observation increment over a long time scale is used to update the current long-term sampling time. The predicted state vector is updated to obtain the current long-timescale sampling time. The optimal estimate of the state vector.

[0045] Furthermore, the specific method for constructing the functional state dynamic model of the battery module based on the functional state dynamic model of a single battery cell is as follows:

[0046] For the first in the battery module For each individual cell, calculate the deviation rate of its rated capacity and the deviation rate of its ohmic internal resistance.

[0047] Based on the module inconsistency parameter matrix, the voltage safety margin deviation of the i-th individual cell is calculated according to the deviation rate of the rated capacity and the deviation rate of the ohmic internal resistance.

[0048] Based on the voltage safety margin deviation and combined with the voltage distribution of individual cells, a voltage constraint model for the battery module is constructed.

[0049] Definition of the first Inconsistency factor of individual cells;

[0050] Based on the battery module terminal voltage constraint model and the inconsistency factors of all individual cells, a functional state dynamic model of the battery module is constructed.

[0051] Furthermore, the specific method for constructing the state estimation model of the battery module under dual time scales based on the state estimation model of a single battery cell under dual time scales and the functional state dynamic model of the battery module is as follows:

[0052] Based on the state estimation model of a single cell under dual time scales, the state vector of the battery module is constructed.

[0053] Define the system topology matrix of the battery modules based on their electrical connection methods;

[0054] Based on the system topology matrix of the battery module, module-level constraint equations are established.

[0055]

[0056] in, This represents the current vector of a single cell. Here is the system current topology matrix; This represents the total current of the module. This is the current constraint error term; This refers to the battery module terminal voltage. The system voltage topology matrix; This represents the voltage vector of a single cell. This is the voltage constraint error term;

[0057] Based on the module-level constraint equation, the state vector of the battery module under the short time scale and the state vector under the long time scale are recursively estimated to obtain the optimal estimated values ​​of the state vector of the battery module under the short time scale and the long time scale.

[0058] Based on the optimal estimate of the state vector of the battery module in a short time scale, obtain The estimated state of charge of each individual cell at any given time;

[0059] Based on the optimal estimate of the state vector of the battery module over a long time scale, obtain The estimated health status of each individual cell at any given time.

[0060] Furthermore, the specific method for recursively estimating the state vector of the battery module under short-term and long-term time scales based on the module-level constraint equations to obtain the optimal estimates of the state vector of the battery module under short-term and long-term time scales is as follows:

[0061] Based on the state prediction equation of a single cell over a short timescale, a state prediction equation of the battery module over a short timescale is constructed to obtain the state prediction equation of the battery module at the current short timescale sampling time. The predicted value of the state vector;

[0062]

[0063] in, For the first Individual cells at the current short-timescale sampling time State vector prediction value; For the first The short-timescale state transition function of a single cell; For the first Individual cells in The optimal estimate of the state vector at time 1; For the i-th single cell at the current long-time sampling time... State vector prediction value; This is the topological interaction gain matrix; For topology state interaction items; In order to be with the first A collection of adjacent individual cells; In order to be with the first The adjacent single cell Individual cells; These are the topological weight coefficients; For the j-th adjacent single cell in State vector estimate at time t;

[0064] Obtain the current short-timescale sampling time of the battery module The actual observation vector, and calculate the observation residual vector of the battery module;

[0065] Based on the short-timescale state update equation of the battery module, and using the observation residual vector of the battery module, the battery module at the current short-timescale sampling time is updated. The state vector prediction value is updated to obtain the battery module at the current short-timescale sampling time. The optimal estimate of the state vector;

[0066]

[0067] in, For the first Individual cells at the current short-timescale sampling time The optimal estimate of the state vector; For the first The Kalman gain matrix of each adjacent cell at time t; For the first The observation residual vector of each individual cell; This is the topology observation coordination matrix; In order to be with the first The observation residual vector of the set of adjacent individual cells of a single cell; This is the topology compatibility strength coefficient; For the first The topological weight matrix of each individual cell;

[0068] Based on the state prediction equation of a single cell over a long time scale, a global state update equation for the battery module is constructed to obtain the battery module's state at the current long-time scale sampling time. The optimal estimate of the state vector;

[0069]

[0070] in, For the battery module at the current long-term sampling time State vector estimate; For the first Current long-term sampling time for individual cells The optimal estimate of the state vector; This is the fusion gain matrix; This is a consistency correction amount; This is the arithmetic mean vector of the long-time-scale state vector estimates for each individual cell. For battery modules State vector estimate at time t; This refers to the number of individual cells connected in series or parallel within the battery module.

[0071] Furthermore, the specific method for calculating the estimated state of charge and state of health of the battery module at the current moment based on the state estimation model of the battery module under dual time scales, and then calculating the functional state of the battery module at the current moment based on the functional state dynamic model of the battery module, is as follows:

[0072] Define the weighted average of the state of charge of each individual cell in the battery module, and calculate the estimated state of charge and state of health of the battery module at the current moment based on the state estimation model of the battery module under dual time scales.

[0073]

[0074] in, For the battery module at the current long-term sampling time Health status estimate; and All are weighting coefficients; For the first Individual cells in Constantly monitor capacity health status; For the first Individual cells in A healthy state based on internal resistance at all times; For the battery module at the current short timescale sampling time The estimated state of charge; For the first Weighting coefficients for individual cells over short time scales; For the first Individual cells in The estimated state of charge at time t;

[0075] Based on the estimated state of charge and state of health of the battery module at the current moment, the maximum charging power and maximum discharging power of the battery module at the current moment are calculated according to the functional state dynamic model of the battery module.

[0076]

[0077] in, The maximum sustainable discharge power of the battery module in its functional state; The maximum sustainable charging power for the battery module's functional state; This is the minimum allowable voltage for a single cell. This is the maximum allowable voltage of a single cell. Duration; In order to be in At that moment, the Individual cells in The maximum allowable discharge current; In order to be in At that moment, the Individual cells in The maximum allowable charging current; For the first Inconsistency factor of individual cells.

[0078] The beneficial effects of adopting the above technical solution are as follows:

[0079] The method of this invention takes into account the key characteristic parameters of the battery, and uses the electrochemical polarization voltage and concentration polarization voltage to characterize the internal dynamic response. It establishes a health state model on a long time scale and a state of charge model on a short time scale, respectively, which effectively reduces the mutual interference between the rapid dynamic process and the slow degradation process in the state estimation.

[0080] The method of this invention constructs a functional state dynamic model based on parameter correlation, couples internal state with external operating conditions, and enhances the adaptability of the state estimation model under actual complex working conditions.

[0081] The method of this invention establishes a dual-time-scale state estimation framework by analyzing the time-scale adaptability, which ensures the real-time performance of the battery dynamic response while avoiding the accumulation of errors caused by single-time-scale estimation, thus achieving synergistic optimization of estimation accuracy and computational efficiency.

[0082] The method of this invention constructs a functional state dynamic model that reflects the internal state of the battery and external operating conditions. It extends the model at the cell level to the module level by considering the inconsistencies between individual battery cells, and completes the joint estimation of the functional state, state of charge and health state at the module level. This improves the reliability of the battery management system and effectively enhances the accuracy of state estimation and dynamic response performance. Attached Figure Description

[0083] Figure 1 This is a flowchart of a battery module state estimation method that takes into account the dynamic characteristics of functional states in this embodiment.

[0084] Figure 2 This is a schematic diagram of a battery module state estimation method that takes into account the dynamic characteristics of functional states in this embodiment.

[0085] Figure 3 The following is a comparison of the battery state estimation results under two time scales in this embodiment; where (a) is the battery state estimation result under a short time scale; and (b) is the battery state estimation result under a long time scale.

[0086] Figure 4 This is a diagram showing the correlation between the functional state, state of charge, and health state of the battery module in this embodiment. Detailed Implementation

[0087] To facilitate understanding of this application, specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. The following embodiments are illustrative of the invention but are not intended to limit its scope. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of this application.

[0088] The basic idea of ​​this invention is as follows: Voltage, current, internal resistance, and capacitance of individual cells are collected as characteristic parameters; electrochemical polarization voltage and concentration polarization voltage are calculated to characterize the internal dynamic response process of the battery; a short-timescale state of charge model adapting to second-level changes and a long-timescale health state model describing monthly / yearly aging evolution are established; considering the relationship between the internal dynamic response of the battery and external electrical behavior, a functional state dynamic model reflecting the internal state of the individual cell and external operating conditions is established; by analyzing the adaptability of battery parameters at different time scales, a state estimation model of the individual cell at dual time scales is constructed, providing accurate individual cell state input for subsequent module-level state estimation; finally, considering the inconsistencies in capacity, internal resistance, and topology of individual cells within the module, the dynamic model and estimation model of the individual cells are extended to the module level, outputting the joint estimation results of the battery module's functional state, state of charge, and health state, thereby improving the reliability of the battery management system.

[0089] Example 1:

[0090] This embodiment provides a battery module state estimation method that takes into account the dynamic characteristics of functional states, such as... Figure 1 and Figure 2 As shown, the method includes the following steps:

[0091] Collect characteristic parameters of individual cells and calculate the electrochemical polarization voltage and concentration polarization voltage inside the cell.

[0092] The characteristic parameters of the single battery cell include: voltage, current, internal resistance, and capacitance.

[0093] In this embodiment, characteristic parameters of a single battery cell, including voltage, current, internal resistance, and capacitance, are collected, and electrochemical polarization voltage and concentration polarization voltage are calculated to characterize the internal dynamic response process of the battery.

[0094]

[0095] Wherein, U1 is the electrochemical polarization voltage of a single cell; U2 is the concentration polarization voltage of a single cell; I is the current of a single cell; R1 is the electrochemical polarization resistance of a single cell; R2 is the concentration polarization resistance of a single cell; C1 is the electrochemical polarization capacitance of a single cell; and C2 is the concentration polarization capacitance of a single cell.

[0096] In this embodiment, the factors affecting module state estimation are identified as: voltage, current, internal resistance, capacitance, state of charge, and state of health. Considering the long and short time dimensions of the battery, a dual-time-scale model for estimating the state of individual cells is constructed. A dynamic model of functional state that reflects the coupling relationship between the internal and external states of the battery is also constructed. By analyzing the inconsistencies between individual cells in the module, the dynamic model and the estimation model are extended to achieve joint estimation of functional state, state of charge, and state of health at the module level, thereby improving the accuracy of battery module state estimation.

[0097] Based on the factors affecting the state estimation of battery modules, a dynamic model of the state of charge characterizing the short-term characteristics of the battery and a health state model characterizing the long-term characteristics of the battery are constructed respectively.

[0098] In this embodiment, a state-of-charge model characterizing the short-term characteristics of the battery is established, and a health state model characterizing the long-term characteristics of the battery is established, taking into account the capacity decay and internal resistance growth of the battery.

[0099] The specific method for constructing a state-of-charge model to characterize the short-term characteristics of the battery and a health state model to characterize the long-term characteristics of the battery, based on the factors affecting the state estimation of the battery module, is as follows:

[0100] Factors affecting battery module state estimation include: voltage, current, internal resistance, capacitance, state of charge, and state of health.

[0101] A dynamic state-of-charge model characterizing the short-timescale properties of a battery is established based on the ampere-hour integral method.

[0102]

[0103] Among them, S OC (t) represents the current time. The state of charge; For the current moment Operating current; This refers to the rated capacity of a single battery cell; For Coulomb efficiency; For the current moment The state of health.

[0104] Establish a health state evolution model that characterizes the long-term characteristics of the battery, including: a health state evolution model that considers battery capacity decay and a health state evolution model that considers battery internal resistance growth.

[0105] The health state evolution model considering battery capacity degradation is as follows:

[0106]

[0107] Among them, S OH,Q For capacity-based health status; The number of complete cycles a single battery cell has undergone; This represents the average operating temperature of a single battery cell. This is the capacity decay coefficient; This is an empirical parameter for how capacity decay varies with the number of cycles; The activation energy for the aging reaction; This is the universal gas constant.

[0108] The health state evolution model considering the increase in battery internal resistance is as follows:

[0109]

[0110] in, The internal resistance in ohms under the current aging condition; The initial ohmic internal resistance of a single cell; This is the internal resistance growth coefficient; This is an empirical parameter showing how the internal resistance increases with the number of cycles. This is the activation energy during the internal resistance growth process; This represents the current state of health based on internal resistance under aging conditions.

[0111] Based on the electrochemical polarization voltage, concentration polarization voltage, state of charge model, and health state model, a dynamic model of the functional state of a single cell is established.

[0112] In this embodiment, based on all the battery parameter models established above, and combined with the correlation between their internal dynamic response and external electrical behavior, a functional state dynamic model reflecting the internal state of the battery and external operating conditions is established.

[0113] The specific method for establishing a dynamic functional state model of a single-cell battery based on the electrochemical polarization voltage, concentration polarization voltage, state of charge model, and health state model is as follows:

[0114] Based on the electrochemical polarization voltage and concentration polarization voltage, a dynamic response equation for the terminal voltage is established according to the polarization effect inside the battery.

[0115] In this embodiment, considering the internal polarization effect of the battery, a dynamic response equation for the terminal voltage is established, expressed as:

[0116]

[0117] in, This is the predicted value for the terminal voltage boundary; This is an open-circuit voltage function used to reflect the nonlinear relationship between the state of charge and the open-circuit voltage; The amplitude of the current pulse; Duration; This is the instantaneous value of the electrochemical polarization voltage; The electrochemical polarization time constant; This represents the instantaneous value of the concentration polarization voltage; The concentration polarization time constant; The internal resistance is ohms, which increases as the temperature decreases and the internal resistance ages.

[0118] Based on the dynamic response equation of the terminal voltage, a prediction model for the terminal voltage is established by iteratively calculating the terminal voltage at the current moment.

[0119] In this embodiment, considering the relationship between the internal dynamic response of a single cell and its external electrical behavior, a prediction model for the terminal voltage is established, expressed as:

[0120]

[0121] in, From the current moment The time elapsed since the start of the moment The predicted value of the terminal voltage; From the current moment The time elapsed since the start of the moment Predicted state of charge; OCV(S) OC (t, t)) is the predicted value based on the state of charge. The estimated open-circuit voltage; It is the internal resistance of the Ohm.

[0122] A maximum current constraint model is established based on the battery's voltage and temperature limitations.

[0123] In this embodiment, considering the external operating conditions of a single battery cell, a maximum current constraint model is established, which is expressed as:

[0124]

[0125] in, Indicates the maximum allowable current; This is the maximum current calculated based on voltage limits; This is the maximum current calculated based on temperature limits.

[0126] The maximum power of the battery under both discharge and charge conditions is defined as follows:

[0127]

[0128] in, This is the maximum dischargeable power; This is the maximum rechargeable power. The minimum permissible voltage for battery operation; The maximum permissible voltage for battery operation; For the duration The terminal voltage predicted by the maximum discharge current; For the duration The terminal voltage predicted by the maximum charging current; The maximum allowable discharge current; This represents the maximum allowable charging current.

[0129] Based on the predicted terminal voltage model, the maximum current constraint model, and the defined limiting power, a functional state dynamic model of a single cell is established, expressed as:

[0130]

[0131] in, This represents the maximum discharge power in the functional state. The maximum charging power for the functional state.

[0132] Based on the aforementioned state of charge model and state of health model, a state estimation model for a single cell under dual time scales is constructed.

[0133] In this embodiment, based on the established time characteristic model, the adaptability of battery parameters under different time scales is analyzed, and a state estimation model of a single battery under dual time scales is constructed.

[0134] The specific method for constructing a state estimation model for a single cell under dual time scales based on the state of charge model and the health model is as follows:

[0135] Based on the aforementioned state of charge model and health model, the state vectors of individual cells are defined on short-term and long-term time scales, respectively.

[0136] In this embodiment, by analyzing the time-scale characteristics of each battery parameter, a state vector for the battery parameter is defined, represented as:

[0137]

[0138] in, For single cells in a short timescale The state vector under; For single cells over a long time scale The state vector under, and exist for Update when the value is an integer multiple of the total value. For a moment The state of charge; For a moment The state of health.

[0139] Based on the definition of the state vector of a single cell under short and long time scales, a time-scale state transition function is established, expressed as:

[0140]

[0141] in, It is a short-timescale state transition function; for The estimated value of the state vector at a short time scale at time t; for The state of charge of the battery at any given time; for Electrochemical polarization voltage at time; for Concentration polarization voltage at time; for The capacity health status at any given time; for Load current at any given moment; For long-term state transition functions; for The estimated value of the state vector at a time scale; The average input vector; and All are regression coefficients, used to reflect the decline trend of health status itself; for The capacity health status at any given time; for The internal resistance and health status at any given moment; and These are all driving coefficients, used to represent the degree of influence of external stress on the rate of health degradation.

[0142] Based on the state transition function of the time scale, the observation residual correction mechanism is used to recursively estimate the state vector of a single cell at a short time scale and at a long time scale, so as to obtain the optimal estimate of the state vector of the single cell at both the short and long time scales, thus completing the dynamic modeling of the functional state of the single cell.

[0143] The method for recursively estimating the state vector of a single cell at a short time scale and at a long time scale using the observation residual correction mechanism based on the state transition function at the time scale is as follows:

[0144] Based on the state transition function of the time scale, state estimation models for individual cells are established under short and long time scales.

[0145]

[0146] in, This refers to noise during a short-term process. This refers to noise generated over a long period of time. for The actual observation vector at time; For observation functions; To observe noise.

[0147] For short time scales, the sampling time is based on the previous short time scale. The optimal estimate of the state vector and the current long-scale sampling time The state vector estimate is obtained from the state prediction equation of a single cell over a short timescale, thus yielding the current short-timescale sampling time. The predicted state vector value.

[0148] In this embodiment, based on The optimal estimate of the state at time step is obtained by predicting the state vector at the current time step using the system model, and establishing the state prediction equation for a single cell over a short time scale, which is expressed as:

[0149]

[0150] in, Based on The current short-timescale sampling time is obtained from the information of the time and previous times. State vector prediction value; The previous short-timescale sampling time The optimal estimate of the state vector; For the current long-term sampling time The estimated state vector.

[0151] Obtain the current short-timescale sampling time of a single cell. The actual observation vector is obtained, and the observation residual is calculated. Then, based on the state update equation of a single cell at a short time scale, the current short time scale sampling time is updated. The predicted state vector value is updated to obtain the current short-timescale sampling moment. The optimal estimate of the state vector.

[0152] In this embodiment, the predicted value is corrected using the actual observed value at time t, the observation residual is calculated, and the short-timescale state estimate is updated. The short-timescale state update equation for a single battery cell is expressed as:

[0153]

[0154] in, To observe the residual vector; Current short-timescale sampling time The optimal estimate of the state vector; This is the Kalman gain matrix.

[0155] For long time scales, based on the sampling time of the previous long time scale The optimal estimate of the state vector and the current long-scale sampling time The average input vector is used to obtain the current long-term sampling time based on the state prediction equation for a single cell over a long time scale. The predicted state vector value.

[0156] In this embodiment, based on an empirical aging model of healthy state, battery parameters adapted to long time scales are predicted, and a state prediction equation for a single cell over a long time scale is established, expressed as:

[0157]

[0158] in, The current long-time scale sampling time is obtained from the predicted value based on time m-1 and information prior to it. State vector prediction value; The previous long-time scale sampling time The optimal estimate of the state vector.

[0159] Using the sampling time from the previous long time scale up to the current long-term sampling time The accumulated observation residual vector is used to construct a virtual observation increment over a long time scale.

[0160] Based on the state update equation of a single cell over a long time scale, the virtual observation increment over a long time scale is used to update the current long-term sampling time. The predicted state vector is updated to obtain the current long-timescale sampling time. The optimal estimate of the state vector.

[0161] In this embodiment, virtual observation increments for long time scales are constructed using the short-term observation residuals accumulated over a long period. This is used to correct and update the state vector values ​​over long time scales, expressed as:

[0162]

[0163] in, It is the attenuation factor; For the current long-term sampling time The optimal estimate of the state vector; This is the update gain matrix over a long time scale.

[0164] In this embodiment, based on the state estimation model of the single cell under dual time scales, it can be calculated that by defining short-time-scale state vectors and long-time-scale state vectors, and utilizing the state transition function and observation residual correction mechanism, collaborative estimation under different time dimensions is achieved. For example... Figure 3 As shown, the estimated and actual values ​​of SOC at a short timescale and SOH at a long timescale are compared using two subgraph systems. The upper part, with time as the horizontal axis and SOC value as the vertical axis, demonstrates the tracking effect of SOC under second-level dynamic changes, verifying the real-time capture capability of the short timescale model for battery transient response; it also presents the slow evolution of SOH with the aging process, reflecting the accurate prediction of the monthly / year-level aging trend by the long timescale model.

[0165] Based on the functional state dynamic model of a single cell, a functional state dynamic model of the battery module is constructed.

[0166] In this embodiment, based on the functional state dynamic model and state estimation model established for individual cells, and considering the inconsistencies between individual cells within the battery module, the functional state dynamic model at the individual cell level is extended to the module level.

[0167] The specific method for constructing the functional state dynamic model of the battery module based on the functional state dynamic model of a single battery cell is as follows:

[0168] For the first in the battery module For each individual cell, calculate the deviation rate of its rated capacity and the deviation rate of its ohmic internal resistance.

[0169] In this embodiment, considering the inconsistency of parameters between modules, the deviation rate between the rated capacity and ohmic internal resistance of each individual battery cell is calculated and expressed as:

[0170]

[0171] in, For the first The deviation rate of the rated capacity of a single cell relative to the average capacity of the module; For the first The deviation rate of the ohmic internal resistance of a single cell relative to the average internal resistance of the module. For the first Rated capacity of each individual battery cell; For the first The internal resistance of a single cell in ohms; n This represents the average module capacity. n This represents the average internal resistance of the module.

[0172] Based on the module inconsistency parameter matrix, the voltage safety margin deviation of the i-th individual cell is calculated according to the deviation rate of the rated capacity and the deviation rate of the ohmic internal resistance.

[0173]

[0174] in, For the first Voltage safety margin deviation of individual cells; This is the contribution coefficient of internal resistance inconsistency to voltage deviation; This represents the total current of the module. This represents the contribution coefficient of capacity inconsistency to voltage deviation. For the first The state of charge of each individual cell.

[0175] Based on the voltage safety margin deviation and the voltage distribution of individual cells, a voltage constraint model for the battery module is constructed.

[0176] In this embodiment, considering the voltage distribution of individual cells, a voltage constraint model for the battery module is constructed, which is expressed as:

[0177]

[0178] in, This is the minimum operating voltage of the battery module; This is the maximum operating voltage of the battery module; This is the minimum allowable voltage for a single cell. This is the maximum allowable voltage of a single cell. For the first Real-time terminal voltage of each individual cell; This refers to the number of individual cells connected in series or parallel within the battery module.

[0179] Definition of the first Inconsistency factor of individual cells.

[0180] In this embodiment, the inconsistency factor between battery modules is defined as follows:

[0181]

[0182] in, For the first Inconsistency factor of individual cells; This is a weighting coefficient for capacity inconsistency; This is a weighting coefficient for the inconsistency of internal resistance; For the first Maximum charging current of each individual battery cell

[0183] Based on the battery module terminal voltage constraint model and the inconsistency factors of all individual cells, a functional state dynamic model of the battery module is constructed.

[0184] In this embodiment, based on the battery module terminal voltage constraint model and the inconsistency factors of all individual cells, the individual cell functional state model is extended to the module level. The dynamic functional state model of the battery module is represented as follows:

[0185]

[0186] in, The maximum sustainable discharge power of the battery module in its functional state; The maximum sustainable charging power for the battery module's functional state; In order to be in At that moment, the Individual cells in The maximum allowable discharge current; In order to be in At that moment, the Individual cells in The maximum charging current that can be allowed inside.

[0187] Based on the state estimation model of a single cell under dual time scales and the functional state dynamic model of the battery module, a state estimation model of the battery module under dual time scales is constructed.

[0188] In this embodiment, considering the topology between individual cells within the battery module, the dual-timescale state estimation model at the individual cell level is extended to the module level, completing the joint estimation of the battery module's functional state, state of charge, and health state. This provides a data foundation for the long-term health management of the battery and improves the reliability of the battery management system.

[0189] The specific method for constructing the state estimation model of the battery module under dual time scales based on the state estimation model of a single battery cell under dual time scales and the functional state dynamic model of the battery module is as follows:

[0190] Based on the state estimation model of a single cell under dual time scales, the state vector of the battery module is constructed.

[0191] In this embodiment, based on the state estimation model of a single cell under dual time scales, a module-level state vector is constructed, represented as:

[0192]

[0193] in, This is the state vector of the battery module; For the first The state vector of a single cell in a short time scale; For the first The state vector of a single cell over a long time scale.

[0194] Based on the electrical connection method of the battery modules, the system topology matrix of the battery modules is defined as follows:

[0195]

[0196] in, Here is the system current topology matrix; The system voltage topology matrix; It is a row vector consisting entirely of 1s; It is a column vector consisting entirely of 1s; for 3D identity matrix.

[0197] Based on the system topology matrix of the battery module, module-level constraint equations are established, expressed as follows:

[0198]

[0199] in, This represents the current vector of a single cell. This is the current constraint error term; This refers to the battery module terminal voltage. This represents the voltage vector of a single cell. This is the voltage constraint error term.

[0200] Based on the module-level constraint equations, the state vectors of the battery module under short-term and long-term time scales are recursively estimated to obtain the optimal estimates of the state vectors of the battery module under both short-term and long-term time scales.

[0201] The specific method for recursively estimating the state vectors of the battery module under short-term and long-term time scales based on the module-level constraint equations to obtain the optimal estimates of the state vectors of the battery module under short-term and long-term time scales is as follows:

[0202] Based on the state prediction equation of a single cell over a short timescale, a state prediction equation of the battery module over a short timescale is constructed to obtain the state prediction equation of the battery module at the current short timescale sampling time. The predicted state vector value.

[0203] In this embodiment, considering the topological relationships between individual battery modules, the state prediction equation for a single battery cell over a short timescale is extended to the module level, as follows:

[0204]

[0205] in, For the first Individual cells at the current short-timescale sampling time State vector prediction value; For the first Individual cells in The optimal estimate of the state vector at time 1; For the i-th single cell at the current long-time sampling time... State vector prediction value; This is the topological interaction gain matrix; For topology state interaction items; In order to be with the first A collection of adjacent individual cells; In order to be with the first The adjacent single cell Individual cells; These are the topological weight coefficients; For the j-th adjacent single cell in The estimated state vector at time t.

[0206] Obtain the current short-timescale sampling time of the battery module The actual observation vector is calculated, and the observation residual vector of the battery module is calculated.

[0207] In this embodiment, the topology is considered when updating the observation vector at the module level, as shown below:

[0208]

[0209] in, This represents the observation residual vector of the battery module; for The actual observation vector of the battery module at any given time; Based on Time information, Predicted state vector value of the battery module at any given time.

[0210] Based on the short-timescale state update equation of the battery module, and utilizing the observation residual vector and topological observation coordination matrix of the battery module, the state update of the battery module at the current short-timescale sampling time is determined. The state vector prediction value is updated to obtain the battery module at the current short-timescale sampling time. The optimal estimate of the state vector;

[0211] In this embodiment, the short-timescale state update equation for a single cell is extended to the module level, and the topology observation coordination matrix is ​​defined as:

[0212]

[0213] in, For the first Individual cells at the current short-timescale sampling time The optimal estimate of the state vector; For the first Adjacent individual cells Kalman gain matrix at time step; For the first The observation residual vector of each individual cell; This is the topology observation coordination matrix; In order to be with the first The observation residual vector of the set of adjacent individual cells of a single cell; This is the topology compatibility strength coefficient; For the first The topological weight matrix of each individual cell.

[0214] Based on the state prediction equation of a single cell over a long time scale, a global state update equation for the battery module is constructed to obtain the battery module's state at the current long-time scale sampling time. The optimal estimate of the state vector.

[0215] In this embodiment, considering the topological connections of the modules, a global state update equation is established, namely, the long-term state prediction equation of the battery module, which is used to correct the consistency value between modules, and is expressed as:

[0216]

[0217] in, For the battery module at the current long-term sampling time State vector estimate; For the first Current long-term sampling time for individual cells The optimal estimate of the state vector; This is the fusion gain matrix; This is a consistency correction amount; This is the arithmetic mean vector of the long-time-scale state vector estimates for each individual cell. For battery modules The estimated state vector at time t.

[0218] Based on the optimal estimate of the state vector of the battery module in a short time scale, obtain The estimated state of charge of each individual cell at any given time.

[0219] In this embodiment, the state vector of each individual battery cell is obtained from the optimal estimate of the state vector of the battery module over a short time scale. The estimated state of charge after topological observation coordination at time is expressed as:

[0220]

[0221] in, For the first Individual cells in The estimated state of charge at time t.

[0222] Based on the optimal estimate of the state vector of the battery module over a long time scale, obtain The estimated health status of each individual cell at any given time.

[0223] In this embodiment, the corrected battery module is sampled at the current long-time scale time. From the optimal estimate of the state vector, the health state components based on capacity and internal resistance are extracted and represented as follows:

[0224]

[0225] in, For the first Individual cells in Constantly monitor capacity health status; For the first Individual cells in A healthy state based on internal resistance at all times.

[0226] In this embodiment, based on the functional state dynamic model of a single battery cell, the dependence of SOF on SOC and SOH is quantified by coupling the internal polarization effect of the battery with external operating conditions, using the terminal voltage prediction equation and maximum power limit. Simultaneously, combined with the functional state dynamic model of the battery module, inconsistency factors and topological constraints are further introduced to ensure the reliability of module-level SOF calculation. Using SOC and SOH as independent variables and SOF as a response index, the system quantifies the dynamic variation of the module's maximum charge and discharge power with SOC and SOH. Figure 4 As shown, when SOH remains constant, an increase in SOC leads to an increase in open-circuit voltage and a relaxation of discharge constraint, thereby significantly improving SOF. Under the condition of fixed SOC, the improvement in SOH simultaneously increases the SOF value by reducing internal resistance and slowing down performance degradation.

[0227] Based on the state estimation model of the battery module under dual time scales, the estimated value of the state of charge and the estimated value of the health state of the battery module at the current moment are calculated. Then, based on the functional state dynamic model of the battery module, the functional state of the battery module at the current moment is calculated.

[0228] In this embodiment, based on the constructed dynamic model and state estimation model of the battery module's functional state, the joint estimation of the battery module's functional state, state of charge, and health state is completed, providing a data foundation for the long-term health management of the battery and improving the reliability of the battery management system.

[0229] The specific method for calculating the estimated state of charge and state of health of the battery module at the current moment based on the state estimation model of the battery module under dual time scales, and then calculating the functional state of the battery module at the current moment based on the functional state dynamic model of the battery module, is as follows:

[0230] Define the weighted average of the state of charge (SOC) of each individual cell in the battery module, and calculate the estimated SOC and health status of the battery module at the current moment based on the state estimation model of the battery module under dual time scales.

[0231] In this embodiment, inconsistency is defined as the weighted average of the state of charge (SOC) of each individual battery cell. The estimated SOC and health status of the battery module are then output as follows:

[0232]

[0233] in, For the battery module at the current long-term sampling time Health status estimate; and These are all weighting coefficients, adjusted according to the application scenario to meet the requirements. ; For the battery module at the current short timescale sampling time The estimated state of charge; For the first The weighting coefficients of individual cells over a short time scale.

[0234] Based on the estimated state of charge and state of health of the battery module at the current moment, the maximum charging power and maximum discharging power of the battery module at the current moment are calculated according to the functional state dynamic model of the battery module.

[0235] In this embodiment, based on the updated module state, the maximum charge and discharge power functional state of the battery module is calculated using the dynamic model of the battery module's functional state, providing a data foundation for the long-term health management of the battery and improving the reliability of the battery management system.

[0236]

[0237] In this embodiment, by using a dynamic model of the functional state of a single battery cell and combining it with a dual-timescale estimation strategy for the single battery cell, the estimation object is extended from the battery cell level to the module level, thereby effectively improving the accuracy and dynamic response capability of battery state estimation and providing a data foundation for the long-term health management of the battery.

[0238] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the present invention.

Claims

1. A method for estimating the state of a battery module considering the dynamic characteristics of its functional state, characterized in that, The method includes the following steps: Collect characteristic parameters of individual cells and calculate the electrochemical polarization voltage and concentration polarization voltage inside the cell; Based on the factors affecting the state estimation of battery modules, a state of charge model characterizing the short-term characteristics of the battery and a health state model characterizing the long-term characteristics of the battery are constructed respectively. Based on the electrochemical polarization voltage, concentration polarization voltage, state of charge model, and health state model, a dynamic model of the functional state of a single cell is established. Based on the state of charge model and the state of health model, a state estimation model for a single cell under dual time scales is constructed. Based on the functional state dynamic model of a single battery cell, a functional state dynamic model of the battery module is constructed. Based on the state estimation model of a single cell under dual time scales and the functional state dynamic model of the battery module, a state estimation model of the battery module under dual time scales is constructed. Based on the state estimation model of the battery module under dual time scales, the estimated value of the state of charge and the estimated value of the health state of the battery module at the current moment are calculated. Then, based on the functional state dynamic model of the battery module, the functional state of the battery module at the current moment is calculated.

2. The battery module state estimation method considering dynamic characteristics of functional states according to claim 1, characterized in that, The characteristic parameters of the single battery cell include: voltage, current, internal resistance, and capacitance.

3. The battery module state estimation method considering dynamic characteristics of functional states according to claim 1, characterized in that, The specific method for constructing a state-of-charge model characterizing the short-term characteristics of the battery and a health state model characterizing the long-term characteristics of the battery, based on the factors affecting the estimation of the battery module's state, is as follows: Identify the factors that affect the state of the battery module, including: voltage, current, internal resistance, capacitance, state of charge, and state of health; A dynamic model of the state of charge (SOC) characterizing the short-timescale properties of a battery is established based on the ampere-hour integral method. ； Among them, S OC (t) represents the current time. The state of charge; For the current moment Operating current; This refers to the rated capacity of a single battery cell; For Coulomb efficiency; For the current moment ; health status; Establish a health state evolution model that characterizes the long-term characteristics of the battery, including: a health state evolution model that considers battery capacity decay and a health state evolution model that considers battery internal resistance growth. The health state evolution model considering battery capacity degradation is as follows: ； Among them, S OH,Q For capacity-based health status; The number of complete cycles a single battery cell has undergone; This represents the average operating temperature of a single battery cell. This is the capacity decay coefficient; This is an empirical parameter for how capacity decay varies with the number of cycles; The activation energy for the aging reaction; This is the universal gas constant; The health state evolution model considering the increase in battery internal resistance is as follows: ； in, The internal resistance in ohms under the current aging condition; The initial ohmic internal resistance of a single cell; This is the internal resistance growth coefficient; This is an empirical parameter showing how the internal resistance increases with the number of cycles. This is the activation energy during the internal resistance growth process; This represents the current state of health based on internal resistance under aging conditions.

4. The battery module state estimation method considering dynamic characteristics of functional states according to claim 3, characterized in that, The specific method for establishing a dynamic functional state model of a single-cell battery based on the electrochemical polarization voltage, concentration polarization voltage, state of charge model, and health state model is as follows: Based on the electrochemical polarization voltage and concentration polarization voltage, a dynamic response equation for the terminal voltage is established according to the polarization effect inside the battery. Based on the dynamic response equation of the terminal voltage, a prediction model for the terminal voltage is established by iteratively calculating the terminal voltage at the current moment. A maximum current constraint model is established based on the battery's voltage and temperature limitations. Define the maximum power of the battery under both discharge and charging conditions; Based on the predicted terminal voltage model, the maximum current constraint model, and the defined limiting power, a functional state dynamic model of a single cell is established, expressed as: ； in, This represents the maximum discharge power in the functional state. The maximum charging power for the functional state; Duration; The minimum permissible voltage for battery operation; The maximum permissible voltage for battery operation; The maximum allowable discharge current; This represents the maximum allowable charging current.

5. The battery module state estimation method considering dynamic characteristics of functional states according to claim 4, characterized in that, The specific method for constructing a state estimation model for a single cell under dual time scales based on the state of charge model and the health model is as follows: Based on the state of charge model and the state of health model, the state vectors of a single cell are defined at short-term and long-term time scales, respectively. Based on the definition of the state vector of a single cell in short and long time scales, a time-scale state transition function is established. Based on the state transition function of the time scale, the observation residual correction mechanism is used to recursively estimate the state vector of a single cell at a short time scale and at a long time scale, so as to obtain the optimal estimate of the state vector of the single cell at both the short and long time scales, thus completing the dynamic modeling of the functional state of the single cell.

6. The battery module state estimation method considering dynamic characteristics of functional states according to claim 5, characterized in that, The method for recursively estimating the state vector of a single cell at a short time scale and at a long time scale using the observation residual correction mechanism based on the state transition function at the time scale is as follows: Based on the state transition function of the time scale, state estimation models for individual cells under short and long time scales are established. For short time scales, the sampling time is based on the previous short time scale. The optimal estimate of the state vector and the current long-scale sampling time The state vector estimate is obtained from the state prediction equation of a single cell over a short timescale, thus yielding the current short-timescale sampling time. The predicted value of the state vector; Obtain the current short-timescale sampling time of a single cell. The actual observation vector is obtained, and the observation residual is calculated. Then, based on the state update equation of a single cell at a short time scale, the current short time scale sampling time is updated. The predicted state vector value is updated to obtain the current short-timescale sampling moment. The optimal estimate of the state vector; For long time scales, based on the sampling time of the previous long time scale The optimal estimate of the state vector and the current long-scale sampling time The average input vector is used to obtain the current long-term sampling time based on the state prediction equation for a single cell over a long time scale. The predicted value of the state vector; Using the sampling time from the previous long time scale up to the current long-term sampling time The accumulated observation residual vector is used to construct virtual observation increments over a long time scale; Based on the state update equation of a single cell over a long time scale, the virtual observation increment over a long time scale is used to update the current long-term sampling time. The predicted state vector is updated to obtain the current long-timescale sampling time. The optimal estimate of the state vector.

7. The battery module state estimation method considering dynamic characteristics of functional states according to claim 6, characterized in that, The specific method for constructing the functional state dynamic model of the battery module based on the functional state dynamic model of a single battery cell is as follows: For the first in the battery module For each individual cell, calculate the deviation rate of its rated capacity and the deviation rate of its ohmic internal resistance. Based on the module inconsistency parameter matrix, the voltage safety margin deviation of the i-th individual cell is calculated according to the deviation rate of the rated capacity and the deviation rate of the ohmic internal resistance. Based on the voltage safety margin deviation and combined with the voltage distribution of individual cells, a voltage constraint model for the battery module is constructed. Definition of the first Inconsistency factor of individual cells; Based on the battery module terminal voltage constraint model and the inconsistency factors of all individual cells, a functional state dynamic model of the battery module is constructed.

8. The battery module state estimation method considering dynamic characteristics of functional states according to claim 7, characterized in that, The specific method for constructing the state estimation model of the battery module under dual time scales based on the state estimation model of a single battery cell under dual time scales and the functional state dynamic model of the battery module is as follows: Based on the state estimation model of a single cell under dual time scales, the state vector of the battery module is constructed. Define the system topology matrix of the battery modules based on their electrical connection methods; Based on the system topology matrix of the battery module, module-level constraint equations are established. ； in, This represents the current vector of a single cell. The system current topology matrix; This represents the total current of the module. This is the current constraint error term; This refers to the battery module terminal voltage. The system voltage topology matrix; This represents the voltage vector of a single cell. This is the voltage constraint error term; Based on the module-level constraint equation, the state vector of the battery module under the short time scale and the state vector under the long time scale are recursively estimated to obtain the optimal estimated values ​​of the state vector of the battery module under the short time scale and the long time scale. Based on the optimal estimate of the state vector of the battery module in a short time scale, obtain The estimated state of charge of each individual cell at any given time; Based on the optimal estimate of the state vector of the battery module over a long time scale, obtain The estimated health status of each individual cell at any given time.

9. The battery module state estimation method considering dynamic characteristics of functional states according to claim 8, characterized in that, The specific method for recursively estimating the state vectors of the battery module under short-term and long-term time scales based on the module-level constraint equations to obtain the optimal estimates of the state vectors of the battery module under short-term and long-term time scales is as follows: Based on the state prediction equation of a single cell over a short timescale, a state prediction equation of the battery module over a short timescale is constructed to obtain the state prediction equation of the battery module at the current short timescale sampling time. The predicted value of the state vector; ； in, For the first Individual cells at the current short-timescale sampling time The predicted value of the state vector; For the first The short-timescale state transition function of a single cell; For the first Individual cells in The optimal estimate of the state vector at time 1; For the i-th single cell at the current long-time sampling time... The predicted value of the state vector; This is the topological interaction gain matrix; For topology state interaction items; In order to be with the first A collection of adjacent individual cells; In order to be with the first The adjacent single cell Individual cells; These are the topological weight coefficients; For the j-th adjacent single cell in State vector estimate at time t; Obtain the current short-timescale sampling time of the battery module The actual observation vector, and calculate the observation residual vector of the battery module; Based on the short-timescale state update equation of the battery module, and using the observation residual vector of the battery module, the battery module at the current short-timescale sampling time is updated. The state vector prediction value is updated to obtain the battery module at the current short-timescale sampling time. The optimal estimate of the state vector; ； in, For the first Individual cells at the current short-timescale sampling time The optimal estimate of the state vector; For the first The Kalman gain matrix of each adjacent cell at time t; For the first The observation residual vector of each individual cell; This is the topology observation coordination matrix; In order to be with the first The observation residual vector of the set of adjacent individual cells of a single cell; This is the topology compatibility strength coefficient; For the first The topological weight matrix of each individual cell; Based on the state prediction equation of a single cell over a long time scale, a global state update equation for the battery module is constructed to obtain the battery module's state at the current long-time scale sampling time. The optimal estimate of the state vector; ； in, For the battery module at the current long-term sampling time State vector estimate; For the first Current long-term sampling time for individual cells The optimal estimate of the state vector; This is the fusion gain matrix; This is a consistency correction amount; This is the arithmetic mean vector of the long-time-scale state vector estimates for each individual cell. For battery modules State vector estimate at time t; This refers to the number of individual cells connected in series or parallel within the battery module.

10. The battery module state estimation method considering dynamic characteristics of functional states according to claim 9, characterized in that, The specific method for calculating the estimated state of charge and state of health of the battery module at the current moment based on the state estimation model of the battery module under dual time scales, and then calculating the functional state of the battery module at the current moment based on the functional state dynamic model of the battery module, is as follows: Define the weighted average of the state of charge of each individual cell in the battery module, and calculate the estimated state of charge and state of health of the battery module at the current moment based on the state estimation model of the battery module under dual time scales. ； in, For the battery module at the current long-term sampling time Health status estimate; and All are weighting coefficients; For the first Individual cells in Constantly monitor capacity health status; For the first Individual cells in A healthy state based on internal resistance at all times; For the battery module at the current short timescale sampling time The estimated state of charge; For the first Weighting coefficients for individual cells over short time scales; For the first Individual cells in The estimated state of charge at time t; Based on the estimated state of charge and state of health of the battery module at the current moment, the maximum charging power and maximum discharging power of the battery module at the current moment are calculated according to the functional state dynamic model of the battery module. ； in, The maximum sustainable discharge power of the battery module in its functional state; The maximum sustainable charging power for the battery module's functional state; This refers to the minimum permissible voltage of a single cell. This is the maximum allowable voltage of a single cell. Duration; In order to be in At that moment, the Individual cells in The maximum allowable discharge current; In order to be in At that moment, the Individual cells in The maximum allowable charging current; For the first Inconsistency factor of individual cells.

Images