A method for detecting performance of a power battery based on double-set filter

Abstract

The application discloses a kind of based on double set member filtering power battery performance detection method, belong to power battery production field.The SOC and SOH joint model of power battery is obtained by using nested double set member algorithm, according to the coupling relationship between SOC and SOH state equation, alternating operation is carried out by full symmetry polytope algorithm, the k time update step result of SOH is taken as known quantity, and is brought into the calculation of SOC prediction step of k+1 time;The result of the SOC prediction step of k+1 time is taken as known quantity, and is brought into the calculation of SOH update step of k+1 time, to realize alternating operation in this way.Different from two independent filtering algorithms, the coupling relationship of SOC and SOH is further strengthened based on joint modeling in the application method, and the filtered result is brought into algorithm alternating operation, the influence of noise is reduced, so that the algorithm result is more accurate and robust.

Description

Technical Field

[0001] This invention relates to a power battery performance testing method based on dual-member filtering, belonging to the field of power battery production. Background Technology

[0002] Power batteries are a key component of the power system of new energy vehicles and a significant factor influencing their development, having a substantial impact on their performance and cost. As new energy vehicles become more high-performance and economical, the requirements for power batteries are increasing. Compared to traditional internal combustion engine vehicles, new energy vehicles, in order to fully leverage their advantages of low fuel consumption and low pollution, must require power batteries with high capacity, high specific energy and specific power, and long lifespan.

[0003] Furthermore, due to their characteristics of high specific energy, high specific power, long lifespan, and low cost, the safety and energy density of power batteries are currently the most important issues they face. To accurately and rationally utilize power batteries, their performance needs to be precisely tested. The most important performance indicators for power batteries are the state of charge (SOC) and the state of health (SOH); where SOC represents the remaining battery capacity, and SOH represents the degree of battery aging.

[0004] Current power battery performance testing methods mostly target only one aspect. However, the State of Charge (SOC) and State of Hypothesis (SOH) of a power battery are interdependent. For example, the degree of battery aging directly affects the battery's storage capacity and fast charge / discharge capability. Testing only one aspect introduces significant errors. To reduce errors and test power battery performance more efficiently and accurately, it is necessary to jointly estimate SOC and SOH. However, current methods for jointly estimating SOC and SOH only consider the impact of the healthy state SOH on the state of charge SOC. That is, during the estimation process, the SOC at the next moment is estimated based only on the current SOH value, while the SOH value at the next moment is only based on the current SOH value, without considering the SOC value. This means that it only considers the SOH at a certain SOC value, without considering the SOH under changing SOC conditions, and it leads to the accumulation of errors during the iteration process. The detection accuracy needs further improvement. Summary of the Invention

[0005] To further improve the accuracy of power battery performance testing, this invention provides a power battery performance testing method based on dual-membership filtering. Using a nested membership algorithm, after jointly modeling the SOC and SOH of the power battery, an alternating computation is performed using a fully symmetric polytope algorithm. The SOH polytope from the update step at time k is used as a known quantity and substituted into the SOC polytope prediction step at time k+1. Similarly, the SOC polytope from time k+1 is used as a known quantity and substituted into the SOH update step at time k+1, thereby achieving alternating computation. The method includes:

[0006] Step 1: Obtain the equivalent circuit model of the power battery based on its electrochemical mechanism;

[0007] Step 2: Obtain the linear expression of SOC of the power battery, and use the battery capacity to characterize SOH;

[0008] Step 3: Based on the coupling relationship between SOC and SOH, establish a joint estimation model for SOC and SOH;

[0009] Step 4: Construct the set of fully symmetric multicellular bodies for the joint estimation model at time k. and Based on the coupling relationship between SOC and SOH, a set of fully symmetric multiple cells corresponding to the state variable prediction set at time k+1 is constructed. and

[0010] Step 5: Construct the band space at time k+1 based on the multicellular constraints. and strip space and The sets of fully symmetric multicells corresponding to the state variable prediction set at time k+1 constructed in step four. and Finding the intersection yields the state variable x at time k+1. c,k+1 and x h,k+1 A collection of fully symmetrical multicellular bodies and

[0011] Step Six: Based on the set of fully symmetric multicellular bodies obtained in Step Five and The upper and lower bounds of SOC and SOH are calculated respectively, and the specific values ​​of SOC and SOH are observed to complete the test of the power battery performance.

[0012] Optionally, step one includes:

[0013] Based on the electrochemical mechanism of the power battery discharge process, the equivalent circuit model of the power battery is established as shown in the following equation:

[0014]

[0015] Where R0 is the ohmic resistance of the power battery, R1 and C1 are the electrochemical polarization internal resistance and capacitance of the power battery, respectively, and the voltage across R1 and C1 connected in parallel is U1; R2 and C2 are the concentration polarization resistance and capacitance of the power battery, respectively, and the voltage across R2 and C2 connected in parallel is U2; I is the discharge current, U... oc U is the internal power source of the battery, and U is the voltage across the battery terminals, i.e., the output voltage.

[0016] Optionally, step two includes:

[0017] Based on the current integral method for power batteries, a state-of-the-art (SOC) model for the power battery is established:

[0018]

[0019] Where SOC and SOC0 represent the battery's state of charge at the current and initial moments, respectively; Q represents the battery's charge at the current moment; η represents the battery's coulombic efficiency; I represents the discharge current; and t and t0 represent the current and initial moments, respectively.

[0020] Characterizing SOH using battery capacity: Based on the relationship between the battery capacity Q and SOH, an SOH model for the power battery is established.

[0021]

[0022] Where SOH represents the current state of battery health; Q and Q N These represent the current battery level and the initial battery level, respectively.

[0023] Optionally, step three includes:

[0024] Taking the battery discharge current I as input, the battery voltage U at time k 1,k U 2,k and SOC k Let U be the state variable, and let U be the voltage across the battery at time k. k For the output, after discretization, the SOC linear model of the battery is established as follows:

[0025]

[0026] U k =U oc (SOC k )-U 1,k -U 2,k -IR0+v k (5)

[0027] Among them, [U1,k U 2,k SOC k ] T The state variable is the discharge current I, which is a constant and represents the input quantity; ΔT represents the sampling time; τ1=R1C1 represents the time constant after R1 and C1 are connected in parallel; τ2=R2C2 represents the time constant after R2 and C2 are connected in parallel.

[0028] U oc (SOC k+1 ) = 0.5158SOC k +3.624 indicates the battery's internal power supply U. oc The linear relationship between and SOC This represents the unknown but bounded perturbation noise in the SOC linear model, i.e. This represents the unknown but bounded measurement noise in the SOC linear model, i.e.

[0029] SOH at time k of the battery k As state variables, after discretization, the SOH linear model of the battery is established as follows:

[0030]

[0031] Among them, SOH k It is a state variable. This represents the unknown but bounded perturbation noise in the SOH linear model, i.e. This represents the unknown but bounded measurement noise in the SOH linear model, i.e. d k This represents the difference between the voltage of a single cell in the battery pack and the average voltage of the single cell.

[0032] Formulas (4)-(6) are the established joint estimation models for SOC and SOH.

[0033] Optionally, step four includes:

[0034] 4.1 Expression of the joint estimation model of SOC and SOH of the power battery;

[0035] 4.1.1 with x c,k Let y represent the state variable at time k. c,k Let k represent the battery output voltage at time k. The SOC linear model of the battery represented by formulas (4) and (5) is expressed as follows:

[0036]

[0037] in,

[0038] xc,k Represents the state variable [U] at time k. 1,k U 2,k SOC k ] T ;u c,k The discharge current I at time k is a constant; y c,k This represents the battery output voltage at time k;

[0039] 4.1.2 with x h,k SOH represents the state variable at time k. k y h,k d represents time k k The linear model of the SOH of the battery represented by formula (6) is expressed as follows:

[0040]

[0041] Among them, A h =1, x h,k SOH represents the state variable at time k. k ;y h,k d represents time k k ;

[0042] 4.2 Constructing a fully symmetric set of multiple cells for the joint estimation model at time k and

[0043] Define the fully symmetric multicell corresponding to the initialization state variable x0. This represents the center point of the fully symmetrical multicell at the initial moment. B represents the shape matrix of the fully symmetric multicell at the initial time step. m Let m be the unit boxes formed by the intervals [-1, 1].

[0044] 4.2.1 Construction

[0045] Assume the state variable x of SOC in the joint estimation model at time k. c,k The corresponding fully symmetrical multicellular body is in, It is the center of the SOC multicellular body. Its shape matrix;

[0046] 4.2.2 Construction

[0047] Assume the state variable x of SOH in the joint estimation model at time k. h,k The corresponding fully symmetrical multicellular body is in, It is the center of the SOH multicellular body. Its shape matrix;

[0048] 4.3 Constructing the set of fully symmetric multiple cells corresponding to the predicted set of state variables at time k+1 and

[0049] Based on the joint estimation model of SOC and SOH of the power battery, the input matrix C of the SOC linear model is... c The model includes the SOH variable at time k, and the output of the SOH model also includes the difference between the SOC at time k and the SOC at time k-1. This indicates a coupling relationship between SOC and SOH; the SOH value at time k affects the estimation result of SOC at time k+1. Therefore, in constructing a fully symmetric multicellular set... and When doing so, the effects of coupling should be considered;

[0050] 4.3.1 Considering the coupling relationship between SOC and SOH, construct a fully symmetric multicellular set corresponding to the state of charge of the power battery at time k+1. for:

[0051]

[0052] Among them, F c For disturbance noise w k The corresponding generating matrix of a fully symmetric multicellular body, SOH k This is the result of the SOH update step at time k;

[0053] 4.3.2 Considering the coupling relationship between SOC and SOH, construct the full set of data corresponding to the health state of the power battery at time k+1.

[0054] Called a multicellular assembly for:

[0055]

[0056] Among them, F h For disturbance noise γ k The generation matrix of the corresponding fully symmetric multicell.

[0057] Optionally, step five includes:

[0058] 5.1 Constructing the strip space at time k+1 and

[0059] 5.1.1 Constructing the strip space corresponding to the state of charge of the power battery at time k+1

[0060]

[0061] Where, x c,k+1 Let k+1 be the state variable of the SOC model. For matrix C c d c =U k +IR0, σ c =v k+1 ;

[0062] 5.1.2 Constructing the strip space corresponding to the health state of the power battery at time k+1

[0063]

[0064] Where, x h,k+1 Let k+1 be the state variable of the SOH model. For matrix C h d h =d k -SOC k+1 +SOC k , σ h =ε k+1 ;

[0065] 5.2 Transforming the strip space and The sets of fully symmetric multiple cells corresponding to the predicted state variables at time k+1. and Finding the intersection yields the state variable x at time k+1. c,k+1 and x h,k+1 A collection of fully symmetrical multicellular bodies and

[0066] The set of fully symmetric multicellular structures of the state variable prediction set at time k+1 and Respectively with ribbon space and The intersection can be represented as:

[0067]

[0068] Using fully symmetrical multicellular structures pack and The intersection of the states is used to calculate the volume of a fully symmetric multicell. The fully symmetric multicell with the smallest volume is selected as the one containing the state variable x at time k+1. k+1 A collection of fully symmetrical multicellular bodies Right now:

[0069] The set of fully symmetric multicellular structures corresponding to the state of charge of a power battery for:

[0070]

[0071] in,

[0072] The set of fully symmetric multicellular structures corresponding to the health status of a power battery for:

[0073]

[0074] in, SOC k+1 This is the SOC prediction step result at time k+1, SOC k This is the result of the SOC update step at time k.

[0075] Optionally, step six includes:

[0076] Based on the fully symmetric multicellular set corresponding to the state of charge of the power battery at time k+1 obtained in step five. The upper and lower bounds of the final predicted value of the power battery SOC are:

[0077]

[0078] Based on the fully symmetric multicellular set corresponding to the health state of the power battery at time k+1 obtained in step five. The upper and lower bounds of the final predicted SOH value of the power battery are:

[0079]

[0080] By observing the upper and lower bounds, we can determine whether the values ​​of SOC and SOH fluctuate within the normal range, thus completing the performance test of the power battery.

[0081] This application also provides a power battery performance testing system based on dual-member filtering. The system includes devices for detecting the current and voltage of the power battery. The system uses the above method to detect the state of charge (SOC) and state of health (SOH) of the power battery.

[0082] This application also provides a fault detection method for power battery performance testing based on dual-member filtering. The method uses the above-mentioned method to obtain the range of the state of charge (SOC) value and the range of the state of health (SOH) value of the power battery, and compares the obtained range of the SOC value and the SOH value with the range of the normal state of charge (SOC) value and the SOH value with the range of the healthy state of charge (SOH). If the difference exceeds a preset threshold, the power battery is judged to have a fault.

[0083] The beneficial effects of this invention are:

[0084] This invention obtains a joint model of the State of Charge (SOC) and State of Health (SOH) of a power battery by using a nested dual-member algorithm. Based on the coupling relationship between the SOC and SOH state equations, it performs alternating calculations using a fully symmetric multicell algorithm. The result of the SOH update step at time k is used as a known quantity and substituted into the SOC prediction step calculation at time k+1; similarly, the result of the SOC prediction step at time k+1 is used as a known quantity and substituted into the SOH update step calculation at time k+1, thus achieving alternating calculations. Unlike two independent filtering algorithms, this method further strengthens the coupling relationship between SOC and SOH based on joint modeling. Simultaneously, by incorporating the filtered results into the alternating calculations, the influence of noise is reduced, resulting in more accurate and robust algorithm results. Attached Figure Description

[0085] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0086] Figure 1 This is a flowchart of a power battery performance testing method based on dual-member filtering, disclosed in one embodiment of the present invention.

[0087] Figure 2 This is a diagram of the equivalent circuit model of a power battery.

[0088] Figure 3 This is a comparison chart of the upper and lower bounds of the SOC value and the true SOC value estimated by the existing detection method and the method proposed in this application under the normal operating state of the power battery disclosed in one embodiment of the present invention.

[0089] Figure 4 This is a comparison chart of the upper and lower bounds of the SOH value and the true SOH value estimated by existing detection methods and the method proposed in this application under the normal operating conditions of a power battery, as disclosed in one embodiment of the present invention. Detailed Implementation

[0090] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0091] Example 1:

[0092] This embodiment provides a power battery performance testing method based on dual-member filtering. (See also...) Figure 1The method includes: establishing a joint estimation model of SOC and SOH based on the coupling relationship between SOC and SOH, and using battery capacity to characterize SOH during the establishment process; based on the joint estimation model of SOC and SOH, performing alternating calculations using a fully symmetric multicell algorithm, taking the SOH update step result at time k as a known quantity and substituting it into the SOC prediction step calculation at time k+1; taking the SOC prediction step result at time k+1 as a known quantity and substituting it into the SOH update step calculation at time k+1, and using this nested dual-member filtering method to achieve alternating calculations to complete the detection of power battery performance; specifically including:

[0093] Step 1: Obtain the equivalent circuit model of the power battery based on its electrochemical mechanism;

[0094] Step 2: Obtain the linear expression of SOC of the power battery, and use the battery capacity to characterize SOH;

[0095] Step 3: Based on the coupling relationship between SOC and SOH, establish a joint estimation model for SOC and SOH;

[0096] Step 4: Construct the set of fully symmetric multicellular bodies for the joint estimation model at time k. and Based on the coupling relationship between SOC and SOH, a set of fully symmetric multiple cells corresponding to the state variable prediction set at time k+1 is constructed. and

[0097] Step 5: Construct the band space at time k+1 based on the multicellular constraints. and strip space and The sets of fully symmetric multicells corresponding to the state variable prediction set at time k+1 constructed in step four. and Finding the intersection yields the state variable x at time k+1. c,k+1 and x h,k+1 A collection of fully symmetrical multicellular bodies and

[0098] Step Six: Based on the set of fully symmetric multicellular bodies obtained in Step Five and The upper and lower bounds of SOC and SOH are calculated respectively, and the specific values ​​of SOC and SOH are observed to complete the test of the power battery performance.

[0099] Example 2:

[0100] This embodiment provides a power battery performance testing method based on dual-member filtering. (See also...) Figure 1 The method includes:

[0101] Step 1: Obtain the equivalent circuit model of the power battery based on its electrochemical mechanism;

[0102] Equivalent circuit model such as Figure 2 As shown.

[0103] Reference Figure 2 Based on the electrochemical mechanism of the power battery discharge process, the following equation can be established:

[0104]

[0105] Where R0 is the ohmic resistance of the power battery, R1 and C1 are the electrochemical polarization internal resistance and capacitance of the power battery, respectively, and the voltage across R1 and C1 connected in parallel is U1; R2 and C2 are the concentration polarization resistance and capacitance of the power battery, respectively, and the voltage across R2 and C2 connected in parallel is U2; I is the discharge current, U... oc U is the internal power source of the battery, and U is the voltage across the battery terminals, i.e., the output voltage.

[0106] Step 2: Obtain the linear expression of the SOC of the power battery using the current integration method, and simultaneously characterize the SOH using the battery capacity.

[0107] Based on the current integral method for power batteries, a state-of-the-art (SOC) model for the power battery is established:

[0108]

[0109] Where SOC and SOC0 represent the state of charge of the battery at the current time and the initial time, respectively; Q represents the battery charge at the current time; η represents the coulombic efficiency of the battery; I represents the discharge current; and t and t0 represent the current time and the initial time, respectively.

[0110] Based on the relationship between the battery charge Q and the state of equilibrium (SOH), a state of equilibrium (SOH) model for the power battery is established:

[0111]

[0112] Where SOH represents the current state of battery health; Q and Q N These represent the current battery level and the initial battery level, respectively.

[0113] Step 3: Based on the coupling relationship between SOC and SOH, establish a joint estimation model for SOC and SOH;

[0114] Taking the battery discharge current I as input, the battery voltage U at time k 1,k U 2,k and SOC k Let U be the state variable, and let U be the voltage across the battery at time k. kFor the output, after discretization, the SOC linear model of the battery is established as follows:

[0115]

[0116] U k =U oc (SOC k )-U 1,k -U 2,k -IR0+v k (5)

[0117] Among them, [U 1,k U 2,k SOC k ] T For state variables, I is a constant representing the input quantity; ΔT represents the sampling time; τ1 = R1C1 represents the time constant after R1 and C1 are connected in parallel; τ2 = R2C2 represents the time constant after R2 and C2 are connected in parallel; U oc (SOC k+1 ) = 0.5158SOC k +3.624 indicates U oc The linear relationship between and SOC This represents the unknown but bounded perturbation noise in the SOC linear model, i.e. This represents the unknown but bounded measurement noise in the SOC linear model, i.e.

[0118] SOH at time k of the battery k As state variables, after discretization, the SOH linear model of the battery is established as follows:

[0119]

[0120] Among them, SOH k It is a state variable. This represents the unknown but bounded perturbation noise in the SOH linear model, i.e. This represents the unknown but bounded measurement noise in the SOH linear model, i.e. d k This represents the difference between the voltage of a single cell in the battery pack and the average voltage of the single cell.

[0121] Step 4: Construct the set of fully symmetric multicellular bodies for the joint estimation model at time k. and Based on the coupling relationship between SOC and SOH, a set of fully symmetric multiple cells corresponding to the state variable prediction set at time k+1 is constructed. and

[0122] 4.1 Expressions for the state-of-charge space equation and the state-of-health space equation of the power battery;

[0123] 4.1.1 with x c,k Let y represent the state variable at time k. c,k Let k represent the battery output voltage at time k. The state-space equation of the power battery represented by formulas (4) and (5) is expressed as follows:

[0124]

[0125] in, x c,k Represents the state variable [U] at time k. 1,k U 2,k SOC k ] T ;u c,k The discharge current I at time k is a constant; y c,k This represents the battery output voltage at time k.

[0126] 4.1.2 with x h,k SOH represents the state variable at time k. k y h,k d represents time k k The state-of-health equation of the power battery, expressed by formula (6), is as follows:

[0127]

[0128] Among them, A h =1, x h,k SOH represents the state variable at time k. k ;y h,k d represents time k k .

[0129] 4.2 Constructing a fully symmetric set of multiple cells for the joint estimation model at time k and

[0130] Define the fully symmetric multicell corresponding to the initialization state variable x0. This represents the center point of the fully symmetrical multicell at the initial moment. B represents the shape matrix of the fully symmetric multicell at the initial time step. m It is a unit box consisting of m unit intervals [-1, 1].

[0131] 4.2.1 Construction

[0132] Assume the state variable x of SOC in the joint estimation model at time k.c,k The corresponding fully symmetrical multicellular body is in, It is the center of the SOC multicellular body. Its shape matrix,

[0133] 4.2.2 Construction

[0134] Assume the state variable x of SOH in the joint estimation model at time k. h,k The corresponding fully symmetrical multicellular body is in, It is the center of the SOH multicellular body. Its shape matrix.

[0135] 4.3 Constructing the set of fully symmetric multiple cells corresponding to the predicted set of state variables at time k+1 and

[0136] Based on the state-space equations of the power battery's SOC and SOH, the input matrix C of the SOC model is... c The model includes the SOH variable at time k, and the output of the SOH model contains the difference between the SOC at time k and the SOC at time k-1. This indicates a coupling relationship between SOC and SOH; the SOH value at time k affects the estimation result of SOC at time k+1. Therefore, in constructing a fully symmetric multicellular set... and When doing so, the effects of coupling should be considered;

[0137] 4.3.1 Considering the coupling relationship between SOC and SOH, construct a fully symmetric multicellular set corresponding to the state of charge of the power battery at time k+1. for:

[0138]

[0139] Among them, F c For disturbance noise w k The corresponding generating matrix of a fully symmetric multicellular body, SOH k This is the result of the SOH update step at time k.

[0140] 4.3.2 Considering the coupling relationship between SOC and SOH, construct the fully symmetric multicellular set corresponding to the health state of the power battery at time k+1. for:

[0141]

[0142]

[0143] Among them, F h For disturbance noise γk The generation matrix of the corresponding fully symmetric multicell.

[0144] Step 5: Based on the multicellular constraints in equations (5) and (6), construct the band space at time k+1. and strip space and The sets of fully symmetric multiple cells corresponding to the predicted state variables at time k+1. and Finding the intersection yields the state variable x at time k+1. c,k+1 and x h,k+1 A collection of fully symmetrical multicellular bodies and

[0145] 5.1 Constructing the strip space at time k+1 and

[0146] 5.1.1 Constructing the strip space corresponding to the state of charge of the power battery at time k+1

[0147]

[0148] Where, x c,k+1 Let k+1 be the state variable of the SOC model. For matrix C c d c =U k +IR0, σ c =v k+1 ;

[0149] 5.1.2 Constructing the strip space corresponding to the health state of the power battery at time k+1

[0150]

[0151] Where, x h,k+1 Let k+1 be the state variable of the SOH model. For matrix C h d h =d k -SOC k+1 +SOC k , σ h =ε k+1 .

[0152] 5.2 Transforming the strip space and The sets of fully symmetric multiple cells corresponding to the predicted state variables at time k+1. and Finding the intersection yields the state variable x at time k+1. c,k+1 and x h,k+1 A collection of fully symmetrical multicellular bodies and

[0153] The set of fully symmetric multicellular structures of the state variable prediction set at time k+1 and Respectively with ribbon space and The intersection can be represented as:

[0154]

[0155] Using fully symmetrical multicellular structures pack and The intersection of the states is used to calculate the volume of a fully symmetric multicell. The fully symmetric multicell with the smallest volume is selected as the one containing the state variable x at time k+1. k+1 A collection of fully symmetrical multicellular bodies Right now:

[0156] The set of fully symmetric multicellular structures corresponding to the state of charge of a power battery for

[0157]

[0158] in,

[0159] The set of fully symmetric multicellular structures corresponding to the health status of a power battery for:

[0160]

[0161] in, SOC k+1 This is the SOC prediction step result at time k+1, SOC k This is the result of the SOC update step at time k.

[0162] Step Six: Based on the fully symmetrical multicellular structure obtained in Step Five and Calculate the upper and lower bounds of the results, observe the specific values, and complete the performance testing of the power battery, including...

[0163] Based on the fully symmetric multicellular set corresponding to the state of charge of the power battery at time k+1 obtained in step five. have to

[0164] The upper and lower bounds of the predicted value after the power battery SOC update are:

[0165]

[0166] Based on the fully symmetric multicellular set corresponding to the health state of the power battery at time k+1 obtained in step five.

[0167] The upper and lower bounds of the updated predicted SOH values ​​for the power battery are:

[0168]

[0169] (The updated forecast values ​​mentioned above are the final forecast values).

[0170] By observing the upper and lower bounds, we can determine whether the values ​​of SOC and SOH fluctuate within the normal range, thus completing the performance test of the power battery.

[0171] In this example, after executing steps one through six within a predetermined time range, a fully symmetric multicellular set of state variables is obtained, thereby realizing the performance detection of the power battery. Figure 3 and Figure 4 The figures show the estimated ranges of SOC and SOH for power batteries using different methods.

[0172] Existing methods for joint state estimation of power batteries' SOC and SOH can be found in "Collaborative Online Prediction of Lithium Battery SOC and SOH Based on Joint Algorithm [J], Journal of Terahertz Science and Electronic Information, 2021, 19(04):739-746". This method establishes a joint model of SOC and SOH based on the battery's equivalent circuit and uses the Kalman algorithm to estimate the battery's state. In this method, the battery's internal resistance is used to characterize the SOH value. However, the temperature change of the battery has a significant impact on the battery's internal resistance, which can lead to deviations in the state estimation. In contrast, the method in this application uses the battery capacity to characterize SOH, which avoids this problem. Furthermore, the joint estimation in this method involves substituting the result into the next calculation after each filtering step. In contrast, this method combines SOC and SOH in both the prediction and update steps of state estimation, resulting in more accurate and robust results.

[0173] from Figure 3 and Figure 4 It can be seen that both the existing estimation methods and the estimation method proposed in this application can achieve state estimation of the SOC and SOH of the power battery, and the true state values ​​are within the estimation range. However, when the system is stable, the method proposed in this application can estimate the SOC and SOH simultaneously, so the estimated value range is smaller, more accurate, and more efficient.

[0174] Some steps in the embodiments of the present invention can be implemented using software, and the corresponding software program can be stored in a readable storage medium, such as an optical disc or a hard disk.

[0175] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for testing the performance of a power battery based on dual-member filtering, characterized in that, The method includes: establishing a joint estimation model of SOC and SOH based on the coupling relationship between SOC and SOH, and using battery capacity to characterize SOH during the establishment process; based on the joint estimation model of SOC and SOH, performing alternating operations through a fully symmetric multi-cell algorithm, taking the update step result of SOH at time k as a known quantity and substituting it into the calculation of the SOC prediction step at time k+1; taking the result of the SOC prediction step at time k+1 as a known quantity and substituting it into the calculation of the SOH update step at time k+1, and using this nested dual-member filtering method to achieve alternating operations to complete the detection of power battery performance; The method includes: Step 1: Obtain the equivalent circuit model of the power battery based on its electrochemical mechanism; Step 2: Obtain the linear expression of SOC of the power battery, and use the battery capacity to characterize SOH; Step 3: Based on the coupling relationship between SOC and SOH, establish a joint estimation model for SOC and SOH; Step 4: Construct the set of fully symmetric multicellular bodies for the joint estimation model at time k. and Based on the coupling relationship between SOC and SOH, a set of fully symmetric multiple cells corresponding to the state variable prediction set at time k+1 is constructed. and Step 5: Construct the band space at time k+1 based on the multicellular constraints. and strip space and The sets of fully symmetric multicells corresponding to the state variable prediction set at time k+1 constructed in step four. and Finding the intersection yields the state variable x at time k+1. c,k+1 and x h,k+1 A collection of fully symmetrical multicellular bodies and Step Six: Based on the set of fully symmetric multicellular bodies obtained in Step Five and The upper and lower bounds of SOC and SOH are calculated respectively, and the specific values ​​of SOC and SOH are observed to complete the test of the power battery performance.

2. The method according to claim 1, characterized in that, Step one includes: Based on the electrochemical mechanism of the power battery discharge process, the equivalent circuit model of the power battery is established as shown in the following equation: Where R0 is the ohmic resistance of the power battery, R1 and C1 are the electrochemical polarization internal resistance and capacitance of the power battery, respectively, and the voltage across R1 and C1 connected in parallel is U1; R2 and C2 are the concentration polarization resistance and capacitance of the power battery, respectively, and the voltage across R2 and C2 connected in parallel is U2; I is the discharge current, U... oc U is the internal power source of the battery, and U is the voltage across the battery terminals, i.e., the output voltage.

3. The method according to claim 2, characterized in that, Step two includes: Based on the current integral method for power batteries, a state-of-the-art (SOC) model for the power battery is established: Where SOC and SOC0 represent the battery's state of charge at the current and initial moments, respectively; Q represents the battery's charge at the current moment; η represents the battery's coulombic efficiency; I represents the discharge current; and t and t0 represent the current and initial moments, respectively. Characterizing SOH using battery capacity: Based on the relationship between the battery capacity Q and SOH, an SOH model for the power battery is established. Where SOH represents the current state of battery health; Q and Q N These represent the current battery level and the initial battery level, respectively.

4. The method according to claim 3, characterized in that, Step three includes: Taking the battery discharge current I as input, the battery voltage U at time k 1,k U 2,k and SOC k Let U be the state variable, and let U be the voltage across the battery at time k. k For the output, after discretization, the SOC linear model of the battery is established as follows: IN k =U oc (SOC k )-IN 1,k -IN 2,k -IR0+v k (5) Among them, [U 1,k U 2,k SOC k ] T The state variable is the discharge current I, which is a constant and represents the input quantity; ΔT represents the sampling time; τ1=R1C1 represents the time constant after R1 and C1 are connected in parallel; τ2=R2C2 represents the time constant after R2 and C2 are connected in parallel. U oc (SOC k+1 ) = 0.5158SOC k +3.624 indicates the battery's internal power supply U. oc The linear relationship between and SOC This represents the unknown but bounded perturbation noise in the SOC linear model, i.e. This represents the unknown but bounded measurement noise in the SOC linear model, i.e. SOH at time k of the battery k As state variables, after discretization, the SOH linear model of the battery is established as follows: Among them, SOH k It is a state variable. This represents the unknown but bounded perturbation noise in the SOH linear model, i.e. This represents the unknown but bounded measurement noise in the SOH linear model, i.e. d k This represents the difference between the voltage of a single cell in the battery pack and the average voltage of the single cell. Formulas (4)-(6) are the established joint estimation models for SOC and SOH.

5. The method according to claim 4, characterized in that, Step four includes: 4.1 Expression of the joint estimation model of SOC and SOH of the power battery; 4.1.1 with x c,k Let y represent the state variable at time k. c,k Let k represent the battery output voltage at time k. The SOC linear model of the battery represented by formulas (4) and (5) is expressed as follows: in, x c,k Represents the state variable [U] at time k. 1,k U 2,k SOC k ] T ;u c,k The discharge current I at time k is a constant; y c,k This represents the battery output voltage at time k; 4.1.2 with x h,k SOH represents the state variable at time k. k y h,k d represents time k k The linear model of the SOH of the battery represented by formula (6) is expressed as follows: Among them, A h =1, x h,k SOH represents the state variable at time k. k ;y h,k d represents time k k ; 4.2 Constructing a fully symmetric set of multiple cells for the joint estimation model at time k and Define the fully symmetric multicell corresponding to the initialization state variable x0. This represents the center point of the fully symmetrical multicell at the initial moment. B represents the shape matrix of the fully symmetric multicell at the initial time step. m Let m be the unit boxes formed by the intervals [-1, 1]. 4.2.1 Construction Assume the state variable x of SOC in the joint estimation model at time k. c,k The corresponding fully symmetrical multicellular body is in, It is the center of the SOC multicellular body. Its shape matrix; 4.2.2 Construction Assume the state variable x of SOH in the joint estimation model at time k. h,k The corresponding fully symmetrical multicellular body is in, It is the center of the SOH multicellular body. Its shape matrix; 4.3 Constructing the set of fully symmetric multiple cells corresponding to the predicted set of state variables at time k+1 and Based on the joint estimation model of SOC and SOH of the power battery, the input matrix C of the SOC linear model is... c The model includes the SOH variable at time k, and the output of the SOH model also includes the difference between the SOC at time k and the SOC at time k-1. This indicates a coupling relationship between SOC and SOH; the SOH value at time k affects the estimation result of SOC at time k+1. Therefore, in constructing a fully symmetric multicellular set... and When doing so, the effects of coupling should be considered; 4.3.1 Considering the coupling relationship between SOC and SOH, construct a fully symmetric multicellular set corresponding to the state of charge of the power battery at time k+1. for: in, F c For disturbance noise w k The corresponding generating matrix of a fully symmetric multicellular body, SOH k This is the result of the SOH update step at time k; 4.3.2 Considering the coupling relationship between SOC and SOH, construct the fully symmetric multicellular set corresponding to the health state of the power battery at time k+1. for: Among them, F h For disturbance noise γ k The generation matrix of the corresponding fully symmetric multicell.

6. The method according to claim 5, wherein step five comprises: 5.1 Constructing the strip space at time k+1 and 5.1.1 Constructing the strip space corresponding to the state of charge of the power battery at time k+1 Where, x c,k+1 Let k+1 be the state variable of the SOC model. For matrix C c d c =U k +IR0, σ c =v k+1 ; 5.1.2 Constructing the strip space corresponding to the health state of the power battery at time k+1 Where, x h,k+1 Let k+1 be the state variable of the SOH model. For matrix C h d h =d k -SOC k+1 +SOC k , σ h =ε k+1 ; 5.2 Transforming the strip space and The sets of fully symmetric multiple cells corresponding to the predicted state variables at time k+1. and Finding the intersection yields the state variable x at time k+1. c,k+1 and x h,k+1 A collection of fully symmetrical multicellular bodies and The set of fully symmetric multicellular structures of the state variable prediction set at time k+1 and Respectively with ribbon space and The intersection can be represented as: Using fully symmetrical multicellular structures pack and The intersection of the states is used to calculate the volume of a fully symmetric multicell. The fully symmetric multicell with the smallest volume is selected as the one containing the state variable x at time k+1. k+1 A collection of fully symmetrical multicellular bodies Right now: The set of fully symmetric multicellular structures corresponding to the state of charge of a power battery for: in, The set of fully symmetric multicellular structures corresponding to the health status of a power battery for: in, SOC k+1 This is the SOC prediction step result at time k+1, SOC k This is the result of the SOC update step at time k.

7. The method according to claim 6, wherein step six comprises: Based on the fully symmetric multicellular set corresponding to the state of charge of the power battery at time k+1 obtained in step five. The upper and lower bounds of the final predicted value of the power battery SOC are: Based on the fully symmetric multicellular set corresponding to the health state of the power battery at time k+1 obtained in step five. The upper and lower bounds of the final predicted SOH value of the power battery are: By observing the upper and lower bounds, we can determine whether the values ​​of SOC and SOH fluctuate within the normal range, thus completing the performance test of the power battery.

8. A power battery performance testing system based on dual-member filtering, characterized in that, The system includes devices for detecting the current and voltage of the power battery, and the system uses the method described in any one of claims 1-7 to detect the state of charge (SOC) and state of health (SOH) of the power battery.

9. A fault detection method for power battery performance testing based on dual-member filtering, characterized in that, The method uses any one of the methods described in claims 1-7 to obtain the range of values ​​for the state of charge (SOC) and the state of health (SOH) of the power battery. The obtained range of values ​​for the state of charge (SOC) and the state of health (SOH) of the power battery are compared with the range of values ​​for the normal state of charge (SOC) and the state of health (SOH) of the power battery. If the difference exceeds a preset threshold, the power battery is judged to be faulty.

Images