Lithium ion battery long-time window power state prediction method and system

Abstract

The invention provides a method and a system for predicting a long-time window power state of a lithium ion battery. The method comprises the following steps of: 1, testing charge and discharge characteristics of the lithium ion battery; 2, constructing a lithium ion battery equivalent circuit model; 3, for long-time window power prediction, dividing a time window into N stages; 4, constructing a long-time window multi-constraint and multi-stage nonlinear optimization model; and step 5, performing charge / discharge SOP estimation based on the detected voltage and current. According to the method, a multi-stage optimization algorithm is adopted, and the influence of nonlinear dynamic change of the battery on power characteristics can be effectively processed, so that the accuracy and reliability of long-time window power state prediction of the lithium ion battery are ensured, and the overall performance of a battery management system is finally improved.

Description

Technical Field

[0001] The present invention relates to the technical field of battery management systems, and in particular, to a method and system for predicting the power state of a lithium-ion battery in a long-time window. Background Art

[0002] Due to advantages such as high energy density, high output power, and long charge-discharge life, lithium-ion batteries have been widely used in fields such as consumer electronics, electric vehicles, and energy storage power stations. The power state is a necessary parameter for implementing power distribution in a lithium-ion battery system and is crucial for ensuring the efficient and safe operation of the system.

[0003] Currently, the power prediction windows in related research are limited to the second level and the minute level, and it is considered that the current remains unchanged within a short-time window. There is still a blank in the long-time window power prediction (prediction of the maximum sustainable charge-discharge power for more than 5 minutes). The difficulty lies in that due to the strong non-linear dynamic characteristics of the charge-discharge curve of lithium-ion batteries, when the time window is long, the current within the window cannot be directly treated as a fixed value, and the maximum sustainable charge-discharge power is difficult to be equivalent to the power at the end or the beginning of the window, resulting in difficult prediction. Summary of the Invention

[0004] Aiming at the defects in the prior art, the purpose of the present invention is to provide a method and system for predicting the power state of a lithium-ion battery in a long-time window.

[0005] According to the method for predicting the power state of a lithium-ion battery in a long-time window provided by the present invention, it includes:

[0006] Step 1: Testing the charge-discharge characteristics of the lithium-ion battery;

[0007] Step 2: Constructing an equivalent circuit model of the lithium-ion battery;

[0008] Step 3: For long-time window power prediction, dividing the time window into N stages;

[0009] Step 4: Constructing a multi-constraint and multi-stage non-linear optimization model for a long-time window;

[0010] Step 5: Based on the detected voltage and current, estimating the charge / discharge SOP.

[0011] Preferably, an open-circuit voltage experiment is performed on the lithium-ion battery to establish an open-circuit voltage model. First, the lithium-ion battery is charged to the cut-off voltage in a constant current and constant voltage manner and left standing for a period of time; then, the battery is continuously discharged at a 1C rate current to a specific SOC and left standing for 4 hours; where the discharge current is defined as a positive value and the charge current is defined as a negative value; during the whole process, the battery terminal voltage and the load current are synchronously collected at a sampling frequency of 1Hz; according to the SOC of each standing point and the corresponding open-circuit voltage measurement value, a functional relationship between the open-circuit voltage and the SOC is established;

[0012] The open-circuit voltage VOC and the SOC are represented by a 16th-order polynomial function relationship:

[0013]

[0014] where n represents the highest degree of the polynomial, VOC k , SOC k respectively represent the open-circuit voltage VOC and the state of charge SOC at the kth sampling point, and β 1j represents the coefficient of the jth term;

[0015] The required SOC is calculated according to the current integration method:

[0016]

[0017] where SOC(0) is the initial SOC value of the battery, C is the battery capacity, and I is the load current;

[0018] Combining relationships (1) and (2), the least squares method is used to identify the parameter β 1j to determine the functional relationship between the open-circuit voltage and the SOC.

[0019] Preferably, the lithium-ion battery is first charged to the fully charged state, and then discharged at a rate of 1C to a specific SOC, and then charged and discharged under a simplified DST condition. Finally, pulse charging and discharging are performed at rates of 0.125C, 0.75C, 1.5C, 2C, -0.125C, -0.75C, -1.5C, and -2C in sequence, and each pulse current lasts for 1 minute. The above steps are continuously repeated until the discharge cut-off voltage is reached;

[0020] The equivalent circuit model includes three parts: the open-circuit voltage VOC, the ohmic internal resistance R0, and the first-order or multi-order RC network, where the RC network is composed of the polarization resistance R1 and the equivalent capacitance C1;

[0021] A first-order equivalent circuit model is adopted, where V b is the battery voltage and I is the load current. This equivalent circuit model conforms to the following voltage-current relationship:

[0022]

[0023] V b = V OC - V1 - IR0 (4)

[0024] where V1 represents the polarization voltage;

[0025] For a sampling period Δt, the discretized forms of relationships (3) and (4) are expressed as:

[0026]

[0027] V b,k =V OC (SOC k ) - V 1,k -I k R 0,k (6)

[0028] Among them, V 1,k+1 represents the polarization voltage at the (k + 1)-th moment; τ 1,k represents the time constant at the k-th moment; I k represents the current at the k-th moment; R 1,k represents the polarization resistance at the k-th moment; V b,k represents the terminal voltage of the battery at the k-th moment; SOC k represents the state of charge SOC at the k-th moment; R 0,k represents the ohmic internal resistance at the k-th moment;

[0029] The time constant τ k =R 1，k C k ; R 1，k represents the polarization resistance at the k-th moment; C k represents the polarization capacitance at the k-th moment;

[0030] In equations (5) and (6), the circuit parameters R0, R1, and C1 are functions of SOC and are expressed in polynomial form:

[0031]

[0032] Among them, the parameters to be identified are the polynomial coefficients β 2j , β 3j , β 4j representing the coefficient of the j-th term. During the identification process, based on equations (5) and (6), the least squares method is used to fit the voltage response curve to obtain β 2j , β 3j , β 4j .

[0033] Preferably, according to the definition of the maximum charge-discharge power, during the charge-discharge process, the peak power P t,peak at any time t is defined as equation (10). Among them, since the charging current is defined as a negative value, the charging peak power is expressed as the minimum charging power:

[0034]

[0035] Among them, P t,max , P t,min respectively represent the maximum and minimum values of the charge-discharge power at time t;

[0036] For the current moment \(t\) k , the power prediction problem is defined as an optimization problem of finding the sustainable peak power within a future time window \(L\), and the optimization objective is expressed as Equation (11):

[0037]

[0038] where \(P\) i represents the power value at the \(i\)-th sampling point;

[0039] When the time window is in seconds and minutes, the current within the \(L\) time window is considered constant. During the discharging process, the battery terminal voltage continuously decreases within the window interval, and the minimum power appears at the end moment; during the charging process, the battery terminal voltage continuously rises within the window interval, and the maximum power within the window interval appears at the initial moment;

[0040] The time window is divided into \(n\) stages, and the number of sampling points in each stage is \(n_1,\cdots,n\) i , \(1\leq n\) i \(\leq L\), taking The time of each stage is expressed as The current within each stage is respectively expressed as

[0041] Adopting a segmented optimization strategy to optimize the power of each stage, the optimization objective of the battery peak power is transformed into Equation (12):

[0042]

[0043] where the optimization variable is the current of each stage. Within each stage, the minimum discharging power appears at the end moment, and it is considered that \(P\) t is the same as ; the maximum charging power appears at the initial moment, and it is considered that \(P\) t is the same as ; \(P\) t represents the power at moment \(t\).

[0044] Preferably, when implementing peak power estimation, the following multiple constraints are considered simultaneously:

[0045]

[0046] where \(V\) b,min and \(V\) b,max are the upper and lower limits of the designed terminal voltage \(V\) b , \(I\) min and \(I\) max are the upper and lower limits of the designed current \(I\), \(SOC\) min and \(SOC\) max are the upper and lower limits of the designed \(SOC\), \(P\) min and \(P\)max Design the upper and lower limits for the sustainable peak P of the battery;

[0047] Combined with the equivalent circuit model and multi-parameter constraints, the SOP estimation problem is transformed into a multi-constrained non-linear optimization problem within the time window L. The discharge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(1), and the charge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(2):

[0048] Discharge process:

[0049]

[0050] Charge process:

[0051]

[0052] Among them, for the discharge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the discharge cut-off SOC, expressed as and not greater than 1, and the peak power constraint is the maximum allowable discharge power For the charge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the charge cut-off SOC, expressed as and greater than 0, and the peak power constraint is the maximum allowable discharge power

[0053] According to the detected voltage and current, adopt a multi-stage optimization algorithm, segment within the time interval [k, k + L], and solve the multi-constrained non-linear optimization problems P(1) and P(2) to obtain the discharge SOP and charge SOP at any k moment.

[0054] According to the lithium-ion battery long-time window power state prediction system provided by the present invention, it includes:

[0055] Module M1: Test the charging and discharging characteristics of lithium-ion batteries;

[0056] Module M2: Construct an equivalent circuit model of lithium-ion batteries;

[0057] Module M3: For long-time window power prediction, divide the time window into N stages;

[0058] Module M4: Construct a long-time window multi-constrained, multi-stage non-linear optimization model;

[0059] Module M5: Estimate charge / discharge SOP based on detected voltage and current.

[0060] Preferably, perform an open-circuit voltage experiment on the lithium-ion battery to establish an open-circuit voltage model. First, charge the lithium-ion battery to the cut-off voltage in a constant current-constant voltage manner and let it stand for a period of time; then discharge the battery continuously at a 1C rate to a specific SOC and let it stand for 4 hours; where the discharge current is defined as positive and the charge current is defined as negative; during the whole process, the battery terminal voltage and the load current are synchronously collected at a sampling frequency of 1Hz; according to the SOC at each standing point and its corresponding open-circuit voltage measurement value, establish the functional relationship between the open-circuit voltage and SOC.

[0061] Use a 16th-order polynomial form to represent the functional relationship between the open-circuit voltage VOC and SOC:

[0062]

[0063] In the formula, n represents the highest degree of the polynomial, VOC k , SOC k respectively represent the open-circuit voltage VOC and the state of charge SOC at the kth sampling point, and β 1j represents the coefficient of the jth term;

[0064] The required SOC is calculated by the current integration method:

[0065]

[0066] In the formula, SOC(0) is the initial SOC value of the battery, C is the battery capacity, and I is the load current;

[0067] Combining relationships (1) and (2), use the least squares method to perform parameter identification on β 1j to determine the functional relationship between the open-circuit voltage and SOC.

[0068] Preferably, first charge the lithium-ion battery to the fully charged state, then discharge it to a specific SOC at a 1C rate first, then perform charge and discharge under a simplified DST condition, and finally perform pulse charge and discharge at rates of 0.125C, 0.75C, 1.5C, 2C, -0.125C, -0.75C, -1.5C, -2C in sequence, each pulse current lasting for 1 minute, and repeat the above steps continuously until the discharge cut-off voltage is reached;

[0069] The equivalent circuit model includes three parts: the open-circuit voltage VOC, the ohmic internal resistance R0, and a first-order or multi-order RC network, where the RC network consists of a polarization resistance R1 and an equivalent capacitance C1;

[0070] Adopt a first-order equivalent circuit model, where V bV is the battery voltage and I is the load current. The equivalent circuit model conforms to the following voltage-current relationship:

[0071]

[0072] V b = V OC - V1 - IR0 (4)

[0073] where V1 represents the polarization voltage;

[0074] For a sampling period Δt, the discretized forms of equations (3) and (4) are expressed as:

[0075]

[0076] V b,k = V OC (SOC k ) - V 1,k - I k R 0,k (6)

[0077] where V 1,k+1 represents the polarization voltage at the (k + 1)-th moment; τ 1,k represents the time constant at the k-th moment; I k represents the current at the k-th moment; R 1,k represents the polarization resistance at the k-th moment; V b,k represents the terminal voltage of the battery at the k-th moment; SOC k represents the state of charge SOC at the k-th moment; R 0,k represents the ohmic internal resistance at the k-th moment;

[0078] The time constant τ k = R 1，k C k ; R 1，k represents the polarization resistance at the k-th moment; C k represents the polarization capacitance at the k-th moment;

[0079] In equations (5) and (6), the circuit parameters R0, R1, and C1 are functions of SOC and are expressed in polynomial form:

[0080]

[0081] where the parameters to be identified are the polynomial coefficients β 2j , β 3j , β 4j representing the coefficient of the j-th term. During the identification process, based on equations (5) and (6), the least squares method is used to fit the voltage response curve to obtain β 2j , β 3j , β4j .

[0082] Preferably, according to the definition of the maximum charge-discharge power, during the charge-discharge process, at any time t, the peak power P t,peak is defined as Equation (10), where, since the charging current is defined as a negative value, the charging peak power is expressed as the minimum charging power:

[0083]

[0084] where P t,max , P t,min respectively represent the maximum and minimum values of the charge-discharge power at time t;

[0085] For the current time t k , the power prediction problem is defined as an optimization problem of finding the sustainable peak power within a future time window L, and the optimization objective is expressed as Equation (11):

[0086]

[0087] where P i represents the power value at the i-th sampling point;

[0088] When the time window is in seconds and minutes, the current is considered to be constant within the L time window. During the discharge process, the battery terminal voltage continuously decreases within the window interval, and the minimum power appears at the end moment; during the charging process, the battery terminal voltage continuously rises within the window interval, and the maximum power within the window interval appears at the initial moment;

[0089] The time window is divided into n stages, and the number of sampling points in each stage is n1,…,n i , 1 ≤ n i ≤ L, take The time of each stage is expressed as The current within each stage is respectively expressed as

[0090] Adopt a segmented optimization strategy to optimize the power of each stage, and the battery peak power optimization objective is transformed into Equation (12):

[0091]

[0092] where the optimization variable is the current of each stage. Within each stage, the minimum discharge power appears at the end moment, and it is considered that P t is the same as ; the maximum charging power appears at the initial moment, and it is considered that P t is the same as ; P t represents the power at time t.

[0093] Preferably, when implementing peak power estimation, the following multiple constraints are considered simultaneously:

[0094]

[0095] Among them, V b,min and V b,max are the upper and lower limits of the designed terminal voltage V b I min and I max are the upper and lower limits of the designed current I, SOC min and SOC max are the upper and lower limits of the designed SOC, P min and P max are the upper and lower limits of the sustainable peak P of the battery;

[0096] Combined with the equivalent circuit model and multiple parameter constraints, the SOP estimation problem is transformed into a multi-constrained non-linear optimization problem within the time window L. The discharge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(1), and the charge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(2):

[0097] Discharge process:

[0098]

[0099] Charge process:

[0100]

[0101]

[0102] Among them, for the discharge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the discharge cut-off SOC, expressed as and not greater than 1, and the peak power constraint is the maximum allowable discharge power For the charge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the charge cut-off SOC, expressed as and greater than 0, and the peak power constraint is the maximum allowable discharge power

[0103] According to the detected voltage and current, using the multi-stage optimization algorithm, segment within the time interval [k, k + L] and solve the multi-constrained non-linear optimization problems P(1) and P(2) to obtain the discharge SOP and charge SOP at any k moment.

[0104] Compared with the prior art, the present invention has the following beneficial effects:

[0105] The method of the present invention adopts a multi-stage optimization algorithm, which can effectively handle the influence of the non-linear dynamic change of the battery on the power characteristics, thereby ensuring the accuracy and reliability of the long-time window power state prediction of the lithium-ion battery, and finally improving the overall performance of the battery management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] Other features, objects and advantages of the present invention will become more apparent by reading the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0107] Figure 1 FIG. 1 is a schematic diagram of a method for estimating the power state of a lithium-ion battery under multiple constraint conditions;

[0108] Figure 2 FIG. 2 is the structure of the power prediction device in the embodiment of the present invention;

[0109] FIG. 3 is a waveform diagram of current excitation and voltage response in the embodiment of the present invention;

[0110] Figure 4 FIG. 4 is an equivalent circuit model diagram of a lithium battery in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0111] The present invention will be described in detail below with reference to specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that those of ordinary skill in the art can make several changes and improvements without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0112] Embodiment 1

[0113] The present invention provides a method for estimating the long-term power state of a lithium-ion battery, which is applied to a lithium battery energy storage device having a microcontroller and a memory. The battery power state estimator includes two parts: model construction and online algorithm implementation. The model construction includes the following two steps: establishing an open-circuit voltage model; establishing an equivalent circuit model. The online algorithm implementation includes the following two steps: based on the definition of the maximum sustainable charge and discharge power, combining the offline model and the multi-stage optimization algorithm, establishing a multi-stage long-time window charge / discharge online prediction model; according to the detected voltage and current, using this online estimation model, realizing long-time window power online prediction.

[0114] As Figure 1, which is the schematic diagram of the battery power state estimation method based on measurement noise tracking of the present invention. The method of the present invention is applied to a lithium battery management system for estimating the state of charge of the lithium battery energy storage device. In the specific embodiment of the present invention, the structure of the lithium battery power state SOP estimation device is as Figure 2 shown, including a microcontroller 100, a memory 102, a current and voltage meter 104, and an SOP estimator 106. The microcontroller 100 generally controls the SOP estimation device and the current and voltage meter. The memory 102 is used to store the programs executed by the controller. The current and voltage meter 104 measures the current and voltage according to the control of the controller 100. The SOP estimator 106 estimates the SOP according to the control of the controller 100 and provides the estimation result to the controller 100. The establishment of the SOP estimator includes offline model construction and online algorithm application.

[0115] The offline model construction of the SOP estimator includes the following two steps:

[0116] Step S11, conduct an open-circuit voltage experiment on the battery to establish an open-circuit voltage model. The battery is a ternary system battery with a capacity of 25 Ah. First, charge the battery to the cut-off voltage in a constant current and constant voltage manner and let it stand for a certain period of time; then discharge the battery continuously at a 1C rate current to a specific SOC and let it stand for 4 hours. Among them, the discharge current is defined as a positive value, and the charge current is defined as a negative value. During the whole process, the battery terminal voltage and the load current are synchronously collected at a sampling frequency of 1 Hz. According to the SOC at each standing point and its corresponding open-circuit voltage measurement value, establish the functional relationship between the open-circuit voltage and the SOC. In the specific implementation of the present invention, the functional relationship between the open-circuit voltage VOC and the SOC is represented in the form of a 16th-order polynomial:

[0117]

[0118] where n represents the highest degree of the polynomial, VOC k and SOC k respectively represent the open-circuit voltage VOC and the state of charge SOC at the kth sampling point, and β 1j represents the coefficient of the jth term.

[0119] The required SOC is calculated by the current integration method:

[0120]

[0121] where SOC(0) is the initial SOC value of the battery, C is the battery capacity, and I is the load current. Combining the relational expressions (1) and (2), the least squares method is used to identify the parameters of β 1j to determine the functional relationship between the open-circuit voltage and the SOC.

[0122] Step S12: Conduct peak power test characteristics test on the battery and establish an equivalent circuit model based on the measured data. The peak power test characteristics test is shown in Figure 3. First, charge the battery to the fully charged state. Then, discharge it at a rate of 1C to a specific SOC, and then conduct charge and discharge under the simplified DST working condition. Finally, conduct pulse charge and discharge at rates of 0.125C, 0.75C, 1.5C, 2C, -0.125C, -0.75C, -1.5C, and -2C in sequence, with each pulse current lasting for 1 minute. Continuously repeat the above steps until the discharge cut-off voltage is reached. In this embodiment, the battery under test is an energy storage battery, the rate range is [-2C, 2C], and the SOC interval of the obtained power test curve is [90%, 10%]. The set charge cut-off voltage is 3.6V, and the discharge cut-off voltage is 2.5V. The equivalent circuit model includes three parts: open circuit voltage VOC, ohmic internal resistance R0, and a first-order or multi-order RC network. The RC network consists of a polarization resistance R1 and an equivalent capacitance C1, and the open circuit voltage VOC is determined by step S11. In the specific implementation of the present invention, a first-order equivalent circuit model is adopted, as Figure 4 shown, where V b is the battery voltage and I is the load current. This equivalent circuit model conforms to the following voltage-current relationship:

[0123]

[0124] V b = V OC - V1 - IR0 (4)

[0125] V1 represents the polarization voltage;

[0126] For a sampling period Δt, the discretized forms of relationships (3) and (4) can be expressed as:

[0127]

[0128] V b,k = V OC (SOC k ) - V 1,k - I k R 0,k (6)

[0129] where, V 1,k+1 represents the polarization voltage at the (k + 1)-th moment; τ 1,k represents the time constant at the k-th moment; I k represents the current at the k-th moment; R 1,k represents the polarization resistance at the k-th moment; V b,k represents the terminal voltage of the battery at the k-th moment; SOC k represents the state of charge SOC at the k-th moment; R 0,kDenote the ohmic internal resistance at time k;

[0130] The time constant τ k = R 1，k C k ; R 1，k Denote the polarization resistance at time k; C k Denote the polarization capacitance at time k;

[0131] In a specific implementation of the present invention, Δt is 1 s.

[0132] In a specific implementation of the present invention, the circuit parameters R0, R1, and C1 in formulas (5) and (6) are functions of SOC and are represented in polynomial form:

[0133]

[0134] Among them, the parameters to be identified are the polynomial coefficients β 2j , β 3j , β 4j Denote the coefficient of the j-th term. During the identification process, based on formulas (5) and (6), the least squares method is used to fit the voltage response curve in Figure 2 to obtain β 2j , β 3j , β 4j . Thus, the construction of the offline model is completed.

[0135] The online implementation of the SOC estimator includes the following steps:

[0136] Step S21, according to the definition of the maximum charge and discharge power, during the charge and discharge process, the peak power P at any time t t,peak can be defined as formula (10). Among them, since the charging current is defined as a negative value, the charging peak power is expressed as the minimum charging power:

[0137]

[0138] P t,max , P t,min respectively represent the maximum and minimum values of the charge and discharge power that can be achieved at time t;

[0139] For the current time t k , the power prediction problem is defined as an optimization problem of finding the sustainable peak power within a future time window L, and the optimization objective can be expressed as formula (11):

[0140]

[0141] P i represents the power value at the i-th sampling point;

[0142] When the time window is at the second level and the minute level, the current within the L time window can be considered constant. During the discharging process, the battery terminal voltage continuously decreases within the window interval, and the minimum power appears at the end moment; during the charging process, the battery terminal voltage continuously rises within the window interval, and the maximum power within the window interval appears at the initial moment. However, when the window is relatively long, if the current within the time window is still treated as a fixed value, it will result in a conservative power prediction. Therefore, the time window is divided into n stages, and the number of sampling points in each stage is n1, …, n i (1 ≤ n i ≤ L). Take Correspondingly, the time of each stage can be expressed as Since the window length of each stage is within the minute level, the current can be considered constant within each stage, and they are respectively expressed as Adopt a segmented optimization strategy to optimize the power of each stage. Correspondingly, the optimization target of the battery peak power can be transformed into Equation (12):

[0143]

[0144] Among them, the optimization variable is the current of each stage. Since the time length of each stage is less than the minute level, within each stage, the minimum discharging power appears at the end moment, and it can be considered that P t is the same as , and the maximum charging power appears at the initial moment, and it can be considered that P t is the same as .

[0145] P t represents the power at time t;

[0146] Step S22. To ensure the safe operation of the battery and slow down the aging speed of the battery, it is necessary to limit the peak power with current, voltage, SOC, etc. Therefore, when implementing the peak power estimation, the following multiple constraints need to be considered simultaneously:

[0147]

[0148] Among them, V b,min and V b,max are the upper and lower limits of the terminal voltage V b design, I min and I max are the upper and lower limits of the design current I, SOC min and SOC max are the upper and lower limits of the SOC design, P min and P max are the upper and lower limits of the sustainable peak P of the battery. For the charging and discharging processes, the upper and lower limits of each parameter are pre-set by the manufacturer or the operating unit.

[0149] At this time, combined with the equivalent circuit model and multi-parameter constraints, the SOP estimation problem can be transformed into a multi-constrained non-linear optimization problem within the time window L. The discharge SOP estimation problem can be expressed as the multi-constrained non-linear optimization problem P(1), and the charge SOP estimation problem can be expressed as the multi-constrained non-linear optimization problem P(2):

[0150] Discharge process:

[0151]

[0152] Charge process:

[0153]

[0154] Among them, for the discharge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the discharge cut-off SOC, expressed as and not greater than 1, and the peak power constraint is the maximum allowable discharge power For the charge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the charge cut-off SOC, expressed as and greater than 0, and the peak power constraint is the maximum allowable discharge power

[0155] Step S23, according to the detected voltage and current, adopt a multi-stage optimization algorithm to segment within the time interval [k, k + L] and solve the multi-constrained non-linear optimization problems P(1) and P(2), then the discharge SOP and charge SOP at any k moment can be obtained.

[0156] Embodiment 2

[0157] The present invention also provides a long-time window power state prediction system for a lithium-ion battery. The long-time window power state prediction system for a lithium-ion battery can be realized by executing the process steps of the long-time window power state prediction method for a lithium-ion battery. That is, those skilled in the art can understand the long-time window power state prediction method for a lithium-ion battery as a preferred implementation manner of the long-time window power state prediction system for a lithium-ion battery.

[0158] The lithium-ion battery long-time window power state prediction system provided by the present invention includes: Module M1: Testing the charge and discharge characteristics of the lithium-ion battery; Module M2: Constructing an equivalent circuit model of the lithium-ion battery; Module M3: For long-time window power prediction, dividing the time window into N stages; Module M4: Constructing a multi-constraint, multi-stage non-linear optimization model for the long-time window; Module M5: Estimating the charge / discharge SOP based on the detected voltage and current.

[0159] Perform an open-circuit voltage experiment on the lithium-ion battery to establish an open-circuit voltage model. First, charge the lithium-ion battery to the cut-off voltage in a constant current and constant voltage manner and let it stand for a period of time; then discharge the battery continuously at a 1C rate to a specific SOC and let it stand for 4 hours; where the discharge current is defined as a positive value and the charge current is defined as a negative value; during the whole process, the battery terminal voltage and the load current are synchronously collected at a sampling frequency of 1Hz; according to the SOC at each standing point and the measured open-circuit voltage corresponding thereto, establish the functional relationship between the open-circuit voltage and the SOC.

[0160] Use a 16th-order polynomial form to represent the functional relationship between the open-circuit voltage VOC and the SOC:

[0161]

[0162] In the formula, n represents the highest degree of the polynomial, VOC k 、SOC k respectively represent the open-circuit voltage VOC and the state of charge SOC at the kth sampling point, and β 1j represents the coefficient of the jth term;

[0163] The required SOC is calculated by the current integration method:

[0164]

[0165] In the formula, SOC(0) is the initial SOC value of the battery, C is the battery capacity, and I is the load current;

[0166] Combining relational expressions (1) and (2), use the least squares method to perform parameter identification on β 1j to determine the functional relationship between the open-circuit voltage and the SOC.

[0167] First charge the lithium-ion battery to the fully charged state, then discharge it to a specific SOC at a 1C rate first, then perform charge and discharge under a simplified DST working condition, and finally perform pulse charge and discharge at rates of 0.125C, 0.75C, 1.5C, 2C, -0.125C, -0.75C, -1.5C, -2C in sequence, each pulse current lasts for 1 minute, and repeat the above steps continuously until reaching the discharge cut-off voltage;

[0168] The equivalent circuit model consists of three parts: the open-circuit voltage VOC, the ohmic internal resistance R0, and a first-order or multi-order RC network. The RC network is composed of the polarization resistance R1 and the equivalent capacitance C1;

[0169] A first-order equivalent circuit model is adopted, where V b is the battery voltage and I is the load current. This equivalent circuit model conforms to the following voltage-current relationship:

[0170]

[0171] V b = V OC - V1 - IR0 (4)

[0172] where V1 represents the polarization voltage;

[0173] For a sampling period Δt, the discretized forms of equations (3) and (4) are expressed as:

[0174]

[0175] V b,k = V OC (SOC k ) - V 1,k - I k R 0,k (6)

[0176] where V 1,k+1 represents the polarization voltage at the (k + 1)-th moment; τ 1,k represents the time constant at the k-th moment; I k represents the current at the k-th moment; R 1,k represents the polarization resistance at the k-th moment; V b,k represents the terminal voltage of the battery at the k-th moment; SOC k represents the state of charge SOC at the k-th moment; R 0,k represents the ohmic internal resistance at the k-th moment;

[0177] The time constant τ k = R 1，k C k ; R 1，k represents the polarization resistance at the k-th moment; C k represents the polarization capacitance at the k-th moment;

[0178] The circuit parameters R0, R1, and C1 in equations (5) and (6) are functions of SOC and are expressed in polynomial form:

[0179]

[0180] where the parameters to be identified are the polynomial coefficients β 2j, β 3j , β 4j denotes the coefficient of the j-th term. During the identification process, based on Equations (5) and (6), the least squares method is used to fit the voltage response curve to obtain β 2j , β 3j , β 4j .

[0181] According to the definition of the maximum charge and discharge power, during the charge and discharge process, the peak power P at any time t t,peak is defined as Equation (10). Among them, since the charging current is defined as a negative value, the charging peak power is expressed as the minimum charging power:

[0182]

[0183] where P t,max , P t,min respectively represent the maximum and minimum values of the charge and discharge power at time t;

[0184] For the current time t k , the power prediction problem is defined as an optimization problem of finding the sustainable peak power within a future time window L, and the optimization objective is expressed as Equation (11):

[0185]

[0186] where P i represents the power value at the i-th sampling point;

[0187] When the time window is in seconds and minutes, the current within the L time window is considered constant. During the discharge process, the battery terminal voltage continuously decreases within the window interval, and the minimum power appears at the end moment; during the charging process, the battery terminal voltage continuously increases within the window interval, and the maximum power within the window interval appears at the initial moment;

[0188] The time window is divided into n stages, and the number of sampling points in each stage is n1,..., n i , 1 ≤ n i ≤ L, take The time of each stage is expressed as The current within each stage is respectively expressed as

[0189] Adopt a segmented optimization strategy to optimize the power of each stage. The optimization objective of the battery peak power is transformed into Equation (12):

[0190]

[0191] where the optimization variable is the current of each stage. Within each stage, the minimum discharge power appears at the end moment, and it is considered that P t and Same; the maximum charging power appears at the initial moment, considered as P t and Same; P t represents the power at time t.

[0192] When implementing peak power estimation, the following multiple constraints are considered simultaneously:

[0193]

[0194] Among them, V b,min and V b,max are the upper and lower limits of the terminal voltage V b The upper and lower limits of the designed current are I min and I max The upper and lower limits of the designed SOC are SOC min and SOC max The upper and lower limits of the designed P are P min and P max are the upper and lower limits of the sustainable peak P of the battery;

[0195] Combined with the equivalent circuit model and multiple parameter constraints, the SOP estimation problem is transformed into a multi-constrained non-linear optimization problem within the time window L. The discharge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(1), and the charge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(2):

[0196] Discharge process:

[0197]

[0198] Charge process:

[0199]

[0200] Among them, for the discharge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the discharge cut-off SOC, expressed as and not greater than 1. The peak power constraint is the maximum allowable discharge power For the charge process, the current constraint is the maximum allowable discharge current The terminal voltage constraint is the discharge cut-off voltage and the charge cut-off voltage The SOC constraint is the charge cut-off SOC, expressed as and greater than 0. The peak power constraint is the maximum allowable discharge power

[0201] According to the detected voltage and current, a multi-stage optimization algorithm is adopted to segment within the time interval [k, k + L] and solve the multi-constraint non-linear optimization problems P(1) and P(2), so as to obtain the discharging SOP and charging SOP at any moment k.

[0202] Those skilled in the art know that in addition to implementing the systems, devices and their respective modules provided by the present invention in the form of pure computer-readable program codes, the method steps can be logically programmed to enable the systems, devices and their respective modules provided by the present invention to be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. to implement the same program. Therefore, the systems, devices and their respective modules provided by the present invention can be regarded as a kind of hardware component, and the modules included therein for implementing various programs can also be regarded as the structures within the hardware component; the modules for implementing various functions can also be regarded as either software programs for implementing the methods or the structures within the hardware component.

[0203] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. Without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for predicting the power state of a lithium-ion battery in a long-time window, characterized in that Including: Step 1: Test the charge and discharge characteristics of the lithium-ion battery; Step 2: Construct an equivalent circuit model of the lithium-ion battery; Step 3: For long-time window power prediction, divide the time window into N stages; Step 4: Construct a multi-constraint, multi-stage non-linear optimization model for the long-time window; Step 5: Based on the detected voltage and current, perform charge / discharge SOP estimation.

2. The method for predicting the long-time window power state of a lithium-ion battery according to claim 1, wherein Conduct an open-circuit voltage experiment on the lithium-ion battery to establish an open-circuit voltage model. First, charge the lithium-ion battery to the cut-off voltage in a constant current and constant voltage manner and let it stand for a period of time; then discharge the battery continuously at a 1C rate to a specific SOC and let it stand for 4 hours; among them, the discharge current is defined as a positive value and the charge current is defined as a negative value; during the whole process, the battery terminal voltage and the load current are synchronously collected at a sampling frequency of 1Hz; according to the SOC of each standing point and its corresponding open-circuit voltage measurement value, establish the functional relationship between the open-circuit voltage and the SOC; Use a 16th-order polynomial form to represent the functional relationship between the open-circuit voltage VOC and the SOC: Where n represents the highest degree of the polynomial, VOC k , SOC k They represent the open circuit voltage VOC and state of charge SOC of the kth sampling point, β 1j represents the coefficient of the j-order term; The required SOC is calculated by the current integration method: In the formula, SOC(0) is the initial SOC value of the battery, C is the battery capacity, and I is the load current; Combining relationship expressions (1) and (2), the least squares method is used to identify the parameter of β 1j to determine the functional relationship between the open-circuit voltage and the SOC.

3. The method for predicting the power state of a lithium-ion battery in a long-time window according to claim 2, wherein, First, charge the lithium-ion battery to the fully charged state. Then, first discharge it to a specific SOC at a 1C rate, and then perform charge and discharge under the simplified DST working condition. Finally, perform pulse charge and discharge at rates of 0.125C, 0.75C, 1.5C, 2C, -0.125C, -0.75C, -1.5C, -2C in turn, and each pulse current lasts for 1 minute. Continuously repeat the above steps until the discharge cut-off voltage is reached; The equivalent circuit model includes three parts: the open-circuit voltage VOC, the ohmic internal resistance R0, and the first-order or multi-order RC network, where the RC network consists of the polarization resistance R1 and the equivalent capacitance C1; Adopt a first-order equivalent circuit model, where V b is the battery voltage and I is the load current. This equivalent circuit model conforms to the following voltage-current relationship: V b = V OC - V1 - IR0 (4) Among them, V1 represents the polarization voltage; For a sampling period Δt, the discretized forms of the relational expressions (3) and (4) are expressed as: V b,k = V OC (SOC k ) - V 1,k - I k R 0,k (6) Among them, V 1,k+1 represents the polarization voltage at the (k + 1)-th moment; τ 1,k represents the time constant at the k-th moment; I k represents the current at the k-th moment; R 1,k represents the polarization resistance at the k-th moment; V b,k represents the terminal voltage of the battery at the k-th moment; SOC k represents the state of charge SOC at the k-th moment; R 0,k represents the ohmic internal resistance at the k-th moment; Time constant τ k = R 1，k C k ; R 1，k represents the polarization resistance at time k; C k represents the polarization capacitance at time k; In the formulas (5) and (6), the circuit parameters R0, R1, and C1 are functions of the SOC, and are expressed in polynomial form: Among them, the required identification parameter is the polynomial coefficient β 2j , β 3j , β 4j represents the coefficient of the j-th term. During the identification process, based on equations (5) and (6), the least squares method is used to fit the voltage response curve to obtain β 2j , β 3j , β 4j .

4. The method for predicting the power state of a lithium-ion battery with a long time window according to claim 3, wherein According to the definition of the maximum charge-discharge power, during the charge-discharge process, the peak power P at any time t t,peak is defined as Equation (10). Among them, since the charging current is defined as a negative value, the charging peak power is expressed as the minimum charging power: Among them, P t,max , P t,min respectively represent the maximum and minimum values of the charge and discharge power at time t; For the current time instant t k , the power prediction problem is defined as an optimization problem of finding the sustainable peak power within a future time window L, and the optimization objective is expressed as Equation (11): Among them, P i represents the power value of the i-th sampling point; When the time window is in seconds and minutes, the current within the L time window is considered to be constant. During the discharge process, the battery terminal voltage continuously decreases within the window interval, and the minimum power appears at the end moment; during the charging process, the battery terminal voltage continuously rises within the window interval, and the maximum power within the window interval appears at the initial moment; The time window is divided into n phases, and the number of sampling points in each phase is n1, …, n i , 1 ≤ n i ≤ L, and take The time of each phase is expressed as The current within each phase is respectively expressed as Adopt a segmented optimization strategy to optimize the power of each stage. The battery peak power optimization target is transformed into formula (12): Among them, the optimization variable is the current at each stage. Within each stage, the minimum discharge power appears at the end moment, and it is considered that P t is the same as ; the maximum charging power appears at the initial moment, and it is considered that P t is the same as ; P t represents the power at time t.

5. The method for predicting the long-time window power state of a lithium-ion battery according to claim 4, wherein When implementing peak power estimation, simultaneously consider the following multiple constraints: Among them, V b,min and V b,max are the upper and lower design limits of the terminal voltage V b I min and I max are the upper and lower design limits of the design current I, SOC min and SOC max are the upper and lower design limits of the SOC design, P min and P max are the upper and lower design limits of the sustainable peak P of the battery; Combined with the equivalent circuit model and multi-parameter constraints, the SOP estimation problem is transformed into a multi-constraint non-linear optimization problem within the time window L. The discharge SOP estimation problem is expressed as the multi-constraint non-linear optimization problem P(1), and the charge SOP estimation problem is expressed as the multi-constraint non-linear optimization problem P(2): Discharge process: Charge process: Among them, for the discharging process, the current constraint is the maximum allowable discharging current The terminal voltage constraint is the discharging cut-off voltage and the charging cut-off voltage The SOC constraint is the discharging cut-off SOC, expressed as and not greater than 1, and the peak power constraint is the maximum allowable discharging power For the charging process, the current constraint is the maximum allowable discharging current The terminal voltage constraint is the discharging cut-off voltage and the charging cut-off voltage The SOC constraint is the charging cut-off SOC, expressed as and greater than 0, and the peak power constraint is the maximum allowable discharging power According to the detected voltage and current, a multi-stage optimization algorithm is adopted to segment within the time interval [k, k + L] and solve the multi-constrained non-linear optimization problems P(1) and P(2), so as to obtain the discharge SOP and charge SOP at any moment k.

6. A long-time window power state prediction system for a lithium-ion battery, characterized in that, Including: Module M1: Testing the charge and discharge characteristics of lithium-ion batteries; Module M2: Constructing an equivalent circuit model of lithium-ion batteries; Module M3: For long-time window power prediction, dividing the time window into N stages; Module M4: Constructing a long-time window multi-constrained, multi-stage non-linear optimization model; Module M5: Estimating the charge / discharge SOP based on the detected voltage and current.

7. The long-time window power state prediction system for a lithium-ion battery according to claim 6, wherein Perform an open-circuit voltage experiment on the lithium-ion battery to establish an open-circuit voltage model. First, charge the lithium-ion battery to the cut-off voltage in a constant current and constant voltage manner and let it stand for a period of time; then discharge the battery continuously at a 1C rate to a specific SOC and let it stand for 4 hours; among them, the discharge current is defined as a positive value and the charge current is defined as a negative value; during the whole process, the battery terminal voltage and load current are synchronously collected at a sampling frequency of 1Hz; according to the SOC of each standing point and its corresponding measured open-circuit voltage value, establish the functional relationship between the open-circuit voltage and SOC. Use a 16th-order polynomial form to represent the functional relationship between the open-circuit voltage VOC and SOC: where n represents the highest degree of the polynomial, VOC k , SOC k respectively represent the open-circuit voltage VOC and the state of charge SOC at the k-th sampling point, and β 1j represents the coefficient of the j-th term; The required SOC is calculated by the current integration method: In the formula, SOC(0) is the initial SOC value of the battery, C is the battery capacity, and I is the load current; Combining relational expressions (1) and (2), the least squares method is used to identify the parameter of β 1j to determine the functional relationship between the open circuit voltage and the SOC.

8. The long-time window power state prediction system for a lithium-ion battery according to claim 7, characterized in that First charge the lithium-ion battery to the fully charged state, then discharge it to a specific SOC at a 1C rate, then charge and discharge it under a simplified DST working condition, and finally perform pulse charge and discharge at rates of 0.125C, 0.75C, 1.5C, 2C, -0.125C, -0.75C, -1.5C, -2C in turn. Each pulse current lasts for 1 minute, and repeat the above steps continuously until the discharge cut-off voltage is reached. The equivalent circuit model includes three parts: the open-circuit voltage VOC, the ohmic internal resistance R0, and the first-order or multi-order RC network, where the RC network is composed of the polarization resistance R1 and the equivalent capacitance C1; Adopt a first-order equivalent circuit model, where V b is the battery voltage and I is the load current. This equivalent circuit model conforms to the following voltage-current relationship: V b = V OC - V1 - IR0 (4) Among them, V1 represents the polarization voltage; For a sampling period Δt, the discretized forms of the relational expressions (3) and (4) are expressed as: V b,k = V OC (SOC k ) - V 1,k - I k R 0,k (6) Among them, V 1,k+1 represents the polarization voltage at the (k + 1)-th moment; τ 1,k represents the time constant at the k-th moment; I k represents the current at the k-th moment; R 1,k represents the polarization resistance at the k-th moment; V b,k represents the terminal voltage of the battery at the k-th moment; SOC k represents the state of charge SOC at the k-th moment; R 0,k represents the ohmic internal resistance at the k-th moment; Time constant τ k = R 1，k C k ; R 1，k represents the polarization resistance at time k; C k represents the polarization capacitance at time k; In the formulas (5) and (6), the circuit parameters R0, R1, and C1 are functions of SOC and are expressed in polynomial form: Among them, the required identification parameter is the polynomial coefficient β 2j , β 3j , β 4j represents the coefficient of the j-th term. During the identification process, based on Equations (5) and (6), the least squares method is used to fit the voltage response curve to obtain β 2j , β 3j , β 4j .

9. The lithium-ion battery long-time window power state prediction system according to claim 8, wherein According to the definition of the maximum charge and discharge power, during the charge and discharge process, the peak power P at any time t t,peak is defined as Equation (10). Among them, since the charging current is defined as a negative value, the charging peak power is expressed as the minimum charging power: Among them, P t,max and P t,min represent the maximum and minimum values of the charge and discharge power at time t, respectively; For the current time instant t k , the power prediction problem is defined as an optimization problem of finding the sustainable peak power within a future time window L, and the optimization objective is expressed as Equation (11): Among them, P i represents the power value of the i-th sampling point; When the time window is in seconds and minutes, the current within the L time window is considered to be unchanged. During the discharge process, the battery terminal voltage continuously decreases within the window interval, and the minimum power appears at the end moment; during the charging process, the battery terminal voltage continuously rises within the window interval, and the maximum power within the window interval appears at the initial moment. The time window is divided into n phases, and the number of sampling points in each phase is n1, …, n i , 1 ≤ n i ≤ L, take The time of each phase is expressed as The current within each phase is respectively expressed as Adopt a segmented optimization strategy to optimize the power of each stage, and the battery peak power optimization target is transformed into formula (12): Among them, the optimized variable is the current at each stage. Within each stage, the minimum discharge power appears at the end moment, and it is considered that P t is the same as ; the maximum charging power appears at the initial moment, and it is considered that P t is the same as ; P t represents the power at time t.

10. The lithium-ion battery long-time window power state prediction system according to claim 9, wherein When implementing peak power estimation, simultaneously consider the following multiple constraints: Among them, V b,min and V b,max are the upper and lower limits of the terminal voltage V b Design, I min and I max are the upper and lower limits of the design current I, SOC min and SOC max are the upper and lower limits of the SOC design, P min and P max are the upper and lower limits of the sustainable peak P of the battery; Combined with the equivalent circuit model and multi-parameter constraints, the SOP estimation problem is transformed into a multi-constrained non-linear optimization problem within the time window L. The discharge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(1), and the charge SOP estimation problem is expressed as the multi-constrained non-linear optimization problem P(2): Discharge process: Charge process: Among them, for the discharging process, the current constraint is the maximum allowable discharging current The terminal voltage constraint is the discharging cut-off voltage and the charging cut-off voltage The SOC constraint is the discharging cut-off SOC, expressed as and not greater than 1. The peak power constraint is the maximum allowable discharging power For the charging process, the current constraint is the maximum allowable discharging current The terminal voltage constraint is the discharging cut-off voltage and the charging cut-off voltage The SOC constraint is the charging cut-off SOC, expressed as and greater than 0. The peak power constraint is the maximum allowable discharging power According to the detected voltage and current, using a multi-stage optimization algorithm, segment within the time interval [k, k + L] and solve the multi-constrained non-linear optimization problems P(1) and P(2) to obtain the discharge SOP and charge SOP at any k moment.

Images