CN106324523B - Lithium battery SOC estimation method based on discrete variable structure observer - Google Patents

Abstract

The invention discloses a method for estimating the SOC of a lithium battery based on a discrete variable structure observer. It includes the following steps: perform a rapid calibration experiment on lithium batteries to obtain the relationship between SOC and open circuit voltage OCV; establish a lithium battery discrete state space model for SOC estimation; perform pulse discharge experiments on lithium batteries to identify lithium battery model parameters; real-time Collect the terminal voltage and charge-discharge current of the lithium battery under working conditions; build a discrete variable structure observer to accurately estimate the SOC of the lithium battery. The method of the invention not only has a good SOC estimation effect, but also can strictly guarantee the convergence, and shows strong robustness to the modeling error of the lithium battery, the perturbation of internal parameters and the external perturbation.

Description

Lithium battery SOC estimation method based on discrete variable structure observer

technical field

The invention belongs to the field of estimation of the state of charge of a lithium battery for vehicles, and in particular relates to a method for estimating the SOC of a lithium battery based on a discrete variable structure observer.

Background technique

The battery management system (BMS) is an important part of electric vehicles, with basic functions such as battery state detection, battery state estimation, battery safety protection, and energy control management.

Battery SOC estimation is the core of the battery management system. SOC is an important parameter to characterize the remaining capacity of the battery. The accurate SOC value is an important basis for battery charge and discharge control, balance control, and energy management strategy formulation. The essential.

SOC is affected by various factors such as battery temperature, charge-discharge rate, self-discharge rate, and lifespan. It cannot be directly measured by sensors. The battery must be modeled and combined with the measured charge-discharge current and terminal voltage when the battery is working. , temperature and other data, select the algorithm to estimate indirectly. During the use of electric vehicle batteries, due to the complex internal electrochemical reactions, the battery characteristics are highly nonlinear, which makes it difficult to accurately estimate the battery SOC.

Compared with traditional electric vehicle batteries, lithium batteries have many advantages in performance, such as high energy density, no memory effect, low environmental pollution, long cycle life, and wide temperature range, so lithium batteries have developed into the most competitive power. Battery.

At present, the commonly used lithium battery SOC estimation algorithms include open circuit voltage method, ampere-hour integration method, Kalman filter algorithm, neural network algorithm, etc.

Chinese invention patent CN 103529398 A published on January 22, 2014 "Online Estimation Method of Lithium-ion Battery SOC Based on Extended Kalman Filtering", which firstly establishes the voltage-current relationship of the first-order RC equivalent circuit of the tested lithium-ion battery and the voltage-current relationship of the second-order RC equivalent circuit; then conduct the charge-discharge experiment of the lithium-ion battery under test, and establish a polynomial fitting function of the initial value SOC ₀ of the Kalman filter of the lithium-ion battery under test; and then obtain the lithium-ion battery under test. The Kalman filter initial value SOC ₀ of the ion battery and the initial error covariance P(0) of the Kalman filter; then the battery SOC estimation based on the extended Kalman filter is performed to realize the online estimation of the SOC of the lithium ion battery. But this method has shortcomings:

1) The algorithm requires that the noise is white noise and the statistical characteristics such as the mean and variance of the noise are known, which is difficult to satisfy in practical applications, not only because the statistical characteristics of the noise are difficult to obtain, but also the white noise is only under ideal conditions. exist;

2) The algorithm has high requirements on the accuracy of the lithium-ion battery performance model. When the model accuracy is low, it will cause a large SOC estimation error;

3) After the nonlinear lithium-ion battery model is linearized, if the deviation is large, the phenomenon of filter divergence will appear, resulting in a large SOC estimation error.

Chinese invention patent CN 105548898 A, published on May 4, 2016, "A Lithium Battery SOC Estimation Method for Offline Data Segmentation Correction" firstly establishes a battery equivalent circuit model; obtains the SOC-OCV curve; The terminal voltage response curve of the battery is used for offline parameter identification of the equivalent circuit model; then the battery state of health SOH is calculated; the current value of SOC is calculated in real time by the ampere-hour integration method; and the SOC value is corrected by the battery state of health; The accumulated error in the ampere-hour integration method is eliminated in stages. Its shortcomings are:

1) The offline identification method is used to identify the parameters of the equivalent circuit model of the lithium battery. During the working process of the lithium battery, due to the influence of temperature, aging, life and other factors, the parameters in the lithium battery model will change, which will cause greater damage. SOC estimation error.

SUMMARY OF THE INVENTION

In order to overcome the above-mentioned shortcomings of the prior art, the present invention provides a method for estimating the SOC of a lithium battery based on a discrete variable structure observer. This method not only has a good SOC estimation effect, but also has strong robustness to lithium battery modeling errors and changes in internal parameters caused by factors such as temperature, aging and lifespan. Estimated divergence occurs.

In order to solve the problems existing in the prior art, the present invention provides a method for estimating the SOC of a lithium battery based on a discrete variable structure observer, including the collection of the terminal voltage and charging and discharging current of the lithium battery under working conditions, and the main steps are as follows:

Step 1, perform a rapid calibration experiment on the lithium battery, and obtain the relationship curve between SOC and open circuit voltage OCV;

Step 1.1, at room temperature, discharge the lithium battery with a charge cut-off voltage of 4.2V, a discharge cut-off voltage of 3V, and a rated capacity of 5Ah at a constant current of 0.2 coulomb until the lithium battery voltage is below 3V, and let it stand for 2 to 3 hours. Wait experimental use;

Step 1.2, use 0.2 coulomb current to charge the lithium battery after standing according to step 1.1 with constant current pulse, after each charge of 10% of the rated capacity of the lithium battery, disconnect the lithium battery and let it stand for 5 minutes, and measure the charging process in real time. The terminal voltage U _{c, OCV} of the lithium battery in each resting time period, and find out the minimum value U _{c, OCV, min} of the terminal voltage of the lithium battery in each resting time period during the charging process, until the lithium battery is fully charged;

Step 1.3, use 0.2 coulomb current to discharge the lithium battery fully charged according to step 1.2 with constant current pulse. After each discharge of 10% of the rated capacity of the lithium battery, disconnect the lithium battery and let it stand for 5 minutes. The terminal voltage U _{d, OCV} of the lithium battery in each resting time period, and find out the maximum value U _{d, OCV, max} of the terminal voltage of the lithium battery in each resting time period during the discharge process, until the lithium battery is empty;

Step 1.4, compare the minimum value U _{c, OCV, min} of the terminal voltage of the lithium battery in each resting time period during the charging process obtained in step 1.2 with the resting time corresponding to the SOC during the discharging process and the charging process obtained in step 1.3. The maximum value U _{d, OCV and max} of the terminal voltage of the lithium battery in the segment are added and averaged, as the open-circuit voltage OCV of the rapid calibration, and a total of 10 open-circuit voltage OCVs are obtained;

Step 1.5, according to the 10 open circuit voltage OCVs obtained in step 1.4, within the entire SOC variation range, that is, within the range of 0% to 100%, perform multi-segment linear fitting on the obtained experimental data, and obtain the lithium battery SOC and open circuit voltage. OCV relationship curve;

In the multi-segment straight line fitting, the length of each segment is ΔSOC=10%, and the expressions of the fitted SOC and open circuit voltage OCV in each segment are:

OCV _i = _ki *SOC _i +d _i i = 1, 2, 3.... 10 (1)

Among them, OCV _i is the open circuit voltage OCV of the i-th lithium battery, SOC _i is the SOC of the i-th lithium battery, ki is the slope of the straight line between the SOC and the open-circuit voltage OCV fitted by the _i -th segment, and d _i is the i-th The intercept of the SOC fitted by the segment and the open circuit voltage OCV straight line;

Step 2: According to the relationship between the lithium battery SOC and the open circuit voltage OCV obtained in step 1, and combined with the lithium battery Thevenin equivalent circuit and the ampere-hour integral formula, establish a lithium battery discrete state space model for SOC estimation;

Discrete equation of state:

Discrete observation equation:

Among them, V _t (k) is the terminal voltage of the lithium battery at time k, SOC(k) is the SOC of the lithium battery at time k, V ₁ (k) is the polarization voltage of the lithium battery at time k, and V _t (k+1) is the terminal voltage of the lithium battery (k+1), SOC(k+1) is the SOC of the lithium battery (k+1), and V ₁ (k+1) is the polarization of the lithium battery (k+1) voltage, T is the sampling time, I _s (k) is the current value flowing through the lithium battery at time k, γ is the disturbance input matrix, ξ(k) is the bounded scalar disturbance input, and y(k) is the lithium battery at time k. Output, a ₁ =1/R ₁ C ₁ , a ₁₁ = _ki a ₁ , a ₂ =1/R _{0 CN} , a ₂₂ = _ki a ₂ , b ₁ = _ki / _CN +1/ C ₁ +R ₀ /R ₁ C ₁ , b ₂ =1/C ₁ , R ₁ is the polarization resistance of the lithium battery, C ₁ is the polarization capacitance of the lithium battery, R ₀ is the ohmic internal resistance of the lithium battery, C _N is the nominal capacity of the lithium battery;

Step 3, perform a pulse discharge experiment on the lithium battery, and identify the parameters of the discrete state space model of the lithium battery in step 2;

Step 3.1, first discharge the lithium battery with a capacity of 5Ah at the current I for 5 minutes, then stop the discharge and let it stand for 10 minutes, then discharge it at the same current I for 5 minutes, use this 20 minutes as a pulse discharge cycle, charge and discharge The device records the change of the terminal voltage U _battery in a pulse discharge cycle of the lithium battery;

Step 3.2, according to the change of the terminal voltage U _battery in one pulse discharge cycle of the lithium battery recorded in step 3.1, analyze the characteristics of the terminal voltage change curve, and identify the parameters in the lithium battery model, the parameters include ohmic internal resistance R ₀ , polarization resistance R ₁ , polarization capacitance C ₁ ;

Step 4: Collect the terminal voltage V _t (k) and the charging and discharging current Is ( _k ) of the lithium battery in real time through the voltage sensor and the current sensor, respectively;

Step 5: Design a discrete variable structure observer based on the discrete state space model of the lithium battery established in step 2, and input the lithium battery terminal voltage V _t (k) and charge-discharge current Is ( _k ) collected in step 4 as signal input , real-time estimation of lithium battery SOC;

The equation of the discrete variable structure observer is as follows:

Among them, x(k) is the state variable of the lithium battery at time k, is the estimated value of x(k), x(k+1) is the state variable at the moment of lithium battery (k+1), is the estimated value of x(k+1), is the estimated value of y(k), λ is the positive feedback input matrix, v(k) is the external positive feedback compensation signal, h is the gain matrix of the discrete variable structure observer, G is the system matrix of the discrete variable structure observer, H is the input matrix of the discrete variable structure observer, C is the output matrix of the discrete variable structure observer, C=[1 0 0], and set Cλ≠0.

Preferably, the Thevenin equivalent circuit model in step 2 is:

V _t =V ₁ +I _s R ₀ +OCV (6)

Among them, V _t is the terminal voltage of the lithium battery, V ₁ is the polarization voltage of the lithium battery, is the differential of V ₁ , _Is is the instantaneous current flowing through the lithium battery, and OCV is the open circuit voltage of the lithium battery.

Preferably, the ampere-hour integral formula in step 2 is:

Among them, SOC ₀ is the initial value of the SOC of the lithium battery, SOC _t is the instantaneous value of the SOC of the lithium battery, _CN is the nominal capacity of the lithium battery, Is is the instantaneous current flowing through the lithium battery, and _{t 0} is the initial moment of the charging and discharging process , t ₁ is the termination time of the charging and discharging process.

Preferably, the bounded scalar perturbation input ξ(k) in step 2 satisfies the following expression:

|ξ(k)|≤ξ ₀ ||e(k)|| (9)

Among them, ξ ₀ is the upper limit parameter, which is a positive number, taking 0.05 to 0.15, and e(k) is the state error of the lithium battery at time k, that is, e(k)=x(k+1)-x(k).

Preferably, the voltage sensor and the current sensor in step 4 are respectively a Hall voltage sensor and a Hall current sensor.

Preferably, the gain matrix h of the discrete variable structure observer in step 5 is determined according to the following formula:

Among them, C ^T is the transpose matrix of the output matrix C of the discrete variable structure observer, and P is the discrete Quincatti equation The positive definite ^symmetric matrix solution of the The identity matrix of 3, α is a positive real number, and its value ranges from 1 to 5.

Preferably, the positive feedback input matrix λ in step 5 is determined according to the following formula:

λ=vH (11)

Among them, v is a positive real number, take v=0.2～0.8.

Preferably, the external positive feedback compensation signal v(k) in step 5 is determined according to the following formula:

Among them, e(k) is the state error of the lithium battery at time k, that is, e(k)=x(k+1)-x(k), W(k) is a positive function and satisfies |W(k)|< 1, v _eq (k) is the equivalent external positive feedback compensation signal, which is determined by the formula v _eq (k)=(Cλ) ^-1 (CGe(k)+Cγξ(k)), in which there is a boundary scalar disturbance input ξ (k) satisfies |ξ(k)|≤ξ ₀ ×||e(k)||, ξ ₀ is the upper limit parameter, which is a positive number and takes 0.05 to 0.15.

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

1. The accuracy of the lithium battery performance model is not high, and the modeling error can be effectively compensated.

2. There is no discrete variable structure observer divergence problem caused by improper linearization of the lithium battery model in the extended Kalman filter algorithm, which can strictly ensure the convergence of the algorithm.

3. When using the offline identification method to identify the parameters of the equivalent circuit model of the lithium battery, when the internal parameters of the lithium battery change due to factors such as temperature, aging, and life, the SOC can still be accurately estimated, showing strong robustness.

Description of drawings

FIG. 1 is a schematic flowchart of a method for estimating the SOC of a lithium battery according to the present invention.

FIG. 2 is a schematic diagram of the relationship between the SOC of the lithium battery and the open circuit voltage OCV.

Figure 3 is the Thevenin equivalent circuit of a lithium battery.

Fig. 4 is the change curve of the terminal voltage of the lithium battery in the pulse discharge experiment.

Detailed ways

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments, but the embodiments of the present invention are not limited thereto.

FIG. 1 is a schematic flowchart of a method for estimating SOC of a lithium battery according to the present invention. It can be seen from the figure that a method for estimating SOC of a lithium battery based on a discrete variable structure observer provided by the present invention includes the following steps:

Step 1, perform a rapid calibration experiment on the lithium battery, and obtain the relationship curve between SOC and open circuit voltage OCV;

Step 1.1, at room temperature, discharge the lithium battery with a charge cut-off voltage of 4.2V, a discharge cut-off voltage of 3V, and a rated capacity of 5Ah at a constant current of 0.2 coulomb until the lithium battery voltage is below 3V, and stand for 2 to 3 hours. Wait for the experiment use;

Step 1.2, use 0.2 coulomb current to charge the lithium battery after standing according to step 1.1 with constant current pulse, after each charge of 10% of the rated capacity of the lithium battery, disconnect the lithium battery and let it stand for 5 minutes, and measure the charging process in real time. The terminal voltage U _{c, OCV} of the lithium battery in each resting time period, and find out the minimum value U _{c, OCV, min} of the terminal voltage of the lithium battery in each resting time period during the charging process, until the lithium battery is fully charged;

Step 1.3, use 0.2 coulomb current to discharge the lithium battery fully charged according to step 1.2 with constant current pulse. After each discharge of 10% of the rated capacity of the lithium battery, disconnect the lithium battery and let it stand for 5 minutes. The terminal voltage U _{d, OCV} of the lithium battery in each resting time period, and find out the maximum value U _{d, OCV, max} of the terminal voltage of the lithium battery in each resting time period during the discharge process, until the lithium battery is empty;

Step 1.4, compare the minimum value U _{c, OCV, min} of the terminal voltage of the lithium battery in each resting time period during the charging process obtained in step 1.2 with the resting time corresponding to the SOC during the discharging process and the charging process obtained in step 1.3. The maximum value U _{d, OCV and max} of the terminal voltage of the lithium battery in the segment are added and averaged, as the open-circuit voltage OCV of the rapid calibration, and a total of 10 open-circuit voltage OCVs are obtained;

Step 1.5, according to the 10 open circuit voltage OCVs obtained in step 1.4, within the entire SOC variation range, that is, within the range of 0% to 100%, perform multi-segment linear fitting on the obtained experimental data, and obtain the lithium battery SOC and open circuit voltage. OCV relationship curve;

In the multi-segment straight line fitting, the length of each segment is ΔSOC=10%, and the expressions of the fitted SOC and open circuit voltage OCV in each segment are:

OCV _i = _ki *SOC _i +d _i i = 1, 2, 3.... 10 (1)

Among them, OCV _i is the open circuit voltage OCV of the i-th lithium battery, SOC _i is the SOC of the i-th lithium battery, ki is the slope of the straight line between the SOC and the open-circuit voltage OCV fitted by the _i -th segment, and d _i is the i-th The intercept of the SOC fitted by the segment and the open circuit voltage OCV line.

In this embodiment, FIG. 2 shows the relationship curve between the SOC of the lithium battery and the open circuit voltage OCV obtained by fitting, and the specific parameters thereof can be seen in Table 1.

Table 1: Lithium battery SOC and open circuit voltage OCV curve parameter table

i 1 2 3 4 5 SOC_i 0%-10% 10%-20% 20%-30% 30%-40% 40%-50% k_i 6.933 0.400 0.178 0.200 0.011 d_i 3.0100 3.6633 3.7077 3.7011 3.7767 i 6 7 8 9 10 SOC_i 50%-60% 60%-70% 70%-80% 80%-90% 90%-100% k_i 0.025 0.025 0.011 0.025 3.092 d_i 3.7697 3.7697 3.7795 3.7683 1.008

Step 2, according to the relationship between the lithium battery SOC and the open circuit voltage OCV obtained in step 1, and combining the lithium battery Thevenin equivalent circuit and the ampere-hour integral formula, a lithium battery discrete state space model for SOC estimation is established. The process is as follows.

Step 2.1, establish a first-order equivalent circuit model based on the Thevenin equivalent circuit of the lithium battery;

Figure 3 is a Thevenin equivalent circuit diagram of a lithium battery, in which OCV is the open circuit voltage of the lithium battery, R ₀ is the ohmic internal resistance of the lithium battery, R ₁ is the polarization resistance of the lithium battery, and C ₁ is the polarization capacitance of the lithium battery.

V _t =V ₁ +I _s R ₀ +OCV (2)

Among them, V _t is the terminal voltage of the lithium battery, V ₁ is the polarization voltage of the lithium battery, Is the differential of V ₁ , I _s is the instantaneous current flowing through the lithium battery.

Step 2.2, according to the first-order equivalent circuit model of the lithium battery established in step 2.1, combined with the ampere-hour integral formula, establish a continuous state space model of the lithium battery;

The ampere-hour integral formula is:

Among them, SOC ₀ is the initial value of the SOC of the lithium battery, SOC _t is the instantaneous value of the SOC of the lithium battery, _CN is the nominal capacity of the lithium battery, Is is the instantaneous current flowing through the lithium battery, and _{t 0} is the initial moment of the charging and discharging process , t ₁ is the termination time of the charging and discharging process.

Simultaneously formula (1), (2), (4), get:

in, is the derivative of the lithium battery SOC.

Combining formulas (1), (2), (3), we get:

in, It is the differential of the terminal voltage V _t of the lithium battery.

The state variable of the lithium battery is selected as The input quantity is the current I, and the output quantity is the terminal voltage V _t , set d _i =0, and establish the continuous state space model of the lithium battery according to formulas (3), (5) and (6):

Continuous state equation:

Continuous observation equation:

in, is the differential of V _t , is the differential of SOC, is the differential of V ₁ , y is the output of the lithium battery, that is, the terminal voltage V _t , a ₁ =1/R ₁ C ₁ , a ₁₁ = _ki a ₁ , a ₂ =1/R _{0 CN} , a ₂₂ = _ki a ₂ , b ₁ = _ki /C _N +1/C ₁ +R ₀ /R ₁ C ₁ , b ₂ =1/C ₁ .

Step 2.3, discretizing the continuous state space model of the lithium battery established in step 2.2 to obtain a discrete state space model of the lithium battery;

Discrete equation of state:

Discrete observation equation:

Among them, V _t (k) is the terminal voltage of the lithium battery at time k, SOC(k) is the SOC of the lithium battery at time k, V ₁ (k) is the polarization voltage of the lithium battery at time k, and V _t (k+1) is the terminal voltage of the lithium battery (k+1), SOC(k+1) is the SOC of the lithium battery (k+1), and V ₁ (k+1) is the polarization of the lithium battery (k+1) Voltage, y(k) is the output of the lithium battery at time k, Is (k) is the current value flowing through the lithium battery at time _k , T is the sampling time, in this example the value T=0.001s.

Step 2.4, adding an uncertainty term γξ(k) to the discrete state space model of the lithium battery established in step 2.3 to compensate for nonlinearity, external disturbances and modeling errors;

The modified discrete equation of state:

Discrete observation equation:

Among them, let the state at time k γ is the perturbation input matrix, in this example, take ξ(k) is the bounded scalar disturbance input, and satisfy |ξ(k)|≤ξ ₀ ||e(k)||, e(k) is the state error of the lithium battery at time k, that is, e(k)= x(k+1)-x(k), ξ ₀ is an upper limit parameter, and in this embodiment, the value ξ ₀ =0.1.

In step 3, a pulse discharge experiment is performed on the lithium battery, and the parameters of the lithium battery model in step 2 are identified. The parameters include ohmic internal resistance R ₀ , polarization resistance R ₁ , and polarization capacitance C ₁ , and the process is:

Step 3.1, first discharge the lithium battery with a capacity of 5Ah at a current of 1C for 5 minutes, then stop the discharge and let it stand for 10 minutes, and then discharge it at a current of 1C for 5 minutes. This 20 minutes is regarded as a pulse discharge cycle, and the charging and discharging equipment records The change of the terminal voltage U _battery in a pulse discharge cycle of the lithium battery;

Step 3.2, according to the change of the terminal voltage U _battery in one pulse discharge cycle of the lithium battery recorded in step 3.1, analyze the characteristics of the terminal voltage change curve, and identify the parameters of the lithium battery model in step 2.

Figure 4 is a schematic diagram of the change curve of the terminal voltage U _battery of the lithium battery obtained in step 3.1 in one pulse discharge cycle, (U ₂ -U ₁ ) is the voltage generated on the ohmic internal resistance R ₀ at the moment t=300s at the end of the pulse discharge drop; (U ₃ -U ₂ ) is the voltage at both ends of the R ₁ and C ₁ loops during the resting stage t=300s～900s; in the resting stage, the lithium battery terminal voltage U _battery rises to (U ₃ - 95% of U ₂ ), and the time 3τ is measured, where the time constant τ=C ₁ R ₁ , then the ohmic internal resistance R ₀ , the polarization resistance R ₁ and the polarization capacitance C ₁ of the lithium battery are calculated according to the following formulas:

In step 4, the terminal voltage V _t (k) and the charging and discharging current Is ( _k ) of the lithium battery under the working conditions are collected in real time by using the Hall voltage sensor and the Hall current sensor respectively.

Step 5: Design a discrete variable structure observer based on the discrete state space model of the lithium battery established in step 2, and input the lithium battery terminal voltage V _t (k) and charge-discharge current Is ( _k ) collected in step 4 as signal input , real-time estimation of lithium battery SOC, the process is as follows:

Step 5.1, design a discrete variable structure observer based on the discrete state space model of the lithium battery established in step 2;

in, is the estimated value of x(k), is the estimated value of x(k+1), is the estimated value of y(k), h is the gain matrix of the discrete variable structure observer, λ is the positive feedback input matrix, v(k) is the external positive feedback compensation signal, and G is the system matrix of the discrete variable structure energy observer , H is the input matrix of the discrete variable structure observer, C is the output matrix of the discrete variable structure observer, C=[1 0 0].

In step 5.2, the lithium battery terminal voltage V _t (k) and the charge and discharge current I (k) collected in step 4 are input as signals, and the designed discrete variable structure observer is used to estimate the lithium battery SOC in real time.

In this embodiment, the gain matrix h of the discrete variable structure observer is determined according to the following formula:

Among them, C ^T is the transpose matrix of the output matrix C of the discrete variable structure observer, and P is the discrete Quincatti equation The positive definite ^symmetric matrix solution of the The identity matrix of 3, α is a positive real number, the value α=1 in this case.

In this embodiment, the positive feedback input matrix λ is determined according to the following formula:

λ=vH (19)

Among them, v is a positive real number, and the value of this case is v=0.5.

In this embodiment, the external positive feedback compensation signal v(k) is determined according to the following formula:

Among them, e(k) is the state error of the lithium battery at time k, that is, e(k)=x(k+1)-x(k), W(k) is a positive function and satisfies |W(k)|< 1. V _eq (k) is the equivalent external positive feedback compensation signal, which is determined by the formula v _eq (k)=(Cλ) ^-1 (CGe(k)+Cγξ(k)), and there is a boundary scalar disturbance input in the formula ξ(k) satisfies |ξ(k)|≤0.1×||e(k)||.

Claims (9)

1. a method for estimating the SOC of a lithium battery based on a discrete variable structure observer, comprising the collection of the terminal voltage and the charging and discharging current of the lithium battery under working conditions, characterized in that the main steps are as follows: Step 1, perform a rapid calibration experiment on the lithium battery, and obtain the relationship curve between the SOC and the open circuit voltage OCV of the lithium battery; Step 1.1, at room temperature, discharge the lithium battery with a charge cut-off voltage of 4.2V, a discharge cut-off voltage of 3V, and a rated capacity of 5Ah at a constant current of 0.2 coulomb until the lithium battery voltage is below 3V, and let it stand for 2 to 3 hours. Wait experimental use; Step 1.2, use 0.2 coulomb current to charge the lithium battery after standing according to step 1.1 with constant current pulse, after each charge of 10% of the rated capacity of the lithium battery, disconnect the lithium battery and let it stand for 5 minutes, and measure the charging process in real time. The terminal voltage U _{c, OCV} of the lithium battery in each resting time period, and find out the minimum value U _{c, OCV, min} of the terminal voltage of the lithium battery in each resting time period during the charging process, until the lithium battery is fully charged; Step 1.3, use 0.2 coulomb current to discharge the lithium battery fully charged according to step 1.2 with constant current pulse. After each discharge of 10% of the rated capacity of the lithium battery, disconnect the lithium battery and let it stand for 5 minutes. The terminal voltage U _{d, OCV} of the lithium battery in each resting time period, and find out the maximum value U _{d, OCV, max} of the terminal voltage of the lithium battery in each resting time period during the discharge process, until the lithium battery is empty; Step 1.4, compare the minimum value U _{c, OCV, min} of the terminal voltage of the lithium battery in each resting time period during the charging process obtained in step 1.2 with the resting time corresponding to the SOC during the discharging process and the charging process obtained in step 1.3. The maximum value U _{d, OCV and max} of the terminal voltage of the lithium battery in the segment are added and averaged, as the open-circuit voltage OCV of the rapid calibration, and a total of 10 open-circuit voltage OCVs are obtained; Step 1.5, according to the 10 open-circuit voltage OCVs obtained in step 1.4, in the entire lithium battery SOC variation range, that is, within the range of 0% to 100%, perform multi-segment linear fitting on the obtained experimental data, and obtain the lithium battery SOC and Open circuit voltage OCV curve; In the multi-segment straight line fitting, the length of each segment is ΔSOC=10%, and the fitted lithium battery SOC and open circuit voltage OCV expressions in each segment are: OCV _i = _ki *SOC _i +d _i , i=1, 2, 3....10 (1) Among them, OCV _i is the open circuit voltage OCV of the i-th lithium battery, SOC _i is the SOC of the i-th lithium battery, ki is the slope of the straight line between the SOC and the open-circuit voltage OCV fitted by the _i -th segment, and d _i is the i-th The intercept of the SOC fitted by the segment and the open circuit voltage OCV straight line; Step 2, according to the relationship between the lithium battery SOC and the open circuit voltage OCV obtained in step 1, and combining the lithium battery Thevenin equivalent circuit and the ampere-hour integral formula, establish a lithium battery discrete state space model for lithium battery SOC estimation; Discrete equation of state: Discrete observation equation: Among them, V _t (k) is the terminal voltage of the lithium battery at time k, SOC(k) is the SOC of the lithium battery at time k, and V ₁ (k) is the polarization voltage of the lithium battery at time k; V _t (k+1) is the terminal voltage of the lithium battery (k+1), SOC(k+1) is the SOC of the lithium battery (k+1), and V ₁ (k+1) is the polarization of the lithium battery (k+1) Voltage; T is the sampling time, I _s (k) is the current value flowing through the lithium battery at time k, γ is the disturbance input matrix, ξ(k) is the bounded scalar disturbance input, and y(k) is the lithium battery at time k. Output, a ₁ =1/R ₁ C ₁ , a ₁₁ = _ki a ₁ , a ₂ =1/R _{0 CN} , a ₂₂ = _ki a ₂ , b ₁ = _ki / _CN +1/ C ₁ +R ₀ /R ₁ C ₁ , b ₂ =1/C ₁ , R ₁ is the polarization resistance of the lithium battery, C ₁ is the polarization capacitance of the lithium battery, R ₀ is the ohmic internal resistance of the lithium battery, C _N is the nominal capacity of the lithium battery; Step 3, perform a pulse discharge experiment on the lithium battery, and identify the parameters of the discrete state space model of the lithium battery in step 2; Step 3.1, first discharge the lithium battery with a capacity of 5Ah at the current I for 5 minutes, then stop the discharge and let it stand for 10 minutes, then discharge it at the same current I for 5 minutes, use this 20 minutes as a pulse discharge cycle, charge and discharge The device records the change of the terminal voltage U _battery in a pulse discharge cycle of the lithium battery; Step 3.2, according to the change of the terminal voltage U _battery in one pulse discharge cycle of the lithium battery recorded in step 3.1, analyze the characteristics of the terminal voltage change curve, and identify the parameters in the discrete state space model of the lithium battery, and the parameters include the ohmic internal resistance R ₀ , polarization resistance R ₁ , polarization capacitance C ₁ ; Step 4: Collect the terminal voltage V _t (k) and the charging and discharging current Is ( _k ) of the lithium battery in real time through the voltage sensor and the current sensor, respectively; Step 5: Design a discrete variable structure observer based on the discrete state space model of the lithium battery established in step 2, and input the lithium battery terminal voltage V _t (k) and charge-discharge current Is ( _k ) collected in step 4 as signal input , real-time estimation of lithium battery SOC; The equation of the discrete variable structure observer is as follows: Among them, x(k) is the state variable of the lithium battery at time k, is the estimated value of x(k), x(k+1) is the state variable at the moment of lithium battery (k+1), is the estimated value of x(k+1), is the estimated value of y(k), λ is the positive feedback input matrix, v(k) is the external positive feedback compensation signal, h is the gain matrix of the discrete variable structure observer, G is the system matrix of the discrete variable structure observer, H is the input matrix of the discrete variable structure observer, C is the output matrix of the discrete variable structure observer, C=[1 0 0], and set Cλ≠0. 2. The method for estimating the SOC of a lithium battery based on a discrete variable structure observer according to claim 1, wherein the Thevenin equivalent circuit model in step 2 is: V _t =V ₁ +I _s R ₀ +OCV (6) Among them, V _t is the terminal voltage of the lithium battery, V ₁ is the polarization voltage of the lithium battery, is the differential of V ₁ , _Is is the instantaneous current flowing through the lithium battery, and OCV is the open circuit voltage of the lithium battery. 3. a kind of lithium battery SOC estimation method based on discrete variable structure observer according to claim 1, is characterized in that, the ampere-hour integral formula in step 2 is: Among them, SOC ₀ is the initial value of the SOC of the lithium battery, SOC _t is the instantaneous value of the SOC of the lithium battery, _CN is the nominal capacity of the lithium battery, Is is the instantaneous current flowing through the lithium battery, and _{t 0} is the initial moment of the charging and discharging process , t ₁ is the termination time of the charging and discharging process. 4. The method for estimating lithium battery SOC based on discrete variable structure observer according to claim 1, wherein the bounded scalar disturbance input ξ(k) in step 2 satisfies the following expression: |ξ(k)|≤ξ ₀ ||e(k)|| (9) Among them, ξ ₀ is the upper limit parameter, which is a positive number, taking 0.05 to 0.15, and e(k) is the state error of the lithium battery at time k, that is, e(k)=x(k+1)-x(k). 5 . The method for estimating the SOC of a lithium battery based on a discrete variable structure observer according to claim 1 , wherein the voltage sensor and the current sensor in step 4 are respectively a Hall voltage sensor and a Hall current sensor. 6 . 6. A lithium battery SOC estimation method based on a discrete variable structure observer according to claim 1, wherein the gain matrix h of the discrete variable structure observer in step 5 is determined according to the following formula: Among them, C ^T is the transpose matrix of the output matrix C of the discrete variable structure observer, and P is the discrete Quincatti equation The positive definite symmetric matrix solution of , G ^T in the discrete Quincatti equation is the transpose matrix of the system matrix G of the discrete variable structure observer, Q is any positive semi-definite symmetric matrix, α is a positive real number, and its value ranges from 1 to 1. 5. 7. The method for estimating the SOC of a lithium battery based on a discrete variable structure observer according to claim 1, wherein the positive feedback input matrix λ in step 5 is determined according to the following formula: λ=vH (11) Among them, v is a positive real number, take v=0.2～0.8. 8. The method for estimating the SOC of a lithium battery based on a discrete variable structure observer according to claim 1, wherein the external positive feedback compensation signal v(k) in step 5 is determined according to the following formula: Among them, e(k) is the state error of the lithium battery at time k, that is, e(k)=x(k+1)-x(k), W(k) is a positive function and satisfies |W(k)|< 1, v _eq (k) is the equivalent external positive feedback compensation signal, which is determined by the formula v _eq (k)=(Cλ) ^-1 (CGe(k)+Cγξ(k)), in which there is a boundary scalar disturbance input ξ (k) satisfies |ξ(k)|≤ξ ₀ ×||e(k)||, ξ ₀ is the upper limit parameter, which is a positive number, and takes 0.05 to 0.15. 9 . The method for estimating the SOC of a lithium battery based on a discrete variable structure observer according to claim 6 , wherein the Q=E, and E is a 3×3 unit matrix. 10 .