CN105277895A - Series battery pack SOP (state of power) on-line estimation method and application thereof - Google Patents

Abstract

The invention discloses an online estimation method of power state SOP of a series battery pack and its application. The method comprises the following steps: recursive online identification of battery parameters and battery dynamic effects; calculation of battery SOP at the next moment based on voltage limitation; The battery SOP calculation based on the next time after the voltage limit; the battery SOP calculation at the next time based on the current limit; the battery SOP calculation based on the next time after the current limit; the next time based on the combination of voltage limit and current limit Online estimation of the battery SOP at the moment and the battery SOP at the moment after the next moment. The present invention considers the impact of the battery voltage and current working window on the peak power, can simultaneously realize high-precision SOP single-step prediction and multi-step prediction, can effectively prevent the battery from being abused during real-time operation, and help other related systems to achieve the best Optimized energy management.

Description

An Online Estimation Method for Power State SOP of Series Battery Pack and Its Application

technical field

The invention relates to the technical field of battery management systems, in particular to an online estimation method and application of the power state SOP of series battery packs.

Background technique

SOP is a parameter describing the maximum charge and discharge capacity of the battery. It is used to determine the maximum input to the battery and the maximum output power of the load, to avoid abuse of the battery, and to determine the acceleration and climbing performance and regenerative braking capabilities of electric vehicles.

Battery SOP prediction is a new field in battery management system research. There are mainly methods for battery SOP estimation at home and abroad: the United States Next Generation Automobile Alliance (PNGV) proposed the use of pulse discharge (HybridPulsePowerCharacteristics, HPPC) method to estimate the maximum charge and discharge capacity of the battery; based on the dynamic electrochemical model of the battery to accurately predict The method of the SOP of the battery at the next sampling point; the method of estimating the SOP of the battery by using the extended Kalman filter, etc. Most of the existing methods only consider the limitation of the battery voltage threshold on the battery SOP, but ignore the limitation of the current threshold on the battery peak power; at the same time, the same method is used to predict the single-step SOP and multi-step SOP of the battery. Real-time sampling values of battery voltage and current can be reasonably used, thus limiting the prediction accuracy of battery single-step SOP.

Contents of the invention

Aiming at the defects of the prior art, the present invention provides an online estimation method of the power state SOP of a series battery pack and its application.

The object of the present invention is achieved through the following technical solutions: an online estimation method of the power state SOP of a series battery pack comprising the steps of:

Step 1. Perform recursive online identification of battery parameters based on battery parameters in the battery equivalent circuit model and battery parameters that are not included in the battery equivalent circuit model for comprehensive simulation of battery effects;

Step 2. Execute the battery SOP calculation at the next moment based on the voltage limit and the battery parameters identified online;

Step 3. Execute the battery SOP calculation based on the voltage limit and the battery parameters identified online at the time after the next time;

Step 4, performing battery SOP calculation based on the current limit and the battery parameters at the next moment identified and calculated;

Step 5. Execute the battery SOP calculation based on the current limit and the battery parameters identified online at the time after the next time;

Step 6, integrating the battery SOP at the next moment calculated in steps 2-5 and the battery SOP at a later moment to realize online estimation of the battery SOP based on voltage limitation and current limitation synthesis.

The battery equivalent circuit model in the step 1 is a Thevenin model, and the battery parameters in the battery equivalent circuit model include the open circuit voltage V _oc of the battery, the DC internal resistance R _in of the battery, and the charge transfer phenomenon used to simulate the battery The resistance R _p and capacitance C _p in the RC loop, at time k, the battery effect is not covered in the battery equivalent circuit model, and the battery parameters for comprehensive simulation are the sliding average value of white noise in the battery equivalent circuit model The colored noise w _k added at the output of .

The recursive online identification method in the step 1 is an online identification method based on the recursive extended least squares method, which specifically includes the following steps:

Step 101. Calculate time k according to the formula Γ ^T _k =[1I _k (I _k -I _k-1 )/Δt(V _{t, k} -V _{t, k-1} )/Δtn _k-1 ...n _k-nc ] The recursive value Γ of the input vector, wherein, Γ ^T ₁ =Γ ^T ₂ =...=Γ ^T _nc =Γ ₀ , Γ ₀ is a given initial value, I, V _t are the current of the battery sampled by the sensor (charging Negative during discharge, positive during discharge), terminal voltage, subscript k represents the kth moment, k-1 represents the k-1st moment, Δt is the time between the kth moment and the k-1th moment, n _k-1 , ..., n _k-nc are the random errors of k-1 at the previous moment and k-nc at the previous nc moment;

Step 102: Update the gain at the kth moment according to the formula P _k =[P _k-1 -P _k-1 Γ _k Γ ^T _k P _k-1 /(λ+Γ ^T _k P _k-1 Γ _k )]/λ Factor P _k , where the subscripts k and k-1 represent the kth moment and k-1 moment respectively, and λ is the forgetting factor (usually the value range is 0.95 to 1);

Step 103. According to the formula 0 _k =0 _k-1 +P _k Γ _k [V _{t, k} -Γ ^T _k 0 _k-1 ], calculate the parameter vector 0 _k to be identified at the kth moment;

Step 104: After updating the sampling values of battery current I and terminal voltage V _t at time k+1, according to the formula n _k+1-i = V _{t, k+1-i} -Γ ^T _k+1-i 0 _{k+ 1-i} (i=1, 2, 3, ..., nc) update the random error of nc times before the current time, replace k with k+1, return to step 101, and realize recursion.

Step 105, using elements 0 _{1, k} , 0 _{2, k} , 0 _{3, k} , 0 _{4, k} in the parameter vector 0 _k to be identified obtained in the recursive calculation in steps 101-104, according to the formula V _oc =0 respectively _{1, k} , R _in =0 _{3, k} /0 _{4, k} , R _p =-0 _{2, k} -0 _{3, k} /0 _{4, k} , C _p =0 _{4, k} ² /(0 _{2, k} 0 _{4, k} +0 _{3, k} ) Calculate the battery open circuit voltage V _oc , DC internal resistance R _in , R _p and C _p in the RC circuit in the battery equivalent circuit model.

The step 2 specifically includes the following steps:

Step 201. Calculate the battery polarization voltage V _p,k at the current moment k according to the formula V _p,k =V _pc -V _t,k -I _k R _in +w _k , where V _oc , R _in and w _k are respectively For the battery open circuit voltage, DC internal resistance, and colored noise identified online in the step 1, V _{t, k} and I _k are the battery terminal voltage and the current through the battery measured by the sensor;

Step 202, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 203: Estimate the colored noise w _k+1 at the next moment according to w _k+1 = Γ ^T _k 0 _k , where Γ ^T _k and 0 _k are respectively the iterations of the input vector at time k calculated in step 1 Inferred value, parameter vector to be identified;

Step 204, according to the formula I ^{chrg, max} _k+1 = (V _oc -V _{p, k+1} -V _max +w _k+1 )/R _in , I ^{dischrg, max} _k+1 = (V _oc -V _{p, k+1} -V _min +w _k+1 )/R _in Calculate the maximum charging current I ^chrg at the next moment k+1 that does not exceed the maximum allowable voltage V _max of the battery, max _k+1 does not exceed the minimum allowable voltage of the battery The maximum discharge current ^{Idischrg of} V _min , max _k+1 ;

Step 205, calculate the battery SOP at the next time k+1 based on the voltage limit according to the following formula:

SOP ^{V, short} _{chargg, k+1} = V _max I ^{chrg, max} _k+1 ;

SOP ^{V, short} _{discharge, k+1} = V _min ^{Idischrg, max} _k+1 .

Described step 3 specifically comprises the following steps:

Step 301, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 302, according to the formula I ^{chrg, max} _k+1 = (V _oc -V _{p, k+1} -V _max )/R _in , I ^{dischrg, max} _k+1 = (V _oc -V _{p, k+1} -V _min )/Rin Calculate the maximum charging current I ^{chrg, max} _k+1 that does not exceed the maximum allowable voltage V _max of the battery at the next moment k+1, and the maximum discharge current I ^{dischrg, max} _k that does not exceed the minimum allowable voltage V _min of the battery ₊₁ ;

Step 303, set k=k+1, repeat step 301 and step 302, then the maximum charging current ^{Ichrg that} does not exceed the maximum allowable voltage V _max of the battery at n times after the current moment can be calculated without other input, ^max _k+i , the maximum discharge current ^Idischrg that does not exceed the minimum allowable voltage V _min of the battery, max _k+i , where i=1~n; and then calculate the battery at the next moment based on the voltage limit according to the following formula SOP:

SOP ^{V, long} _{charge, k+j} = V _max I ^{chrg, max} _k+j ;

SOP ^{V, long} _{discharge, k+j} = V _min I ^{discharge, max} _k+j ;

Wherein, k in the subscript represents the current time, and k+j represents j (j=1, 2, . . . , n) after the current time.

The step 4 specifically includes the following steps:

Step 401. Calculate the battery polarization voltage V _p,k at the current moment k according to the formula V _p,k =V _oc -V _t,k -I _k R _in +w _k , where V _oc , R _in and w _k are respectively For the battery open circuit voltage, DC internal resistance, and colored noise identified online in the step 1, V _{t, k} and I _k are the battery terminal voltage and the current through the battery measured by the sensor;

Step 402, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 403: Estimate the colored noise w _k+1 at the next moment according to w _k+1 = Γ ^T _k 0 _k , where Γ ^T _k and 0 _k are respectively the iterations of the input vector at time k calculated in step 1 Inferred value, parameter vector to be identified;

Step 404, according to the formulas V ^{chrg, max} _k+1 = V _oc - V _{p, k+1} - I _min R _in + w _k+1 , V ^{dischrg, min} _k+1 = V _oc - V _{p, k+ respectively 1} -I _max R _in +w _k+1 Calculate the highest voltage V ^chrg at the next moment k+1 that does not exceed the maximum allowable charging current I _min of the battery, max _k+1 , and the minimum voltage that does not exceed the maximum allowable discharge current I _max of the battery ^{Vdischrg, min} _k+1 ;

Step 405, calculate the battery SOP at the next time k+1 based on the current limit according to the following formula:

SOP ^{I, short} _{charge, k+1} = I _min V ^{chrg, max} _k+1 ;

SOP ^{I, short} _{discharge, k+1} = I _max ^{Vdischrg, min} _k+1 .

Described step 5 specifically comprises the following steps:

Step 501, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 502, according to the formulas V ^{chrg, max} _k+1 = V _oc -V _{p, k+1} -I _min R _in , V ^{dischrg, min} _k+1 = V _oc -V _{p, k+1} -I _max R _in calculates the highest voltage V ^chrg,max _k+1 that does not exceed the maximum allowable charging current I _min of the battery at the next moment k+1, and the minimum voltage V ^dischrg,min _k+ 1 that does not exceed the maximum allowable discharge current I _max of the battery;

Step 503, let k=k+1, repeat the steps 501 and 502, then can calculate the maximum voltage V that does not exceed the maximum battery charging current _Imin at n moments after the current moment without other input ^{chrg, max} _k+i , the minimum voltage V ^{dischrg, min} _k+ i that does not exceed the maximum allowable discharge current I _max of the battery, where i=1~n; and then calculate the time after the next time based on the current limit according to the following formula Battery SOP:

SOP ^{I, long} _{charge, k+j} = I _min V ^{chrg, max} _k+j ;

SOP ^{I, long} _{discharge, k+j} = I _max ^{Vdischrg, min} _k+j ;

Wherein, k in the subscript represents the current time, and k+j represents j (j=1, 2, . . . , n) after the current time.

Described step 6 specifically comprises the following steps:

Step 601, calculate the battery SOP at the next moment according to the following formula:

SOP ^short _charge.k+1 = max[SOP ^{V, short} _{charge, k+1} , SOP ^{I, short} _{charge, k+1} ];

SOP ^short _{discharge.k+1} = max[SOP ^{V, short} _{discharge, k+1} , SOP ^{I, short} _{discharge, k+1} ];

Among them, k in the subscript represents the current moment, k+1 represents the next moment, SOP ^short _charge.k+1 and SOP ^short _{didcharge.k+1} are the battery SOP at the next moment during the charging process and discharging process respectively;

Step 602, according to the following formula, the battery SOP at the time after the next time

SOP ^long _charge.k+j = max[SOP ^{V, long} _{charge, k+j} , SOP ^{I, long} _{charge, k+j} ];

SOP ^long _{discharge.k+j} = max[SOP ^{V, long} _{discharge, k+j} , SOP ^{I, long} _{discharge, k+j} ];

Among them, k in the subscript represents the current moment, k+j represents the j (j=1, 2,...,n) moment after the current moment, SOP ^long _charge.k+j and SOP ^long _{discharge.k+j} are respectively The SOP of the battery at the time after the next time during the charging process and the discharging process.

The battery SOP during the discharge process of the series battery pack is the battery SOP with the lowest cell voltage in the series battery pack, and the battery SOP during the charging process of the series battery pack is the battery SOP with the highest cell voltage in the series battery pack.

The battery SOP can be calculated by the above method, which is the battery temperature limit used to prevent accelerated aging of the battery or even spontaneous combustion and explosion.

Compared with the prior art, the present invention has the following advantages:

1) The present invention simultaneously considers the impact of the battery voltage and current working window on the peak power, thereby improving the reliability of SOP calculation and ensuring that the battery can work efficiently and for a long time.

2) The present invention can simultaneously realize single-step prediction and multi-step prediction of SOP according to different input conditions. Among them, SOP single-step prediction can effectively prevent the battery from being abused during real-time operation, while SOP multi-step prediction can help other related systems to achieve optimal energy management.

3) Through experimental verification, the present invention has the high precision of SOP single-step prediction value and the actual value are almost completely consistent, and the maximum error of multi-step prediction within 15s is -3.27%.

Description of drawings

FIG. 1 is a schematic flowchart of an online estimation method for the power state SOP of a series battery pack according to an embodiment of the present invention.

detailed description

The present invention will be described in detail below in conjunction with specific embodiments. The following examples 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 skilled in the art can make several modifications and improvements without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

As shown in FIG. 1 , an embodiment of the present invention provides an online estimation method for the power state SOP of a series battery pack, including the following steps:

Step 1. Perform recursive online identification of battery parameters based on battery parameters in the battery equivalent circuit model and battery parameters that are not included in the battery equivalent circuit model for comprehensive simulation of battery effects;

Step 2. Execute the battery SOP calculation at the next moment based on the voltage limit and the battery parameters identified online;

Step 3. Execute the battery SOP calculation based on the voltage limit and the battery parameters identified online at the time after the next time;

Step 4, performing battery SOP calculation based on the current limit and the battery parameters at the next moment identified and calculated;

Step 5. Execute the battery SOP calculation based on the current limit and the battery parameters identified online at the time after the next time;

Step 6, integrating the battery SOP at the next moment calculated in steps 2-5 and the battery SOP at a later moment to realize online estimation of the battery SOP based on voltage limitation and current limitation synthesis.

The battery equivalent circuit model in the step 1 is a Thevenin model, and the battery parameters in the battery equivalent circuit model include the open circuit voltage V _oc of the battery, the DC internal resistance R _in of the battery, and the charge transfer phenomenon used to simulate the battery The resistance R _p and capacitance C _p in the RC loop, at time k, the battery effect is not covered in the battery equivalent circuit model, and the battery parameters for comprehensive simulation are the sliding average value of white noise in the battery equivalent circuit model The colored noise w _k added at the output of .

The recursive online identification method in the step 1 is an online identification method based on the recursive extended least squares method, which specifically includes the following steps:

Step 101. Calculate time k according to the formula Γ ^T _k =[1I _k (I _k -I _k-1 )/Δt(V _{t, k} -V _{t, k-1} )/Δtn _k-1 ...n _k-nc ] The recursive value Γ of the input vector, wherein, Γ ^T ₁ =Γ ^T ₂ =...=Γ ^T _nc =Γ ₀ , Γ ₀ is a given initial value, I, V _t are the current of the battery sampled by the sensor (charging Negative during discharge, positive during discharge), terminal voltage, subscript k represents the kth moment, k-1 represents the k-1st moment, Δt is the time between the kth moment and the k-1th moment, n _k-1 , ..., n _k-nc are the random errors of k-1 at the previous moment and k-nc at the previous nc moment;

Step 102: Update the gain at the kth moment according to the formula P _k =[P _k-1 -P _k-1 Γ _k Γ ^T _k P _k-1 /(λ+Γ ^T _k P _k-1 Γ _k )]/λ Factor P _k , where the subscripts k and k-1 represent the kth moment and k-1 moment respectively, and λ is the forgetting factor (usually the value range is 0.95 to 1);

Step 103. According to the formula 0 _k =0 _k-1 +P _k Γ _k [V _{t, k} -Γ ^T _k 0 _k-1 ], calculate the parameter vector 0 _k to be identified at the kth moment;

Step 104: After updating the sampling values of battery current I and terminal voltage V _t at time k+1, according to the formula n _k+1-i = V _{t, k+1-i} -Γ ^T _k+1-i 0 _{k+ 1-i} (i=1, 2, 3, ..., nc) update the random error of nc times before the current time, replace k with k+1, return to step 101, and realize recursion.

Step 105, using elements 0 _{1, k} , 0 _{2, k} , 0 _{3, k} , 0 _{4, k} in the parameter vector 0 _k to be identified obtained in the recursive calculation in steps 101-104, according to the formula V _oc =0 respectively _{1, k} , R _in =0 _{3, k} /0 _{4, k} , R _p =-0 _{2, k} -0 _{3, k} /0 _{4, k} , C _p =0 _{4, k} ² /(0 _{2, k} 0 _{4, k} +0 _{3, k} ) Calculate the battery open circuit voltage V _oc , DC internal resistance R _in , R _p and C _p in the RC circuit in the battery equivalent circuit model.

The step 2 specifically includes the following steps:

Step 201. Calculate the battery polarization voltage V _p,k at the current moment k according to the formula V _p,k =V _oc -V _t,k -I _k R _in +w _k , where V _oc , R _in and w _k are respectively For the battery open circuit voltage, DC internal resistance, and colored noise identified online in the step 1, V _{t, k} and I _k are the battery terminal voltage and the current through the battery measured by the sensor;

Step 202, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 203: Estimate the colored noise w _k+1 at the next moment according to w _k+1 = Γ ^T _k 0 _k , where Γ ^T _k and 0 _k are respectively the iterations of the input vector at time k calculated in step 1 Inferred value, parameter vector to be identified;

Step 204, according to the formula I ^{chrg, max} _k+1 = (V _oc -V _{p, k+1} -V _max +w _k+1 )/R _in , I ^{dischrg, max} _k+1 = (V _oc -V _{p, k+1} -V _min +w _k+1 )/R _in Calculate the maximum charging current I ^chrg at the next moment k+1 that does not exceed the maximum allowable voltage V _max of the battery, max _k+1 does not exceed the minimum allowable voltage of the battery The maximum discharge current ^{Idischrg of} V _min , max _k+1 ;

Step 205, calculate the battery SOP at the next time k+1 based on the voltage limit according to the following formula:

SOP ^{V, short} _{charge, k+1} = V _max I ^{chrg, max} _k+1 ;

SOP ^{V, short} _{discharge, k+1} = V _min ^{Idischrg, max} _k+1 .

Described step 3 specifically comprises the following steps:

Step 301, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 302, according to the formula I ^{chrg, max} _k+1 = (V _oc -V _{p, k+1} -V _max )/R _in , I ^{dischrg, max} _k+1 = (V _oc -V _{p, k+1} -V _min )/Rin Calculate the maximum charging current I ^{chrg, max} _k+1 that does not exceed the maximum allowable voltage V _max of the battery at the next moment k+1, and the maximum discharge current I ^{dischrg, max} _k that does not exceed the minimum allowable voltage V _min of the battery ₊₁ ;

Step 303, set k=k+1, repeat step 301 and step 302, then the maximum charging current ^{Ichrg that} does not exceed the maximum allowable voltage V _max of the battery at n times after the current moment can be calculated without other input, ^max _k+i , the maximum discharge current ^Idischrg that does not exceed the minimum allowable voltage V _min of the battery, max _k+i , where i=1~n; and then calculate the battery at the next moment based on the voltage limit according to the following formula SOP:

SOP ^{V, long} _{charge, k+j} = V _max I ^{chrg, max} _k+j ;

SOP ^{V, long} _{discharge, k+j} = V _min I ^{discharge, max} _k+j ;

Wherein, k in the subscript represents the current time, and k+j represents j (j=1, 2, . . . , n) after the current time.

The step 4 specifically includes the following steps:

Step 401. Calculate the battery polarization voltage V _p,k at the current moment k according to the formula V _p,k =V _oc -V _t,k -I _k R _in +w _k , where V _oc , R _in and w _k are respectively For the battery open circuit voltage, DC internal resistance, and colored noise identified online in the step 1, V _{t, k} and I _k are the battery terminal voltage and the current through the battery measured by the sensor;

Step 402, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 403: Estimate the colored noise w _k+1 at the next moment according to w _k+1 = Γ ^T _k 0 _k , where Γ ^T _k and 0 _k are respectively the iterations of the input vector at time k calculated in step 1 Inferred value, parameter vector to be identified;

Step 404, according to the formulas V ^{chrg, max} _k+1 = V _oc - V _{p, k+1} - I _min R _in + w _k+1 , V ^{dischrg, min} _k+1 = V _oc - V _{p, k+ respectively 1} -I _max R _in +w _k+1 Calculate the highest voltage V ^chrg at the next moment k+1 that does not exceed the maximum allowable charging current I _min of the battery, max _k+1 , and the minimum voltage that does not exceed the maximum allowable discharge current I _max of the battery ^{Vdischrg, min} _k+1 ;

Step 405, calculate the battery SOP at the next time k+1 based on the current limit according to the following formula:

SOP ^{I, short} _{charge, k+1} = I _min V ^{chrg, max} _k+1 ;

SOP ^{I, short} _{discharge, k+1} = I _max ^{Vdischrg, min} _k+1 .

Described step 5 specifically comprises the following steps:

Step 501, according to the formula V _p,k+1 =e ^-Δt/Rp/Cp V _p,k +(1-e ^-Δt/Rp/Cp )R _p I _k to estimate the battery polarization at the next moment k+1 Voltage V _{p, k+1} , where R _p and C _p are the resistance and capacitance in the RC circuit identified online in step 1 to simulate the charge transfer phenomenon of the battery respectively, and Δt is k+1 at the next moment The time between and the current moment k;

Step 502, according to the formulas V ^{chrg, max} _k+1 = V _oc -V _{p, k+1} -I _min R _in , V ^{dischrg, min} _k+1 = V _oc -V _{p, k+1} -I _max R _in calculates the highest voltage V ^chrg,max _k+1 that does not exceed the maximum allowable charging current I _min of the battery at the next moment k+1, and the minimum voltage V ^dischrg,min _k+ 1 that does not exceed the maximum allowable discharge current I _max of the battery;

Step 503, let k=k+1, repeat the steps 501 and 502, then can calculate the maximum voltage V that does not exceed the maximum battery charging current _Imin at n moments after the current moment without other input ^{chrg, max} _k+i , the minimum voltage V ^{dischrg, min} _k+ i that does not exceed the maximum allowable discharge current I _max of the battery, where i=1~n; and then calculate the time after the next time based on the current limit according to the following formula Battery SOP:

SOP ^{I, long} _{charge, k+j} = I _min V ^{chrg, max} _k+j ;

SOP ^{I, long} _{discharge, k+j} = I _max ^{Vdischrg, min} _k+j ;

Wherein, k in the subscript represents the current time, and k+j represents j (j=1, 2, . . . , n) after the current time.

Described step 6 specifically comprises the following steps:

Step 601, calculate the battery SOP at the next moment according to the following formula:

SOP ^short _charge.k+1 = max[SOP ^{V, short} _{charge, k+1} , SOP ^{I, short} _{charge, k+1} ];

SOP ^short _{discharge.k+1} = max[SOP ^{V, short} _{discharge, k+1} , SOP ^{I, short} _{discharge, k+1} ];

Among them, k in the subscript represents the current moment, k+1 represents the next moment, and SOP ^short _charge.k+1 and SOP ^short _{discharge.k+1} are the battery SOP at the next moment during the charging process and discharging process, respectively;

Step 602, according to the following formula, the battery SOP at the time after the next time

SOP ^long _charge.k+j = max[SOP ^{V, long} _{charge, k+j} , SOP ^{I, long} _{charge, k+j} ];

SOP ^long _{discharge.k+j} = max[SOP ^{V, long} _{discharge, k+j} , SOP ^{I, long} _{discharge, k+j} ];

Among them, k in the subscript represents the current moment, k+j represents the j (j=1, 2,...,n) moment after the current moment, SOP ^long _charge.k+j and SOP ^long _{discharge.k+j} are respectively The SOP of the battery at the time after the next time during the charging process and the discharging process.

The battery SOP during the discharge process of the series battery pack is the battery SOP with the lowest cell voltage in the series battery pack, and the battery SOP during the charging process of the series battery pack is the battery SOP with the highest cell voltage in the series battery pack.

This specific implementation also considers the influence of the battery voltage and current working window on the peak power, thereby improving the reliability of SOP calculation and ensuring that the battery can work efficiently and continuously; according to different input conditions, the single-step prediction of SOP can be realized at the same time and multi-step forecasting. Among them, SOP single-step prediction can effectively prevent the battery from being abused during real-time operation, while SOP multi-step prediction can help other related systems to achieve optimal energy management. Through experimental verification, the present invention has the high precision that the SOP single-step prediction value is almost completely consistent with the actual value, and the maximum error of multi-step prediction within 15s is -3.27%.

Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention.

Claims (10)

1. an On-line Estimation method of series battery power rating SOP, is characterized in that, comprise the steps:

Step 1, perform based on the battery parameter in battery equivalent-circuit model and to the recursion on-line identification not containing battery effect in battery equivalent-circuit model and carry out the battery parameter of comprehensive simulation;

Step 2, the battery SOP performed based on the subsequent time of the battery parameter that voltage limits and on-line identification goes out calculate;

Step 3, the battery SOP performed based on the subsequent time later moment of the battery parameter that voltage limits and on-line identification goes out calculate;

Step 4, perform and calculate based on current limit and the battery SOP of the subsequent time of battery parameter that goes out at identification meter;

Step 5, perform the battery parameter gone out based on current limit and on-line identification subsequent time after the battery SOP in moment calculate;

The battery SOP of the subsequent time that step 6, combining step 2-5 calculate and the battery SOP in later moment, realizes the On-line Estimation based on the battery SOP that voltage limits and current limit is comprehensive.

2. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, it is characterized in that, battery equivalent-circuit model in described step 1 is Thevenin model, and the battery parameter in described battery equivalent-circuit model comprises the open-circuit voltage V of battery _oc, battery DC internal resistance R _in, for the resistance R in the RC loop of the charge transfer phenomenon of simulated battery _pwith electric capacity C _p, in battery equivalent-circuit model described in moment k, do not contain the coloured noise w that battery effect carries out output terminal at the battery equivalent-circuit model interpolation of battery parameter constructed by the sliding average of white noise of comprehensive simulation _k.

3. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, is characterized in that, the method for the recursion on-line identification in described step 1 is the on-line identification method based on recursion ELS method, specifically comprises the steps:

Step 101, by formula Γ ^t _k=[1I _k(I _k-I _k-1)/Δ t (V _{t, k}-V _{t, k-1})/Δ tn _k-1n _k-nc] calculate the recursion value Γ of moment k input vector, wherein, Γ ^t ₁=Γ ^t ₂=...=Γ ^t _nc=Γ ₀, Γ ₀for given initial value, I, V _tfor electric current (being negative during charging, is just during electric discharge), the terminal voltage of the battery by sensor sample, subscript k represents the kth moment, k-1 represents kth-1 moment, and Δ t is the time between kth moment and kth-1 moment, n _k-1..., n _k-ncbe respectively the stochastic error of previous moment k-1, front nc moment k-nc;

Step 102, by formula P _k=[P _k-1-P _k-1Γ _kΓ ^t _kp _k-1/ (λ+Γ ^t _kp _k-1Γ _k)]/λ upgrades the gain factor P in kth moment _k, wherein, subscript k, k-1 represent kth moment and k-1 moment respectively, and λ is forgetting factor (usual interval is 0.95 ~ 1);

Step 103, by formula 0 _k=0 _k-1+ P _kΓ _k[V _{t, k}-Γ ^t _k0 _k-1] calculate the parameter vector to be identified 0 in kth moment _k;

Step 104, at the electric current I of k+1 moment battery and terminal voltage V _tafter sampled value upgrades, by formula n _k+1-i=V _{t, k+1-i}-Γ ^t _k+1-i0 _k+1-i(i=1,2,3 ..., nc) and upgrade the stochastic error in nc moment before current time, k k+1 is replaced, returns step 101, realize recursion.

Step 105, utilize obtain parameter vector 0 to be identified in the recurrence calculation of step 101 ~ 104 _kin element 0 _{1, k}, 0 _{2, k}, 0 _{3, k}, 0 _{4, k}, press formula V respectively _oc=0 _{1, k}, R _in=0 _{3, k}/ 0 _{4, k}, R _p=-0 _{2, k}-0 _{3, k}/ 0 _{4, k}, C _p=0 _{4, k} ²/ (0 _{2, k}0 _{4, k}+ 0 _{3, k}) calculate battery open circuit voltage V in battery equivalent-circuit model _oc, DC internal resistance R _in, R in RC circuit _pand C _p.

4. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, it is characterized in that, described step 2 specifically comprises the steps:

Step 201, by formula V _{p, k}=V _oc-V _{t, k}-I _kr _in+ w _kcalculate the battery polarization voltage V of current time k _{p, k}, wherein, V _oc, R _inand w _kbe respectively the battery open circuit voltage of on-line identification in described step 1, DC internal resistance, coloured noise, V _{t, k}and I _kfor the battery terminal voltage that obtained by sensor measurement and the electric current by battery;

Step 202, by formula V _{p, k+1}=e ^{-Δ t/Rp/Cp}v _{p, k}+ (1-e ^{-Δ t/Rp/Cp}) R _pi _kestimate the battery polarization voltage V of subsequent time k+1 _{p, k+1}, wherein, R _p, C _pbe respectively resistance, the electric capacity in the RC loop of the charge transfer phenomenon for simulated battery of ONLINE RECOGNITION in described step 1, Δ t is the time between subsequent time k+1 and current time k;

Step 203, by w _k+1=Γ ^t _k0 _kestimate the coloured noise w of subsequent time _k+1, wherein, Γ ^t _k, 0 _kbe respectively recursion value, the parameter vector to be identified of the moment k input vector calculated in described step 1;

Step 204, press formula I respectively ^{chrg, max} _k+1=(V _oc-V _{p, k+1}-V _max+ w _k+1)/R _in, I ^{dischrg, max} _k+1=(V _oc-V _{p, k+1}-V _min+ w _k+1)/R _incalculate subsequent time k+1 and be no more than battery permission ceiling voltage V _maxmaximum charging current I ^{chrg, max} _k+1, be no more than battery and allow minimum voltage V _minmaximum discharge current I ^{dischrg, max} _k+1;

Step 205, be calculated as follows out the battery SOP of subsequent time k+1 based on voltage restriction:

SOP ^V，short _charge，k+1＝V _maxI ^chrg，max _k+1；

SOP ^V，short _{discharge，k+1}＝V _minI ^{dischrg，max} _k+1。

5. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, it is characterized in that, described step 3 specifically comprises the steps:

Step 301, by formula V _{p, k+1}=e ^{-Δ t/Rp/Cp}v _{p, k}+ (1-e ^{-Δ t/Rp/Cp}) R _pi _kestimate the battery polarization voltage V of subsequent time k+1 _{p, k+1}, wherein, R _p, C _pbe respectively resistance, the electric capacity in the RC loop of the charge transfer phenomenon for simulated battery of ONLINE RECOGNITION in described step 1, Δ t is the time between subsequent time k+1 and current time k;

Step 302, press formula I respectively ^{chrg, max} _k+1=(V _oc-V _{p, k+1}-V _max)/R _in, I ^{dischrg, max} _k+1=(V _oc-V _{p, k+1}-V _min)/Rin calculates subsequent time k+1 and is no more than battery permission ceiling voltage V _maxmaximum charging current I ^{chrg, max} _k+1, be no more than battery and allow minimum voltage V _minmaximum discharge current I ^{dischrg, max} _k+1;

Step 303, make k=k+1, repeat step 301 and step 302, then what can calculate n moment after current time when not having other input is no more than battery permission ceiling voltage V _maxmaximum charging current I ^{chrg, max} _k+i, be no more than battery and allow minimum voltage V _minmaximum discharge current I ^{dischrg, max} _k+i, wherein, i=1 ~ n; And then be calculated as follows out the battery SOP in the subsequent time later moment based on voltage restriction:

SOP ^V，long _charge，k+j＝V _maxI ^chrg，max _k+j；

SOP ^V，long _{discharge，k+j}＝V _minI ^{dischrg，max} _k+j；

Wherein, the k in subscript represents current time, k+j represents the later j of current time (j=1,2 ..., n).

6. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, it is characterized in that, described step 4 specifically comprises the steps:

Step 401, by formula V _{p, k}=V _oc-V _{t, k}-I _kr _in+ w _kcalculate the battery polarization voltage V of current time k _{p, k}, wherein, V _oc, R _inand w _kbe respectively the battery open circuit voltage of on-line identification in described step 1, DC internal resistance, coloured noise, V _{t, k}and I _kfor the battery terminal voltage that obtained by sensor measurement and the electric current by battery;

Step 402, by formula V _{p, k+1}=e ^{-Δ t/Rp/Cp}v _{p, k}+ (1-e ^{-Δ t/Rp/Cp}) R _pi _kestimate the battery polarization voltage V of subsequent time k+1 _{p, k+1}, wherein, R _p, C _pbe respectively resistance, the electric capacity in the RC loop of the charge transfer phenomenon for simulated battery of ONLINE RECOGNITION in described step 1, Δ t is the time between subsequent time k+1 and current time k;

Step 403, by w _k+1=Γ ^t _k0 _kestimate the coloured noise w of subsequent time _k+1, wherein, Γ ^t _k, 0 _kbe respectively recursion value, the parameter vector to be identified of the moment k input vector calculated in described step 1;

Step 404, press formula V respectively ^{chrg, max} _k+1=V _oc-V _{p, k+1}-I _minr _in+ w _k+1, V ^{dischrg, min} _k+1=V _oc-V _{p, k+1}-I _maxr _in+ w _k+1calculate subsequent time k+1 and be no more than battery permission maximum charging current I _minceiling voltage V ^{chrg, max} _k+1, be no more than battery and allow maximum discharge current I _maxminimum voltage V ^{dischrg, min} _k+1;

Step 405, be calculated as follows out the battery SOP of the subsequent time k+1 based on current limit:

SOP ^I，short _charge，k+1＝I _minV ^chrg，max _k+1；

SOP ^I，short _{discharge，k+1}＝I _maxV ^{dischrg，min} _k+1。

7. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, it is characterized in that, described step 5 specifically comprises the steps:

Step 501, by formula V _{p, k+1}=e ^{-Δ t/Rp/Cp}v _{p, k}+ (1-e ^{-Δ t/Rp/Cp}) R _pi _kestimate the battery polarization voltage V of subsequent time k+1 _{p, k+1}, wherein, R _p, C _pbe respectively resistance, the electric capacity in the RC loop of the charge transfer phenomenon for simulated battery of ONLINE RECOGNITION in described step 1, Δ t is the time between subsequent time k+1 and current time k;

Step 502, press formula V respectively ^{chrg, max} _k+1=V _oc-V _{p, k+1}-I _minr _in, V ^{dischrg, min} _k+1=V _oc-V _{p, k+1}-I _maxr _incalculate subsequent time k+1 and be no more than battery permission maximum charging current I _minceiling voltage V ^{chrg, max} _k+1, be no more than battery and allow maximum discharge current I _maxminimum voltage v ^{dischrg, min} _k+1;

Step 503, make k=k+1, repeating said steps 501 and step 502, then can calculate the battery that is no more than in n moment after current time when not having other to input and allow maximum charging current I _minceiling voltage V ^{chrg, max} _k+i, be no more than battery and allow maximum discharge current I _maxminimum voltage V ^{dischrg, min} _k+i, wherein, i=1 ~ n; And then be calculated as follows out the battery SOP in the subsequent time later moment based on current limit:

SOP ^I，long _charge，k+j＝I _minV ^chrg，max _k+j；

SOP ^I，long _{discharge，k+j}＝I _maxV ^{dischrg，min} _k+j；

Wherein, the k in subscript represents current time, k+j represents the later j of current time (j=1,2 ..., n).

8. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, it is characterized in that, described step 6 specifically comprises the steps:

Step 601, be calculated as follows the battery SOP of subsequent time:

Sop ^short _charge.k+1＝max[SOP ^V，short _charge，k+1，SOP ^I，short _charge，k+1]；

SOP ^short _{discharge.k+1}＝max[SOP ^V，short _{discharge，k+1}，SOP ^I，short _{discharge，k+1}]；

Wherein, the k in subscript represents current time, k+1 represents subsequent time, SOP ^short _charge.k+1, SOP ^short _dischrge.k+1be respectively the battery SOP of subsequent time in charging process and discharge process;

Step 602, battery SOP by the moment after following formula subsequent time

SOP ^long _charge.k+j＝max[SOP ^V，long _charge，k+j，SOP ^I，long _charge，k+j]；

sop ^long _{discharge.k+j}＝max[SOP ^V，long _{discharge，k+j}，SOP ^I，long _{discharge，k+j}]；

Wherein, the k in subscript represents current time, k+j represents the later j of current time (j=1,2 ..., n) moment, sop ^long _charge.k+j, SOP ^long _{discharge.k+j}be respectively the battery SOP in moment after subsequent time in charging process and discharge process.

9. the On-line Estimation method of a kind of series battery power rating SOP as claimed in claim 1, it is characterized in that, battery SOP in series battery discharge process is the battery SOP that the monomer voltage in series battery is minimum, and the battery SOP in series battery charge process is the battery SOP that the monomer voltage in series battery is the highest.

10. the application of the On-line Estimation method of a kind of series battery power rating SOP as described in any one of claim 1-9, is characterized in that, is limited for preventing battery the accelerated deterioration even battery temperature of spontaneous combustion blast by the battery SOP calculated.