WO2018112881A1

WO2018112881A1 - Rapid prediction method for battery charging performance and system thereof

Info

Publication number: WO2018112881A1
Application number: PCT/CN2016/111703
Authority: WO
Inventors: 马艳辉
Original assignee: 深圳中兴力维技术有限公司
Priority date: 2016-12-23
Filing date: 2016-12-23
Publication date: 2018-06-28
Also published as: CN107735691B; CN107735691A

Abstract

Provided are a rapid prediction method for battery charging performance and a system thereof. The method comprises the following steps: obtaining at least two charging characteristic curves between charging time and characteristic value at different depths of discharge (S11, S21), the characteristic value comprising at least one of battery capacity, electric current, and voltage; selecting a plurality of points in the charging characteristic curves and converting the points into actual point coordinates (S12, S22); fitting the charging characteristic curves using a preset function, and obtaining a fitting coefficient of the preset function according to the actual point coordinates (S13, S23); calculating a relational expression between the depth of discharge and the fitting coefficient (S14, S24); and predicting an unknown status by using a semi-supervised method in machine learning to fit charging characteristic curves at various depths of discharge (S25). The charging characteristic curves at various depths of discharge fitted by the method and the system can satisfy the relationship between capacity, voltage, electric current, and depth of discharge; and an error between a predicted value and an actual value is basically kept within 2%, which is very small to meet actual needs.

Description

Method and system for quickly predicting battery charging performance

Technical field

The present invention relates to the technical field of battery charging performance prediction, and in particular, to a method and system for quickly predicting battery charging performance.

Background technique

The battery charging characteristic curve is an important indicator of battery performance. The target lead-acid battery charging characteristic curve prediction is mainly tested by connecting complicated control modules. This test method is complex and easily affected by external factors, which is not conducive to grasping the battery. Charging characteristics.

The Chinese patent application of the publication No. CN101639521 describes a battery charging performance test method, comprising: connecting a test charger to a test battery, using a first digital multimeter to read a voltage value of the test battery, and reading the read voltage value Importing a test computer for display, wherein the method further comprises: inserting a resistor between the test charger and the test battery, using a second digital multimeter to test the voltage value on the resistor, and reading The voltage value is imported into the test computer for data conversion, and the corresponding performance parameters are obtained. This method requires a series of components to be connected, which is susceptible to external influences and has a certain impact on the test results.

Chinese Patent Application No. CN100392942 describes a device for controlling a battery charging process, including: a microprocessor module, a voltage stabilizing current module, a power management module, a trickle charging module, and a microprocessor module according to a battery type. Select the corresponding charging characteristic curve, periodically sample the charging parameter, and the microprocessor module compares the returned sampling signal with the corresponding charging characteristic curve to determine whether the charging parameter needs to be changed. If the charging parameter needs to be changed, according to the charging characteristic curve Reset the current charging parameters and adjust the output of the regulated current regulator module. This method requires more modules and is computationally complex.

Summary of the invention

The main object of the present invention is to provide a method and a system for quickly predicting the charging performance of a battery, and aim at a method for predicting the charging performance of a battery which is simple in calculation, can quickly obtain a result, and has a small error.

To achieve the above object, the present invention provides a method for quickly predicting battery charging performance, The method includes the steps of:

Obtaining at least two charging characteristic curves between charging time and characteristic values at different depths of discharge, the characteristic values including at least one of battery capacity, current, and voltage;

Taking some points in the charging characteristic curve and converting it into actual point coordinates;

And fitting a charging characteristic curve by using a preset function, and acquiring a fitting coefficient of the preset function according to the actual point coordinate;

Calculating a relationship between the depth of discharge and the fitting coefficient;

The semi-supervised method in machine learning is used to predict the unknown state and to fit the charging characteristic curves at various depths of discharge.

Optionally, taking the points in the charging characteristic curve and converting them into actual point coordinates includes:

If the coordinates of several points on a charging characteristic curve are (x' ₁ , y' ₁ ), ..., (x' _i , y' _i ), ..., (x' _n , y' _n ); (x' ₁ , y ' ₁ ) corresponds to the actual point coordinates (x ₁ , y ₁ ); (x' _n , y ' _n ) corresponds to the actual point coordinates (x _n , y _n ); then (x' _i , y' _i ) corresponds to the actual point coordinates (x _i , y _i ):

x _i =(x _i '-x ₁ ')/(x' _n -x ₁ ')*(x _n -x ₁ ); y _i =(y _i '-y ₁ ')/(y' _n -y ₁ ')*(y _n -y ₁ ).

Optionally, the charging characteristic curve is matched by using a preset function, and the fitting coefficient of the preset function is obtained according to the actual point coordinate, and at least one of the following:

A voltage curve is fitted using a polynomial function y=p ₁ *x ² +p ₂ *x+p ₃ , and a fitting coefficient of the polynomial function is obtained according to the actual point coordinates; wherein x represents charging time and y represents voltage , p ₁ , p ₂ , p ₃ are fitting coefficients;

A current curve is fitted using an exponential function y=a*e ^b*x +c*e ^d*x , and a fitting coefficient of the exponential function is obtained according to the actual point coordinates; wherein x represents charging time and y represents current , a, b, c, d are the fitting coefficients.

Optionally, the calculating a relationship between the depth of the discharge and the fitting coefficient comprises:

Under the same charging time, obtaining the depth of discharge _{DOD. 1} is a characteristic value of the corresponding actual point coordinates y ₁ and an actual depth of discharge DOD characteristic point coordinates corresponding to the value of ₂ Y _2;

Calculating an actual point coordinate y _i corresponding to a characteristic value of a discharge depth i at the same charging time; wherein DOD ₁ <i<DOD ₂ :

y _i = y ₁ - (y _{1 -} y ₂ ) * (i - DOD ₁ ) / DOD ₁ .

Optionally, after the method for predicting an unknown state by using a semi-supervised method in machine learning, and predicting a charging characteristic curve at various depths of discharge, the method further includes:

The charging characteristic curve of each new fitting is used by the sampling point under the least squares criterion; if the error exceeds the preset threshold, the fitting coefficient is readjusted.

In addition, in order to achieve the above object, the present invention also provides a rapid prediction system for battery charging performance, including:

An obtaining unit configured to obtain at least two charging characteristic curves between charging times and characteristic values at different depths of discharge, the characteristic values including at least one of battery capacity, current, and voltage;

a coordinate conversion unit configured to take a plurality of points in the charging characteristic curve and convert the same into actual point coordinates;

a function fitting unit configured to fit the charging characteristic curve by using a preset function, and obtain a fitting coefficient of the preset function according to the actual point coordinate;

a calculating unit configured to calculate a relationship between the depth of discharge and the fitting coefficient;

The prediction unit is configured to predict an unknown state using a semi-supervised method in machine learning, and to fit a charging characteristic curve at various depths of discharge.

Optionally, the coordinate conversion unit is configured to: if the coordinates of the points on a certain charging characteristic curve are (x' ₁ , y' ₁ ), ..., (x' _i , y' _i ), ..., (x' _n , y ' _n ); and (x' ₁ , y' ₁ ) corresponds to the actual point coordinates (x ₁ , y ₁ ); (x' _n , y' _n ) corresponds to the actual point coordinates (x _n , y _n); if _{_{(x 'i, y' i}} ) corresponding to the actual point coordinates (x _{_i,} y _i) is:

Optionally, the function fitting unit is set to at least one of the following:

Optionally, the computing unit is configured to:

y _i = y ₁ - (y _{1 -} y ₂ ) * (i - DOD ₁ ) / DOD ₁ .

Optionally, a test adjustment unit is further configured to set a charging characteristic curve for each new fitting with the sampling point under the least squares criterion; if the error exceeds the preset threshold, the fitting coefficient is readjusted.

The rapid prediction method and system for charging performance of the battery proposed by the invention are especially suitable for predicting the charging performance of the lead-acid battery, and the newly fitted charging characteristic curve can well satisfy the relationship between capacity, voltage, current and DOD. The error between the predicted value and the actual value is kept within 2%, and the error is small, which satisfies the actual needs.

DRAWINGS

1 is a schematic flow chart of a method for quickly predicting battery charging performance according to an embodiment of the present invention;

2 is a schematic diagram of a charging characteristic curve of a battery obtained by an experiment;

Figure 3 is a partial actual value of the battery capacity curve in the charging characteristic curve of Figure 2 at four different depths of discharge;

4 is a schematic flow chart of a method for quickly predicting a charging performance of a battery according to another embodiment of the present invention;

5 is a schematic diagram of a fitting curve of a voltage in a state where a discharge depth is 100% using a polynomial function according to an embodiment of the present invention;

6 is a schematic diagram of a fitting curve of a current using an exponential function in a state where a discharge depth is 100% according to an embodiment of the present invention;

7 is a schematic diagram showing a capacity curve of a charging characteristic curve of a battery under different depths of discharge according to an embodiment of the present invention;

8 is a schematic diagram of a voltage curve in a charging characteristic curve of a battery under different depths of discharge according to an embodiment of the present invention;

9 is a schematic diagram of current curves in a charging characteristic curve of a battery under different depths of discharge according to an embodiment of the present invention;

FIG. 10 is a schematic structural diagram of a rapid prediction system for battery charging performance according to an embodiment of the present invention; FIG.

11 is a schematic structural diagram of a rapid prediction system for battery charging performance according to another embodiment of the present invention;

The implementation, functional features and advantages of the object of the present invention will be further described with reference to the accompanying drawings. Description.

detailed description

It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

In the following description, the use of suffixes such as "module", "component" or "unit" for indicating an element is merely an explanation for facilitating the present invention, and does not have a specific meaning per se. Therefore, "module" and "component" can be used in combination.

As shown in FIG. 1 , a first embodiment of the present invention provides a method for quickly predicting battery charging performance, including the steps of:

S11: Obtain at least two charging characteristic curves between charging time and characteristic values at different depths of discharge, the characteristic values including at least one of battery capacity, current, and voltage;

The above charging characteristic curve can be obtained through experiments. As shown in FIG. 2, it can be seen that the battery capacity curve in the figure is not smooth, and all the curves of the battery capacity cannot be well fitted and predicted by the function fitting method; in this embodiment Using the weighting method to collect the points of the charging characteristic curve given in the figure, and then using the averaging method to predict the points of other charging characteristic curves, the charging characteristic curves at different depths of discharge can be obtained;

In the specific implementation, the charging characteristic curves of the discharge depths of 25%, 50%, 75%, and 100% can be respectively obtained;

S12. Take some points in the charging characteristic curve and convert them into actual point coordinates;

Please refer to FIG. 3 at the same time, FIG. 3 is a partial actual value of the battery capacity curve in the charging characteristic curve of FIG. 2 at a depth of discharge of 25%, 50%, 75% and 100%;

S13. Fit the charging characteristic curve by using a preset function, and obtain a fitting coefficient of the preset function according to the actual point coordinate;

It should be noted that different preset functions may be adopted according to different characteristic values. For example, a polynomial function may be used to fit the voltage curve in the charging characteristic curve, and an exponential function is used to fit the current curve in the charging characteristic curve;

S14. Calculate a relationship between the depth of the discharge and the fitting coefficient.

S15. Using a semi-supervised method in machine learning to predict an unknown state and fitting a charging characteristic curve at various depths of discharge;

More specifically, using the semi-supervised method in machine learning to predict the unknown state, the capacity characteristic curve of other storage depths of the battery can be obtained, as shown in FIG. 7; using the above steps to fit the coefficient and the fitting coefficient and the depth of discharge The relationship between the voltage curve and the current curve in the charging characteristic curve at other depths of discharge can be predicted, as shown in Figs. 8 and 9, respectively. It has been verified by experiments that the predicted charging characteristic curve can well satisfy the charging performance of the battery. The error between the predicted value and the actual value is kept within 2%, and the error is small, which satisfies the actual needs.

As shown in FIG. 4, a second embodiment of the present invention provides a fast prediction method for battery charging performance, which includes steps S21 to S25 which are the same as S11 to S15 mentioned in the first embodiment, as described above, specifically No longer.

The difference is that, in this embodiment, after the step S25, the method further includes:

S26. Using a sampling point for each newly fitted charging characteristic curve under a least squares criterion; if the error exceeds a preset threshold, re-adjust the fitting coefficient.

The preset threshold is, for example, 2%.

A third embodiment of the present invention provides a method for quickly predicting battery charging performance, comprising the same steps as S11 to S15 mentioned in the first embodiment or steps S21 to S26 mentioned in the second embodiment, specifically as above The description will not be repeated here.

The difference is that, in this embodiment, the step S12 or S22, that is, the taking of the points in the charging characteristic curve and converting the actual points to the actual point coordinates specifically includes:

If the coordinates of several points on the charging characteristic curve at a certain depth of discharge are (x' ₁ , y' ₁ ), ..., (x' _i , y' _i ), ..., (x' _n , y' _n ); and (x' ₁ , y' ₁ ) corresponds to the actual point coordinates (x ₁ , y ₁ ); (x' _n , y ' _n ) corresponds to the actual point coordinates (x _n , y _n ); x _{_'i,} y' _i) corresponding to the actual point coordinates (x _{_i,} y _i) is:

x _i =(x _i '-x ₁ ')/(x' _n -x ₁ ')*(x _n -x ₁ ); y _i =(y _i '-y ₁ ')/(y' _n -y ₁ ')*(y _n -y ₁ );

x _i represents the charging time, and y _i represents the voltage or current.

Wherein, when determining the coordinates of the above points, the 0-9 on the axis corresponding to the charging time can be equally divided by the order of magnitude 0.1, and after 9, the values 9.1, 9.4, 9.8, 10.1, 11.6, 12.1, 14.4, 17.4 will be These actual values are converted into the values of coordinate points (x ₁ '), ..., (x _i '), ..., (x _n '), and the corresponding ordinate values are acquired and composed of coordinate points (x ₁ ', y ₁ '), ..., (x _i ', y _i '), ..., (x _n ', x _n '). The corresponding vertical coordinate value may be a current value or a voltage value.

A fourth embodiment of the present invention provides a method for quickly predicting battery charging performance, comprising the same steps as S11 to S15 mentioned in the first embodiment or steps S21 to S26 mentioned in the second embodiment, specifically as above The description will not be repeated here.

Differently, in this embodiment, in step S13 or S23, the charging characteristic curve is fitted by using a preset function, and the fitting coefficient of the preset function is obtained according to the actual point coordinate, including at least the following One:

A voltage curve is fitted using a polynomial function y=p ₁ *x ² +p ₂ *x+p ₃ , and a fitting coefficient of the polynomial function is obtained according to the actual point coordinates; wherein x represents charging time and y represents voltage , p ₁ , p ₂ , and p ₃ are fitting coefficients; please refer to FIG. 5 , which is a fitting curve of the voltage in a state where the depth of discharge is 100%;

A current curve is fitted using an exponential function y=a*e ^b*x +c*e ^d*x , and a fitting coefficient of the exponential function is obtained according to the actual point coordinates; wherein x represents charging time and y represents current , a, b, c, d are fitting coefficients; please refer to FIG. 6, which is a fitting curve of the current in a state where the depth of discharge is 100%.

Since the fitting coefficients at different depths of discharge are different, the fitting coefficients at each depth of discharge need to be calculated separately.

In addition, the actual value corresponding to the time point of other discharge depths (excluding 25%, 50%, 75%, 100%) can be obtained by further matching the actual battery capacity value with the depth of discharge (DOD).

A fifth embodiment of the present invention provides a method for quickly predicting a charging performance of a battery, comprising the same steps as S11 to S15 mentioned in the first embodiment or steps S21 to S26 mentioned in the second embodiment, specifically as described above The description will not be repeated here.

Differently, in this embodiment, in step S14 or S24, the calculating a relationship between the depth of discharge and the fitting coefficient includes:

y _i = y ₁ - (y _{1 -} y ₂ ) * (i - DOD ₁ ) / DOD ₁ .

For example, at the same charging time, the actual value of the ordinate of the voltage curve in the charging characteristic curve with a depth of discharge of 25% is y ₂₅ , and the depth of discharge is 50%. The actual value of the ordinate of the voltage curve in the charging characteristic curve. For y ₅₀ , the actual value of the ordinate of the voltage curve with a depth of discharge between 26% and 49% can be calculated using the following formula.

y _i =y ₂₅ -(y ₂₅ -y ₅₀ )*(i-25)/25(i=26, 27, 28...49)

Using the above formula, the actual value corresponding to the depth of discharge of 26% to 49% can be obtained, and the actual values at other depths of discharge can be obtained by the same reason. There is a functional relationship between the voltage and the depth of discharge, and the relationship between the curve fitting coefficient value and the depth of discharge can be obtained at 25%, 50%, 75%, and 100%. Similarly, the relationship between the current fitting curve value and the depth of discharge can be obtained.

The method for quickly predicting the charging performance of the battery in the embodiment of the present invention has been described above. The following describes the rapid prediction system for the charging performance of the battery in the embodiment of the present invention.

As shown in FIG. 10, a sixth embodiment of the present invention provides a fast prediction system for battery charging performance, including an acquisition unit 10, a coordinate conversion unit 20, a function fitting unit 30, a calculation unit 40, and a prediction unit 50.

The obtaining unit 10 is configured to acquire at least two charging characteristic curves between charging time and characteristic values at different depths of discharge, the characteristic values including at least one of a battery capacity, a current, and a voltage; the charging characteristic curve may be Obtained by experiment, as shown in Fig. 2, it can be seen that the battery capacity curve in the figure is not smooth, and all the curves of the battery capacity cannot be well fitted and predicted by the function fitting method; in this embodiment, the weighting method is used to collect The point of the charging characteristic curve given in the figure, and then using the averaging method to predict the points of other charging characteristic curves, the charging characteristic curves at different depths of discharge can be obtained; in the specific implementation, the depth of discharge can be respectively obtained as 25%, 50%, 75% and 100% charging characteristics.

The coordinate conversion unit 20 is configured to take several points in the charging characteristic curve and convert it into actual point coordinates; please refer to FIG. 3 at the same time, and FIG. 3 is the battery capacity curve in the charging characteristic curve of FIG. 2 at a discharge depth of 25%. Partial actual values at 50%, 75% and 100%.

The function fitting unit 30 is configured to fit the charging characteristic curve by using a preset function, and obtain a fitting coefficient of the preset function according to the actual point coordinate; The same characteristic value can adopt different preset functions. For example, a polynomial function can be used to fit the voltage curve in the charging characteristic curve, and an exponential function is used to fit the current curve in the charging characteristic curve.

The calculating unit 40 is configured to calculate a relationship between the depth of discharge and the fitting coefficient;

The prediction unit 50 is arranged to predict an unknown state using a semi-supervised method in machine learning, fitting a charging characteristic curve at various depths of discharge. More specifically, using the semi-supervised method in machine learning to predict the unknown state, the capacity characteristic curve of other storage depths of the battery can be obtained, as shown in FIG. 7; using the above steps to fit the coefficient and the fitting coefficient and the depth of discharge The relationship between the voltage curve and the current curve in the charging characteristic curve at other depths of discharge can be predicted, as shown in Figs. 8 and 9, respectively. It has been verified by experiments that the predicted charging characteristic curve can well satisfy the charging performance of the battery. The error between the predicted value and the actual value is kept within 2%, and the error is small, which satisfies the actual needs.

As shown in FIG. 11, a seventh embodiment of the present invention provides a fast prediction system for battery charging performance, including an acquisition unit 10, a coordinate conversion unit 20, a function fitting unit 30, a calculation unit 40, and a prediction unit 50.

The acquiring unit 10, the coordinate converting unit 20, the function fitting unit 30, the calculating unit 40, the predicting unit 50, and the obtaining unit 10, the coordinate converting unit 20, the function fitting unit 30 in the sixth embodiment, The calculation unit 40 and the prediction unit 50 are the same, as described above, and are not described herein again.

The difference is that, in this embodiment, the test adjustment unit 60 is further configured to set a charging characteristic curve for each new fitting with the sampling point under the least squares criterion; if the error exceeds the preset threshold, re-adjust the The fitting coefficient; the preset threshold is, for example, 2%.

An eighth embodiment of the present invention provides a fast prediction system for battery charging performance, including an acquisition unit 10, a coordinate conversion unit 20, a function fitting unit 30, a calculation unit 40, and a prediction unit 50.

The difference is that, in this embodiment, the coordinate conversion unit 20 is set to if the coordinates of several points on a certain charging characteristic curve are (x' ₁ , y' ₁ ), ..., (x' _i , y' _i ), ..., (x' _n , y' _n ); and (x' ₁ , y' ₁ ) corresponds to the actual point coordinates (x ₁ , y ₁ ); (x' _n , y ' _n ) corresponds to the actual point coordinates (x _{_n,} y _n); if _{_{(x 'i, y' i}} ) corresponding to the actual point coordinates (x _{_i,} y _i) is:

x _i represents the charging time, and y _i represents the voltage or current.

In another embodiment of the present invention, the test adjustment unit as described in the seventh embodiment may also be included, as specifically shown above, and details are not described herein again.

A ninth embodiment of the present invention provides a fast prediction system for battery charging performance, including an acquisition unit 10, a coordinate conversion unit 20, a function fitting unit 30, a calculation unit 40, and a prediction unit 50.

The difference is that, in this embodiment, the function fitting unit 30 is set to at least one of the following:

In addition, the corresponding relationship between the actual battery capacity and the depth of discharge (DOD) can be obtained, and the time points corresponding to other discharge depths (excluding 25%, 50%, 75%, and 100%) can be obtained. Actual value.

A tenth embodiment of the present invention provides a fast prediction system for battery charging performance, comprising an acquisition unit 10, a coordinate conversion unit 20, a function fitting unit 30, a calculation unit 40, and a prediction unit 50.

The difference is that, in this embodiment, the calculating unit is configured to:

y _i = y ₁ - (y _{1 -} y ₂ ) * (i - DOD ₁ ) / DOD ₁ .

y _i =y ₂₅ -(y ₂₅ -y ₅₀ )*(i-25)/25(i=26, 27, 28...49)

It should be noted that, in this document, the terms "including", "comprising", or any other variations thereof are intended to encompass a non-exclusive inclusion, such that a process, method, article It includes not only those elements, but also other elements that are not explicitly listed, or elements that are inherent to such a process, method, item, or device. An element that is defined by the phrase "comprising a ..." does not exclude the presence of additional equivalent elements in the process, method, item, or device that comprises the element.

The serial numbers of the embodiments of the present invention are merely for the description, and do not represent the advantages and disadvantages of the embodiments.

Through the description of the above embodiments, those skilled in the art can clearly understand that the foregoing embodiment method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be through hardware, but in many cases, the former is better. Implementation. Based on such understanding, the technical solution of the present invention, which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM, disk, The optical disc includes a number of instructions for causing a terminal device (which may be a cell phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the methods described in various embodiments of the present invention.

The above are only the preferred embodiments of the present invention, and are not intended to limit the scope of the invention, and the equivalent structure or equivalent process transformations made by the description of the present invention and the drawings are directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of the present invention.

Industrial applicability

The invention provides a rapid prediction method and system for charging performance of a battery, and adopts a preset function for different characteristic values by coordinate conversion, and calculates a relationship between a fitting coefficient and a fitting coefficient and a depth of discharge, combined with machine learning. The semi-supervised method predicts the unknown state and then fits the charging characteristic curves at various depths of discharge. It is not only simple to test, but also has small error. It can well satisfy the relationship between capacity, voltage, current and depth of discharge. Predicted value and actual The error of the value is kept within 2%, and the error is small to meet the actual needs.

Claims

A method for quickly predicting battery charging performance, the method comprising the steps of:

Obtaining at least two charging characteristic curves between charging time and characteristic values at different depths of discharge, the characteristic values including at least one of battery capacity, current, and voltage;

Taking some points in the charging characteristic curve and converting it into actual point coordinates;

And fitting a charging characteristic curve by using a preset function, and acquiring a fitting coefficient of the preset function according to the actual point coordinate;

Calculating a relationship between the depth of discharge and the fitting coefficient;

The semi-supervised method in machine learning is used to predict the unknown state and to fit the charging characteristic curves at various depths of discharge.
The method for rapidly predicting battery charging performance according to claim 1, wherein said taking a plurality of points in said charging characteristic curve and converting it into actual point coordinates comprises:

If the coordinates of several points on a charging characteristic curve are (x' 1 , y' 1 ), ..., (x' i , y' i ), ..., (x' n , y' n ); and (x' 1 , y' 1 ) corresponds to the actual point coordinates (x 1 , y 1 ); (x' n , y' n ) corresponds to the actual point coordinates (x n , y n ); if (x 'i, y' i ) corresponding to the actual point coordinates (x i, y i) is:

x i =(x' i -x' 1 )/(x' n -x' 1 )*(x n -x 1 ); y i =(y' i -y' 1 )/(y' n -y ' 1 )*(y n -y 1 ).
The method for quickly predicting the charging performance of a battery according to claim 1, wherein the fitting of the charging characteristic curve by using a preset function, and acquiring the fitting coefficient of the preset function according to the actual point coordinates includes At least one of the following:

A voltage curve is fitted using a polynomial function y=p 1 *x 2 +p 2 *x+p 3 , and a fitting coefficient of the polynomial function is obtained according to the actual point coordinates; wherein x represents charging time and y represents voltage , p 1 , p 2 , p 3 are fitting coefficients;

A current curve is fitted using an exponential function y=a*e b*x +c*e d*x , and a fitting coefficient of the exponential function is obtained according to the actual point coordinates; wherein x represents charging time and y represents current , a, b, c, d are the fitting coefficients.
The method of rapidly predicting battery charging performance according to claim 1, wherein said calculating a relationship between said depth of discharge and said fitting coefficient comprises:

Under the same charging time, obtaining the depth of discharge DOD. 1 is a characteristic value of the corresponding actual point coordinates y 1 and an actual depth of discharge DOD characteristic point coordinates corresponding to the value of 2 Y 2;

Calculating an actual point coordinate y i corresponding to a characteristic value of a discharge depth i at the same charging time; wherein DOD 1 <i<DOD 2 :

y i = y 1 - (y 1 - y 2 ) * (i - DOD 1 ) / DOD 1 .
The method of claim 1, wherein the method further comprises: predicting an unknown state by using a semi-supervised method in machine learning, and predicting a charging characteristic curve at various depths of discharge, the method further comprising:

The charging characteristic curve of each new fitting is used by the sampling point under the least squares criterion; if the error exceeds the preset threshold, the fitting coefficient is readjusted.
A rapid prediction system for battery charging performance, comprising:

An obtaining unit configured to obtain at least two charging characteristic curves between charging times and characteristic values at different depths of discharge, the characteristic values including at least one of battery capacity, current, and voltage;

a coordinate conversion unit configured to take a plurality of points in the charging characteristic curve and convert the same into actual point coordinates;

a function fitting unit configured to fit the charging characteristic curve by using a preset function, and obtain a fitting coefficient of the preset function according to the actual point coordinate;

a calculating unit configured to calculate a relationship between the depth of discharge and the fitting coefficient;

The prediction unit is configured to predict an unknown state using a semi-supervised method in machine learning, and to fit a charging characteristic curve at various depths of discharge.
A rapid prediction system for battery charging performance according to claim 6, wherein said coordinate conversion unit is arranged such that coordinates of a plurality of points on a certain charging characteristic curve are (x' 1 , y' 1 ), .... .., (x' i , y' i ), ..., (x' n , y' n ); and (x' 1 , y' 1 ) corresponds to the actual point coordinates (x 1 , y 1); (x 'n, y' n) corresponding to the actual point coordinates (x n, y n); if (x 'i, y' i ) corresponding to the actual point coordinates (x i, y i) is:

x i =(x' i -x' 1 )/(x' n -x' 1 )*(x n -x 1 ); y i =(y' i -y' 1 )/(y' n -y ' 1 )*(y n -y 1 ).
A rapid prediction system for battery charging performance according to claim 6, wherein said function fitting unit is set to at least one of the following:

A voltage curve is fitted using a polynomial function y=p 1 *x 2 +p 2 *x+p 3 , and a fitting coefficient of the polynomial function is obtained according to the actual point coordinates; wherein x represents charging time and y represents voltage , p 1 , p 2 , p 3 are fitting coefficients;

A current curve is fitted using an exponential function y=a*e b*x +c*e d*x , and a fitting coefficient of the exponential function is obtained according to the actual point coordinates; wherein x represents charging time and y represents current , a, b, c, d are the fitting coefficients.
A rapid prediction system for battery charging performance according to claim 6, wherein said calculation unit is configured to:

Under the same charging time, obtaining the depth of discharge DOD. 1 is a characteristic value of the corresponding actual point coordinates y 1 and an actual depth of discharge DOD characteristic point coordinates corresponding to the value of 2 Y 2;

Calculating an actual point coordinate y i corresponding to a characteristic value of a discharge depth i at the same charging time; wherein DOD 1 <i<DOD 2 :

y i = y 1 - (y 1 - y 2 ) * (i - DOD 1 ) / DOD 1 .
A rapid prediction system for battery charging performance according to claim 6, further comprising a test adjustment unit configured to use a sampling point for each newly fitted charging characteristic curve under a least squares criterion; if the error exceeds a preset threshold Then, the fitting coefficient is readjusted.