Battery state linear estimation system for establishing model based on interval data

Abstract

The invention discloses a battery state linear estimation system for establishing a model based on interval data. According to the battery state estimation system, only part of data in the process that charging and discharging cycle times of the battery reach specified charging and discharging times are sampled and collected to a training set. The estimation model building module learns a linear change rule among a plurality of numerical values in a training set by using an artificial intelligence machine learning algorithm so as to train a model. The battery state estimation system uses the model to estimate other battery health state data of the battery.

Description

Technical Field

[0001] The invention relates to a battery, and in particular to a battery state linear estimation system based on interval data modeling. Background Art

[0002] Batteries are widely used in all aspects of people's daily lives. Whether it is for new energy vehicles using power batteries or backup power supplies for wireless communication base stations, battery life estimation has become a key factor in the development of various industries. In particular, rapid and accurate estimation of battery health status is the core of safety technology and also a difficult problem currently faced. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a battery state linear estimation system based on interval data modeling in view of the shortcomings of the prior art, comprising a battery cycle indication module, a segment data sampling module, an estimation model establishment module and a battery state estimation module. The battery cycle indication module is configured to determine a specified number of charge and discharge times, indicating that the number of times the battery is charged and discharged reaches the specified number of charge and discharge times. The specified number of charge and discharge times is less than a rated number of cycles of the battery. The segment data sampling module is connected to the battery cycle indication module. The segment data sampling module is configured to divide a battery voltage range when the battery is charged each time into multiple battery voltage segment ranges, and sample one of the battery voltage segment ranges from the multiple battery voltage segment ranges when the battery is charged each time. The estimation model establishment module is connected to the segment data sampling module. The estimation model establishment module is configured to collect multiple values ​​of a battery charging time range corresponding to the battery voltage segment range sampled by the segment data sampling module and multiple values ​​of battery capacity data when the battery is discharged multiple times as battery segment sampling data into a training set. The estimation model building module is configured to use an artificial intelligence machine learning algorithm to learn the linear change rules between the multiple values ​​in the training set to train a model, and define the model as a battery health state estimation model. The battery state estimation module is connected to the estimation model building module. The battery state estimation module is configured to use the battery health state estimation model of the battery to estimate other state data of the battery.

[0004] In an embodiment, the battery state linear estimation system based on interval data model building further includes a linear correlation analysis module. The linear correlation analysis module is connected to the segment data sampling module. The linear correlation analysis module is configured to analyze the linear correlation between the length of charging time when the battery is charged multiple times and the battery capacity after the battery is discharged multiple times, and the segment data sampling module samples one of the battery voltage segment ranges from the multiple battery voltage segment ranges when the battery is charged each time. The estimation model building module collects the battery charging time range and the battery capacity after discharge corresponding to the sampled battery voltage segment range when the battery is charged into the training set based on the linear correlation between the length of charging time when the battery is charged multiple times and the battery capacity after the battery is discharged multiple times.

[0005] In an embodiment, the estimation model building module selects a group of battery charging and discharging whose linear correlation degree best meets a specified linear correlation degree from among the selected batteries during multiple charging and discharging, and collects the battery charging time length and capacity during multiple charging and discharging into the training set based on the linear correlation degree between the charging time length of the battery during multiple charging and discharging and the battery capacity after the battery is discharged multiple times.

[0006] In an embodiment, the battery status estimation module uses the battery health status estimation model to estimate the maximum capacity that can be released after the current battery is fully charged based on the current charging time of the battery, or further converts the maximum capacity that can be released after the current battery is fully charged to obtain battery health data.

[0007] In an embodiment, the battery state estimation module converts the charging time of each battery charge into battery health state data. The battery state estimation module establishes a battery linear prediction model based on the converted battery health state data and the corresponding number of charge or discharge cycles. The battery state estimation module estimates the battery health state data after the battery undergoes more charge and discharge cycles based on the battery linear prediction model.

[0008] In an embodiment, the battery state linear estimation system based on interval data modeling further includes a linear curve establishment module. The linear curve establishment module is connected to the segment data sampling module. The linear curve establishment module is configured to establish a curve as a battery curve on a curve graph based on the battery voltage and charging time of the battery each time it is charged, wherein the vertical axis of the curve graph is the battery voltage, and the horizontal axis data of the curve graph is the charging time of the battery. The segment data sampling module samples a battery voltage range from the horizontal axis data on the curve graph as the battery voltage segment range, and uses the horizontal axis data corresponding to the battery voltage segment range on the curve graph as the battery segment sampling data.

[0009] In an embodiment, the battery state linear estimation system based on interval data modeling further includes a moving pane construction module. The moving pane construction module is connected to the linear curve establishment module. The moving pane construction module is configured to construct a moving pane on the curve graph, and move the moving pane along the vertical axis direction on the curve graph to one of the multiple curve segments that frame the battery curve. The range of the vertical axis data of the curve segment framed by the moving pane on the curve graph is the battery voltage segment range. The estimation model establishment module collects the battery charging time on the horizontal axis of the curve graph corresponding to the battery voltage segment range of the curve segment framed by the moving pane as the battery segment sampling data into the training set.

[0010] As described above, the present invention provides a battery state linear estimation system based on interval data modeling. The battery state linear estimation system based on interval data modeling of the present invention only samples a portion of the battery state data of the battery for short-term charging and discharging and collects it into a training set. The battery state linear estimation system based on interval data modeling of the present invention uses an artificial intelligence machine learning algorithm to learn the linear change rules of the battery state data of the training set to train a battery health state estimation model for linearly estimating other (health) state data of the current battery and other batteries with the same battery cell material ratio. Unlike traditional battery state evaluation systems that need to wait for the battery to be fully charged and discharged, the battery state linear estimation system based on interval data modeling of the present invention can effectively save time and quickly and accurately estimate other state data of the battery.

[0011] To further understand the features and technical contents of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are only for reference and description and are not intended to limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a block diagram of a battery state linear estimation system based on interval data modeling according to an embodiment of the present invention.

[0013] Figure 2 The figure is a flowchart of the steps of a battery state linear estimation system based on interval data model building according to an embodiment of the present invention.

[0014] Figure 3 The graph is a graph of battery voltage versus charging time after a battery is charged, which is established by the battery state linear estimation system based on interval data modeling according to an embodiment of the present invention.

[0015] Figure 4It is a schematic diagram of constructing moving panes of different sizes on a curve graph in a battery state linear estimation system based on interval data model building according to an embodiment of the present invention.

[0016] Figure 5 It is a schematic diagram of constructing a moving pane on a curve graph in a battery state linear estimation system based on interval data modeling according to an embodiment of the present invention.

[0017] Figure 6 for Figure 5 A graph of the converted data framed by the moving pane.

[0018] Figure 7 It is a schematic diagram of a battery state linear estimation system based on interval data modeling according to an embodiment of the present invention, with a moving pane.

[0019] Figure 8 for Figure 7 A graph of the converted data framed by the moving pane.

[0020] Fig. 9 The graph is a curve diagram of a battery state linear estimation system for establishing a model based on interval data in an embodiment of the present invention, which collects training sets for linear estimation.

[0021] Fig.10 The graph is a graph of battery health versus charge and discharge cycle number established by a battery state linear estimation system based on interval data modeling according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following is an explanation of the embodiments of the present invention through specific embodiments. Those skilled in the art can understand the advantages and effects of the present invention from the contents disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and the details in this specification can also be modified and changed in various ways based on different viewpoints and applications without departing from the concept of the present invention. In addition, the drawings of the present invention are only simple schematic illustrations and are not depicted according to actual sizes. It is stated in advance. The following embodiments will further explain the relevant technical contents of the present invention in detail, but the disclosed contents are not intended to limit the scope of protection of the present invention.

[0023] Please refer to Figures 1 to 10 ,in Figure 1 is a block diagram of a battery state linear estimation system based on interval data modeling according to an embodiment of the present invention, Figure 2 The figure is a flowchart of the steps of a battery state linear estimation system based on interval data model building according to an embodiment of the present invention.

[0024] The battery state estimation system of the present invention may include: Figure 1The battery cycle indication module 10, the segment data sampling module 20, the estimation model establishment module 30, the battery state estimation module 40, the linear curve establishment module 50, the moving window construction module 60 and the linear correlation analysis module 70 are shown, but the present invention is not limited thereto and some of them may be omitted in practice.

[0025] The battery cycle indication module 10 determines a specified number of charge and discharge times, and indicates that the battery is charged and discharged for each of the specified number of charge and discharge times (e.g. Figure 2 In step S11 shown in the figure, the specified number of charge and discharge times can be determined according to actual needs, but is less than a rated number of charge and discharge times of the battery (i.e., the maximum number of charge and discharge times that can be performed when the battery is brand new).

[0026] That is to say, the battery status estimation system of the present invention does not need to leave the battery still for a long time or perform full charge and discharge for a long time like the traditional battery status detection system, but only needs to charge and discharge the battery for a short time to obtain the segment data in the currently known battery short-time charge and discharge related data, which is used to estimate the unknown status data of the battery or to predict the status data of other batteries.

[0027] The linear curve building module 50 can build a curve graph, such as Figure 3 As shown, the horizontal axis of the curve graph is the battery charging time, and the vertical axis of the curve graph is the battery voltage. The linear curve establishment module 50 can establish a curve on the curve graph as a battery curve according to the battery voltage and charging time of each battery charge, such as Figure 3 As shown in the graph, a plurality of battery curves are respectively established when the battery is charged multiple times.

[0028] The moving pane construction module 60 can create a moving pane on a graph for each battery curve, such as Figure 4 A moving pane of any size among a plurality of moving panes of different sizes indicated by dashed lines is shown.

[0029] The moving pane construction module 60 can move a moving pane along the vertical axis direction on the curve graph, and during the movement process, the moving pane can sequentially frame a plurality of curve segments of a battery curve established according to each battery charging data.

[0030] If necessary, the moving pane construction module 60 can change the size of a moving pane created on a graph (during the moving pane movement process) according to actual needs to frame data of different ranges, such as Figure 5 and Figure 7 The shown movable panes have different sizes and respectively frame different curve segments of a battery curve.

[0031] The linear correlation analysis module 70 can collect data on a curve graph of multiple curve segments framed sequentially during the movement of a moving pane, including battery charging time data (such as Figure 2 Step S13 shown).

[0032] The linear correlation analysis module 70 can obtain the battery charging time length data according to the battery charging time data of the curve segment framed by the moving pane each time on the curve graph. The linear curve establishment module 50 can establish a curve on a curve graph as a battery capacity curve according to the battery charging time length data obtained from the linear correlation analysis module 70. For example, the linear curve establishment module 50 can establish a curve on a curve graph according to the battery charging time length data obtained from the linear correlation analysis module 70. Figure 5 The battery charging time length data and battery capacity of the curve segment framed by a movable pane are shown in FIG. Figure 6 A battery capacity curve is established on a curve graph shown in FIG. Figure 7 The battery charging time length data and battery capacity of the curve segment framed by a movable pane are shown in FIG. Figure 8 Another battery capacity curve is established on the one graph shown.

[0033] The linear correlation analysis module 70 can analyze the linearity of a battery capacity curve established based on the battery charging time length data and the battery capacity framed by the movable pane during the movement of the curve graph, so as to obtain multiple linear correlation levels (such as the linear correlation between the battery charging time length and the battery capacity corresponding to the multiple curve line segment data of the battery curve framed sequentially during the movement of the movable pane) respectively. Figure 2 Step S15 shown). Figure 6 The linear correlation between the battery charging time and the battery capacity of a battery capacity curve shown in FIG. 1 is different from that of FIG. Figure 8 A battery capacity curve is shown showing the linear correlation between the battery charging time and the battery capacity.

[0034] The segment data sampling module 20 determines which curve segment of the battery curve to sample from among the multiple curve segments of the battery curve when the battery is charged each time, i.e., the battery voltage range, as a battery voltage segment range, and uses the horizontal axis data corresponding to the battery voltage segment range on the curve graph, i.e., the battery charging time, as a battery segment sampling data (e.g., Figure 2 Step S15 shown).

[0035] According to the above, the battery status prediction system of the present invention only needs to obtain partial segment data of the battery, and does not need to continuously detect the complete battery-related data of each battery from zero voltage to full charge like the traditional battery status detection system.

[0036] For example, the linear correlation analysis module 70 may use the following formula to calculate the linear correlation between the charging time length of each battery charge and the battery capacity after multiple discharges of the battery:

[0037]

[0038] Among them, ρ X,Y , Corr(X,Y) represents the linear correlation between the charging time and battery capacity when the battery is charged, x i Represents the length of battery charging time for the i-th battery, μ X Represents the average length of multiple battery charging times, y i Represents the battery capacity of the i-th record, μ Y Represents the average value of the battery capacity of the i-pen battery after multiple discharges.

[0039] The estimation model building module 30 can collect multiple values ​​of the charging time within the battery charging time range corresponding to the battery voltage segment range sampled for multiple battery charging and the battery capacity data during multiple battery discharges into a training set, and use an artificial intelligence machine learning algorithm to learn the linear change rules between the multiple values ​​in the training set to train a model, and define the model as a battery health status estimation model (such as Figure 2 Step S17 shown).

[0040] The linear curve building module 50 can be based on a battery health status estimation model such as Fig. 9 As shown, a curve is established on a curve graph as a battery health status estimation model curve.

[0041] The battery state estimation module 40 can use a battery health state estimation model established by the estimation model establishment module 30 to estimate the maximum capacity that can be released after the battery is fully charged based on the current charging time of the battery, or further convert the maximum capacity that can be released after the battery is fully charged to obtain battery health data (such as Figure 2 Step S17 shown).

[0042] The linear curve building module 50 can be based on the data estimated by using a battery health state estimation model to Fig. 9 As shown, a curve is established on a curve graph as a battery health state estimation curve, which is connected to the above-mentioned battery health state estimation model curve.

[0043] If necessary, the battery state estimation module 40 may convert the charging time of each battery charge into a battery health state data, and may establish a battery linear prediction model according to the converted battery health state data.

[0044] The battery state estimation module 40 can use a battery linear prediction model to predict the battery health state data after the battery undergoes more charge and discharge cycles (e.g. Figure 2 In this article, when the accumulated amount of charge or discharge of the battery reaches the maximum amount of charge that can be released after the battery is fully charged, it is regarded as one accumulated cycle number.

[0045] The battery health status data is expressed in the following equation:

[0046] SOH=(Q max / Q n ) · 100%

[0047] Among them, SOH represents the battery health status data, Qmax refers to the maximum amount of electricity that can be released when the battery is fully charged, and Qn is the rated capacity of the battery.

[0048] The linear curve building module 50 can be based on the actual detected battery health status data, the battery health status data converted from the charging time data estimated using a battery health status estimation model, and the battery health status data predicted using a battery linear prediction model, such as Fig.10 As shown, multiple curves are established on a curve graph.

[0049] In summary, the present invention provides a battery state linear estimation system based on interval data modeling. The battery state linear estimation system based on interval data modeling of the present invention only samples a portion of the battery state data of the battery for short-term charging and discharging and collects it into a training set. The battery state linear estimation system based on interval data modeling of the present invention uses an artificial intelligence machine learning algorithm to learn the linear change rules of the battery state data of the training set to train a battery health state estimation model for linear estimation of other (health) state data of the current battery and other batteries with the same battery cell material ratio. Unlike traditional battery state evaluation systems that need to wait for the battery to be fully charged and discharged, the battery state linear estimation system based on interval data modeling of the present invention can effectively save time and quickly and accurately estimate other state data of the battery.

[0050] The contents disclosed above are only preferred feasible embodiments of the present invention, and are not intended to limit the claims of the present invention. Therefore, all equivalent technical changes made using the contents of the present invention description and drawings are included in the claims of the present invention.

Claims

1. A battery state linear estimation system based on interval data modeling, characterized in that: The battery state linear estimation system based on interval data modeling includes: A battery cycle indication module is configured to determine a specified number of charge and discharge times, and to indicate that the battery is charged and discharged for a number of times each reaching the specified number of charge and discharge times, and the specified number of charge and discharge times is less than a rated number of cycles of the battery; a segment data sampling module, connected to the battery cycle indication module, configured to divide a battery voltage range of the battery each time it is charged into a plurality of battery voltage segment ranges, and sample one of the battery voltage segment ranges from the plurality of battery voltage segment ranges of the battery each time it is charged; an estimation model building module, connected to the segment data sampling module, configured to collect multiple values ​​of a battery charging time range corresponding to the battery voltage segment range sampled by the segment data sampling module and multiple values ​​of battery capacity data when the battery is discharged multiple times as a battery segment sampling data into a training set, use an artificial intelligence machine learning algorithm to learn the linear change rules between the multiple values ​​in the training set to train a model, and define the model as a battery health status estimation model; and The battery state estimation module is connected to the estimation model establishment module and is configured to use the battery health state estimation model of the battery to estimate other state data of the battery.

2. The battery state linear estimation system based on interval data modeling according to claim 1, characterized in that: The battery state linear estimation system based on interval data modeling further comprises: a linear correlation analysis module connected to the segment data sampling module and configured to analyze the linear correlation between the charging time length of the battery when it is charged multiple times and the battery capacity after the battery is discharged multiple times, and the segment data sampling module samples one of the battery voltage segment ranges from the multiple battery voltage segment ranges when the battery is charged each time; The estimation model establishment module collects the battery charging time range corresponding to the battery voltage segment range when the sampled battery is charged and the battery capacity after discharge into the training set based on the linear correlation between the charging time length when the battery is charged multiple times and the battery capacity after multiple discharges.

3. The battery state linear estimation system based on interval data modeling according to claim 1, characterized in that: The estimation model building module collects the charging time and capacity of the battery during multiple charging and discharging of the battery, which has the most linear correlation degree that best meets a specified linear correlation degree, into the training set based on the linear correlation degree between the charging time length during multiple charging cycles and the battery capacity after multiple discharging cycles.

4. The battery state linear estimation system based on interval data modeling according to claim 1, characterized in that: The battery status estimation module uses the battery health status estimation model to estimate the maximum capacity that can be released after the current battery is fully charged based on the current charging time of the battery, or further converts the maximum capacity that can be released after the current battery is fully charged to obtain battery health data.

5. The battery state linear estimation system based on interval data modeling according to claim 1, characterized in that: The battery status estimation module converts the charging time of each battery charge into battery health status data, establishes a battery linear prediction model based on the converted battery health status data and the corresponding number of charge or discharge cycles, and estimates the battery health status data after the battery undergoes more charge and discharge cycles based on the battery linear prediction model.

6. The battery state linear estimation system based on interval data modeling according to claim 1, characterized in that: The battery state linear estimation system based on interval data model building further includes: a linear curve establishing module connected to the segment data sampling module and configured to establish a curve on a curve graph as a battery curve according to the battery voltage and charging time of each battery charging, wherein the vertical axis of the curve graph is the battery voltage and the horizontal axis of the curve graph is the charging time of the battery; The segment data sampling module samples a battery voltage range from the horizontal axis data on the curve graph as the battery voltage segment range, and uses the horizontal axis data corresponding to the battery voltage segment range on the curve graph as the battery segment sampling data.

7. The battery state linear estimation system based on interval data modeling according to claim 6, characterized in that: The battery state linear estimation system based on interval data model building further includes: a moving pane construction module, connected to the linear curve establishment module, configured to construct a moving pane on the curve graph, and move the moving pane along the longitudinal axis direction on the curve graph to one of the plurality of curve segments framing the battery curve, wherein the range of the longitudinal axis data of the curve segment framed by the moving pane on the curve graph is the battery voltage segment range; The estimation model building module collects the battery charging time on the horizontal axis of the curve graph corresponding to the battery voltage segment range of the curve segment framed by the moving pane as the battery segment sampling data into the training set.

Images