Battery diagnostic device

Abstract

The invention provides a battery diagnosis device which can analyze the use state of a vehicle with high precision and can be effectively used for diagnosis of a vehicle-mounted battery. A battery diagnostic device that diagnoses a battery mounted in a vehicle is provided with: a first processing unit that acquires, from the vehicle, parking data indicating a period during which the vehicle is in a parking state; a second processing unit that performs frequency analysis of parking data using a plurality of specific periods of different lengths as window functions; a third processing unit that maps each of a plurality of results obtained by frequency analysis performed on each of the plurality of window functions to a four-dimensional map expressed by elapsed time, frequency, specific period, and intensity; and a fourth processing unit that uses the result mapped to the four-dimensional map in the diagnosis of the battery.

Description

Technical Field

[0001] This invention relates to a battery diagnostic device capable of diagnosing batteries installed in vehicles. Background Technology

[0002] Patent Document 1 discloses an apparatus that improves the prediction accuracy when forecasting the remaining lifespan of a vehicle battery. This apparatus includes a technique for predicting the remaining lifespan of a battery by learning time-series data of battery degradation indicators.

[0003] Patent Document 1: Japanese Patent Application Publication No. 2020-148560 Summary of the Invention

[0004] In the device described in Patent Document 1, the neural network used for battery degradation estimation extracts values ​​daily and interpolates previous values ​​on days when no vehicles were driven (traveled), processing the data as one value per day for the past 30 days without missing values. However, this method of processing values ​​has the following problem: even if there are multiple trips on the same day, only one value is extracted. Therefore, for example, even if there are 10 trips in one day, only the value of any one of the trips is considered, and the values ​​of the other 9 trips, even if they are values ​​that affect the degradation estimation, are not reflected at all.

[0005] Therefore, further research is needed on the analysis methods for data used in the diagnosis of vehicle battery degradation to improve accuracy.

[0006] The present invention was made in view of the above-mentioned problems, and its object is to provide a battery diagnostic device that can analyze the usage status of a vehicle with high accuracy and be effectively used for the diagnosis of vehicle batteries.

[0007] To address the aforementioned issues, one aspect of the present invention is a battery diagnostic device for diagnosing a battery mounted in a vehicle. The battery diagnostic device comprises: a first processing unit that acquires parking data from the vehicle representing periods during which the vehicle is in a parked state; a second processing unit that performs frequency analysis on the parking data using multiple specific periods of different lengths as window functions; a third processing unit that maps multiple results obtained from the frequency analysis performed on the multiple window functions onto a four-dimensional graph expressed by elapsed time, frequency, specific period, and intensity; and a fourth processing unit that utilizes the results mapped to the four-dimensional graph for battery diagnosis.

[0008] Invention Effects

[0009] According to the battery diagnostic apparatus of the present invention, it is possible to appropriately extract and analyze, with high precision, the usage conditions of a vehicle that are repeatedly subjected to conditions presumed to cause significant damage to the vehicle battery. Therefore, the results of this high-precision analysis can be effectively utilized for battery diagnostics. Attached Figure Description

[0010] Figure 1 This is a schematic structural diagram of a battery diagnostic device according to an embodiment of the present invention.

[0011] Figure 2A This is a flowchart of the battery diagnostic control process executed by the battery diagnostic device.

[0012] Figure 2B This is a flowchart of the battery diagnostic control process executed by the battery diagnostic device.

[0013] Figure 3 It is an example of an ON / OFF binary sequence.

[0014] Figure 4 This is a diagram showing how the ON / OFF ratio changes based on the location during a specific period.

[0015] Figure 5 This is a schematic diagram of a four-dimensional graph. Detailed Implementation

[0016] Vehicle batteries are prone to cumulative damage and accelerated degradation due to repeated daily ignition-on / off operations (IG-ON / OFF). Furthermore, the number of trips per day is variable, and the timing of these trips depends on the driver and is therefore uncertain. Therefore, this invention proposes a method using frequency analysis techniques such as Fast Fourier Transform (FFT) to represent situations involving repeated short-term IG-ON / OFF driving or long-term parking on a single axis.

[0017] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0018] <Implementation Method>

[0019] [structure]

[0020] Figure 1 This is a functional block diagram illustrating the schematic structure of a battery diagnostic device 20 and a vehicle 10 according to an embodiment of the present invention. Figure 1 The battery diagnostic device 20 illustrated herein is communicatively connected to the vehicle 10.

[0021] (1) Vehicle

[0022] Vehicle 10 is, for example, a hybrid electric vehicle (HEV), a plug-in hybrid electric vehicle (PHEV), or a battery electric vehicle (BEV). This vehicle 10 includes at least a data acquisition unit 11 and a data transmission unit 12. Furthermore, in Figure 1 The example shown is of one vehicle 10 that is communicatively connected to the battery diagnostic device 20, but it is also possible for multiple vehicles 10 to be communicatively connected to the battery diagnostic device 20.

[0023] The data acquisition unit 11 continuously acquires data related to the state of the vehicle 10. In this embodiment, the data acquisition unit 11 acquires at least "parking data" indicating the period during which the vehicle 10 is in a parked state as data related to the state of the vehicle 10. As this parking data, information such as the IG signal indicating the on / off state (IG-ON) / off state (IG-OFF) of the ignition switch of the vehicle 10 and the time when the on / off state of the IG signal switches can be exemplified.

[0024] The data transmission unit 12 performs communication between the vehicle 10 and the battery diagnostic device 20. This data transmission unit 12 transmits data related to the state of the vehicle 10, acquired by the data acquisition unit 11, to the battery diagnostic device 20. In this embodiment, the data transmission unit 12 transmits parking data (e.g., one day's worth) for at least a predetermined period as data related to the state of the vehicle 10. The timing of transmission can be exemplified by when the ignition switch of the vehicle 10 transitions to IG-ON or IG-OFF (i.e., each trip).

[0025] (2) Battery diagnostic device

[0026] The battery diagnostic device 20 is installed, for example, in a central hub that centrally manages the status of multiple vehicles 10. The battery diagnostic device 20 includes a first processing unit 21, a second processing unit 22, a third processing unit 23, and a fourth processing unit 24.

[0027] The first processing unit 21 receives data related to the state of the vehicle 10 (including parking data for a specified period) sent from the vehicle 10. The parking data received by the first processing unit 21 is stored in a storage unit (not shown). The second processing unit 22 performs frequency analysis on the parking data received and stored by the first processing unit 21, using multiple specific periods of different lengths (sizes) as window functions. The third processing unit 23 maps the multiple results obtained by the frequency analysis performed by the second processing unit 22 on the multiple window functions onto a four-dimensional graph expressed by elapsed time, frequency, specific period, and intensity. The fourth processing unit 24 uses the results mapped by the third processing unit 23 to the four-dimensional graph for on-board battery diagnosis. Details of the processing performed by the first processing unit 21, the second processing unit 22, the third processing unit 23, and the fourth processing unit 24 will be described below.

[0028] [control]

[0029] Next, refer to Figure 2A , Figure 2B , Figure 3 , Figure 4 and Figure 5 The control performed by the battery diagnostic device 20 according to one embodiment of the present invention will be described.

[0030] Figure 2A and Figure 2B This is a flowchart illustrating the processing sequence of battery diagnostic control performed by each structure of the battery diagnostic device 20. Figure 2A processing and Figure 2B The processing is handled by connectors X, Y, and Z, respectively. Figure 2A and Figure 2B The battery diagnostic control shown is initiated, for example, by receiving parking data from the vehicle 10 via the first processing unit 21 of the battery diagnostic device 20. In the case where there are multiple vehicles 10 communicatively connected to the battery diagnostic device 20, battery diagnostic control is implemented for each vehicle 10.

[0031] (Step S201)

[0032] The second processing unit 22 extracts a past period's worth of parking data from the parking data sent from the vehicle 10 and stored in a storage unit (not shown). The specified period is not particularly limited and can be set to any value (several months or one year, etc.) depending on the computational load or required precision. If the second processing unit 22 extracts a past period's worth of parking data, the process proceeds to step S202.

[0033] (Step S202)

[0034] The second processing unit 22 converts the extracted parking data for a specified past period into a binary sequence of ON / OFF values. Specifically, the second processing unit 22 converts the parking data for a specified past period into a binary sequence where the driving state (where the ignition switch of vehicle 10 is in the ON state, IG-ON) is represented by a value of "1", and the parking state (where the ignition switch of vehicle 10 is in the OFF state, IG-OFF) is represented by a value of "0". An example of this binary sequence is shown below. Figure 3 The switching unit for 1 / 0 in the binary sequence can be set to, for example, 1 second. If the parking data for a predetermined period of time is converted into an ON / OFF binary sequence by the second processing unit 22, the process proceeds to step S203.

[0035] (Step S203)

[0036] The second processing unit 22 extracts data for a predetermined period, or "specific period," from the binary sequence of ON / OFF obtained by converting parking data. Here, the specific period (initial value) when initially extracting data is preferably a period (e.g., 1 day) that is sufficiently shorter than the predetermined past period extracted in step S201 above. If the second processing unit 22 extracts data for the specific period from the binary sequence of ON / OFF, the process proceeds to step S204.

[0037] (Step S204)

[0038] The second processing unit 22 calculates the ratio of the value to the period (driving state period) during which a specific period is "1 (ON)" for data extracted from the binary ON / OFF sequence (extracted data). This calculated ratio (= total ON periods / specific period) is used to determine whether effective frequency analysis can be performed. For example, as... Figure 4 As shown in (a), when the driving and parking of vehicle 10 are appropriately repeated and the ratio increases, the data exhibits periodicity, and the intensity of frequencies close to the spectrum increases, thus enabling effective frequency analysis. On the other hand, as... Figure 4 As shown in (b), when the ratio decreases due to the long-term parking of vehicle 10, the spectrum becomes zero and effective frequency analysis cannot be expected. If the ratio of the value "1 (ON)" to the period of a specific period is calculated by the second processing unit 22, the process proceeds to step S205.

[0039] (Step S205)

[0040] The second processing unit 22 determines whether the ratio of the calculated value "1 (ON)" to the period of a specific period exceeds a first threshold. This determination is made to determine whether frequency analysis should be performed. Therefore, the first threshold is set to any value that can be determined to be sufficient for effective frequency analysis using data from the specific period. If the second processing unit 22 determines that the ratio of the value "1 (ON)" to the period of the specific period exceeds the first threshold (step S205, Yes), the process proceeds to step S206. On the other hand, if the second processing unit 22 determines that the ratio of the value "1 (ON)" to the period of the specific period does not exceed the first threshold (step S205, No), the process proceeds to step S209.

[0041] (Step S206)

[0042] The second processing unit 22 performs frequency analysis by using a specific period of data from which the extraction ratio exceeds a first threshold as a window function. In this frequency analysis, known methods such as the Fast Fourier Transform (FFT) are used. If the second processing unit 22 performs frequency analysis by using a specific period as a window function, the process proceeds to step S207.

[0043] (Step S207)

[0044] The third processing unit 23 maps the results obtained through the frequency analysis performed in step S206 above onto a four-dimensional graph expressed by elapsed time, frequency, specific period and intensity. Figure 5 The diagram shows a schematic of a four-dimensional graph. Figure 5 In the example, by representing elapsed time as the first axis, frequency as the second axis, and a specific period as the third axis, the spectral intensity configuration (distribution) over each specific period is expressed as the fourth axis, forming a four-dimensional graph. In other words, the state with intensity (magnitude) values ​​existing on three-dimensional coordinates is expressed as four-dimensional. The mapped values ​​become indicators representing the characteristics of the IG-ON / OFF mode of vehicle 10. If the second processing unit 22 maps the results obtained through frequency analysis onto the four-dimensional graph, the process proceeds to step S208.

[0045] (Step S208)

[0046] The second processing unit 22 determines whether to end the frequency analysis. That is, the second processing unit 22 determines whether the required number of frequency analysis results for diagnosing the vehicle battery of the vehicle 10 have been obtained. This determination can be arbitrarily set according to the desired number of results or the sampling range, etc. If the second processing unit 22 determines that the frequency analysis should end (step S208, Yes), the process proceeds to step S212. On the other hand, if the second processing unit 22 determines that the frequency analysis should not end (step S208, No), the process proceeds to step S209.

[0047] (Step S209)

[0048] The second processing unit 22 sets a specific period for extracting data from the ON / OFF binary sequence to be longer (larger) than the current setting. For example, if the current specific period is set to 1 day, the second processing unit 22 will reset the specific period to 2 days or 1 week, etc. The length (size) of the reset specific period is not particularly limited. If the second processing unit 22 sets the specific period to be longer than it is now, the process proceeds to step S210.

[0049] (Step S210)

[0050] The second processing unit 22 determines whether the reset specific period exceeds a second threshold. This determination is made to determine whether the situation is such that the reset specific period cannot be specified, or other conditions that prevent accurate processing. Therefore, for example, a previously specified period is set as the second threshold. If the second processing unit 22 determines that the specific period exceeds the second threshold (step S210, yes), the process proceeds to step S211. On the other hand, if the second processing unit 22 determines that the specific period does not exceed the second threshold (step S210, no), the process proceeds to step S203.

[0051] (Step S211)

[0052] Since the second processing unit 22 is unable to perform accurate processing, it performs the prescribed error handling. Examples of this error handling include failing to display the specified period on the designated screen. If the second processing unit 22 performs error handling, the process proceeds to step S212.

[0053] (Step S212)

[0054] The fourth processing unit 24 determines, based on the multiple results (indicators representing the characteristics of the IG-ON / OFF mode of vehicle 10) mapped to the four-dimensional graph in step S207 above, whether the result of the frequency analysis performed in step S201 on the extracted parking data for a predetermined period of the past is standard. This determination can be performed, for example, as follows.

[0055] These results (indicators) can be further categorized into normal / abnormal groups using machine learning or discriminant methods. Furthermore, four-dimensional maps of a standard model (presumably based on the usage state of a vehicle 10 with minimal damage to the battery) and an error model (presumably based on the usage state of a vehicle 10 with significant damage to the battery) can be prepared in advance. The distances (e.g., Euclidean distances) between the obtained four-dimensional map and both the standard and error model maps can be calculated, and the closer one is determined. This allows for a high-precision determination of whether the vehicle battery is in a standard state or close to an error state. The results of this determination can be effectively used for vehicle battery diagnosis.

[0056] If the result of the frequency analysis of the parking data determined by the fourth processing unit 24 is standard, then the battery diagnostic control ends.

[0057] <Function / Effect>

[0058] According to the battery diagnostic device 20 of the present invention described above, parking data of vehicle 10 is acquired from vehicle 10, and the length (size) of a specific period is changed according to the ratio of driving (ON) / parking (OFF) of parking data. Multiple specific periods of different lengths are used as window functions to perform frequency analysis of parking data respectively, and multiple results obtained by frequency analysis performed on multiple window functions are mapped to a four-dimensional graph (time elapsed, frequency, specific period and intensity).

[0059] This process allows for the extraction and high-precision analysis of the usage conditions of a vehicle 10 that are repeatedly presumed to cause significant damage to the on-board battery. The results of this high-precision analysis (mapped to a four-dimensional image) can then be effectively used for the diagnosis of the on-board battery.

[0060] Furthermore, in this embodiment, the durability limit zone can be further determined from the operation (market usage) or development data of other vehicles, and a fourth value is used for comparison along the dimensions of the first to third axes. Alternatively, based on normal conditions, the distance (Euclidean distance, etc.) between the location and a map showing frequent short-distance trips involving repeated IG-ON / OFF cycles or prolonged parking can be calculated to indicate the likelihood of future malfunctions. Alternatively, it can be connected to a system that urges changes in driving frequency or maintenance (confirmation) at a dealership to prevent battery depletion.

[0061] In addition, the results of the mapping frequency analysis can be plotted in more than 5 dimensions.

[0062] The above describes one embodiment of the technology of the present invention. However, the present invention can be understood not only as a battery diagnostic device, but also as a method performed by a battery diagnostic device, a program of the method, a computer-readable non-transitory storage medium storing the program, a vehicle equipped with a battery diagnostic device, etc.

[0063] The battery diagnostic device of the present invention can be used in situations such as diagnosing the condition of a battery installed in a vehicle.

[0064] Symbol Explanation

[0065] 10-Vehicle, 11-Data acquisition unit, 12-Data transmission unit, 20-Battery diagnostic device, 21-First processing unit, 22-Second processing unit, 23-Third processing unit, 24-Fourth processing unit.

Claims

1. A battery diagnostic device for diagnosing a battery mounted in a vehicle, characterized in that it comprises: The first processing unit acquires parking data from the vehicle indicating the period during which the vehicle is in a parked state; The second processing unit uses multiple specific periods of different lengths as window functions to perform frequency analysis on the parking data. The third processing unit maps the multiple results obtained by frequency analysis performed on the multiple window functions onto a four-dimensional graph expressed by elapsed time, frequency, the specific period, and intensity. and The fourth processing unit uses the results mapped to the four-dimensional graph for the diagnosis of the battery.

2. The battery diagnostic device according to claim 1, characterized in that, The second processing unit uses the plurality of specific periods in the parking data where the ratio of driving status exceeds the first threshold as the window function.

3. The battery diagnostic device according to claim 1 or 2, characterized in that, The second processing unit gradually extends the length of the specific period from the initial value, and performs the prescribed error handling if the specific period exceeds the second threshold.

Images