KR20210063965A - Secondary battery performance evaluation support apparatus, performance evaluation data storing and analysis system, and using method thereof - Google Patents

Abstract

A device for supporting evaluation of the performance of a secondary battery using a cycler evaluation equipment is provided. The device includes a profile PC which receives a profile, which is result data of inspection and evaluation by unit from the evaluation equipment, and the profile PC further includes an agent module for processing the profile. The agent module performs data monitoring processing for checking whether the profile has been transmitted from the charger/discharger evaluation equipment, and performs data storage process for determining whether a format of the profile is standardized, standardizes it if necessary, and outputting data to the outside so that the profile can be stored in another external device. Therefore, the present invention contributes to the improvement of work efficiency and accuracy of existing researchers/developers or producers.

Description

A device for supporting performance evaluation of secondary batteries, a system for storing and analyzing performance evaluation data, and a method for using the same

The present invention relates to an apparatus for supporting performance evaluation of a secondary battery, a system for storing and analyzing performance evaluation data, and a method of using the same. Specifically, it is a process for research/development or producers to evaluate product performance and obtain results through charger/discharger evaluation equipment for products manufactured through product development or manufacturing processes in the secondary battery manufacturing industry site, It is about an automated system including a process for collecting/storing an evaluation result profile and a process for analyzing data.

In the existing secondary battery manufacturing industry site quality evaluation data collection, the evaluator directly accesses the system of the evaluation equipment and individually copies/acquires the original data through a portable storage medium such as USB or a shared directory copy on the network. There is a hassle, and a method is needed to solve it.

In the case of using evaluation equipment from several equipment manufacturers, the data interface is different for each manufacturer, which makes work inefficient, and there are several difficulties such as the possibility of making errors in data interpretation because the data is not standardized. is needed

To this end, it is necessary to suggest standardization items for the result data when evaluating secondary battery quality, and guide the creation of a standardized profile based on this to facilitate automatic data collection and storage.

Data generated from a plurality of evaluation devices vary in the number and capacity of data generation according to the evaluation method, and the generation time is also varied, so a data management system for integrated management is required.

Existing evaluators perform data analysis tasks using Excel formulas or macros on the original data collected offline. If the amount of such original data is large, repeated analysis tasks cost a lot of money and time. Due to difficulties, methods for automated data summarization and analysis are needed.

Charger/discharger evaluation for secondary batteries consists of evaluation steps such as charging and discharging, and cycles including multiple steps, and these multiple time-series original files are linked. Therefore, when analyzing data, it takes a lot of time for preprocessing for data separation and combining. Content-based business processing processes such as quick retrieval of such data and content confirmation are required, and software functions that include functions such as data movement, deletion, separation, and combination through data editing tools are required.

According to the present invention, as a device supporting evaluation of the performance of a secondary battery using a charger/discharger evaluation equipment, a profile PC that receives a profile, which is the result data of the unit inspection and evaluation from the evaluation equipment, is transmitted Including, wherein the profile PC further comprises an agent module for processing the profile, the agent module, in order to check whether the profile has been transmitted from the charger/discharger evaluation equipment, a file for new file creation data monitoring processing of detecting creation event information, checking whether the newly created file is a profile, and adding it to a work list if it is a profile; and parsing the content of the received profile, judging whether the format of the profile is standardized, performing standardization transformation on the profile if it is not standardized, and converting the profile if it is standardized There is provided an apparatus for supporting performance evaluation of a secondary battery, characterized in that it performs a data storage process that is output to the outside so that it can be stored in another external device.

Preferably, the agent module uses data to be output to the other external device, or receives data already output to the external device from the other device, classifies the type of inspection and evaluation, , performs a summary algorithm for each classified type on the parsed profile, and additionally performs data summary processing for outputting the summarized data of the profile to the outside so that it can be stored in another external device, and the type is charged ( charge), discharge (discharge), and includes at least one of rest (rest).

Preferably, the data storage process further comprises: parsing the file name of the received profile to generate basic information for evaluation and inspection, and generating a database key value combining the basic information The basic information includes at least one data of evaluation/inspection name, equipment name, channel, step, cycle, and time, and the database key value is uniquely determined in association with the profile content.

Preferably, the agent module performs deep learning learning of normal and abnormal summation record values for several steps constituting a cycle, and according to the learning, at the end of the real-time four-step summary, the deep learning By processing the learning result, it immediately detects whether there is an error in the corresponding cycle.

Preferably, the agent module (a) performs deep learning learning to predict a pattern error whether cycle evaluation is in progress normally, (b) processing the deep learning learning result at the time when a predetermined number of cycle summaries are finished, , predicts whether the evaluation is being performed properly without errors, and (a) is a summary of consecutive 'predetermined number-n' cycles when normal, and summary of consecutive 'predetermined number + n' cycles when normal By learning as an output, deep learning learns the pattern for the '2n' cycles, which is the difference, to generate an evaluation error prediction AI (artificial intelligence), and for the n flows so far, the n+1th, n + 2, ..., n + n, respectively, when the difference from the learned correct answer is farther away, the probability of error prediction increases, and (b) is, the real-time 'predetermined number - n' cycle summary ends At a point in time, the error prediction AI is processed.

According to the present invention, as a performance evaluation data storage and analysis system for a secondary battery, the external output data from the above-described performance evaluation support device is received as an input, and the profile, which is an evaluation result from a plurality of evaluation equipment, is integrated an integrated aggregation/storage database server that collects and stores; A data analysis server that receives data from the integrated collection/storage database server as input and performs statistical processing on the received data, wherein the statistical processing is performed on at least one of a data collection amount, a data-based evaluation equipment operation rate, and a data processing rate. Data analysis server, including processing for; and a management/visualization server that receives data from the data analysis server as an input, and provides a search function or data visualization function of the collected data for the received data storage and analysis of performance evaluation data system is provided.

Preferably, the data analysis server performs deep learning learning of normal and abnormal summation record values for several steps constituting a cycle, and according to the learning, at the time when the real-time four-step summary is finished. By processing the deep learning learning result, it immediately detects whether there is an error in the corresponding cycle.

Preferably, the data analysis server (a) performs deep learning learning to predict a pattern error whether cycle evaluation is in progress normally, (b) processing the deep learning learning result at the end of a predetermined number of cycle summarization Thus, it is predicted whether the evaluation is properly executed without error, and in (a), STEP 1) Enter a summary of consecutive 'predetermined number-n' cycles when normal, and continuous 'predetermined number + n' when normal By learning the 'dog cycle summary as an output, deep learning learning is performed on the pattern for the '2n' cycles, which is the difference, to generate an evaluation error prediction AI (artificial intelligence).

According to the present invention, as a method of using the apparatus for supporting performance evaluation of a secondary battery described above, in order to check whether the profile has been transmitted from the charger/discharger evaluation equipment, file creation event information is detected for new file creation a data monitoring processing step of checking whether the newly created file is a profile or not, and adding it to a work list if it is a profile; Parsing the content of the received profile, judging whether the format of the profile is standardized or not, performing standardization transformation on the profile if it is not standardized, and converting the profile to an external source if it is standardized data storage processing step of outputting to the outside so that it can be stored in another device of the; and using data to be output to the external device, or by receiving data already output to the external device from the other device, classifying the type of inspection and evaluation, and performing the parsed profile with respect to the parsed profile. Use of a device for supporting performance evaluation of secondary batteries, comprising a data summary processing step of performing a summary algorithm for each classified type and outputting the summarized data of the profile to the outside so that it can be stored in another external device A method is provided.

According to the present invention, as a method of using the above-described secondary battery performance evaluation data storage and analysis system, the integrated collection/storage database server receives the external output data from the performance evaluation support device as an input, integrally collecting and storing the profile, which is an evaluation result from the evaluation equipment; receiving, by the data analysis server, data from the integrated collection/storage database server as an input, and performing statistical processing on the received data; and the management/visualization server receiving data from the data analysis server as input, and providing a search function or data visualization function of the collected data for the received data. Performance evaluation of a secondary battery comprising the steps of: A method of using a data storage and analysis system is provided.

If a series of processes such as data collection, storage, analysis, and visualization of the result data of such evaluation equipment are online systemized and data standardization for linking multiple equipment is carried out, it will contribute to the improvement of work efficiency and accuracy of existing researchers/developers or producers. think that

1 is a schematic diagram showing the configuration of a system according to the present invention.
2 is a diagram schematically illustrating a flow of data generation and storage.
FIG. 3 is a diagram showing the flow of data monitoring performed by the profile PCs 310 and 320 .
4 is a diagram for explaining the data storage process in detail.
Fig. 5 is a diagram for explaining the data summary processing in detail.
6 is a diagram for a process for error detection in the charging/discharging evaluation/inspection process.

Advantages and features of the present invention, and a method of achieving them will become apparent with reference to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various forms different from each other, and only these embodiments make the disclosure of the present invention complete, and common knowledge in the technical field to which the present invention pertains. It is provided to completely inform the scope of the invention to the possessor, and the invention is only defined by the scope of the claims. The same reference numerals refer to the same elements throughout the specification.

1 is a schematic diagram showing the configuration of a system according to the present invention.

The evaluation equipment closed network 100 is a network in which the existing evaluation equipments 110, 120, 130 exist, and is configured to interwork with the evaluation data automatic storage/analysis system 400, and one evaluation equipment 110, 120 , 130), there are a plurality of slots for evaluation (this is also called a channel or cell, for example, the number of slots may be 128), and the control PC per evaluation equipment (each of 110, 120, 130) is, for example, 1 Each (210, 220, 230) exists, and the measurement data generated from the plurality of evaluation devices 110, 120, 130 is transmitted to the profile PCs 310 and 320 through the control PCs 210, 220, 230. has a structure to be For reference, the profile PCs 310 and 320 are sometimes expressed as 'a device for supporting performance evaluation of a secondary battery' in the claims.

However, since the control PC (210-230) plays a mediating role in delivering the evaluation information (also referred to as profile, original profile, and original profile) from the evaluation equipment (110-130), the control PC is not necessarily described, In FIG. 1 , it may be expressed that evaluation information is transmitted from the evaluation equipment 110 to 130 to the profile PCs 310 and 320 . That is, as shown in FIG. 2 to be described later, the control PC may be omitted, the control PC may be included in the evaluation/inspection equipment 110', 120', or at least its function may be included.

In the present invention, the start of data collection corresponds to the data collection agent software module (or also called the agent module, 'agent S/W' described in FIG. 1 for the evaluation original profile transmitted to the profile PCs 310 and 320). ) is responsible for loading it into the integrated collection/storage database 410, and such agent software (agent S/W) and the storage server 410 may have a many: 1 relationship. 1 shows, as an example, that the agent S/W and the storage server are 2:1, but is not limited thereto.

The agent S/W (agent software module) of the profile PCs 310 and 320 functions as a data monitoring process, a data collection process, a data summary process, and the like, and performs evaluation original data storage, summary data storage, and the like.

Using the stored data, a data collection status analysis process, a data search process, a data editing process, etc. are performed through the data analysis server 420 .

Visualization output of data charts, data tables, etc. and system management work logic by the user's request are performed through the management/visualization server 430 .

Hereinafter, it will be described in more detail with reference to FIG. 1 .

A plurality of evaluation equipment is shown, such as evaluation equipment 1-1 (110), evaluation equipment 1-2 (120), and evaluation equipment #1-3 (130). These evaluation devices 110 , 120 , and 130 are devices for testing and evaluating the performance of charging and discharging secondary batteries. When it is necessary to clearly distinguish between evaluation equipment, it is indicated as evaluation equipment 1-1, evaluation equipment 1-2, and evaluation equipment 1-3. If any evaluation equipment is acceptable, it can be simply described as evaluation equipment. The same applies to control PCs 1-1, 1-2, 1-3, profile PCs 1 and 2, etc., which will be described below.

Although only three evaluation devices 110 , 120 , and 130 are shown in FIG. 1 , another evaluation device may additionally exist as indicated as “extended” in FIG. 1 . In general, it can be seen that there are N evaluation devices.

Each of the evaluation equipment (110, 120, 130) has a corresponding control PC (210, 220, 230), respectively. For example, the control PC 1-1 (210) is matched to the evaluation equipment 1-1 (110), the control PC 1-2 (220) is matched to the evaluation equipment 1-2 (120), and the evaluation equipment 1- 3 (130) is matched with the control PC 1-3 (230).

These control PCs (210, 220, 230) transmit the profile to the profile PC 1 (310). Here, the profile refers to a result of completion of inspection/evaluation for each unit according to a predetermined schedule. In other words, the profile refers to the multi-channel evaluation result original file that comes out through the cycler of the evaluation equipment.

In FIG. 1 , the control PCs 210 , 220 , 230 of the evaluation equipment 110 , 120 , 130 are connected to the profile PC 1 310 as an example, but not shown in addition to the evaluation equipment 110 , 120 , 130 . Other evaluation equipment of may exist, and other evaluation equipment not shown may be connected to the profile PC 2 (320). In addition, another evaluation device (not shown) may be connected to another profile PC (not shown).

And, the data from the profile PC 1 310 and the profile PC 2 320 may be stored in the database 410 (integrated collection/storage DB server). Of course, if another profile PC exists, profile PC 1 310 and profile PC 2 320 as well as data from the other profile PC may also be transferred to the database 410 . This can also be understood from the marking "extended" under the evaluation equipment 1-3 (130) and the control PC 1-3 (230) and the marking "extended" under the profile PC 2 (320).

In addition, the three evaluation devices 110 , 120 , 130 correspond to one profile PC 310 , but this is an example and is not limited thereto.

Agent software (S/W) is installed in each of the profile PCs 310 and 320, and this may be called an agent software module in the profile PC. This agent software can perform data monitoring, data storage processing, and data summary processing, each of which will be described in detail later.

It can be seen that the evaluation equipment 110 to 130, the control PC 210 to 230, the profile PC 310, 320 and the like are included in the evaluation facility closed network 100.

Data from the profile PCs 310 , 320 (eg, profile raw data or profile summary data) is sent to the evaluation data automatic storage/analysis system 400 . The evaluation data automatic storage/analysis system 400 includes, as shown in FIG. 1 , an integrated collection/storage DB server 410 , a data analysis WAS server 420 , and a management/visualization WEB server 430 . The configuration and function of each of these servers 410, 420, 430 will be described later.

Data from the evaluation data automatic storage/analysis system 400 (evaluation result data) is transmitted to the user PC 1 510 , the user PC 520 , and the user PC 530 . These user PCs 510 to 530 may be considered to be included in the development/manufacturing evaluation business network 500 .

As can be seen from the description of "extended" under the user PC 530 in FIG. 1, the number of user PCs is not limited to three, and a necessary number of user PCs may additionally exist.

2 is a diagram schematically illustrating a flow of data generation and storage.

The evaluation equipment 1-1 ( 110 ) and the control PC 1-1 ( 210 ) shown in FIG. 1 are referred to as evaluation/inspection equipment 1 ( 110 ′) in FIG. 2 for convenience. Similarly, the evaluation equipment 1-2 120 and the control PC 1-2 220 shown in FIG. 1 are referred to as evaluation/inspection equipment 2 120 ′ in FIG. 2 for convenience.

The result of the evaluation/inspection for each unit completed by the schedule according to each of the evaluation/inspection equipment 110' and 120' is referred to as a 'profile', and is transmitted to the designated profile PCs 310 and 320 .

At this time, the relationship between the evaluation/inspection equipment 110', 120' and the profile PC 310 is N:1 (3:1 in the example of FIG. 1, 2:1 in the example of FIG. 2), and one evaluation The /inspection equipment has M evaluation/inspection channels. That is, the profiles transmitted to the profile PCs 310 and 320 are generated at a maximum of M×N profiles every few seconds and minutes. Since the evaluation/inspection schedule varies depending on the usage method, the profile generation cycle becomes non-linear. Therefore, data is collected through file monitoring on whether a profile is created.

First, the data generation will be described as follows.

[Data Generation]

STEP 1) N evaluation/inspection equipments 110' and 120' are connected to one profile PC 310 (M evaluation/inspection channels exist inside one evaluation/inspection equipment 110') .

STEP 2) Conduct evaluation/inspection according to the schedule.

STEP 3) Transmit the evaluation/inspection results for each unit to the profile PC 310 (eg, several seconds, minutes)

STEP 4) Inspection 1 is completed only after evaluation/inspection for each designated unit is completed.

SETP 5) Repeat these tests dozens or hundreds of times.

That is, as shown in FIG. 1 or FIG. 2 , evaluation/inspection is performed by the evaluation/inspection equipment 110 ′, 120 ′ and evaluation result data (profile) is generated.

The following data monitoring and data processing (data storage processing and data summary processing) are performed in the profile PCs 310 and 320 of FIGS. 1 and 2 .

3 is a diagram showing the flow of data monitoring performed by the profile PCs 310 and 320 .

[Data monitoring]

STEP 1) As described above (see STEP 3 of [Data Generation] above), the profile is transmitted from the evaluation/inspection equipment 110', 120' to the profile PC 310 (step 3010).

STEP 2) 'File creation' event information for a newly created file is notified from the system (steps 3020, 3030, and 3040).

STEP 3) It is checked whether the file is a profile by filtering the file information (step 3050).

STEP 4) The file location and file name information is added to the work list (step 3060).

STEP 5) Periodically data processing is performed on the data added to the work list (step 3070).

The notification from the system in STEP 2) may be, for example, notification from the file processing system of the profile PCs 310 and 320 of the profile PC itself (the profile PC detects itself). For example, an event such as file creation, change, or deletion of a file system provided by the Windows OS or Linux OS may be received. In addition, in such a system, if the number of files to be monitored is large, frequent, or the search depth is deep (in the tree structure, when a lower node is deep), the system load increases, and the profile PC (310, 320) Since processing may become difficult to be real-time (single monitoring process logic), in order to solve this, multi-threaded, multi-core processing that can be processed by dividing the monitored node into several (that is, load balancing) , system resources can be effectively maximized and used.

The data processing in STEP 5) may include, for example, data storage processing (step 3080) and data summary processing (step 3090), the details of which will be described later.

4 is a diagram for explaining the data storage process in detail.

[Data storage processing]

This data storage processing is indicated as data storage processing (step 3080) in FIG. 3 as an example of the data processing (step 3070) of FIG. The details are as follows.

STEP 1) Data parsing is performed on the profile file name (step 4010).

STEP 2) Create evaluation/inspection basic information such as evaluation/inspection name, equipment name, channel, step, cycle, and time (steps 4020 and 4030).

STEP 3) A database key value combining basic information is generated (step 4040).

STEP 4) Parsing the profile file contents (step 4050).

STEP 5) Normalization conversion is performed on the original file content (steps 4060 and 4070).

STEP 6) It is stored in the database (step 4080).

In STEP 3) above, a database key value is generated through file name parsing, which is stored in association with the profile content (original). In processing big data, setting such a key value makes it possible to perform fast search and storage efficiency. In order to create a unique combination through file name parsing, for example, if equipment manufacturer code, equipment ID, channel ID in equipment, inspection name, sample number, step number, cycle number, etc. are combined, the corresponding inspection/evaluation file You can create a unique key value.

It can be seen that standardized transformation is being performed in STEP 5). In this regard, a profile is transmitted from the control PCs 210 , 220 , 230 of FIG. 1 to the profile PCs 310 and 320 . In FIG. 1 , it is described that a “standardized profile” is transmitted. It is preferable to transmit the standardized profile as described above, but depending on the circumstances, it is not a "standardized profile" as described in FIG. 1, but only a "profile (in a format that may be slightly different for each device)" is transmitted. There may be cases. In this regard, two cases can be divided and processed as shown in (1) and (2) below.

(1) A standardized profile is processed by the normal logic of the present invention. That is, if the system of the present invention determines that the profile transmitted from the control PCs 210, 220, 230 of FIG. 1 to the profile PCs 310 and 320 is standardized, standard profile ⇒ profile parsing (step 4050, etc.) ( ⇒ If necessary or optionally, profile summary to be described later in FIG. 5) ⇒ Save (steps 4080, 5040, etc.).

(2) For a profile that is not standardized, it is processed by the normal logic of the present invention after normalization conversion. That is, if the system of the present invention determines that the profile transmitted from the control PCs 210, 220, 230 of FIG. 1 to the profile PCs 310 and 320 is not standardized, non-standard profile ⇒ standardized conversion (step 4070, etc.) ⇒ Profile parsing (steps 4050, etc.) (⇒ profile summary to be described later in FIG. 5, if necessary or optionally) ⇒ storage (steps 4080, 5040, etc.).

In addition, as shown in FIG. 4 , steps 4010 to 4040 and steps 4050 to 4070 may be performed in parallel.

In Fig. 4, step 4090 means processing every line of the profile content (file content). That is, if it is the last row, the processing is complete. If it is not the last row, it returns to step 4050 and repeats the processing (until the last row).

Fig. 5 is a diagram for explaining the data summary processing in detail.

[Data summary processing]

This data summary processing is indicated as data summary processing (step 3090) in FIG. 3 as an example of the data processing (step 3070) of FIG. The details are as follows.

STEP 1) Input a data source (step 5010).

STEP 2) Classify the test/evaluation type (step 5020).

STEP 3) A summary algorithm suitable for the type is performed (steps 5031, 5032, and 5033).

STEP 4) It is stored in the database (step 5040).

STEP 5) Repeat steps 2 to 4 above while moving from the first row to the last row of the data source (step 5050).

In STEP 1), the data source is parsed in step 4080 of FIG. 4 (described in [Data storage processing]) and stored in the database. That is, the record of the database 410 can be considered as a data source.

In the above STEP 2) and STEP 3), the type (evaluation type) is, for example, charge, discharge, and rest described in FIG. 5 . The summarization algorithm is different according to these types.

The test of the charge, discharge, and rest once (or less frequently) is a step (or step summary), and when several steps are gathered, it becomes a cycle (or cycle summary).

In step 4080 of FIG. 4 and step 5040 of FIG. 5 , the operation of saving to the database has been performed. At this time, the database corresponds to the database 410 of FIG. 1 or 2 (integrated collection/storage DB server).

Although it has been described previously that the profile PCs 310 and 320 perform data summarization processing, according to an embodiment, the profile PCs 310 and 320 perform only the data storage processing shown in FIG. 4, and the database 410 by the processing It is also possible that the data analysis WAS server 420 performs data summary processing using the profile stored in the , and the summary data is stored in the database 410 .

Turning back to Figure 1, which is the most general representation of operation according to the present invention, the description will be made.

In FIG. 1, in collecting the evaluation result file data generated from a plurality of charger/discharger evaluation equipment 110 to 130, the data is automatically integrated and collected through the server-client communication-based software, and the original file and the file contents are collected. It is automatically stored in the database 410 in real time.

The evaluation result original data is processed into a plurality of summary items through a data summary process (step 3090, steps 5010 to 5050), and is summarized/stored in units of steps and cycles of evaluation.

In Fig. 1, through the data statistical process in the data analysis WAS (Web Application Server) server 420, statistical-based analysis of data processing such as data collection amount, data-based evaluation equipment operation rate, data processing rate, etc. is performed and the result is stored do. The storage of this result may be done back to the database 410 .

In FIG. 1 , the management/visualization WEB server 430 provides visualization software in the form of a dashboard for system management such as collection data status, equipment status, and collection process status information through user PCs 510 to 530 .

In FIG. 1, the management/visualization WEB server 430 provides a data visualization screen such as a search method of collected data, a result data view, a chart graph through selection of the X-axis and Y-axis of data items, and the like through software.

Through the system ( 100 , 400 , 500 ) of the present invention, the user is provided with automatic data collection result provision for the original evaluation result, automatic summary data provision by a summary algorithm, and various data visualization results according to user definition, It can be used as a system for evaluation measurement and quality result judgment.

According to the present invention, these problems can be solved for problems such as the inconvenience and difficulty in procedures for using evaluation data of products in the existing secondary battery development/manufacturing site, and the problem of a decrease in the productivity of the company due to the inability to share data. A system and a related method are provided to achieve the efficiency of work by configuring the automated systems 100 , 400 , 500 such as data collection, storage, analysis, and visualization by approaching from a storage and utilization point of view.

Specifically, for evaluators who have previously used evaluation equipment data offline, a system for automatically collecting, storing, and analyzing evaluation result original files is provided.

In Figure 1, the system of the present invention is a 'profile PC (310, 320)' in which the original file (profile) of the evaluation result of the multi-channel coming out through the charger and discharger (Cycler; 110 to 130 (evaluation equipment)) is stored, data Agent software module (agent S/W) that collects data in real time, database server 410 for integrated data storage, data analysis WAS (Web Application Server) server 420 with data analysis algorithm, data source and analysis results It is composed of a service WEB server 430 and the like for services such as visualization.

In FIG. 1, a number of equipment companies that supply chargers and dischargers 110 to 130, which are evaluation equipment 110 to 130, have their own data formats and items. In order to collect these different data, standardized evaluation items are designed and protocols are defined so that different equipment makers can generate standardized evaluation results.

The standardized evaluation items are sequential number by time order, measurement time, repeated measurement number, measurement type classification, measurement value 1, measurement value 2, … , measured value n, etc.

In Figure 1, in the profile PC (310, 320), the real-time file monitoring function of the agent software module (agent S / W) for data collection is a process of real-time monitoring of a specific directory and sub-directory where original profile data is collected, Configures a file list of profile data that is newly created.

1 , the task of the agent software module (agent S/W) reads the generated file list periodically, obtains basic data information through standardized file name parsing, and stores it in the memory.

In addition, by sequentially accessing the files in the above file list, the contents of the file are opened and the contents are parsed according to the standardized protocol. Here, the content of the file means the original content of the evaluation result.

The contents of the evaluation result file are the result values of the corresponding measurement/inspection according to the time sequence, and data are recorded for a time of at least several seconds to a maximum of several days, and the evaluation result storage process stores these data in the database 410 .

The time series data of the result source is required for detailed data analysis, but general analysis requires data summary, and the evaluation result summary process (step 3090 in Fig. 3, see Fig. 5) is a standardized summary of this source data A result is generated and stored in the database 410 through application of not only an algorithm for calculating , but also a personalized summary algorithm of the researcher/developer.

The stored original data (step 3080 of FIG. 3 , see FIG. 4 ) and summary analysis result data (step 3090 of FIG. 3 , see FIG. 5 ) are delivered to the user through a visualization process.

In FIG. 1, the management/visualization WEB server 430 provides information on daily data collection status, collection software module status, evaluation equipment status, etc. in the form of tables and charts that can be grasped at a glance from an evaluation/administrator's point of view. Create a Dashboard visualization.

In FIG. 1 , the management/visualization WEB server 430 creates a table data visualization for a data search process and search results through various search options such as equipment, channel, and date.

1 , the management/visualization WEB server 430 generates an agent management visualization capable of real-time understanding and management of the activity status of the agent software module (agent S/W) that collects the original evaluation result.

In FIG. 1 , the management/visualization WEB server 430 detects the activity status of a plurality of evaluation equipment 110 to 130 in real time, and manages the evaluation equipment that can accurately provide a measurement of the equipment utilization rate based on the collected data. Create a visualization.

The collected evaluation raw data is time-series data, and it is divided into several files according to the evaluation method. Existing users were analyzing the data by grouping the data in units of evaluation steps and cycles for data analysis. Management/visualization WEB server 430 provides a data editing module for facilitating connection/addition between data, separation/deletion of data, etc. for data searched by a user from dozens or hundreds of data sources by systemizing it, and For visualization of , edited data, it provides an editor module that provides functions such as data range selection, horizontal axis selection, vertical axis selection, and chart type selection.

In FIG. 1, the results of such user-defined data editing and analysis in the management/visualization WEB server 430 include chart visualization for data, data table visualization, and the like, and a function to enable the user to generate a number of visualization results. provides

In addition, it provides a screen layout configuration in the form of a report for the visualization result of the management/visualization WEB server 430, a document storage function such as pdf, a print function, and the like.

The present invention solves the existing difficulties and inefficiencies of users by systematizing and providing a series of operations ranging from the collection, storage, analysis, and visualization of charger/discharger data for product evaluation in the manufacture of secondary batteries, and further By sharing and utilizing data, it is expected that it will be helpful in improving business capabilities of companies and individuals.

6 is a diagram illustrating a process for error detection in a charging/discharging evaluation/inspection process.

The data in the present invention will be further described as follows.

(1) Original (original profile, original profile): Time series data (in minutes or seconds) as an evaluation source for charging and discharging, and one file is defined as one step.

(2) step summary: Summarize one file (step) into one record (start/end of data, average, sum, value under specific conditions, etc.).

(3) Cycle summary: There are three types of steps: 'charge', 'discharge', and 'rest', and the definition of one cycle may be different for each evaluation method, but basically Four steps of 'charge step - rest step - discharge step - rest step' are one cycle. These steps are summarized in one record.

(4) Evaluation is to perform this cycle, for example, hundreds of times.

Error detection according to the present invention includes (A) cyclic error detection and (B) evaluation error detection. Hereinafter, each is demonstrated.

(A) Cycle error AI

It will be described with reference to Fig. 6(A). This is to develop artificial intelligence so that it can determine whether the cycle is an error in real time by learning when it is normal and when the summary record values for several steps constituting a cycle are normal and abnormal.

[Learning Step]

STEP 1) A summary of four consecutive steps when normal is input, and the resulting output is set to 1 (step 6010).

STEP 2) A summary of four consecutive steps when abnormal is input, and the resulting output is set to 0 (step 6020).

STEP 3) Deep learning is performed by repeating steps 1) and 2) above to generate an error detection AI (step 6030).

[Inference phase]

At the end of the real-time four-step summary, by processing the error detection AI that went through STEP 1), 2), and 3) of the [learning step], it is possible to immediately detect whether there is an error in the corresponding cycle (step 6040).

(B) Evaluation error AI

It will be described with reference to Fig. 6(B). This is to develop an artificial intelligence that predicts a pattern error whether cycle evaluation is in progress normally, and prevents unnecessary evaluation.

[Learning Step]

STEP 1) A summary of t-5 consecutive cycles when normal is input, and a summary of consecutive t+5 cycles when normal is learned as an output (step 7010).

STEP 2) Deep learning is performed on the pattern for the 10 cycles, which is the difference, to generate an evaluation error prediction AI. That is, with respect to the five flows so far, it can be seen that the probability of error prediction increases when the difference from the learned correct answer increases during the 6th, 7th...10th, respectively (step 7020).

[Inference phase]

STEP 1) At the end of the real-time t-5 cycle summary, the error prediction AI is processed (step 7030).

STEP 2) t+5 cycles are predicted (step 7040).

STEP 3) By calculating the difference between the predicted value and the t+5 measured value in real time, it is possible to predict whether the evaluation is being performed properly without errors (step 7050).

In the above description, t-5, t+5, and the like are arbitrarily determined examples, and are not necessarily limited to the number 5. In 't-5' of the above example, it may be generalized that t is a predetermined number and 5 is n.

The above-described cycle error AI and evaluation error AI may be executed by the profile PCs 310 and 320 of FIG. 1 and/or the data analysis WAS server 420 of FIG. 1 according to processing capabilities or circumstances.

On the other hand, each device to which a separate reference numeral is given in FIG. 1 does not necessarily have to be a physically separate device. For example, the servers 410 , 420 , 430 may be one device, and two or more devices. Functionality may be distributed across devices.

Although the embodiments of the present invention have been described with reference to the accompanying drawings, the present invention is not limited to the above embodiments, but may be manufactured in various different forms, and those skilled in the art to which the present invention pertains. It will be understood that the present invention can be implemented in other specific forms without changing the technical spirit or essential features of the present invention. Therefore, it should be understood that the embodiments described above are illustrative and non-limiting in all respects.

110, 120, 130: evaluation equipment
210, 220, 230: Control PC
110', 120': evaluation/inspection equipment
310, 320: Profile PC
410: Integrated collection/storage DB server (database)
420: data analysis WAS server
430: management/visualization web server
510, 520, 530 : User PC

Claims (10)

A device that supports evaluating the performance of a secondary battery using a cycler evaluation equipment, comprising:
It includes a profile PC that receives a profile that is the result data of inspection and evaluation for each unit from the evaluation equipment,
The profile PC further comprises an agent module for processing the profile,
The agent module,
In order to check whether the profile has been transmitted from the charger/discharger evaluation device, file creation event information is detected for a new file creation, it is checked whether the newly created file is a profile, and if it is a profile, a task list In addition to data monitoring processing; and
Parsing the content of the received profile, judging whether the format of the profile is standardized or not, performing standardization transformation on the profile if it is not standardized, and converting the profile to an external source if it is standardized Data storage processing that is output to the outside so that it can be saved in other devices of the
A device for supporting performance evaluation of a secondary battery, characterized in that it performs. The method of claim 1,
The agent module,
By using the data to be output to the external device or by receiving data already output to the external device from the other device, the type of inspection and evaluation is classified, and the parsed profile is classified Data summary processing that executes a summary algorithm for each type of data and outputs the summarized data of the profile to an external device to store
do additionally,
The type is an apparatus for supporting performance evaluation of a secondary battery, characterized in that it includes at least one of a charge, a discharge, and a rest. The method of claim 1,
The data storage process is
Parsing the file name of the received profile to generate basic information for evaluation and inspection, and generating a database key value combining the basic information,
The basic information includes at least one data of evaluation/inspection name, equipment name, channel, step, cycle, and time,
The database key value is associated with the profile content and is a device for supporting performance evaluation of a secondary battery, characterized in that it is uniquely determined. The method of claim 1,
The agent module,
Deep learning learns when the summary record values for multiple steps constituting a cycle are normal and when it is abnormal,
According to the learning, the performance evaluation support apparatus of a secondary battery, characterized in that by processing the deep learning learning result at the end of the real-time four-step summary, and immediately detecting whether there is an error in the corresponding cycle. The method of claim 1,
The agent module,
(a) Perform deep learning learning to predict pattern error whether cycle evaluation is in progress normally,
(b) processing the deep learning learning result at the end of the summary of a predetermined number of cycles, predicting whether the evaluation is properly executed without error,
In (a), a continuous 'predetermined number-n' cycle summary when normal is input, and a continuous 'predetermined number + n' cycle summary when normal is learned as an output, and the difference is '2n' Deep learning learns patterns for dog cycles to generate evaluation error prediction AI (artificial intelligence), and for n flows so far, n+1th, n+2th, ..., n+n Second, when the difference from the learned correct answer increases during each process, the probability of error prediction increases,
The (b) is, the real-time 'predetermined number-n' cycle summary, the performance evaluation support apparatus for secondary batteries, characterized in that processing the error prediction AI at the end of the summary. As a secondary battery performance evaluation data storage and analysis system,
Integrated collection/receiving the external output data from the performance evaluation support device according to any one of claims 1 to 5 as an input, and collectively collecting and storing the profile as an evaluation result from a plurality of evaluation equipment storage database server;
A data analysis server that receives data from the integrated collection/storage database server as input and performs statistical processing on the received data, wherein the statistical processing is performed on at least one of a data collection amount, a data-based evaluation equipment operation rate, and a data processing rate. Data analysis server, including processing for; and
A management/visualization server that receives data from the data analysis server as an input, and provides a search function or data visualization function of collected data for the received data
A secondary battery performance evaluation data storage and analysis system comprising a. The method of claim 6,
The data analysis server,
Deep learning learns when the summary record values for multiple steps constituting a cycle are normal and when it is abnormal,
According to the learning, the performance evaluation data storage and analysis system of the secondary battery, characterized in that by processing the deep learning learning result at the end of the real-time four-step summary, and immediately detecting whether there is an error in the corresponding cycle. The method of claim 6,
The data analysis server,
(a) Perform deep learning learning to predict pattern error whether cycle evaluation is in progress normally,
(b) processing the deep learning learning result at the end of the summary of a predetermined number of cycles, predicting whether the evaluation is properly executed without error,
In (a), STEP 1) input a summary of consecutive 'predetermined number-n' cycles when normal, and learn a summary of consecutive 'predetermined number + n' cycles when normal as an output, and the difference is A system for storing and analyzing performance evaluation data of secondary batteries, characterized in that deep learning learns patterns for '2n' cycles, and generates evaluation error prediction AI (artificial intelligence). A method of using the apparatus for supporting performance evaluation of a secondary battery according to any one of claims 1 to 5, comprising:
In order to check whether the profile has been transmitted from the charger/discharger evaluation device, file creation event information is detected for a new file creation, it is checked whether the newly created file is a profile, and if it is a profile, a task list In addition to the data monitoring processing step;
Parsing the content of the received profile, judging whether the format of the profile is standardized or not, performing standardization transformation on the profile if it is not standardized, and converting the profile to an external source if it is standardized data storage processing step of outputting to the outside so that it can be stored in another device of and
By using the data to be output to the external device or by receiving data already output to the external device from the other device, the type of inspection and evaluation is classified, and the parsed profile is classified A data summary processing step of executing a summary algorithm for each type of data and outputting the summarized data of the profile to the outside so that it can be stored in another external device
A method of using a device for supporting performance evaluation of a secondary battery, comprising: A method of using the system for storing and analyzing performance evaluation data of a secondary battery according to claim 6,
receiving, by the integrated collection/storage database server, the external output data from the performance evaluation support device as an input, and integrally collecting and storing the profile, which is an evaluation result from a plurality of evaluation equipment;
receiving, by the data analysis server, data from the integrated collection/storage database server as an input, and performing statistical processing on the received data; and
receiving, by the management/visualization server, data from the data analysis server as input, and providing a search function or data visualization function of collected data for the received data;
A method of using a secondary battery performance evaluation data storage and analysis system comprising a.

Images