TWM588251U - Battery health state prediction device - Google Patents

Abstract

The present invention relates to a device for predicting the health of a battery. The main structure includes a storage box that can be connected to any test equipment with a battery charge and discharge function, and a server with a predictive calculation function. As long as the user connects the storage box to the battery to be tested, the storage box can automatically collect the first feature information of the battery to be tested, compare it with the built-in feature database, and extract the matching part as the second feature information, and then It is transmitted to the server through the communication module to search for the second feature information in the parameter database and apply the model calculation of the model database to obtain multiple sets of corresponding first feature parameters and second feature parameters. An estimated value is obtained by a comprehensive calculation using a prediction algorithm module and displayed on the human-machine interface.

Description

Battery health state prediction device

The present invention provides a device for predicting the health of a battery, especially a device that can be safely, quickly, and low-demand externally connected between any test equipment and the battery under test, and automatically collects data and simulates calculations to provide stable battery charge and discharge status Device for monitoring battery health.

According to the increasing demand for modern electric vehicles and electrical energy storage devices, one of the main application components, the "battery", has received more and more attention to its condition detection. In order to indicate the indicator of battery status, the concept of SOH is used. State of Health (SOH) is a parameter that reflects the overall performance of the battery and the ability to release electricity under certain conditions, that is, the ratio of the total amount of battery discharge capacity to the available capacity of a new factory battery under certain conditions. With the use of batteries, there will be many irrecoverable physical or chemical factors that cause battery aging, resulting in a decline in battery health. At present, the SOH of most batteries is determined by the aging state of the battery. The main parameters that characterize battery aging include Decline in battery capacity, increase in battery internal resistance, etc.

Because battery internal resistance is the parameter that has the greatest influence on battery SOH, there are methods to estimate battery SOH by measuring internal resistance, such as the internal resistance, temperature, and battery state of charge (SOC) and battery SOH. The relationship is built into a reference table, and the battery's SOH is estimated by measuring the battery's internal resistance lookup table. However, the process of building the SOH table of various factors and batteries requires a lot of experimental data to support, which means that a lot of equipment, time and manpower are needed to collect the data, which greatly increases the cost.

Leaving aside time and labor costs, in the past, in order to measure battery SOH, special measuring equipment was required. Even different testing equipment and batteries to be tested would have different measuring equipment. The measuring equipment is sometimes Only one kind of parameter can be measured. Therefore, it is limited by the equipment specifications and its application range is limited. Some measurement systems need to be intrusively connected to the test equipment or additional system integration is required, which will affect the existing test equipment. Problems with functioning or causing data outflow Love affairs.

Therefore, how to solve the above-mentioned problems and shortcomings, that is, where the creators of the new model and related manufacturers engaged in this industry are eager to study and improve.

Therefore, in view of the above-mentioned shortcomings, the creators of this new model have collected relevant information, evaluated and considered from various parties, and based on years of experience accumulated in this industry, and have continued to make trials and modifications to design such a safe, fast, and fast A new patenter of a low-battery plug-in device that automatically collects data and simulates calculations between any test equipment and the battery to be tested, while stably providing a battery health status monitoring device that monitors the charge and discharge status of the battery.

The main purpose of this new model is: simply use the storage box to quickly connect to the test equipment and the battery to be tested. It has the advantages of no installation and high applicability. It can pre-fetch feature information in the storage box, and the server predicts the waiting time. Measure the health of the battery.

In order to achieve the above purpose, the main structure of the novel model includes: a storage box for power supply connected to at least one battery under test, a human-machine interface provided on the storage box, and a data collection provided in the storage box. A module, at least one feature database provided in the storage box, a feature comparison module provided on the side of the data collection module, and a communication provided on the side of the feature comparison module and linked to its information Module, a server with information linking the communication module, at least one parameter database provided in the server, at least one model database provided in the server, and a feature search provided in the server A module, a model calculation module provided in the server, a prediction calculation module provided in the server with information linking the feature search module and the model calculation module, and a prediction calculation module provided in the server Test information management platform on the group side.

As long as the user connects the storage box to at least one test device with battery charging and discharging function and the battery to be tested, the storage box can automatically use the data collection module to collect the first characteristic information of the battery to be tested, and the built-in characteristics After the database is compared, the matching part is extracted as the second characteristic information, and then transmitted to the server through the communication module, so as to find the matching person in the parameter database as the first characteristic parameter for the second characteristic information, and apply The model database is used to perform model operations to obtain the second characteristic parameters, and then the prediction algorithm module is used to comprehensively evaluate the first characteristic parameters and the second characteristic parameters, and an estimated value is obtained by the operation and displayed on the human-machine interface. In this way, because the storage box can simply capture important features in advance, and the server performs prediction calculations, it is possible to realize the prediction of any test equipment without the need for installation. Make.

With the above technology, it is possible to address the problems of conventional battery health measurement equipment that can only perform analysis on special equipment, have limited application scope, and have inconvenient connections with test equipment and the battery to be tested. Breakthrough to achieve the practical and progressive nature of the above advantages.

1.1a‧‧‧Storage Box

11‧‧‧ HMI

12‧‧‧Data Collection Module

121‧‧‧First feature information

13‧‧‧ Feature Database

14‧‧‧ Feature Comparison Module

141‧‧‧Second feature information

15‧‧‧Communication Module

2.2a‧‧‧Server

21‧‧‧parameter database

22‧‧‧model database

23‧‧‧Feature Search Module

231‧‧‧The first characteristic parameter

24‧‧‧ Model Calculation Module

241‧‧‧Second characteristic parameter

25‧‧‧ Forecast Calculus Module

251‧‧‧Test Information Management Platform

252‧‧‧Battery Model Management Module

3.3a‧‧‧test equipment

31‧‧‧Test battery

The first figure is a perspective view of a preferred embodiment of the present invention.

The second figure is a structural block diagram of the preferred embodiment of the present invention.

The third figure is a state diagram of the preferred embodiment of the present invention.

The fourth figure is a block flow chart (1) of the preferred embodiment of the present invention.

The fifth figure is a block diagram (2) of the action of the preferred embodiment of the present invention.

The sixth figure is a block diagram (c) of the action of the preferred embodiment of the present invention.

The seventh diagram is a schematic diagram of the implementation of yet another preferred embodiment of the present invention.

In order to achieve the above-mentioned purpose and effect, the technical means and structure adopted by the present invention are described in detail below with reference to the features and functions of the preferred embodiment of the present invention.

Please refer to the first figure and the second figure, which are perspective views and structural block diagrams of the preferred embodiment of the present invention. It can be clearly seen from the figure that the novel system includes: a storage box 1, which is a power supply link At least one battery 31 to be tested of the test device 3; a human-machine interface 11 provided on the storage box 1; and a data collection module 12 provided in the storage box 1 for reading the battery 31 to be tested First feature information 121, which is voltage, current, and time information; at least one feature database 13 provided in the storage box 1; a feature ratio provided on one side of the data collection module 12 For the module 14, the second feature information 141 is extracted from the first feature information 121 in accordance with the feature database 13. The second feature information 141 is an open circuit voltage, a DC internal resistance, or a time change. One of the constants; A communication module 15 provided on one side of the feature comparison module 14 and connected with information thereof; a server 2 connected with information by the communication module 15; at least one parameter database 21 provided in the server 2; At least one model database 22 provided in the server 2; a feature search module 23 provided in the server 2 searches the parameter database 21 for a portion that matches the second feature information 141, A corresponding first characteristic parameter 231 is obtained; a model calculation module 24 provided in the server 2 is matched with the model database 22 and is operated to obtain a corresponding second characteristic parameter 241 according to the second characteristic information 141. The first characteristic parameter 231 and the second characteristic parameter 241 are the capacity of the capacitor, the energy capacity, the residual charge, the equivalent internal impedance, the Coulomb efficiency, the conversion efficiency, the number of remaining cycles, the self-discharge law, and the cell charge balance. Or the internal resistance balance of the battery cell; a prediction calculation module 25 provided in the server 2 and information-linked to the feature search module 23 and the model calculation module 24 is integrated with the first Feature parameter 231 and the second feature parameter 2 41 calculation obtains an estimated value and displays it on the human-machine interface 11; and a test information management platform 251 provided on the side of the prediction calculation module 25, and a battery model is provided on the side of the test information management platform 251 Management module 252.

With the above description, we can understand the structure of this technology, and according to the corresponding cooperation of this structure, it can be safely, quickly and low-demand plug-in between any test equipment 3 and the battery 31 to be tested, and automatically collect data, The simulation operation has the advantages of providing stable monitoring of the battery charge and discharge status, and a detailed explanation will be described below.

Please refer to Figures 1 to 6 at the same time. This is a perspective view from the perspective of the preferred embodiment of the present invention to the flow chart of the action block (3). When the above components are assembled, it can be clearly seen from the figure In the present invention, the battery health status is mainly predicted by the cooperation between a storage box 1 and a server 2. Specifically, as shown in the third and fourth figures, the storage box 1 is first used to connect a battery 31 to be tested. The connection method can be electrically connected through the test device 3, and the test device 3 and the battery 31 to be tested are connected. Test line, or directly connected to the battery 31 under test 3, and then the data collection module 12 automatically and continuously reads the first characteristic information 121 of the battery 31 under test, including voltage, current and time information, and then uses the characteristic ratio For the module 14, the first feature information 121 and the features in the storage box 1 The database 13 performs a comparison to obtain a portion of the first feature information 121 that matches the feature database 13 as the second feature information 141. For example, when the first characteristic information 121 is presented as a waveform diagram of voltage and current over time, the section in which the current characteristic matches the characteristic database 13 is used as the time point at which the data collection module 12 starts to collect. The characteristics and voltage characteristics coincide with the characteristics database 13 at the same time as the time point at which the data collection module 12 ends the collection, and the data collection result is organized into the second characteristic information 141 for transmission to the servo via the communication module 15 The processor 2 performs subsequent processing (as shown in the fifth figure).

Since the storage box 1 can operate independently and collect and organize the characteristic information of the battery in advance, the data collection only needs to be electrically connected with the battery 31 to be tested, and there is no need to perform software installation or system integration. Therefore, the storage box 1 Compatibility with various test equipment 3 is very strong, can be directly plug-in or post-installed in any form of charge and discharge equipment (test equipment 3), has the advantage of simple and fast, and the storage box 1 itself only needs to collect data, compare data The action of data transmission has a very simple built-in structure, so the storage box 1 itself is small in size, which is convenient for users to carry or does not take up space on the test device 3 side for a long time, which is more conducive to the health of the battery. Perform long-term stable self-monitoring.

When the server 2 (which may be the cloud server 2 or the cabinet server 2) receives the second characteristic information 141 through the communication module 15, it is divided into two parts to obtain the specified parameters and their probability distributions. First, as shown in FIG. 6, according to the content of the second feature information 141, the feature search module 23 is used to search the corresponding parameter data table in the parameter database 21. For example, some feature values in the second feature information 141 are extracted. Take it out, such as the open circuit voltage. When the curve of the characteristic value matches the data in the parameter database 21, the first characteristic parameter 231 (such as SOC) can be obtained by referring to the data table. The first characteristic parameter 231 also includes the probability of specifying the parameter. distributed.

In the second part, the second characteristic information 141 is first converted into a normalized mixed sequence, and the interval required for the calculation action is separated from the mixed sequence, for example, the DC internal resistance and the time constant are separated, and then separated. The data is applied to the corresponding operation model (such as a neural network model) in the model database 22, and the model calculation module 24 is used to obtain the specified parameters and their probability distributions as the second characteristic parameter 241. Finally, the prediction algorithm module 25 is used to integrate the first characteristic parameter 231 and the second characteristic parameter 241, and the probability is summed up. The summing method is not limited. For example, it may be an average method (weight of all the summed parameters). All are the same), anomaly elimination method (anomaly data judgment mechanism is added before the algorithm calculation, if the data judgment is abnormal, the calculation result is not included in the total parameters), or regression adjustment method (the correctness of the algorithm result is verified through the verification mechanism, depending on the calculation Method to adjust weight parameters) . In this embodiment, the regression adjustment method is used to integrate the interval probability densities of the first characteristic parameter 231 and the second characteristic parameter 241 to obtain an estimated value, and the peak value of the estimated value can be interpreted as the characteristic parameter. The correct rate that falls into this probability density (for example, SOC is 33% ± 0.5%, and the correct rate is 78%). Finally, the result of the prediction calculation is transmitted back to the human-machine interface 11 for user's review.

Due to the prediction calculation method of the server 2, the estimated value of the specified parameter is derived by searching from the parameter database 21 (comparison table) and the model database 22 (operation model) with at least one second characteristic information 141. The multi-parameter comprehensive evaluation calculation method such as the first characteristic parameter 231 and the second characteristic parameter 241 can more quickly and accurately predict the health status of the battery. In addition, the test management platform can automatically record the process of each prediction, including the content of the tested data, test methods, test results, etc., can be fully recorded in the test information management platform 251 for users to view, or can cooperate with the battery model The management module 252 revises, manages, improves, or creates a new model to accumulate the amount of inventory data of the server 2, which is more conducive to responding to different battery test scenarios. In addition, if the server 2 is a type such as a cabinet server that is not externally connected, the storage box 1, the server 2 and the test equipment 3 can be limited to work in the local area network, thereby ensuring data security and confidentiality.

Please also refer to the seventh figure at the same time, which is a schematic diagram of another preferred embodiment of the present invention. As can be clearly seen from the figure, this embodiment is similar to the above embodiment and is a multi-channel storage box 1a. Implementation of detection mode. Because the connection between the storage box 1a and the test equipment 3a is simple, fast, and highly compatible, and the storage box 1a only performs data collection and feature comparison operations, the rest of the prediction calculation operations are performed by the server 2a, and will not be performed. The hardware causes an excessive load, so one-to-many detection operations can be easily achieved. Of course, one storage box 1a can be connected to a variety of test equipment 3a.

However, the above description is only the preferred embodiment of the new model, and it does not limit the patent scope of the new model. Therefore, all simple modifications and equivalent structural changes made by using the new model's description and diagram contents should be the same. It is included in the patent scope of this new model and is conferred to Chen Ming.

In summary, when the new battery health state prediction device is used, it can truly achieve its efficacy and purpose. Therefore, this new model is a new model with excellent practicability. In order to meet the application requirements of the new patent, it is required to apply according to law. I hope that the judges will grant the new model as soon as possible to protect the hard work of the creators. If the jury members of the Bureau have any suspicions, please follow the letter and instruct the creators to cooperate as hard as possible.

Claims (5)

A battery health state prediction device mainly includes: a storage box, which is a battery to be tested that is electrically connected to at least one test device; a human-machine interface provided on the storage box; and data collection provided in the storage box The module is used to read the first characteristic information of the battery to be tested; at least one characteristic database set in the storage box; a characteristic comparison module set on one side of the data collection module is set by the The part matching the feature database is extracted from the first feature information as the second feature information; a communication module provided on one side of the feature comparison module and linked to its information; a server whose information is linked to the communication module Device; at least one parameter database set in the server; at least one model database set in the server; and a feature search module set in the server, which is searched by the parameter database. A corresponding first characteristic parameter is obtained when the second characteristic information matches; a model calculation module set in the server is matched with the model database to obtain a corresponding first characteristic parameter according to the second characteristic information operation; two A prediction parameter set in the server and information linking the feature search module and the model calculation module, which are integrated with the calculation of the first characteristic parameter and the second characteristic parameter to obtain an estimated value, And displayed on the man-machine interface; and a test information management platform set on the side of the prediction calculation module. The battery health state prediction device described in item 1 of the scope of patent application, wherein a battery model management module is provided on one side of the test information management platform. The device for predicting battery health according to item 1 of the scope of patent application, wherein the first characteristic information is voltage, current, and time information. The battery health state prediction device according to item 1 of the scope of patent application, wherein the second characteristic information is one of an open circuit voltage, a DC internal resistance, or a time variation constant. The device for predicting the health of a battery according to item 1 of the scope of patent application, wherein the first characteristic parameter and the second characteristic parameter are a capacitance amount, an electric energy capacity, a charged residual amount, an equivalent internal impedance, a coulomb efficiency, and a conversion efficiency. , The remaining number of cycles, the self-discharge law, the cell charge balance, or the cell internal resistance balance.