Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

system and method to estimate remaining battery life using similar battery usage pattern matching algorithm

IP.com Disclosure Number: IPCOM000238313D
Publication Date: 2014-Aug-18
Document File: 7 page(s) / 112K

Publishing Venue

The IP.com Prior Art Database

Abstract

We propose a system and method to estimate remaining battery life using similar battery usage pattern matching algorithm, which firstly decides the battery decay stage, then computes the similarity based on the battery clustering method to find the most similar using status of other batteries in database, finally predict the battery’s end of life. Compared with some current methods that predict the end of life for a battery energy storage system by relying on data provided by the manufacturer and described in cycles, our method could greatly save the experimental cost and equipment cost, make more precise prediction of rechargeable battery residual service life.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 7

system and method to estimate remaining battery life using similar battery usage pattern matching algorithm


1) Problem background


Power grids often rely on renewable energy resources, which lead to fluctuations in electric power generation causing a blackout in a worst case scenario, especially if wind power or photo-voltaic can't meet the requirements during peak power demand. Thus energy storage systems are required to balance this load to satisfy the demand of the grid and stabilize it. Besides, unexpected end of the battery life will lead to the functional failure of the whole system. However, the system performance of the lithium ion battery with the best quality will be reduced along with the passage of time, aging, and impact factors like environmental condition, operation, etc. Therefore, to set up a reliable and safe grid a reliable prediction system for the battery capacity is needed.


2) Currentmethods


Currently, there are methods predict the end of life for a battery energy storage system, based on historical data of the battery compared to a normal use prediction, which is based on the manufacturers specified life. However, this technique may result in an imprecise estimation of the battery's end of life, because the model relies on the data provided by the manufacturer and is described in cycles. A remaining battery life function which is based on cycles is error-prone, because the life cycles aren't constant during the battery's life.

Some methods focus on building electrochemical or physical battery model to forecast the remaining useful life of the battery. However, this method based on models has obvious shortage, which results in difficulty of wide practical application in industry: the estimation relied on digital circuit virtual component requires expensive equipment like electrochemical impedance spectrometry (EIS), which cost high.


3) Our method


The following Fig1. is our system chart and Fig2 is the process description of our method.

1


Page 02 of 7

Fig.1. System Chart

2


Page 03 of 7

Fig.2. Process description of the method proposed

Figure 1 and Figure 2 show there are 3 steps for our system process:


Step1. Use "battery stage detection model " to decide the stage of the battery capacity decline, and find the most influential factors at this stage. Step2. Use ""battery usage pattern matching method" to compute similarity of usage status among target battery and other batteries in database, in order to find the most similarity battery using status.

Step3. Predict the t...