Analysis and verification of suitable li-ion models as basis for a self-adaptive battery observer
View/Open
REPORT_475.pdf (1,991Mb) (Restricted access)
Budget_54.pdf (552,6Kb) (Restricted access)
Annex1.xlsx (20,22Kb) (Restricted access)
Annex2.txt (17,79Kb) (Restricted access)
Cita com:
hdl:2117/183099
Document typeBachelor thesis
Date2018-01-10
Rights accessRestricted access - author's decision
All rights reserved. This work is protected by the corresponding intellectual and industrial
property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public
communication or transformation of this work are prohibited without permission of the copyright holder
Abstract
The purpose of this project is to create and simulate a self-adaptive battery
observer using Matlab Simulink. The first phase of the project involves the
research of a suitable lithium-ion cell model to cover a wide range of battery
based technologies. After selecting the optimum one, the same real lithium-ion
cell must undergo experimentation to acquire real current and terminal voltage
data to be later on used in the simulation.
Afterwards, a model to describe the internal dynamics of the lithium-ion
processes must be chosen. The parameters of the model state-space equation
must be estimated at every selected state of charge point, preferably more
populated at high and low state of charge to improve its accuracy. To estimate
the parameters of the selected lithium-ion cell, a pulse discharge test is
performed on the cell simulation in Matlab Simulink. The acquired data is
processed using the Curve Fitting tool. Look-up-tables are populated using the
acquired parameter data as a function of the state of charge.
A main model named Self-adaptive battery observer contains 3 different blocks:
Coulomb Counting + Extended Kalman Filter correction block, Lithium Cell block
and Extended Kalman Filter block. Coulomb Counting + Extended Kalman Filter
correction block calculates the actual state of charge using the experimentally
measured current while the output state of charge will be calibrated using the
estimated state of charge provided by the Extended Kalman Filter.
The actual state of charge enters in the Lithium Cell referenced model and is
used as the breakpoints at each sample time for the look-up-table mentioned
before. Within this block there is another subsystem that uses the parameter
estimations provided by the look-up-tables to simulate the transient voltage. This
transient voltage is subtracted to the experimentally measured terminal voltage to
obtain the calculated open circuit voltage of the system which is compared to the
measured from the current pulse discharge test and the error is used in the
covariance settings of the Extended Kalman Filter. The Kalman Filter block has
two function inputs. The state space function and the measurement function. The
first one provides the actual state of charge and the second one the calculated
open circuit voltage. Due to the non-linear relationship between OCV and SOC
the Extended Kalman Filter has to be used instead of the regular Kalman Filter.
Extended Kalman Filter estimates the SOC that is used in the Coulomb Counting
block to calibrate the actual SOC.
In the end, the code generated should be able to be integrated in the basic
software of a Battery Management System (BMS).
SubjectsLithium ion batteries, Computer algorithms, SIMULINK, Kalman filtering, Bateries d'ió liti, Algorismes computacionals, SIMULINK, Kalman, Filtratge de
DegreeGRAU EN ENGINYERIA EN TECNOLOGIES INDUSTRIALS (Pla 2010)
Files | Description | Size | Format | View |
---|---|---|---|---|
REPORT_475.pdf![]() | 1,991Mb | Restricted access | ||
Budget_54.pdf![]() | 552,6Kb | Restricted access | ||
Annex1.xlsx![]() | 20,22Kb | Microsoft Excel 2007 | Restricted access | |
Annex2.txt![]() | 17,79Kb | Text file | Restricted access |