battery soc estimation using EKF

Question

Rihane boutaina 2021-4-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/790704-battery-soc-estimation-using-ekf

回答： Sabin 2023-9-22

hello,

please i need help ! i'm trying to estimate the battery state of charge using kalman filtering ! but actually i have no clue :/

i've written this programme but it doesnt work ?!

any one can help me please ! i'm feeling really stuck :(

clc
tic
% Enter input data
i=tryfilekalmanS1(:,3);
vt=tryfilekalmanS1(:,2);
soc=tryfilekalmanS5(:,1);
t=tryfilekalman(:,1);
OCV=tryfilekalmanS2(:,2);
OCV=table2array(OCV);
current=table2array(i);
voltage=table2array(vt);
SOC=table2array(soc);
Time=table2array(t);
%battery parameters 
Cn=2.6; %battery nominal capacity
deltat=5; %time step
R=0.1174; %internal resistance
taud=4.4925;
eff=1; %coulomb effeiciency (for discharging case it takes 1)
sigmaW =30; %process noise covariance
sigmaV=20; %sensor noise covariance
F=(-2.6/taud);A=-2.6;B=1;C=0.1;D=1; %plant definition matrices
maxIter=16; %number of iterations to execute simulation
% Providing the value of battery parameters (ocv(soc) curve fitting
% equation)
M0=277512.69155; % Coefficient from OCV fitting curve
M1=-110959.5326; % Coefficient from OCV fitting curve
M2=14761.90214; % Coefficient from OCV fitting curve
M3=-653.33323; % Coefficient from OCV fitting curve
I  = current;
V=voltage;
for k=1:maxIter
    %initialization
xtrue(1)=1; %initialise true system initial state
xhat(1)=0; %initialise kalman filter initial estimate
sigmaX(1)=0.005; %initialise kalman filter covariance
u=7.7783; %unknown initial driving input
    %time update
    xhat=A*(M0 + M1.*OCV(k) + M2.*OCV(k)^2+ M3.*OCV(k)^3)+w; 
    sigmaX=A*sigmaX*A'+sigmaW;
    %
    u=OCV; 
    w=0.012; v=0.023; %noises
    ytrue = C.*xhat+D.*u+v; 
    xtrue = A*xtrue+B*u+w; 
    %Estimate system output
    yhat=C*xhat+D*u;
    %Compute kalman gain matrix
    sigmaY=C*sigmaX*C'+sigmaV;
    L=sigmaX*C'/sigmaY;
    %measurement update
    xhat=xhat+L*(ytrue-yhat);
    sigmaX=sigmaX-L*sigmaY*L';
    %store
    xstore(1:16,k)= xtrue(1:16);
    xhatstore(1:16,k)=xhat(1:16);
    sigmaXstore(1:1,k)=sigmaX(1:1);
    ytruestore(1:17,k)=ytrue(1:17);
    yhatstore(1:17,k)=yhat(1:17);
end
toc
%code plot
figure(1);
plot(1:maxIter,xstore(1:maxIter)','r',1:maxIter,xhatstore',1:maxIter,xhatstore(1:maxIter)-xstore(1:maxIter)','b');
grid;
legend('true','estimate','error');
title('soc estimation');
xlabel('time'); ylabel('soc');