How to put a set of pre-registered values as input for a kalman filter ?

Hello,

I have a set of pre-recorded values here

X = [10, 15, 25]

vX = [3, 11, 12]

I have my A B C D Q and R matrices already.

I have a system and kalman filter invoked as followed

sys = ss(A,B,C,D,1,'StateName',{'Position' 'Velocity'},'InputName','Force');

[kalmf,L,P] = kalman(sys,Q,R);

How can I 'plug' my input in the kalman filter ?

Answer 1

[Anshuman]

Hi Elouan,

To "plug" your input into the Kalman filter, use the lsim function in MATLAB, which simulates the output of a linear system given an input signal. In this case, you will use lsim to simulate the output of the Kalman filter given your input signal.

Here's an example code that shows how to simulate the output of the Kalman filter with your input signal:

% Define the input signal
t = linspace(0, 1, 100); % Time vector
u = interp1(X, vX, t, 'linear'); % Interpolate input signal using pre-recorded values

% Simulate the output of the Kalman filter
y = lsim(kalmf, u, t);

% Plot the input and output signals
figure;
plot(t, u, 'b', t, y, 'r');
legend('Input', 'Output');
xlabel('Time');
ylabel('Position');

In this code, the interp1 function is used to interpolate the input signal u using the pre-recorded values X and vX. The lsim function is then used to simulate the output of the Kalman filter (y) given the input signal u and the time vector t. Finally, the input and output signals are plotted for visualization.

Note that the Kalman filter assumes that the input signal is a zero-mean white noise process, so you may need to preprocess your input signal to remove any biases or trends before using it with the Kalman filter.

Hope it helps!