Detail Description of Kalman Filter | Real Time System | BSc.CSIT Sixth Semester

Download our Android App from Google Play Store and start reading Reference Notes Offline.

kalman filterKalman Filter – Detail Explanation,
Real Time System Notes | Sixth Semester,
BSc.CSIT | Tribhuvan University (TU)

Kalman Filter
Kalman filter is a commonly used means to improve the accuracy of measurements and to estimate model parameters in the presence of noise and uncertainty. Consider a simple monitor system that takes a measured value yk every sampling period ‘k’ in order to estimate the value xk of a state variable. Suppose starting from time 0, the value of this state variable is equal to a constant x. Because of noise, the measured value yk is equal to x + \epsilon_{k} where \epsilon_{k} is a random variable whose average value is 0 and standard deviation is \sigma_{k}. The Kalman filter starts with the initial estimate \tilde{x}_{1}=y_1 and computes a new estimate each sampling period. Specifically, for k>1, the filter computes the estimate \tilde{x}_{k} as follows:

\tilde{x}_{k}=\tilde{x}_{k-1}+K_k(y_{k}-\tilde{x}_{k-1}) —————(i)
where K_{k}=\frac{P_k}{\sigma_{k}^2+P_k} ———————–(ii)
Here, K= Kalman gain and Pk = variance of the estimation error \tilde{x}_{k}-x
and P_{k}=E[(\tilde{x}_{k}-x)^2]=(1-K_{k-1})P_{k-1} ——————(iii)

This value of kalman gain gives the best compromise between the rate at which Pk decreases with k and the steady-state variance i.e; Pk for large k.

In a multivariate system, the state variable xk is an n-dimensional vector where ‘n’ is the no. of variables whose values define the state of the plant. The measured value yk is an n' -dimensionalvector, if during each sampling period, the readings of n' sensors are taken. Let A denote measurement matrix; which is n \times n' matrix that relates n' measured variables to the n state variables. In other words,

y= Axk + ek
where ek = additive noise in each of the n' measured values
Equation (i) becomes an n-dimensional vector equation
\tilde{x}_{k}=\tilde{x}_{k-1}+K_{k}(y_{k}-A\tilde{x}_{k-1})
Similarly kalman gain Kk and variance Pk are given by matrix version of equation (ii) and (iii). So the computation involves a few matrix multiplication and addition and one matrix inversion

(Visited 192 times, 1 visits today)

Posted By : Digvijay | Comment RSS | Category : Sixth Semester
Tag :

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

Wordpress DMCA
Community | Toolbar | Android App | Founder/Developer : Hari Prasad Chaudhary | CSIT Portal Manager : Digvijay Chaudhary