This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision | ||
cs:localization:kalman:introduction:start [2017/01/16 23:39] Brian Moore First revision complete |
cs:localization:kalman:introduction:start [2017/01/31 02:06] (current) Brian Moore [Section 1 - Linear Least Squares Estimation] |
||
---|---|---|---|
Line 17: | Line 17: | ||
In more formal terms, for some $m$ measurements $Y$ that are linear functions of a system with $n$ unknown states $X$ where $m>=n$. Such systems are said to be //over-determined,// whereby it is impossible to choose values of $X$ that will satisfy every measurement perfectly, and thus a compromise of values of $X$ is chosen that minimizes the total sum of the squares of the error between each measurement | In more formal terms, for some $m$ measurements $Y$ that are linear functions of a system with $n$ unknown states $X$ where $m>=n$. Such systems are said to be //over-determined,// whereby it is impossible to choose values of $X$ that will satisfy every measurement perfectly, and thus a compromise of values of $X$ is chosen that minimizes the total sum of the squares of the error between each measurement | ||
- | $$ X_{est} = \text{arg}\,\min\limits_{\beta}\sum\|y-X\beta||^2 $$ | + | $$ X_{est} = \text{arg}\,\min\limits_{X}\sum\|y-\beta X||^2 $$ |
Given the matrix format: | Given the matrix format: | ||
Line 206: | Line 206: | ||
\end{bmatrix} \\ | \end{bmatrix} \\ | ||
- | X_{est} = (\beta'\beta)^{-1}\beta Y | + | X_{est} = (\beta'\beta)^{-1}\beta' Y |
$$ | $$ | ||