Join 60,000+ competitive exam aspirants
As per least square estimation, the estimated value of X (XтВСтВЫтВЬ) is represented by (Where Z is measurement vector and H is the linear relation between Z and X)
XestтАЛ=(HTH)тИТ1HTHZ
XestтАЛ=(HTH)тИТ1HTZ
XestтАЛ=(HTH)тИТ1HZ
None of above
XestтАЛ=(HTH)тИТ1HTZ
In least square estimation for a linear system defined by Z=HX+╧╡, the goal is to minimize the sum of squared residuals J=╧╡T╧╡=(ZтИТHX)T(ZтИТHX). Setting the derivative of J with respect to X to zero yields the normal equations, which result in the estimator XestтАЛ=(HTH)тИТ1HTZ.
In least square estimation for a linear system defined by Z=HX+╧╡, the goal is to minimize the sum of squared residuals J=╧╡T╧╡=(ZтИТHX)T(ZтИТHX). Setting the derivative of J with respect to X to zero yields the normal equations, which result in the estimator XestтАЛ=(HTH)тИТ1HTZ.
J=(ZтИТHX)T(ZтИТHX) тАФ Objective function to minimize
XestтАЛ=(HTH)тИТ1HTZ тАФ Least square estimator solution
The principle relies on projecting the measurement vector Z onto the column space of the matrix H. By minimizing the Euclidean norm of the residual vector, we obtain the Moore-Penrose pseudoinverse (HTH)тИТ1HT. This method assumes that HTH is non-singular and invertible, which corresponds to the system being observable.
The estimator is based on the minimization of the sum of the squares of the errors (residuals).
It assumes the errors are independent and identically distributed with zero mean.
If the matrix H is not of full column rank, HTH is singular and the unique inverse does not exist.
This method is widely used in power system state estimation for processing SCADA data.
Provides a mathematically optimal solution when errors are Gaussian.
Computational efficiency for linear systems.
Sensitive to outliers (bad data) in the measurement vector Z.
Requires high computational overhead for extremely large sparse matrices.
Power system state estimation (monitoring bus voltages and angles).
Signal processing and sensor fusion.
The term (HTH)тИТ1HT is known as the left Moore-Penrose pseudoinverse of H.
Option C is dimensionally inconsistent as (HTH)тИТ1 is n├Чn and H is m├Чn, making their product undefined unless m=n.
B is correct тАФ The least square estimator for a linear model Z=HX is derived from the normal equations, resulting in XestтАЛ=(HTH)тИТ1HTZ.
Always ensure dimensions match in matrix algebra; in (HTH)тИТ1HTZ, if H is m├Чn, then (HTH) is n├Чn, and the total expression simplifies to an n├Ч1 vector.