A set of equations used in the solution of n linear algebraic equations in n unknowns, having the general form:
Cramer’s rule states that the jth unknown is given directly by
where Δ is the determinant of the matrix of the coefficients aij (i,j = 1, 2, 3,…,n) and Δj is the determinant of the matrix obtained from the matrix of the coefficients aij by replacing the coefficients of the jth column with the coefficients bj (j = 1, 2, 3,…,n). This assumes that the equations are linearly independent, i.e. that Δ ≠ 0.
The Gaussian elimination technique repeatedly combines different rows such that the original system of equations is transformed into a system of the type shown below. This is achieved by taking the first equation of the original system and multiplying it by a factor such that when added to the second equation the coefficient of x1 is zero. This is repeated but with the third equation and so on until all the coefficients of x1 become zero. With this new set of equations the technique is repeated so that the coefficients of x2 become zero. This is repeated until xn can be solved directly, i.e. αnnxn = βn. Then by a process of back-substitution all the other unknowns can be determined.
Gaussian elimination and Cramer’s rule are used extensively in circuit analysis, where the simultaneous equations are formulated using either node or loop analysis: x1 through xn are the unknown node voltages (or mesh currents), aij (i,j = 1, 2, 3…,n) are known admittances (or resistances), and b1 through bn are known source voltages (or currents).