摘要
For a long time,the observations’adjustment concepts are doing based on the principle of Least Squares(LS),which so far,is the most popular technique that is widely used for such purposes.In the current study,the author introduces a new vision of the measurements’adjustment that depends on the principle of the artificial neural network backpropagation(ANNB).However,the novel approach of the adjustment in the current study is based upon the principle of the measurement’s precision,which will be represented through the value of the variance(σ^(2)).Thus,the determination of the required residuals for the adjustment procedure can be derived from the variances of observations and the value of geometrical closure error.The evaluation of the proposed method of adjustment ANNB is done according to the parametric method of LS adjustment for two selected cases of observations,representing the equal and non-equal precision observations.The obtained results of adjustment using the new technique(ANNB)are fully agreed with the results of LS for the two evaluation cases.These results indicate good compatibility between the ANNB and the LS as two approaches for observations’adjustment.
