Abstract

The Mixed Exponential Distribution is one of the important probability distributions used in the celebrated engineering and statistical fields, and this research focused on the comparison between some common estimation methods represented by the method of Maximum likeihood, method of least squares and method of of Percentiles estimations, and that the research methodology depends on a study Theoretically, the estimation methods were derived in detail to arrive at formulas for the estimations of these methods for the parameters of the mixed exponential distribution. The data of the energy consumed from electricity was used, which represents the actual monthly consumption of electric energy in Muqdadiya district for the years (2017,2018,2019) in megawatts, knowing that the consumption in all Its types (commercial and governmental) and then using the mean square error (MSE) criterion to compare these methods. ) in the year (2017) and (0.1221) in the year (2018), and in the year (2019), the fractional estimators method was the best as it had the lowest mean error (0.0625) This indicates that it is the best method of estimation for this year, and after testing the data used for the consumed amount of electrical energy in the station and using the Boxplot in the (spss) program, it was reached that there was only one abnormal value in the year (2017) and the absence of any abnormal values in The years (2018 and 2019), the researcher recommended that the use of ordinary methods such as the moment and new Bayes method and the use of different distributions and comparison with them and the use of good-matching tests to reach the best distribution among them to find out the lowest consumed energies of electrical energy.