تطوير نماذج التنبؤ الذكي لمؤشرات القيمة المكتسبة في مشاريع محطات معالجة مياة الصرف الصحي

الملخص EN

Despite the difficulty of estimating in the early stages of the life of the projects in light of the limited data and information in the contracting documents, and for the purpose of overcoming this problem and making high-accuracy decisions, traditional and smart techniques have been applied to estimate the earned value indicators with a high degree of accuracy and with minimal errors, in order to empower the stakeholders accurate forecasting of time and cost performance.

therefore, main objective of this research was to compare the accuracy of three techniques for estimating earned value indicators, which are the multiple regression analysis technique, artificial neural networks, and the support vector machine technique for the selected research sample, which are wastewater treatment plants projects in the republic of Iraq.

accordingly, nine intelligent models were developed to estimate scheduling performance index, cost performance index, and overall performance index.

data and information in this study were collected from wastewater treatment plants projects in Iraq.

the number of these projects that were used in building of intelligent models was (32) projects, these data was collected using various methods such as questionnaires, historical data, work study, and field cohabitation.

nine intelligent prediction models were developed using three techniques, namely, artificial neural networks models (ANNM), support vector machines models (svmm), and multiple linear regression models (MLRM)), to predict earned value indicators.

three acquisitions, namely scheduling performance index (SPI), (CPI) and (CSI) in early life cycle of WWTP projects in the republic of Iraq.

results of the comparison between the three techniques, MLR, ANN, and SVM, showed that best techniques used to predict three earned value indicators based on average degree of accuracy (AA%) are ANN, as degree of accuracy for the SPI was (98.93%) and CPI was (98.29%).

and CSI is (95.63%).