Uncertainty analysis model Since the model of the system only seeks to show the reality, then inevitable existence of simplifications, assumptions and idealizations of the complex processes and phenomena occurring in the system. The consequence of these simplifications and idealizations are uncertainties in the final results obtained in the process application of the model. Uncertainties due to incomplete models arise from the fact that in constructing models of systems analyst had not foreseen by some of the simulated processes in system. The second type of uncertainty is associated with inadequate models. Even in those cases where the model takes into account all the peculiarities of the existence and development of systems, the logical sequence of events and features functioning of the systems incorporated into the model does not accurately reflect reality. The third type of uncertainty – the uncertainty of initial parameters. Parameters of different models are uncertain.
The reason for this is failure data used in the statistical estimation of input parameters, the inability to accurately describe the behavior of staff working within the system under analysis, the presence of the assumptions made in compilation model. Sensitivity analysis of model sensitivity analysis model, called the procedure for evaluating the effect of tolerances on the input parameters of its output characteristics. Conduct sensitivity analysis as follows manner: define input parameter deviation to the right and left side of its mean value and fixed, as this changes the output values of the characteristics of the model. As the deviation is usually taken standard deviation. Set a realistic model of realism, then, to answer the question whether the model is so special cases for which there is already evidence.
The efficiency goal performance analysis model to determine how the model is practical and easy to use. First, the model should provide a result within a reasonable time. Second, labor costs and resources required for operation model must fit within the limits of computer time and the salary fund. The condition of practical expediency. The next aspect of model validation involves analyzing the assumptions and hypotheses, adopted in the model. At this stage, verify that the quality of the estimated model and its properties in terms of real impact of external disturbances and parameters. Source: Student Portal – StudProspekt. A lot of laboratory work papers.