Comparison of NEVADA Simulation to Monte Carlo Simulation

A Monte Carlo simulation on a business jet engine thermodynamic performance was repeated using NEVADA Simulation in order to compare the accuracy and computing requirements of the different techniques. NEVADA Simulation is a quadrature technique for calculating functions of random variables. In NEVADA (NumErical integration of Variance And probabilistic Dependence Analyzer) Simulation, variables are modeled with mixtures of 4-parameter random variables, called "Continuous Trees". Functions of random variables are calculated using gaussian quadrature. NEVADA Simulation can take advantage of the probabilistic independence in a decision problem while allowing for probabilistic dependence to achieve close to polynomial computational time complexity. The results show that NEVADA Simulation was superior to Monte Carlo Simulation in terms of computational speed and accuracy. To achieve the accuracy of the NEVADA Simulation on specific fuel consumption, which took 3.9 seconds of 80486 time for NEVADA Simulation to solve, 187,000 Monte Carlo runs would be required. These 187,000 runs would require about 5.2 days of Cyber mainframe computer time.