Estimation method of simulation model parameters based on credibility optimization

The ranges of simulation model parameters can be obtained, but their accurate values are difficult to determine sometimes. To solve the problem, an optimization method of simulation model parameters based on the credibility optimization and the work of simulation model validation was proposed. The Latin hypercube experiment was carried out on the simulation model, and the credibility of simulation model was evaluated under each experimental point. Then, the values of simulation model parameters and credibility of simulation models under all experimental points were taken as training samples. The relational model between the simulation model parameters and the credibility were obtained using support vector machine. So, the dynamic optimization problem of simulation model parameters was transformed into a static optimization problem of function. Finally, the estimation values of simulation model parameters optimizing the credibility were obtained via the genetic algorithm. In the application, the accurate parameters values of the simulation model of a missile guidance system were determined effectively.