Aircraft MDO Techonology Research Based On Polynomial Interpolation Surrogate Model

Today, Multidisciplinary Design Optimization (MDO) technology is one of the hotissues in the field of aerospace, in particular, hypersonic vehicle advanced aircraft MDOis particularly important. In this paper, solving the high-dimensional, high complexity,high coupling problems of aircraft MDO is a starting point, from the surrogate modeltechnology, global optimization algorithm two sides, make research in the polynomialinterpolation surrogate model technology, carry out based on local multi-point strategyin high order polynomial interpolation surrogate modeling techniques, quadraticprogramming algorithm based on ant colony algorithm/sequence of the geneticalgorithm/optimization algorithm, surrogate-based model of collaborative optimizationalgorithm and the above method is applied to the concept design of hypersonic vehiclesand hypersonic waverider aircraft aerodynamic shape optimization design, access tovaluable results. The following papers:First, paper analyzes existing surrogate modeling techniques, study polynomialinterpolation surrogate model technology. The basic theories and concepts ofmultivariate polynomial interpolation model, Lagrange polynomial interpolation modelis introduced, and the properly posed quality and error is discussed here. Grobner basistool in the theory of algebraic geometry is used to construct multivariate polynomialinterpolation basis function, to complete structure of multivariate polynomialinterpolation model. Multiple high-order polynomial interpolation surrogate modelingtechniques based on local plus point strategy is carried out, examples tests showedMultiple high-order polynomial interpolation surrogate modeling technology has a highaccuracy of approximation.Secondly, in order to improve the efficiency of the optimization algorithm searchin high-dimensional complex models, as well as optimize accuracy, in the analysis ofthe genetic algorithm, ant colony algorithm, the sequential quadratic programmingalgorithm based on their own characteristics, here proposed, collaborative optimizationalgorithm based on genetic algorithm/ant colony algorithm/sequential quadraticprogramming algorithm. Using global optimization function to test, and validate, globaloptimal characteristics of the collaborative optimization algorithm, as well as theadvantages of a high dimensional complex optimization objective function.Finally, the multivariate polynomial interpolation surrogate model technology andgenetic algorithm/ant colony algorithm/sequential quadratic programming algorithmthree collaborative optimization algorithm, are used to the hypersonic vehicle conceptoptimization design and hypersonic waverider aircraft aerodynamic shape designoptimization, research shows that, based on local add strategy multiple high orderpolynomial interpolation agent model and three collaborative optimization algorithm achieved a very good approximation effect and the optimization effect, the obtainedresults of optimization for hypersonic aircraft multidisciplinary design optimization hassome reference value.In short, in this paper the aircraft MDO surrogate model technology and globaloptimization algorithm is studied, the surrogate model and algorithm for globaloptimization to make improvement. The proposed method is applied to hypersonicflight vehicles multidisciplinary design optimization, obtained valuable results, on flightMDO technology research has a certain theoretical and practical significance.