Research On Simulation Model Validation Methods Based On Data Consistency Analysis

The model validation in the simulation process is of great significance to ensure the reliability of the simulation results.In the practical occasion,the model verification of large complex system is based on the system evaluation method.Due to the diversity of the actual node model and the corresponding verification method,the data consistency analysis results are varied on form and connotations,which makes it difficult to synthesize and obtain a general credibility of superior nodes.In order to solve the problem,this paper focuses and starts from the point of view of system analysis and establishes an ideal index to describe the similarity of the system.By analyzing the relationship between it and other indexes,the mathematical compatibility and physical meaning are used to derive the conversion algorithm for unity and achieve the conversion of all result of data consistency analysis.The first step is to establish the ideal index.By analyzing the similarity of the simple geometric system and the conditions of similarity description for the general system is proposed.The Hausdorff distance is introduced for describing the similarity between systems and a normalized one for linear system is proposed in this paper to solve the problem of uniform.Later it has been extended to be System Angle Metric(SAM)for general systems with unchanged meaning and a kind of algorithm for SAM under the restriction of inputs has been proposed for practical calculation.Then,for the unity and transform of various results of data consistency analysis,the appropriateness of SAM as the standard credibility has been discussed and so is the criteria of the consistency of different data types.The nature of TIC(Theil Inequality Coefficient)has been analyzed and its conversion to the SAM has been completed.For the window estimation and the maximum entropy spectral estimation,with their assumptions analyzed respectively,their expected energy differences are obtained from the statistics and their conversions are done by physical way.Finally,the design and implementation of the auxiliary tool to achieve the former function is done with software programming.Through the analysis of the specific validation data,the results show that the SAM can reflect the system difference preferably and turn the consistency of different method results to an appropriate range for unity.