| System simulation technology is increasingly being used in various fields. The users are all rightly concerned with whether a model and its results are"correct". The method to address that concern is validation of models, which requires validation methodology and tools to support. This paper dose further research on validation methodology and tools, the main contents are as follows:Firstly, the dissertation's background and purposes are expatiated. Three problems in current validation fields of simulation models are proposed. The history and status of validation techniques and tools in the home and abroad are analyzed and the structure of this paper is presented.Secondly, validation approaches used in both static and dynamic validation are concluded. Theories, advantages, disadvantages and using conditions for each category of methods are highly summarized.Thirdly, the application of knowledge engineering used in validation is analyzed. With theory knowledge, rules used to select methods in the process of static data validation are designed. With experience knowledge, aiming at dynamic validation when the real data are unavailable, a new method, namely, features extract and match is proposed and detailed.Fourthly, a validation system is designed on the basis of the above theory. The framework of the system is established and a brief introduction for the key components is given. Validation database, expert knowledge base, methods base and rules base are constructed. And six validation flows are designed and described in details.Finally, the validation system is developed with the tool of Visual C + +6.0. Experiments show that the validation tool is effective.In conclusion, theories and using conditions for each category of methods are summarized. The application of knowledge engineering is analyzed. Besides, a validation system is designed and developed. This paper's research findings make a positive contribution to the validation of simulation models. |
PDF Full Text Request