【摘要】 在复杂仿真系统模型验证的有效方法的研究中,神经网络可以做为复杂仿真系统校核、验证和确认的主要应用工具和手段。首先介绍了传统仿真系统的验证方法,指出了传统验证方法存在的问题。提出了基于神经网络的复杂仿真系统验证方法。采用神经网络模型识别和验证,根据神经网络时间序列预测和神经网络的最大熵谱估计验证方法。最后给出了具体仿真应用示例并分析了几种验证方法的优缺点,可以为解决复杂仿真系统验证问题提供了一条新的途径。更多还原

【Abstract】 The applications of neural network in VV&A for complex simulation system are discussed and studied in this paper.Firstly,the classical validation methods of simulation system are introduced and its disadvantages are pointed out.Secondly,the validation methods of simulation system based on the neural network are proposed.There are validation methods of model identifying,time sequence forecasting and estimating of max entropy chart.Finally,the examples are given and advantages and disadvantages of the validation methods are analyzed.The new approach for solving validation problem of complex simulation system is provided.更多还原