摘要
针对型号AIT合理性不能定量评价的现状,提出将其作为马尔可夫过程来处理的方法。以经典马氏四元组为基本框架建模。采用领域知识训练模型,提高模型优化能力和适用性。通过报酬的期望平均来表征AIT的优劣,以此判断其合理性。应用实例表明,方法效果明显,结合自动化手段,具有可观的应用前景。
Aimed at the actuality that project AIT rationality can't be evaluated quantificationally,put for- ward the method that dealing with it as a Markov process. Constructed the model with the frame of classi- cal Markov four- element group. Trained the model with domain knowledge to enhance its optimizing abili- ty and applicability. Reflected the advantage or disadvantage of AIT with expectation average of reward and evaluated its rationality. The example showed that the method is effective. If running the model auto- matically ,it's coming.
出处
《航空计算技术》
2016年第2期72-75,共4页
Aeronautical Computing Technique