为进一步提高效能评估的可信性及智能化水平,提出了一种基于灰色AHP理论的数据链系统BP神经网络效能评估模型。从数据链系统性能结构出发,针对数据链信息优势的特点,建立了数据链系统作战效能评估指标体系。通过专家组依据标准打分的方式,采用AHP确定指标的权重向量,考虑借助灰色理论白化指标,最后与不同灰类分效能值聚合生成综合效能评价值从而建立了效能输出模型。构造三层BP神经网络模型,以专家的知识与经验作为输入,以灰色AHP综合效能评价值作为输出训练网络。最后通过实例验证,由于该模型运用BP神经网络非线性拟合的优势继承并有效聚合了专家的知识与经验,结果与专家评定一致,从而论证了模型的有效性及可靠性,为数据链系统效能评估提供了智能化的算子,在数据链的发展和应用中发挥重要作用。

In order to improve the level of credibility and intelligence of efficiency evaluation, a BP neural network effectiveness evaluation frame of data link system on the basis of gray AHP theory was proposed.Considering performance and structure of data link system, according to the characteristics of advantage of data link information, the operational effectiveness assessment index system of data link was established.Through expert marking by standard, using AHP to determine the weight vectors of index, considering to use the albino indicators index of gray theory aggregate with different gray class efficiency value to generate comprehensive efficiency evaluation so that the output efficiency model was established.A three-layer BP neural network model was structured with the expert's knowledge and experience as input, and the grey AHP comprehensive efficiency evaluation value as the output of the training network.Finally, the experiment validated that the result was consistent with the expert evaluation.Since the model uses the advantage of nonlinear fitting inheritance of BP neural network and aggregates the expert's knowledge and experience effectively, therefore, it demonstrates validity and reliability of the model and provides intelligent operator for the data link system effectiveness evaluation.The model plays an important role in development and application of the data link.