多形态卷积并行神经网络建立效能评估指标体系  

An Effectiveness Evaluation Indicator System Based on Multi-Scale Parallel Convolution Neural Network

作　　者：李辰 陈浩 李建勋[1] LI Chen;CHEN Hao;LI Jianxun(Department of Automation,Shanghai Jiao Tong University,Shanghai 200000,China;Military Representative Bureau of Naval Equipment Department in Shanghai,Shanghai 201000,China;Institute of Aerospace Engineering,Beijing 100000,China)

机构地区：[1]上海交通大学自动化系,上海200000 [2]海装驻上海地区军事代表局,上海201000 [3]航天工程研究所,北京100000

出　　处：《电光与控制》2021年第11期31-34,93,共5页Electronics Optics & Control

基　　金：国家自然科学基金(61673265);民机专项(MJ-2017-S-38);航空科学基金(20170157001);CEMEE重点实验室开放课题(2019G 0302)。

摘　　要：针对复杂装备建立效能评估指标体系的难题,引入了一维卷积神经网络模型,构建了多形态卷积核并行处理框架;从多个观测角度自适应学习装备原始运行数据,再对数据蕴含的特征加以整合,进而建立新的效能评估指标体系,避免了主观因素和计算难度的束缚,为研究效能评估智能化方法奠定了基础。Towards the establishment of effectiveness evaluation indicator system for complex equipment, the one-dimensional convolution neural network is introduced, and a parallel processing framework with multi-scale convolution kernels is constructed.Then, the original data of equipment is learned adaptively from multiple dimensions, then the characteristics are integrated into a new effectiveness evaluation indicator system, which avoids limitation of subjective factors and computational difficulty, and lays a foundation for intelligent methods of effectiveness evaluation.

关 键 词：一维卷积神经网络 多形态卷积核 指标体系 效能评估 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]