Journal of Computer Applications
Next Articles
Received:
Revised:
Online:
Published:
Supported by:
刘艺1,杨国利2,郑奇斌2,李翔2,周扬森2,陈德鹏2
通讯作者:
基金资助:
Abstract: Sensors are the basis for unmanned systems to perform intelligent actions. The fusion of multi-sensor data can enhance intelligent perception and autonomous decision-making capabilities of unmanned systems, and improve their reliability and robustness. Data fusion of unmanned systems encounters many challenges such as diverse sensor types, heterogeneous data formats, real-time fusion needs, analysis of varied and complex data, fast evolving models. Traditional methods of developing fusion models through customization on edge equipment and approaches based on fusion platform running on cloud devices are difficult to apply in these cases. To address these problems, the paper proposes a data pipeline platform for data fusion of intelligent sensors on un-manned systems. The architecture has capabilities to support automatic data transformation, seamless algorithm integration, dynamic model configuration, and agile service evolution to achieve quick response to different tasks. Based on the analysis of data fusion process and techniques, it characterizes the key components and functions of the proposed pipeline architecture, discusses relevant significant technologies, gives its running way and actual case, and points out research directions for future development.
Key words: data fusion, unmanned systems, pipeline, sensing data, machine learning
摘要: 传感器是无人系统智能化行动的基础,通过融合多传感器的数据能够增强无人系统的智能感知和自主决策能力,提升无人系统的可靠性和鲁棒性。无人系统的数据融合面临传感器类型多样,数据格式异构,数据融合分析的实时性强和算法模型种类复杂、更新演化快等挑战,传统定制化开发前端融合模型和基于后端融合平台的方法难以适用。为此,提出一种面向数据融合的流水线平台,以支持数据自动转换、算法灵活组合、模型高可配置、功能快速迭代,且能面向任务,动态、快速构建数据融合模型,提供信息服务。在剖析无人系统数据融合流程和技术体系的基础上,设计流水线框架及其关键功能构件,分析亟待突破的关键技术,给出框架的运行方式和实际案例,探讨未来的发展方向。
关键词: 数据融合, 无人系统, 流水线, 感知数据, 人工智能
CLC Number:
TP391
刘艺 杨国利 郑奇斌 李翔 周扬森 陈德鹏. 无人系统数据融合流水线架构设计[J]. 《计算机应用》唯一官方网站, DOI: 10.11772/j.issn.1001-9081.2023081184.
0 / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: https://www.joca.cn/EN/10.11772/j.issn.1001-9081.2023081184