基于物联网的农机作业轨迹时空数据建模与挖掘方法

The rapid development of modern agriculture in China has put forward higher requirements for agricultural machinery operation and intelligent management. The application of Internet of things (IOT) as the representative of the new telematics in agricultural machinery equipment produces a great amount of agricultural machinery operation trajectory data. In order to reveal practical value of these data and improve the quality and efficiency of agricultural machinery operation, this project proposed to carry out the research on spatio-temporal data modeling and data mining method for agricultural machinery operation trajectories based on IOT. In the project spatio-temporal data model will be built based on temporal and spatial variation process of agricultural machinery operation. According to the characteristics of spatial agglomeration, time series and spatio-temporal density, trajectory data mining will be investigated including automatic segmentation algorithm of field operation area, in-field operation path splitting algorithm and automatic recognition algorithm of operation state. The scientific results from the research will be evaluated by the development and testing agricultural machinery operation monitoring management and data analysis IOT system in demonstration application. The project intends to enhance the value of agricultural machinery IOT system, provide theoretic basis and data support for precise management and scheduling service, accelerate practicability and industrialization of IOT technology in the field of agricultural machinery intelligent equipment.

我国现代农业快速发展对农机集群化作业和智能化管理提出了更高的要求，以物联网为代表的新一代信息技术与农机装备技术的融合运用，产生了海量农机作业轨迹信息。本项目提出开展基于物联网的农机作业轨迹时空数据建模与挖掘方法研究，旨在揭示这些数据的应用价值，为农机作业提质增效。通过分析抽象农机作业时空变化过程，建立适用的时空数据模型。通过分析农机作业轨迹数据的空间聚集特征、时间序列特征和时空密度特征，研究作业区域自动识别与分割算法、地块内作业路径分割与作业指标测算方法、农机运营状态识别与效率评估方法。此外，在理论研究基础，研发农机作业监控管理与数据分析物联网系统，并通过系统试验和示范应用进行模型修正和算法检验。本研究的成功开展，将有助于提升农机物联网的应用价值，为农机精准管理和调度服务提供理论依据和数据支持，推动物联网技术在农机智能装备领域的实用化和产业化进程。

基于物联网的农机集群化作业和智能化管理产生了海量作业轨迹数据。本项目开展了农机运行轨迹时空数据建模与挖掘方法研究。主要研究成果包括：（1）建立农机作业轨迹数据分析架构，针对轨迹大数据集中存在的重复数据、缺失属性、属性范围异常、丢失数据、漂移轨迹、停歇轨迹等异常情况，进行相应的数据清洗算法研究，以保证数据质量；（2）提出了农机作业轨迹与道路转移轨迹分割、异常轨迹识别、地块内直行作业轨迹与转弯轨迹分割、作业面积测算等轨迹数据挖掘方法，并进行了海量数据的分布式计算测试，为科学测算有效作业面积、作业达标率，客观评估运营效率、时间利用率等指标提供了理论依据；（3）研发了农机作业轨迹数据分析系统，能够完善现有的农机作业监管物联网系统，实现了农机有效作业面积测算、作业质量评价、运营状态识别与效率综合评估。本项目的研究成果有助于提升农机物联网的应用价值，为农机精准管理和调度服务提供理论依据和数据支持。