摘要
为减少农业物联报警中门限阈值报警配置任务的工作量,通过对历史数据的样本抽采、分组计算,获得报警门限阈值参考值计算前期的样本,结合正态数值分布情况获得+3σ、+2σ、+1σ、-1σ、-2σ、-3σ,并作为初步报警门限阈值参考值;通过SPC(统计过程控制)过程统计中对均值控制上下限、极差控制上下限的计算获得UCLx、LCLx、UCL_(r)、LCL_(r),并作为进一步报警门限参考阈值,最终生成10档报警门限阈值参考值,再对参考值采取直接投入使用、按平均分权投入使用和结合ISA 18.2报警标准分权投入使用3种方式进行实际报警门限阈值落实工作,对于过大量的历史数据通过预统计的方式来提升计算效率。以此方式智能化、自动化地生成农业物联报警门限阈值,辅助农业物联报警阈值录入配置等工作提升工作效率、减少所需时长。
In order to reduce the workload of threshold alarm configuration tasks in agricultural Internet of Things,the sample of previous stage was calculated based on the reference value of alarm threshold value obtained through sampling of the historical data samples,+3σ,+2σ,+1σ,-1σ,-2σand-3σwere obtained as reference value for preliminary alarm threshold value based on the normal distribution situation;UCLx,LCLx,UCL_(r),and LCL_(r)were obtained as further alarm threshold reference value through the calculation method of controlling upper and lower limit and range control over the average value in the SPC process statistics and the 10 levels of alarm threshold reference value were finally generated.Then the reference value is put into use directly,according to the average decentralization or combined with the ISA 18.2 alarm standard decentralization to implement the actual alarm threshold,for a large number of historical data through pre-statistics to improve the calculation efficiency.In this way,agricultural Internet of Things alarm threshold value was generated in a smart and automatic way.The working efficiency was improved by assisting agricultural Internet of Things alarm value to be input into configuration so as to reduce required time.
出处
《智慧农业导刊》
2023年第5期9-12,19,共5页
JOURNAL OF SMART AGRICULTURE
关键词
农业物联网
报警系统
农业信息化
正态分布
SPC过程统计
报警门限阈值
agricultural Internet of Things
alarm system
agricultural information
normal distribution
SPC process statistics
alarm threshold value