Biofloc Farming with IoT and Machine Learning Predictive Water Quality System

2022 IEEE International RF and Microwave Conference (RFM) Pub Date : 2022-12-19 DOI:10.1109/RFM56185.2022.10065258

Abdelmoneim A. Bakhit, M. Jamlos, Nura A. Alhaj, R. Mamat

引用次数: 0

Abstract

Biofloc fish farming system depends on full-time monitoring of water quality. The Internet of Things (IoT) can play a vital role in promoting development. However, only a few are able to do stream or real-time predictive analytics at a high cost. Therefore, This article introduces a Biofloc monitoring system based on IoT., which is proficient in performing stream analytics and predictive at a lower cost. This paper evaluates the predictive analytics of the Autoregressive Integrated Moving Average (ARIMA) based on Percentage Error (PE) and Prediction Accuracy (PA). Findings show that ARIMA’s PE is 1.96%, 7.83 %, 1.78%, 12.17%, 4.52% and 0.58%, for DO, EC, pH TDS, Temperature and water volume, respectively which led to achieving higher prediction accuracy (PA) percentage of 98.03%, 92.16%, 98.21%, 87.82%, 95.47% and 99.41% correspondingly.

查看原文本刊更多论文

生物絮团农业与物联网和机器学习预测水质系统

生物絮团养鱼系统依赖于对水质的全天候监测。物联网(IoT)在促进发展方面可以发挥至关重要的作用。然而，只有少数公司能够以高昂的成本进行流或实时预测分析。因此，本文介绍了一种基于物联网的Biofloc监测系统。该公司以较低的成本精通流分析和预测。本文评价了基于百分比误差(PE)和预测精度(PA)的自回归综合移动平均(ARIMA)预测分析方法。结果表明，ARIMA对DO、EC、pH TDS、温度和水量的预测准确率分别为1.96%、7.83%、1.78%、12.17%、4.52%和0.58%，预测准确率分别为98.03%、92.16%、98.21%、87.82%、95.47%和99.41%。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

2022 IEEE International RF and Microwave Conference (RFM)

自引率

0.00%

发文量