IoT 센서 데이터 기반으로 산불 발생을 예측하는 머신러닝 알고리즘 = Machine Learning Algorithm to Predict Wildfire Occurrence Based on IoT Sensor Data

다국어 초록 (Multilingual Abstract)

Wildfires destroy forests, a resource for mankind, and cause great damage to wild animals and plants. These wildfires are caused by weather conditions and human activities. Various researches are being conducted in each country to detect and effectively suppress wildfires. In this study, a machine learning model that can predict wildfires was proposed. In the study, the proposed model was validated using a reliable forest fire database. For verification, data set collection, data preprocessing, model training, and model testing were performed. ANN was used to create a forest fire prediction model. A website was built to predict wildfires and take quick action using the learned model. As a result of the experiment using the proposed model, it was confirmed that the accuracy was 87%.

국문 초록 (Abstract)

산불은 인류의 자원인 산림을 파괴하고 야생 동식물에 많은 피해를 주고 있다. 이와 같은 산불은 기상 조건 및 인간의 활동에 의해서 발생하고 있다. 각 국가별로 산불을 감지하고 효율적으로 진압하기 위해서 다양한 연구를 진행하고 있다. 본 연구에서는 산불을 예측할 수 있는 머신 러닝 모델을 제안하였다. 연구에서는 신뢰할 수 있는 산불 데이터베이스를 사용하여 제안하는 모델을 검증하였다. 검증을 위해서 데이터 세트 수집, 데이터 전 처리, 모델 학습, 모델 테스트 등을 수행하였다. 산불 예측 모델을 만들기 위해서 ANN을 활용하였다. 학습한 모 델을 이용하여 산불을 예측하고 빠른 조치를 하기 위해서 웹 사이트를 구축하였다. 제안하는 모델을 이용하여 실 험한 결과 87%의 정확도를 보이는 것을 확인하였다.

참고문헌 (Reference)

1 유영중 ; 문상호 ; 심상진 ; 박성호, "딥러닝 기법을 이용한 웹 카메라 입력 자동차 번호판 인식" 국제차세대융합기술학회 4 (4): 565-572, 2020

2 송다은 ; 전동빈 ; 하태성 ; 이형원 ; 김경이, "딥러닝 기반 이미지 필터 간 한국인 표정 분류 성능 비교" 국제차세대융합기술학회 6 (6): 767-774, 2022

3 "scikit-learn"

4 "pandas"

5 R. Nair, "Wildfire: Approximate synchronization of parameters in distributed deep learning" IBM 61 (61): 7:1-7:9, 2017

6 Sarah Masoumi, "Reflection Measurement of Fire Over Microwave Band: A Promising Active Method for Forest Fire Detection" Institute of Electrical and Electronics Engineers (IEEE) 21 (21): 2891-2898, 2021

7 K. Doyle, "New fire danger rating system set to be trialled this summer to cope with new extremes"

8 Canadian National Fire Database, "National burned area composite"

9 Michael L. Mann, "Incorporating Anthropogenic Influences into Fire Probability Models: Effects of Human Activity and Climate Change on Fire Activity in California" Public Library of Science (PLoS) 11 (11): e0153589-, 2016

10 K. Bonsor, "How wildfires work"

11 Paul D. Pickell, "FuelNet: An Artificial Neural Network for Learning and Updating Fuel Types for Fire Research" Institute of Electrical and Electronics Engineers (IEEE) 59 (59): 7338-7352, 2021

12 P. Moore, "Forests and wildfires:Fixing the future by avoiding the past" World Forestry Congress

13 Josué Toledo-Castro, "Forest Fire Prevention, Detection, and Fighting Based on Fuzzy Logic and Wireless Sensor Networks" Hindawi Limited 2018 : 1-17, 2018

14 Casey A. Graff, "Forecasting Daily Wildfire Activity Using Poisson Regression" Institute of Electrical and Electronics Engineers (IEEE) 58 (58): 4837-4851, 2020

15 VIRGINIA H. DALE, "Climate change and forest disturbances: Climate change can affect forests by altering the frequency, intensity, duration, and timing of fire, drought, introduced species, insect and pathogen outbreaks, hurricanes, windstorms, ice storms, or landslides" Oxford University Press (OUP) 51 (51): 723-734, 2001

16 Rob Skakun, "Area burned adjustments to historical wildland fires in Canada" IOP Publishing 16 (16): 064014-, 2021

17 Jianmei Zhang, "ATT Squeeze U-Net: A Lightweight Network for Forest Fire Detection and Recognition" Institute of Electrical and Electronics Engineers (IEEE) 9 : 10858-10870, 2021

18 Hao Liang, "A Neural Network Model for Wildfire Scale Prediction Using Meteorological Factors" Institute of Electrical and Electronics Engineers (IEEE) 7 : 176746-176755, 2019

19 H. S. Lim, "2020 forestry statistics"