Development of a prediction model of particulate matter less than 10 um using hybrid support vector regression with fuzzy logic theory based on quantitative decision algorithm
Abstract:
This thesis proposes a prediction model for forecasting Particulate Matter with
aerodynamic diameter up to 10 μm (PM10). PM10 is targeted because it constitutes a major concern for the air quality of urban in Thailand. This thesis describes the development of a prediction model based-on Hybrid Support Vector Regression Method with Fuzzy Logic for
forecasting PM10 in Bangkok. This research compares the performance of the model developed by
Support Vector Regression (SVR) and Fuzzy Logic Theory Based on Quantitative Decision
Algorithm (FQDA) with the SVR, Back-propagation Neural Network and ANFIS (Adaptive
Network-Based Fuzzy Inference System) Model. The developed models are used to establish the
relationships of PM10 with meteorological variables including global radiation, net radiation, air
pressure, rainfall, relative humidity, temperature, wind direction, wind speed, the air quality
concentrations of Carbon monoxide, Ozone, Nitrogen dioxide, and Sulfur dioxide. The data sets
examined in the current study were collected by monitoring station operated by Pollution Control
Department of Thailand corresponding to PM10 concentrations for the years 20072011 from Din Dang Station in Bangkok.
The experimental results indicate that the proposed FQDASVR (Fuzzy Logic Theory
Based on Quantitative Decision Algorithm with Support Vector Regression) model has better
results than SVR, Back-propagation Neural Network and ANFIS model in the PM10 forecasting.