Performance Evaluation of Some Noise Prediction Modelling Approaches: Multiple Linear Regressions and a Hybrid Approach
Journal Title: American journal of Engineering Research - Year 2018, Vol 7, Issue 7
Abstract
Noise prediction models are very useful for urban planning and environmental management. As a result, researchers are always searching for methods that are practically applicable in predicting noise levels accurately. It is therefore important to implement systems that could be employed to predict accurately, noise pollution levels in an urban area. In this study, two land-use regression methods was used to formulate two models namely, multiple linear regression (MLR) and a hybrid of analytical hierarchy process-multiple linear regression (AHP-MLR) to predict noise pollution levels in the Tarkwa mining community. The performance of the two models was evaluated using statistical indicators. The statistical findings indicate that the MLR model achieved a RMSE, standard deviation (SD), R 2 and R values of 1.569 dB(A), 1.585 dB(A), 0.961 and 0.980, respectively. The hybrid model (AHP-MLR), on the other hand, produced 1.774 dB(A), 1.758 dB(A), 0.955 and 0.977 as its corresponding RMSE, SD, R 2 and R values. Plotted box-and-whisker and range plots further confirmed the performance of the two models. The overall analysis showed that the MLR outperformed the AHP-MLR. The resulting noise map based on the noise predictions from the two models suggested that with the appropriate data and useful tools the noise pollution levels of an urban area could be well predicted and mapped for urban planning and environmental management.
Authors and Affiliations
Baffoe1 , P. E. , Duker2 , A. A.
Keywords
Related Articles
- EP ID EP398002
- DOI -
- Views 57
- Downloads 0
Reducing Asphalt Pavement Swell and Loss in Stability Due To Submergence in Moisture Using Polystyrene As Hydro-Carbon Additive: Study On Nigerian Roads In The Niger Delta Region
Laboratory investigation of asphalt pavement was carried out in order to ascertain the properties of swell and loss in stability that will occur when under submergence of water between 1 – 5 days. Secondly, the study was...
Dynamic Modeling Of A Dual Winding Induction Motor Using Rotor Reference Frame
A dynamic modelling method is needed for a better understanding of the transient behavior of dual winding induction motor. A good mathematical model using Park’s Transformation equation to describe the behavior of the mo...
A Wireless Process Monitoring And Control System With Zigbee
Recently wireless technology is getting popularity day by day in IT sector. Wireless Sensor Network (WSN) provides reduced cost, better power management, easy controls, high reliability and safety in industrial process m...
Diagnostics of Single and Multiple Outliers on Likelihood Distance
A problem often resist in statistical analysis is that there are may exist some extremely small or large observations. Diagnostics of these observations is a crucial tool of statistical analysis. In this paper, we propos...
Large-Scale Data-Centric Systems – A Research Analysis
:Vast scale information driven frameworks enable associations to store, control, and get an incentive from expansive volumes of information. They comprise of dispersed segments spread over an adaptable number of associat...