Calderón, Giovanni Salini; Jara, Patricio Pérez
September 2006
INGENIARE - Revista Chilena de Ingeniería;Sep-Dec2006, Vol. 14 Issue 3, p284
Academic Journal
An artificial neural network for the forecasting of concentrations of fine particulate matter in the atmosphere was designed. The data set analyzed corresponds to three years of pm2.5 time series (particulate matter in suspension with aerodynamic diameter less than 2, 5 microns), measured in a station that belongs to Santiago's monitoring network (Red MACAM) and is located near downtown. We consider measurements of concentrations between May and August for years between 1994 and 1996. In order to find the optimal time spacing between data and the number of values into the past necessary to forecast a future value, two standard tests were performed, Average Mutual Information (AMI) and False Nearest Neighbours (FNN). The results of these tests suggest that the most convenient choice for modelling was to use 4 data with 6 hour spacing on a given day as input in order to forecast the value at 6 AM on the following day. Once the number and type of input and output variables are fixed, we implemented a forecasting model based on the neural network technique. We used a feedforward multilayer neural network and we trained it with the backpropagation algorithm. We tested networks with none, one and two hidden layers. The best model was one with one hidden layer, in contradiction with a previous study that found that minimum error was obtained with a net without hidden layer. Forecasts with the neural network are more accurate than those produced with a persistence model (the value six hours ahead is the same as the actual value).


Related Articles

  • Prediction of Particlulate Air Pollution using Neural Techniques. Perez, Patricio; Reyes, Jorge // Neural Computing & Applications;2001, Vol. 10 Issue 2 

    We have analysed the possibility of predicting hourly average concentrations of suspended atmospheric particulate matter with aerodynamic diameter less than 2.5 microns (PM2.5) several hours in advance using data obtained in downtown Santiago, Chile. By performing some standard tests used in the...

  • Prediction Model of Traffic Based on Wavelet Analysis and BP Neural Network. Yongbin Li // International Journal of Digital Content Technology & its Applic;Dec2012, Vol. 6 Issue 22, p775 

    By research of traditional traffic prediction model it is found that the defects will decrease the performance of network and the methods are unsuitable for modeling of self-similar network traffic. So we put forward a novel model for predicting based on the wavelet transform and BP neural...

  • Feedforward Neural Nets as Models for Time Series Forecasting. Zaiyong Tang; Fishwick, Paul A. // ORSA Journal on Computing;Fall93, Vol. 5 Issue 4, p374 

    We have studied neural networks as models for time series forecasting, and our research compares the Box-Jenkins method against the neural network method for long and short term memory series Our work was inspired by previously published works that yielded inconsistent results about comparative...

  • Optimization of Neural Predictor for Air Pollution. Mechgoug, Raihane; Ahmed, A. Taleb; Cherroun, Lakhmissi // Proceedings of the World Congress on Engineering 2012 Volume II;2012, p1 

    This paper presents a neural network modeling applied to predict monthly ozone and carbon dioxide concentration time series. The performances of neural network are critically depending on learning parameters, on its structure and control parameters. Even when a suitable setting of parameters...

  • Forecasting high waters at Venice Lagoon using chaotic time series analysis and nonlinear neural networks. Zaldívar, J.M.; Gutiérrez, E.; Galván, I.M.; Strozzi, F.; Tomasin, A. // Journal of Hydroinformatics;Jan2000, Vol. 2 Issue 1, p61 

    Examines the application of time series analysis using nonlinear dynamics systems theory and multilayer neural networks models to the time sequence of water level data from Venice Lagoon in Italy. Reconstruction of the state space attractor using time delay embedding vectors; Characterization...

  • Laser pointing stabilization and control in the submicroradian regime with neural networks. Breitling, F.; Weigel, R. S.; Downer, M. C.; Tajima, T. // Review of Scientific Instruments;Feb2001, Vol. 72 Issue 2, p1339 

    The possibility of controling the pointing stability of a slowly pulsed Ti:Sapphire laser system by lowpass filters and artificial neural networks (NN) is investigated by performing time series analysis and computer simulations on experimentally measured datasets. The simulations show that at...

  • Quantifying Structural Time Varying Changes in Helical Data. Singh, Sameer // Neural Computing & Applications;2001, Vol. 10 Issue 2 

    Spiral structures are one of the most difficult patterns to classify. Spiral time series data has a helical movement with time that is both difficult to predict as well as classify. This paper focusses on how structural information about spirals can be useful in providing critical information to...

  • Artificial Neural Network Model for a Low Cost Failure Sensor: Performance Assessment in Pipeline Distribution. Khan, Asar; Widdop, Peter D.; Day, Andrew J.; Wood, Aliaster S.; Mounce, Steve R.; Machell, John // Enformatika;2006, Vol. 15, p195 

    This paper describes an automated event detection and location system for water distribution pipelines which is based upon low-cost sensor technology and signature analysis by an Artificial Neural Network (ANN). The development of a low cost failure sensor which measures the opacity or...

  • Non-linear time series analysis of combustion pressure data for neural network training with the concept of mutual information. Heister, F; Froehlich, M // Proceedings of the Institution of Mechanical Engineers -- Part D;2001, Vol. 215 Issue 2, p299 

    In recent years, after a period of disillusion in the field of neural processing and adaptive algorithms, neural networks have been reconsidered for solving complex technical tasks. The problem of neural network training is the presentation of input/output data showing an appropriate information...


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics