Search results for: forecasting model

Authors: Amir Hossein Hejazi, Nima Amjady

Abstract:

In recent years, due to environmental issues traditional energy sources had been replaced by renewable ones. Wind energy as the fastest growing renewable energy shares a considerable percent of energy in power electricity markets. With this fast growth of wind energy worldwide, owners and operators of wind farms, transmission system operators, and energy traders need reliable and secure forecasts of wind energy production. In this paper, a new forecasting strategy is proposed for short-term wind power prediction based on Echo State Networks (ESN). The forecast engine utilizes state-of-the-art training process including dynamical reservoir with high capability to learn complex dynamics of wind power or wind vector signals. The study becomes more interesting by incorporating prediction of wind direction into forecast strategy. The Big Bang-Big Crunch (BB-BC) evolutionary optimization algorithm is adopted for adjusting free parameters of ESN-based forecaster. The proposed method is tested by real-world hourly data to show the efficiency of the forecasting engine for prediction of both wind vector and wind power output of aggregated wind power production.

Keywords: wind power forecasting, echo state network, big bang-big crunch, evolutionary optimization algorithm

Procedia PDF Downloads 540

Authors: M. Cizmesija, N. Erjavec, V. Bahovec

Abstract:

While the European Union (EU) harmonized methodology is a benchmark of worldwide used business survey (BS) methodology, the choice of variables that are components of the confidence indicators, as the leading indicators, is not strictly determined and unique. Therefore, the aim of this paper is to investigate and to quantify the relationship between all business survey variables in manufacturing industry and industrial production as a reference macroeconomic series in Croatia. The assumption is that there are variables in the business survey, that are not components of Industrial Confidence Indicator (ICI) and which can accurately (and sometimes better then ICI) predict changes in Croatian industrial production. Empirical analyses are conducted using quarterly data of BS variables in manufacturing industry and Croatian industrial production over the period from the first quarter 2005 to the first quarter 2013. Research results confirmed the assumption: three BS variables which is not components of ICI (competitive position, demand and liquidity) are the best leading indicator then ICI, in forecasting changes in Croatian industrial production instantaneously, with one, two or three quarter ahead.

Keywords: balance, business survey, confidence indicators, industrial production, forecasting

Procedia PDF Downloads 448

Authors: Georgia Pozoukidou

Abstract:

TELUM software is a land use model designed specifically to help metropolitan planning organizations (MPOs) prepare their transportation improvement programs and fulfill their numerous planning responsibilities. In this context obtaining, preparing, and validating socioeconomic forecasts are becoming fundamental tasks for an MPO in order to ensure that consistent population and employment data are provided to travel demand models. Chittenden County Metropolitan Planning Organization of Vermont State was used as a case study to test the applicability of TELUM land use model. The technical insights and lessons learned from the land use model application have transferable value for all MPOs faced with land use forecasting development and transportation modelling.

Keywords: calibration data requirements, land use models, land use planning, metropolitan planning organizations

Procedia PDF Downloads 270

Authors: Karim Hamidi Machekposhti, Hossein Sedghi, Abdolrasoul Telvari, Hossein Babazadeh

Abstract:

Floods have huge environmental and economic impact. Therefore, flood prediction is given a lot of attention due to its importance. This study analysed the annual maximum streamflow (discharge) (AMS or AMD) of Karkheh River in Karkheh River Basin for flood predicting using ARIMA model. For this purpose, we use the Box-Jenkins approach, which contains four-stage method model identification, parameter estimation, diagnostic checking and forecasting (predicting). The main tool used in ARIMA modelling was the SAS and SPSS software. Model identification was done by visual inspection on the ACF and PACF. SAS software computed the model parameters using the ML, CLS and ULS methods. The diagnostic checking tests, AIC criterion, RACF graph and RPACF graphs, were used for selected model verification. In this study, the best ARIMA models for Annual Maximum Discharge (AMD) time series was (4,1,1) with their AIC value of 88.87. The RACF and RPACF showed residuals’ independence. To forecast AMD for 10 future years, this model showed the ability of the model to predict floods of the river under study in the Karkheh River Basin. Model accuracy was checked by comparing the predicted and observation series by using coefficient of determination (R²).

Keywords: time series modelling, stochastic processes, ARIMA model, Karkheh river

Procedia PDF Downloads 270

Authors: Hidir S. Nogay

Abstract:

In this study, two systems were created to predict interior temperature in induction motor. One of them consisted of a simple ANN model which has two layers, ten input parameters and one output parameter. The other one consisted of eight ANN models connected each other as cascaded. Cascaded ANN system has 17 inputs. Main reason of cascaded system being used in this study is to accomplish more accurate estimation by increasing inputs in the ANN system. Cascaded ANN system is compared with simple conventional ANN model to prove mentioned advantages. Dataset was obtained from experimental applications. Small part of the dataset was used to obtain more understandable graphs. Number of data is 329. 30% of the data was used for testing and validation. Test data and validation data were determined for each ANN model separately and reliability of each model was tested. As a result of this study, it has been understood that the cascaded ANN system produced more accurate estimates than conventional ANN model.

Keywords: cascaded neural network, internal temperature, inverter, three-phase induction motor

Procedia PDF Downloads 322

Authors: Sami Mestiri, Abdeljelil Farhat

Abstract:

The aim of this current paper is to predict the credit risk of banks in Tunisia, over the period (2000-2005). For this purpose, two methods for the estimation of the logistic regression model with random effects: Penalized Quasi Likelihood (PQL) method and Gibbs Sampler algorithm are applied. By using the information on a sample of 528 Tunisian firms and 26 financial ratios, we show that Bayesian approach improves the quality of model predictions in terms of good classification as well as by the ROC curve result.

Keywords: forecasting, credit risk, Penalized Quasi Likelihood, Gibbs Sampler, logistic regression with random effects, curve ROC

Procedia PDF Downloads 513

Authors: Ivana Čavlina Tomašević, Višnjica Vučetić, Maja Telišman Prtenjak, Barbara Malečić

Abstract:

The Split wildfire on the mid-Adriatic Coast in July 2017 is one of the most severe wildfires in Croatian history, given the size and unexpected fire behavior, and it is used in this research as a case study to run the Weather Research and Forecasting Spread Fire (WRF SFIRE) model. This coupled fire-atmosphere model was successfully run for the first time ever for one Croatian wildfire case. Verification of coupled simulations was possible by using the detailed reconstruction of the Split wildfire. Specifically, precise information on ignition time and location, together with mapped fire progressions and spotting within the first 30 hours of the wildfire, was used for both – to initialize simulations and to evaluate the model’s ability to simulate fire’s propagation and final fire scar. The preliminary simulations were obtained using high-resolution vegetation and topography data for the fire area, additionally interpolated to fire grid spacing at 33.3 m. The results demonstrated that the WRF SFIRE model has the ability to work with real data from Croatia and produce adequate results for forecasting fire spread. As the model in its setup has the ability to include and exclude the energy fluxes between the fire and the atmosphere, this was used to investigate possible fire-atmosphere interactions during the Split wildfire. Finally, successfully coupled simulations provided the first numerical evidence that a wildfire from the Adriatic coast region can modify the dynamical structure of the surrounding atmosphere, which agrees with observations from fire grounds. This study has demonstrated that the WRF SFIRE model has the potential for operational application in Croatia with more accurate fire predictions in the future, which could be accomplished by inserting the higher-resolution input data into the model without interpolation. Possible uses for fire management in Croatia include prediction of fire spread and intensity that may vary under changing weather conditions, available fuels and topography, planning effective and safe deployment of ground and aerial firefighting forces, preventing wildland-urban interface fires, effective planning of evacuation routes etc. In addition, the WRF SFIRE model results from this research demonstrated that the model is important for fire weather research and education purposes in order to better understand this hazardous phenomenon that occurs in Croatia.

Keywords: meteorology, agrometeorology, fire weather, wildfires, couple fire-atmosphere model

Procedia PDF Downloads 57

Authors: A. Hakimiyan, H. Namazi

Abstract:

Human behavior is defined as a range of behaviors exhibited by humans who are influenced by different internal or external sources. Human behavior is the subject of much research in different areas of psychology and neuroscience. Despite some advances in studies related to forecasting of human behavior, there are not many researches which consider the effect of the time delay between the presence of stimulus and the related human response. Analysis of EEG signal as a fractal time series is one of the major tools for studying the human behavior. In the other words, the human brain activity is reflected in his EEG signal. Artificial Neural Network has been proved useful in forecasting of different systems’ behavior especially in engineering areas. In this research, a time delay neural network is trained and tested in order to forecast the human EEG signal and subsequently human behavior. This neural network, by introducing a time delay, takes care of the lagging time between the occurrence of the stimulus and the rise of the subsequent action potential. The results of this study are useful not only for the fundamental understanding of human behavior forecasting, but shall be very useful in different areas of brain research such as seizure prediction.

Keywords: human behavior, EEG signal, time delay neural network, prediction, lagging time

Procedia PDF Downloads 634

Authors: Vikrant Gupta, Amrit Goswami

Abstract:

The fixed income market forms the basis of the modern financial market. All other assets in financial markets derive their value from the bond market. Owing to its over-the-counter nature, corporate bonds have relatively less data publicly available and thus is researched upon far less compared to Equities. Bond price prediction is a complex financial time series forecasting problem and is considered very crucial in the domain of finance. The bond prices are highly volatile and full of noise which makes it very difficult for traditional statistical time-series models to capture the complexity in series patterns which leads to inefficient forecasts. To overcome the inefficiencies of statistical models, various machine learning techniques were initially used in the literature for more accurate forecasting of time-series. However, simple machine learning methods such as linear regression, support vectors, random forests fail to provide efficient results when tested on highly complex sequences such as stock prices and bond prices. hence to capture these intricate sequence patterns, various deep learning-based methodologies have been discussed in the literature. In this study, a recurrent neural network-based deep learning model using long short term networks for prediction of corporate bond prices has been discussed. Long Short Term networks (LSTM) have been widely used in the literature for various sequence learning tasks in various domains such as machine translation, speech recognition, etc. In recent years, various studies have discussed the effectiveness of LSTMs in forecasting complex time-series sequences and have shown promising results when compared to other methodologies. LSTMs are a special kind of recurrent neural networks which are capable of learning long term dependencies due to its memory function which traditional neural networks fail to capture. In this study, a simple LSTM, Stacked LSTM and a Masked LSTM based model has been discussed with respect to varying input sequences (three days, seven days and 14 days). In order to facilitate faster learning and to gradually decompose the complexity of bond price sequence, an Empirical Mode Decomposition (EMD) has been used, which has resulted in accuracy improvement of the standalone LSTM model. With a variety of Technical Indicators and EMD decomposed time series, Masked LSTM outperformed the other two counterparts in terms of prediction accuracy. To benchmark the proposed model, the results have been compared with traditional time series models (ARIMA), shallow neural networks and above discussed three different LSTM models. In summary, our results show that the use of LSTM models provide more accurate results and should be explored more within the asset management industry.

Keywords: bond prices, long short-term memory, time series forecasting, empirical mode decomposition

Procedia PDF Downloads 109

Authors: Milan Joshi, Harsh Agrawal, Pallaw Mishra, Sanand Sule

Abstract:

Predicting electricity usage is a crucial aspect of organizing and controlling sustainable energy systems. The task of forecasting electricity load is intricate and requires a lot of effort due to the combined impact of social, economic, technical, environmental, and cultural factors on power consumption in communities. As a result, it is important to create strong models that can handle the significant non-linear and complex nature of the task. The objective of this study is to create and compare three machine learning techniques for predicting electricity load for both the day ahead and intraday, taking into account various factors such as meteorological data and social events including holidays and festivals. The proposed methods include a LightGBM, FBProphet, combination of FBProphet and LightGBM for day ahead and Motifs( Stumpy) based on Mueens algorithm for similarity search for intraday. We utilize these techniques to predict electricity usage during normal days and social events in the Himachal Region. We then assess their performance by measuring the MSE, RMSE, and MAPE values. The outcomes demonstrate that the combination of FBProphet and LightGBM method is the most accurate for day ahead and Motifs for intraday forecasting of electricity usage, surpassing other models in terms of MAPE, RMSE, and MSE. Moreover, the FBProphet - LightGBM approach proves to be highly effective in forecasting electricity load during social events, exhibiting precise day ahead predictions. In summary, our proposed electricity forecasting techniques display excellent performance in predicting electricity usage during normal days and special events in the Himachal Region.

Keywords: feature engineering, FBProphet, LightGBM, MASS, Motifs, MAPE

Procedia PDF Downloads 44

Authors: Selin Guney, Andres Riquelme

Abstract:

Among the main purposes of optimal and efficient forecasts of agricultural commodity prices is to guide the firms to advance the economic decision making process such as planning business operations and marketing decisions. Governments are also the beneficiaries and suppliers of agricultural price forecasts. They use this information to establish a proper agricultural policy, and hence, the forecasts affect social welfare and systematic errors in forecasts could lead to a misallocation of scarce resources. Various empirical approaches have been applied to forecast commodity prices that have used different methodologies. Most commonly-used approaches to forecast commodity sectors depend on classical time series models that assume values of the response variables are precise which is quite often not true in reality. Recently, this literature has mostly evolved to a consideration of fuzzy time series models that provide more flexibility in terms of the classical time series models assumptions such as stationarity, and large sample size requirement. Besides, fuzzy modeling approach allows decision making with estimated values under incomplete information or uncertainty. A number of fuzzy time series models have been developed and implemented over the last decades; however, most of them are not appropriate for forecasting repeated and nonconsecutive transitions in the data. The modeling scheme used in this paper eliminates this problem by introducing Markov modeling approach that takes into account both the repeated and nonconsecutive transitions. Also, the determination of length of interval is crucial in terms of the accuracy of forecasts. The problem of determining the length of interval arbitrarily is overcome and a methodology to determine the proper length of interval based on the distribution or mean of the first differences of series to improve forecast accuracy is proposed. The specific purpose of this paper is to propose and investigate the potential of a new forecasting model that integrates methodologies for determining the proper length of interval based on the distribution or mean of the first differences of series and Fuzzy Time Series- Markov Chain model. Moreover, the accuracy of the forecasting performance of proposed integrated model is compared to different univariate time series models and the superiority of proposed method over competing methods in respect of modelling and forecasting on the basis of forecast evaluation criteria is demonstrated. The application is to daily corn and soybean prices observed at three commercially important North Carolina markets; Candor, Cofield and Roaring River for corn and Fayetteville, Cofield and Greenville City for soybeans respectively. One main conclusion from this paper is that using fuzzy logic improves the forecast performance and accuracy; the effectiveness and potential benefits of the proposed model is confirmed with small selection criteria value such MAPE. The paper concludes with a discussion of the implications of integrating fuzzy logic and nonarbitrary determination of length of interval for the reliability and accuracy of price forecasts. The empirical results represent a significant contribution to our understanding of the applicability of fuzzy modeling in commodity price forecasts.

Keywords: commodity, forecast, fuzzy, Markov

Procedia PDF Downloads 198

Authors: Yichao Liu, Peter Fransson, Julian Heidecke, Jonas Wallin, Joacim Rockloev

Abstract:

Dengue, a mosquito-borne viral disease, poses a significant public health challenge in endemic tropical or subtropical countries, including Sri Lanka. To reveal insights into the complexity of the dynamics of this disease and study the drivers, a comprehensive model capable of both robust forecasting and insightful inference of drivers while capturing the co-circulating of several virus strains is essential. However, existing studies mostly focus on only one aspect at a time and do not integrate and carry insights across the siloed approach. While mechanistic models are developed to capture immunity dynamics, they are often oversimplified and lack integration of all the diverse drivers of disease transmission. On the other hand, purely data-driven methods lack constraints imposed by immuno-epidemiological processes, making them prone to overfitting and inference bias. This research presents a hybrid model that combines machine learning techniques with mechanistic modelling to overcome the limitations of existing approaches. Leveraging eight years of newly reported dengue case data, along with socioeconomic factors, such as human mobility, weekly climate data from 2011 to 2018, genetic data detecting the introduction and presence of new strains, and estimates of seropositivity for different districts in Sri Lanka, we derive a data-driven vector (SEI) to human (SEIR) model across 16 regions in Sri Lanka at the weekly time scale. By conducting ablation studies, the lag effects allowing delays up to 12 weeks of time-varying climate factors were determined. The model demonstrates superior predictive performance over a pure machine learning approach when considering lead times of 5 and 10 weeks on data withheld from model fitting. It further reveals several interesting interpretable findings of drivers while adjusting for the dynamics and influences of immunity and introduction of a new strain. The study uncovers strong influences of socioeconomic variables: population density, mobility, household income and rural vs. urban population. The study reveals substantial sensitivity to the diurnal temperature range and precipitation, while mean temperature and humidity appear less important in the study location. Additionally, the model indicated sensitivity to vegetation index, both max and average. Predictions on testing data reveal high model accuracy. Overall, this study advances the knowledge of dengue transmission in Sri Lanka and demonstrates the importance of incorporating hybrid modelling techniques to use biologically informed model structures with flexible data-driven estimates of model parameters. The findings show the potential to both inference of drivers in situations of complex disease dynamics and robust forecasting models.

Keywords: compartmental model, climate, dengue, machine learning, social-economic

Procedia PDF Downloads 48

Authors: Alexandros Pournaras, Anastasia Papadopoulou, Serafim Kontos, Anastasios Karakostas

Abstract:

The air quality of modern cities is an important concern, as poor air quality contributes to human health and environmental issues. Reliable air quality forecasting has, thus, gained scientific and governmental attention as an essential tool that enables authorities to take proactive measures for public safety. In this study, the potential of Machine Learning (ML) models to forecast PM2.5 at local scale is investigated in the city of Thessaloniki, the second largest city in Greece, which has been struggling with the persistent issue of air pollution. ML models, with proven ability to address timeseries forecasting, are employed to predict the PM2.5 concentrations and the respective Air Quality Index 5-days ahead by learning from daily historical air quality and meteorological data from 2014 to 2016 and gathered from two stations with different land use characteristics in the urban fabric of Thessaloniki. The performance of the ML models on PM2.5 concentrations is evaluated with common statistical methods, such as R squared (r²) and Root Mean Squared Error (RMSE), utilizing a portion of the stations’ measurements as test set. A multi-categorical evaluation is utilized for the assessment of their performance on respective AQIs. Several conclusions were made from the experiments conducted. Experimenting on MLs’ configuration revealed a moderate effect of various parameters and training schemas on the model’s predictions. Their performance of all these models were found to produce satisfactory results on PM2.5 concentrations. In addition, their application on untrained stations showed that these models can perform well, indicating a generalized behavior. Moreover, their performance on AQI was even better, showing that the MLs can be used as predictors for AQI, which is the direct information provided to the general public.

Keywords: Air Quality, AQ Forecasting, AQI, Machine Learning, PM2.5

Procedia PDF Downloads 45

Authors: Namwinwelbere Dabire, Eugene C. Ezin, Adandedji M. Firmin

Abstract:

The prediction of hydrological flows (rainfall-depth or rainfall-discharge) is becoming increasingly important in the management of hydrological risks such as floods. In this study, the Long Short-Term Memory (LSTM) network, a state-of-the-art algorithm dedicated to time series, is applied to predict the daily water level of Nokoue Lake in Benin. This paper aims to provide an effective and reliable method enable of reproducing the future daily water level of Nokoue Lake, which is influenced by a combination of two phenomena: rainfall and river flow (runoff from the Ouémé River, the Sô River, the Porto-Novo lagoon, and the Atlantic Ocean). Performance analysis based on the forecasting horizon indicates that LSTM can predict the water level of Nokoué Lake up to a forecast horizon of t+10 days. Performance metrics such as Root Mean Square Error (RMSE), coefficient of correlation (R²), Nash-Sutcliffe Efficiency (NSE), and Mean Absolute Error (MAE) agree on a forecast horizon of up to t+3 days. The values of these metrics remain stable for forecast horizons of t+1 days, t+2 days, and t+3 days. The values of R² and NSE are greater than 0.97 during the training and testing phases in the Nokoué Lake basin. Based on the evaluation indices used to assess the model's performance for the appropriate forecast horizon of water level in the Nokoué Lake basin, the forecast horizon of t+3 days is chosen for predicting future daily water levels.

Keywords: forecasting, long short-term memory cell, recurrent artificial neural network, Nokoué lake

Procedia PDF Downloads 33

Authors: Pruethipong Xinghatiraj, Rajwanlop Kumpoopong

Abstract:

Considering the current situation of road safety, Thailand has the world’s second-highest road fatality rate. Therefore, decreasing the road accidents in Thailand is a prime policy of the Thai government seeking to accomplish. One of the approaches to reduce the accident rate is to improve road environments to fit with the local behavior of the road users. The Department of Highways ensures that choosing the road median types right to the road characteristics, e.g. roadside characteristics, traffic volume, truck traffic percentage, etc., can decrease the possibility of accident occurrence. Presently, raised median, depressed median, painted median and median barriers are typically used in Thailand Highways. In this study, factors affecting road accident for each median type will be discovered through the analysis of the collecting of accident data, death numbers on sample of 600 Kilometers length across the country together with its roadside characteristics, traffic volume, heavy vehicles percentage, and other key factors. The benefits of this study can assist the Highway designers to select type of road medians that can match local environments and then cause less accident prone.

Keywords: highways, road safety, road median, forecasting model

Procedia PDF Downloads 241

Authors: Karim Hamidi Machekposhti, Hossein Sedghi, Abdolrasoul Telvari, Hossein Babazadeh

Abstract:

Rainfall and runoff phenomenon is a chaotic and complex outcome of nature which requires sophisticated modelling and simulation methods for explanation and use. Time Series modelling allows runoff data analysis and can be used as forecasting tool. In the paper attempt is made to model river runoff data and predict the future behavioural pattern of river based on annual past observations of annual river runoff. The river runoff analysis and predict are done using ARIMA model. For evaluating the efficiency of prediction to hydrological events such as rainfall, runoff and etc., we use the statistical formulae applicable. The good agreement between predicted and observation river runoff coefficient of determination (R²) display that the ARIMA (4,1,1) is the suitable model for predicting Karkheh River runoff at Iran.

Keywords: time series modelling, ARIMA model, river runoff, Karkheh River, CLS method

Procedia PDF Downloads 313

Authors: Shih-Pin Chen, Shih-Syuan You

Abstract:

Fuzzy regression models are useful for investigating the relationship between explanatory variables and responses in fuzzy environments. To overcome the deficiencies of previous models and increase the explanatory power of fuzzy data, the graded mean integration (GMI) representation is applied to determine representative crisp regression coefficients. A fuzzy regression model is constructed based on the modified dissemblance index (MDI), which can precisely measure the actual total error. Compared with previous studies based on the proposed MDI and distance criterion, the results from commonly used test examples show that the proposed fuzzy linear regression model has higher explanatory power and forecasting accuracy.

Keywords: dissemblance index, fuzzy linear regression, graded mean integration, mathematical programming

Procedia PDF Downloads 412

Authors: Hamid Rostami Jaz, Kamran Ameri Siahooei

Abstract:

Up to now different methods have been used to forecast the index returns and the index rate. Artificial intelligence and artificial neural networks have been one of the methods of index returns forecasting. This study attempts to carry out a comparative study on the performance of different Radial Base Neural Network and Feed-Forward Perceptron Neural Network to forecast investment returns on the index. To achieve this goal, the return on investment in Tehran Stock Exchange index is evaluated and the performance of Radial Base Neural Network and Feed-Forward Perceptron Neural Network are compared. Neural networks performance test is applied based on the least square error in two approaches of in-sample and out-of-sample. The research results show the superiority of the radial base neural network in the in-sample approach and the superiority of perceptron neural network in the out-of-sample approach.

Keywords: exchange index, forecasting, perceptron neural network, Tehran stock exchange

Procedia PDF Downloads 424

Authors: Atish Bagchi, Siva Chandrasekaran

Abstract:

Background: According to World Bank national accounts data, the estimated global manufacturing value-added output in 2020 was 13.74 trillion USD. These manufacturing processes are monitored, modelled, and controlled by advanced, real-time, computer-based systems, e.g., Industrial IoT, PLC, SCADA, etc. These systems measure and manipulate a set of physical variables, e.g., temperature, pressure, etc. Despite the use of IoT, SCADA etc., in manufacturing, studies suggest that unplanned downtime leads to economic losses of approximately 864 billion USD each year. Therefore, real-time, accurate detection, classification and prediction of machine behaviour are needed to minimise financial losses. Although vast literature exists on time-series data processing using machine learning, the challenges faced by the industries that lead to unplanned downtimes are: The current algorithms do not efficiently handle the high-volume streaming data from industrial IoTsensors and were tested on static and simulated datasets. While the existing algorithms can detect significant 'point' outliers, most do not handle contextual outliers (e.g., values within normal range but happening at an unexpected time of day) or subtle changes in machine behaviour. Machines are revamped periodically as part of planned maintenance programmes, which change the assumptions on which original AI models were created and trained. Aim: This research study aims to deliver a Graph Neural Network(GNN)based hybrid forecasting model that interfaces with the real-time machine control systemand can detect, predict machine behaviour and behavioural changes (anomalies) in real-time. This research will help manufacturing industries and utilities, e.g., water, electricity etc., reduce unplanned downtimes and consequential financial losses. Method: The data stored within a process control system, e.g., Industrial-IoT, Data Historian, is generally sampled during data acquisition from the sensor (source) and whenpersistingin the Data Historian to optimise storage and query performance. The sampling may inadvertently discard values that might contain subtle aspects of behavioural changes in machines. This research proposed a hybrid forecasting and classification model which combines the expressive and extrapolation capability of GNN enhanced with the estimates of entropy and spectral changes in the sampled data and additional temporal contexts to reconstruct the likely temporal trajectory of machine behavioural changes. The proposed real-time model belongs to the Deep Learning category of machine learning and interfaces with the sensors directly or through 'Process Data Historian', SCADA etc., to perform forecasting and classification tasks. Results: The model was interfaced with a Data Historianholding time-series data from 4flow sensors within a water treatment plantfor45 days. The recorded sampling interval for a sensor varied from 10 sec to 30 min. Approximately 65% of the available data was used for training the model, 20% for validation, and the rest for testing. The model identified the anomalies within the water treatment plant and predicted the plant's performance. These results were compared with the data reported by the plant SCADA-Historian system and the official data reported by the plant authorities. The model's accuracy was much higher (20%) than that reported by the SCADA-Historian system and matched the validated results declared by the plant auditors. Conclusions: The research demonstrates that a hybrid GNN based approach enhanced with entropy calculation and spectral information can effectively detect and predict a machine's behavioural changes. The model can interface with a plant's 'process control system' in real-time to perform forecasting and classification tasks to aid the asset management engineers to operate their machines more efficiently and reduce unplanned downtimes. A series of trialsare planned for this model in the future in other manufacturing industries.

Keywords: GNN, Entropy, anomaly detection, industrial time-series, AI, IoT, Industry 4.0, Machine Learning

Procedia PDF Downloads 119

Authors: Lukas Reznak, Maria Reznakova

Abstract:

Recession of an economy has a profound negative effect on all involved stakeholders. It follows that timely prediction of recessions has been of utmost interest both in the theoretical research and in practical macroeconomic modelling. Current mainstream of recession prediction is based on standard OLS models of continuous GDP using macroeconomic data. This approach is not suitable for two reasons: the standard continuous models are proving to be obsolete and the macroeconomic data are unreliable, often revised many years retroactively. The aim of the paper is to explore a different branch of recession forecasting research theory and verify the findings on real data of the Czech Republic and Germany. In the paper, the authors present a family of discrete choice probit models with parameters estimated by the method of maximum likelihood. In the basic form, the probits model a univariate series of recessions and expansions in the economic cycle for a given country. The majority of the paper deals with more complex model structures, namely dynamic and bivariate extensions. The dynamic structure models the autoregressive nature of recessions, taking into consideration previous economic activity to predict the development in subsequent periods. Bivariate extensions utilize information from a foreign economy by incorporating correlation of error terms and thus modelling the dependencies of the two countries. Bivariate models predict a bivariate time series of economic states in both economies and thus enhance the predictive performance. A vital enabler of timely and successful recession forecasting are reliable and readily available data. Leading indicators, namely the yield curve and the stock market indices, represent an ideal data base, as the pieces of information is available in advance and do not undergo any retroactive revisions. As importantly, the combination of yield curve and stock market indices reflect a range of macroeconomic and financial market investors’ trends which influence the economic cycle. These theoretical approaches are applied on real data of Czech Republic and Germany. Two models for each country were identified – each for in-sample and out-of-sample predictive purposes. All four followed a bivariate structure, while three contained a dynamic component.

Keywords: bivariate probit, leading indicators, recession forecasting, Czech Republic, Germany

Procedia PDF Downloads 228

Authors: S. Chuenchooklin, S. Taweepong, U. Pangnakorn

Abstract:

This research was conducted in the Mae Sot Watershed whereas located in the Moei River Basin at the Upper Salween River Basin in Tak Province, Thailand. The Mae Sot Municipality is the largest urbanized in Tak Province and situated in the midstream of the Mae Sot Watershed. It usually faces flash flood problem after heavy rain due to poor flood management has been reported since economic rapidly bloom up in recently years. Its catchment can be classified as ungauged basin with lack of rainfall data and no any stream gaging station was reported. It was attached by most severely flood event in 2013 as the worst studied case for those all communities in this municipality. Moreover, other problems are also faced in this watershed such shortage water supply for domestic consumption and agriculture utilizations including deterioration of water quality and landslide as well. The research aimed to increase capability building and strengthening the participation of those local community leaders and related agencies to conduct better water management in urban area was started by mean of the data collection and illustration of appropriated application of some short period rainfall forecasting model as the aim for better flood relief plan and management through the hydrologic model system and river analysis system programs. The authors intended to apply the global rainfall data via the integrated data viewer (IDV) program from the Unidata with the aim for rainfall forecasting in short period of 7 - 10 days in advance during rainy season instead of real time record. The IDV product can be present in advance period of rainfall with time step of 3 - 6 hours was introduced to the communities. The result can be used to input to either the hydrologic modeling system model (HEC-HMS) or the soil water assessment tool model (SWAT) for synthesizing flood hydrographs and use for flood forecasting as well. The authors applied the river analysis system model (HEC-RAS) to present flood flow behaviors in the reach of the Mae Sot stream via the downtown of the Mae Sot City as flood extents as water surface level at every cross-sectional profiles of the stream. Both models of HMS and RAS were tested in 2013 with observed rainfall and inflow-outflow data from the Mae Sot Dam. The result of HMS showed fit to the observed data at dam and applied at upstream boundary discharge to RAS in order to simulate flood extents and tested in the field, and the result found satisfied. The result of IDV’s rainfall forecast data was compared to observed data and found fair. However, it is an appropriate tool to use in the ungauged catchment to use with flood hydrograph and river analysis models for future efficient flood relief plan and management.

Keywords: global rainfall, flood forecast, hydrologic modeling system, river analysis system

Procedia PDF Downloads 323

Authors: Taisiia Gorshkova

Abstract:

Since the 1980s, spatial correlation models have been used more often to model regional indicators. An increasingly popular method for studying regional indicators is modeling taking into account spatial relationships between objects that are part of the same economic zone. In 2000s the new class of model – spatial vector autoregressions was developed. The main difference between standard and spatial vector autoregressions is that in the spatial VAR (SpVAR), the values of indicators at time t may depend on the values of explanatory variables at the same time t in neighboring regions and on the values of explanatory variables at time t-k in neighboring regions. Thus, VAR is a special case of SpVAR in the absence of spatial lags, and the spatial panel data model is a special case of spatial VAR in the absence of time lags. Two specifications of SpVAR were applied to Russian regional data for 2000-2017. The values of GRP and regional CPI are used as endogenous variables. The lags of GRP, CPI and the unemployment rate were used as explanatory variables. For comparison purposes, the standard VAR without spatial correlation was used as “naïve” model. In the first specification of SpVAR the unemployment rate and the values of depending variables, GRP and CPI, in neighboring regions at the same moment of time t were included in equations for GRP and CPI respectively. To account for the values of indicators in neighboring regions, the adjacency weight matrix is used, in which regions with a common sea or land border are assigned a value of 1, and the rest - 0. In the second specification the values of depending variables in neighboring regions at the moment of time t were replaced by these values in the previous time moment t-1. According to the results obtained, when inflation and GRP of neighbors are added into the model both inflation and GRP are significantly affected by their previous values, and inflation is also positively affected by an increase in unemployment in the previous period and negatively affected by an increase in GRP in the previous period, which corresponds to economic theory. GRP is not affected by either the inflation lag or the unemployment lag. When the model takes into account lagged values of GRP and inflation in neighboring regions, the results of inflation modeling are practically unchanged: all indicators except the unemployment lag are significant at a 5% significance level. For GRP, in turn, GRP lags in neighboring regions also become significant at a 5% significance level. For both spatial and “naïve” VARs the RMSE were calculated. The minimum RMSE are obtained via SpVAR with lagged explanatory variables. Thus, according to the results of the study, it can be concluded that SpVARs can accurately model both the actual values of macro indicators (particularly CPI and GRP) and the general situation in the regions

Keywords: forecasting, regional data, spatial econometrics, vector autoregression

Procedia PDF Downloads 111

Authors: Salako Rotimi, Oshungade Stephen, Ojewoye Opeyemi

Abstract:

The maintenance of price stability is one of the macroeconomic challenges facing Nigeria as a nation. This paper attempts to build a co-integration multivariate time series model for inflation movement in Nigeria using data extracted from the abstract of statistics of the Central Bank of Nigeria (CBN) from 2008 to 2017. The Johansen cointegration test suggests at least one co-integration vector describing the long run relationship between Consumer Price Index (CPI), Food Price Index (FPI) and Non-Food Price Index (NFPI). All three series show increasing pattern, which indicates a sign of non-stationary in each of the series. Furthermore, model predictability was established with root-mean-square-error, mean absolute error, mean average percentage error, and Theil’s unbiased statistics for n-step forecasting. The result depicts that the long run coefficient of a consumer price index (CPI) has a positive long-run relationship with the food price index (FPI) and non-food price index (NFPI).

Keywords: economic, inflation, model, series

Procedia PDF Downloads 219

Authors: Saba Arif, Anam Nadeem, Roman Kalvin, Tanzeel Rashid, Burhan Ali, Juntakan Taweekun

Abstract:

Currently, the energy crisis is taking serious attention. Globally, industries and building are major share takers of energy. 72% of total global energy is consumed by residential houses, markets, and commercial building. Additionally, in appliances air conditioners are major consumer of electricity; about 60% energy is used for cooling purpose in houses due to HVAC units. Energy demand will aid in determining what changes will be needed whether it is the estimation of the required energy for households or instituting conservation measures. Bottom-up model is one of the most famous methods for forecasting. In current research bottom-up model of air conditioners' energy consumption in all income classes in comparison with seasonal variation and hourly consumption is calculated. By comparison of energy consumption of all income classes by usage of air conditioners, total consumption of actual demand and current availability can be seen.

Keywords: air conditioning, bottom up model, income classes, energy demand

Procedia PDF Downloads 217

Authors: Byung hyun Bae

Abstract:

In power transformer, leakage magnetic flux generate temperature rise of inner steel plate. Sometimes, this temperature rise can be serious problem. If temperature of steel plate is over critical point, harmful gas will be generated in the tank. And this gas can be a reason of fire, explosion and life decrease. So, temperature rise forecasting of steel plate is very important at the design stage of power transformer. To improve accuracy of forecasting of temperature rise, comparison between simulation and experiment achieved in this paper.

Keywords: power transformer, steel plate, temperature rise, experiment, simulation

Procedia PDF Downloads 468

Authors: Nor Zila Abd Hamid, Mohd Salmi M. Noorani

Abstract:

This study is focused on the development of prediction models of the Ozone concentration time series. Prediction model is built based on chaotic approach. Firstly, the chaotic nature of the time series is detected by means of phase space plot and the Cao method. Then, the prediction model is built and the local linear approximation method is used for the forecasting purposes. Traditional prediction of autoregressive linear model is also built. Moreover, an improvement in local linear approximation method is also performed. Prediction models are applied to the hourly ozone time series observed at the benchmark station in Malaysia. Comparison of all models through the calculation of mean absolute error, root mean squared error and correlation coefficient shows that the one with improved prediction method is the best. Thus, chaotic approach is a good approach to be used to develop a prediction model for the Ozone concentration time series.

Keywords: chaotic approach, phase space, Cao method, local linear approximation method

Procedia PDF Downloads 302

Authors: Jang kyun Cho, Jeong-dong Lee

Abstract:

The advent of the Internet, mobile communications, and social network services has stimulated social interactions among consumers, allowing people to affect one another’s innovation adoptions by exchanging information more frequently and more quickly. Previous diffusion models, such as the Bass model, however, face limitations in reflecting such recent phenomena in society. These models are weak in their ability to model interactions between agents; they model aggregated-level behaviors only. The agent based model, which is an alternative to the aggregate model, is good for individual modeling, but it is still not based on an economic perspective of social interactions so far. This study assumes the presence of social utility from other consumers in the adoption of innovation and investigates the effect of individual interactions on innovation diffusion by developing a new model called the interaction-based diffusion model. By comparing this model with previous diffusion models, the study also examines how the proposed model explains innovation diffusion from the perspective of economics. In addition, the study recommends the use of a small-world network topology instead of cellular automata to describe innovation diffusion. This study develops a model based on individual preference and heterogeneous social interactions using utility specification, which is expandable and, thus, able to encompass various issues in diffusion research, such as reservation price. Furthermore, the study proposes a new framework to forecast aggregated-level market demand from individual level modeling. The model also exhibits a good fit to real market data. It is expected that the study will contribute to our understanding of the innovation diffusion process through its microeconomic theoretical approach.

Keywords: innovation diffusion, agent based model, small-world network, demand forecasting

Procedia PDF Downloads 314

Authors: Isaac Mugume, Charles Basalirwa, Daniel Waiswa, Mary Nsabagwa, Triphonia Jacob Ngailo, Joachim Reuder, Sch¨attler Ulrich, Musa Semujju

Abstract:

The Numerical weather prediction (NWP) models are considered powerful tools for guiding quantitative rainfall prediction. A couple of NWP models exist and are used at many operational weather prediction centers. This study considers two models namely the Consortium for Small–scale Modeling (COSMO) model and the Weather Research and Forecasting (WRF) model. It compares the models’ ability to predict rainfall over Uganda for the period 21st April 2013 to 10th May 2013 using the root mean square (RMSE) and the mean error (ME). In comparing the performance of the models, this study assesses their ability to predict light rainfall events and extreme rainfall events. All the experiments used the default parameterization configurations and with same horizontal resolution (7 Km). The results show that COSMO model had a tendency of largely predicting no rain which explained its under–prediction. The COSMO model (RMSE: 14.16; ME: -5.91) presented a significantly (p = 0.014) higher magnitude of error compared to the WRF model (RMSE: 11.86; ME: -1.09). However the COSMO model (RMSE: 3.85; ME: 1.39) performed significantly (p = 0.003) better than the WRF model (RMSE: 8.14; ME: 5.30) in simulating light rainfall events. All the models under–predicted extreme rainfall events with the COSMO model (RMSE: 43.63; ME: -39.58) presenting significantly higher error magnitudes than the WRF model (RMSE: 35.14; ME: -26.95). This study recommends additional diagnosis of the models’ treatment of deep convection over the tropics.

Keywords: comparative performance, the COSMO model, the WRF model, light rainfall events, extreme rainfall events

Procedia PDF Downloads 237

Authors: Andres F. Ramirez, Carlos F. Valencia

Abstract:

The increasing interest in renewable energies strategies application and the path for diminishing the use of carbon related energy sources have encouraged the development of novel strategies for integration of solar energy into the electricity network. A correct inclusion of the fluctuating energy output of a photovoltaic (PV) energy system into an electric grid requires improvements in the forecasting and simulation methodologies for solar energy potential, and the understanding not only of the mean value of the series but the associated underlying stochastic process. We present a methodology for synthetic generation of solar irradiance (shortwave flux) and air temperature bivariate time series based on copula functions to represent the cross-dependence and temporal structure of the data. We explore the advantages of using this nonlinear time series method over traditional approaches that use a transformation of the data to normal distributions as an intermediate step. The use of copulas gives flexibility to represent the serial variability of the real data on the simulation and allows having more control on the desired properties of the data. We use discrete zero mass density distributions to assess the nature of solar irradiance, alongside vector generalized linear models for the bivariate time series time dependent distributions. We found that the copula autoregressive methodology used, including the zero mass characteristics of the solar irradiance time series, generates a significant improvement over state of the art strategies. These results will help to better understand the fluctuating nature of solar energy forecasting, the underlying stochastic process, and quantify the potential of a photovoltaic (PV) energy generating system integration into a country electricity network. Experimental analysis and real data application substantiate the usage and convenience of the proposed methodology to forecast solar irradiance time series and solar energy across northern hemisphere, southern hemisphere, and equatorial zones.

Keywords: copula autoregressive, solar irradiance forecasting, solar energy forecasting, time series generation

Procedia PDF Downloads 295

Authors: J. P. Panda, K. Sasmal, H. V. Warrior

Abstract:

Eddy viscosity models in turbulence modeling can be mainly classified as linear and nonlinear models. Linear formulations are simple and require less computational resources but have the disadvantage that they cannot predict actual flow pattern in complex geophysical flows where streamline curvature and swirling motion are predominant. A constitutive equation of Reynolds stress anisotropy is adopted for the formulation of eddy viscosity including all the possible higher order terms quadratic in the mean velocity gradients, and a simplified model is developed for actual oceanic flows where only the vertical velocity gradients are important. The new model is incorporated into the one dimensional General Ocean Turbulence Model (GOTM). Two realistic oceanic test cases (OWS Papa and FLEX' 76) have been investigated. The new model predictions match well with the observational data and are better in comparison to the predictions of the two equation k-epsilon model. The proposed model can be easily incorporated in the three dimensional Princeton Ocean Model (POM) to simulate a wide range of oceanic processes. Practically, this model can be implemented in the coastal regions where trasverse shear induces higher vorticity, and for prediction of flow in estuaries and lakes, where depth is comparatively less. The model predictions of marine turbulence and other related data (e.g. Sea surface temperature, Surface heat flux and vertical temperature profile) can be utilized in short term ocean and climate forecasting and warning systems.

Keywords: Eddy viscosity, turbulence modeling, GOTM, CFD

Procedia PDF Downloads 174