Quantification and Forecasting of Cumulonimbus (Cb) Clouds Direction, Nebulosity and Occurrence With Autoregression Using 2018-2020 Dataset From Yaounde-nsimalen

This works reports the quantification and forecasting of Cumulonimbus (Cb) clouds direction, nebulosity and occurrence with auto regression using 2018-2020 dataset from Yaoundé –Nsimalen of Cameroon. Data collected for October 2018-2020 consisted of 2232 hourly observations. Codes were written to automatically align, multi-find and replace data points in excel to facilitate treating big datasets. The approach included quantification of direction generating time series from data and determination of model orders using the correlogram. The coefficients of the SARIMA model were determined using Yule-Walker equations in matrix form, the Augmented Dickey Fuller test (ADF) was used to check for stationarity assumption, Portmanteau test to check for white noise in residuals and Shapiro-Wilk test to check normality assumptions. After writing several algorithms to test different models, an Autoregressive Neural Network (ANN) was fitted and used to predict the parameters over window of 2 weeks. Autocorrelation Function (ACF) shows no correlation between residuals, with p ≤ 0.05, fitting the stationarity assumption. Average performance is 80%. A stationary wavelike occurrence of the direction has been observed, with East as the most frequented sector. Forecast of Cb parameters is important in effective air traffic management, creating situational awareness and could serve as reference for future research. The method of decomposition could be made applicable in future research to quantify/forecast cloud directions.

Automatic time series forecasting using nonlinear autoregressive neural network model with exogenous input

This study aims to determine an automatic forecasting method of univariate time series, using the nonlinear autoregressive neural network model with exogenous input (NARX). In this automatic setting, users only need to supply the input of time series. Then, an automatic forecasting algorithm sets up the appropriate features, estimate the parameters in the model, and calculate forecasts, without the users’ intervention. The algorithm method used include preprocessing, tests for trends, and the application of first differences. The time series were tested for seasonality, and seasonal differences were obtained from a successful analysis. These series were also linearly scaled to [−1, +1]. The autoregressive lags and hidden neurons were further selected through the stepwise and optimization algorithms, respectively. The 20 NARX models were fitted with different random starting weights, and the forecasts were combined using the ensemble operator, in order to obtain the final product. This proposed method was applied to real data, and its performance was compared with several available automatic models in the literature. The forecasting accuracy was also measured by mean squared error (MSE) and mean absolute percent error (MAPE), and the results showed that the proposed method outperformed the other automatic models.

Modeling the Levelized Cost of Energy for Concentrating Solar Thermal Power Systems Based on a Nonlinear AutoRegressive Neural Network With Exogenous Inputs

This article presents the results of the development of a mathematical model for predicting the levelized cost of energy (LCOE) for solar concentrating thermal power systems (CSP systems) based on a nonlinear autoregressive neural network with exogenous inputs (NARX). A two-layer NARX network with sigmoid hidden neurons and linear output neurons has been developed. The input layer is made up of the following variables: the volume of input power of CSP systems in the world, the total world energy consumption, domestic energy consumption, domestic gas consumption, domestic consumption of coal and lignite, domestic energy consumption, the share of renewable energy in electricity generation, the share of wind and solar energy in the production of electricity, carbon dioxide emissions from fuel combustion, the price of Brent oil against the US dollar, and the average price for natural gas auctions. The output layer specifies LCOE values for CSP systems.

Autoregressive Neural Network EURO STOXX 50 Forecasting Model Based on Principal Component Stock Selection

Purpose: The given study looks into forecast accuracy of a traditional ARIMA model while comparing it to Autoregressive Neural Network (AR-NN) model for 984 trading days on EURO STOXX 50 Index. Methodology: A hybrid model is constructed by combining ARIMA model and feed-forward neural network model aiming to attain linear and non-linear price fluctuations. The study also incorporates the investigation of component stock prices of the index, that can be selected to improve the predictability of the hybrid model. Findings:The reached ARIMA (1,1,3) model showed higher scores than AR-NN model however integrating selected exogenous stock prices from the index components gave much notable accuracy results. The selected exogenous stocks were extracted after conducting PCA and model scores were compared via MAPE and RMSE. Unique contribution to theory, practice and policy: The major contribution of this work is to provide the researcher and fnancial analyst a systematic approach for development of intelligent methodology to forecast stock market. This paper also presents the outlines of proposed work with the aim to enhance the performance of existing techniques. Therefore, Empirical analysis is employed along with a hybrid model based on a feed-forward Neural Network. Lesser error is attained on the test set of Index stock price by comparing the performance of ARIMA and AR-NN while forecasting. Hence, The components of extracted Index stock price like exogenous features are added to make an influence from the AR-NN model.

SOH estimation method for lithium-ion batteries based on nonlinear autoregressive neural network model with exogenous input

The working state of lithium-ion batteries must be estimated accurately and efficiently in the battery management system. Building a model is the most prevalent way of predicting the battery's working state. Based on the variable order equivalent circuit model, this paper examines the attenuation curve of battery capacity with the number of cycles. It identifies the order of the equivalent circuit model using Bayesian Information Criterion (BIC). Based on the correlation between capacity and resistance, the paper concludes that there is a nonlinear correlation between model parameters and state of health (SOH). The nonlinear autoregressive neural network with exogenous input (NARX) is used to fit the nonlinear correlation for capacity regeneration. Then, the self-adaptive weight particle swarm optimization (SWPSO) method is suggested to train the neural network. Finally, single-battery and multi-battery tests are planned to validate the accuracy of the SWPSO-NARX estimate of SOH. The experimental findings indicate that the SOH estimate effect is significant.