Predicting Slump Values of Concrete Made by Pozzolans and Manufactured Sand using ANN

Large amounts of natural fine aggregate (NFA) and cement are used in building, which has major environmental consequences. This view of industrial waste can be used in part as an alternative to cement and part of the sand produced by the crusher as fine aggregate, similar to slag sand (GGBFS), fly ash, metacaolin, and silica fume. Many times, there are issues with the fresh characteristics of concrete when using alternative materials. The ANN tool is used in this paper to develop a Matlab software model that collapses concrete made with pozzolanic material and partially replaces natural fine aggregate (NFA) with manufactured sand (MS). Predict. The slump test was carried out in reference with I.S11991959, and the findings were used to create the artificial neural network (ANN) model. To mimic the formation, a total of 131 outcome values are employed, with 20% being used for model testing and 80% being used for model training. 25 enter the material properties to determine the concrete slump achieved by partially substituting pozzolan for cement and artificial sand (MS) for natural fine aggregate (NFA). According to studies, the workability of concrete is critically harmed as the amount of artificial sand replacing natural sand grows. The ANN model's results are extremely accurate, and they can forecast the slump of concrete prepared by partly substituting natural fine aggregate (NFA) and artificial sand (MS) with pozzolan.

Prediction of Rate of Penetration for wells at Nam Con Son basin using Artificial Neural Networks models

The rate of penetration (ROP) is an important parameter that affects the success of a drilling operation. In this paper, the research approach is based on different artificial neural network (ANN) models to predict ROP for oil and gas wells in Nam Con Son basin. The first is the process of collecting and evaluating drilling parameters as input data of the model. Next is to find the network model capable of predicting ROP most accurately. After that, the study will evaluate the number of input parameters of the network model. The ROP prediction results obtained from different ANN models are also compared with traditional models such as the Bingham model, Bourgoyne & Young model. These results have shown the competitiveness of the ANN model and its high applicability to actual drilling operations.

Artificial neural network model for precipitation forecast over Western Himalaya using satellite images

Kalpana satellite images in real time available by India meteorological department (IMD), contain relevant inputs about the cloud in infra-red (IR), water vapor (WV), and visible (VIS) bands. In the present study an attempt has been made to forecast precipitation at six stations in western Himalaya by using extracted grey scale values of IR and WV images. The extracted pixel values at a location are trained for the corresponding precipitation at that location. The precipitation state at 0300 UTC is considered to train the model for precipitation forecast with 24 hour lead time. The satellite images acquired in IR (10.5 - 12.5 µm) and WV (5.7 - 7.1 µm) bands have been used for developing Artificial Neural Network (ANN) model for qualitative as well as quantitative precipitation forecast. The model results are validated with ground observations and skill scores are computed to check the potential of the model for operational purpose. The probability of detection at the six stations varies from 0.78 for Gulmarg in Pir-Panjal range to 0.95 for Dras in Greater Himalayan range. Overall performance for qualitative forecast is in the range from 61% to 84%. Root mean square error for different locations under study is in the range 5.81 to 8.7.

Application of Artificial Neural Networks for Mangrove Mapping Using Multi-Temporal and Multi-Source Remote Sensing Imagery

Mangroves, as unique coastal wetlands with numerous benefits, are endangered mainly due to the coupled effects of anthropogenic activities and climate change. Therefore, acquiring reliable and up-to-date information about these ecosystems is vital for their conservation and sustainable blue carbon development. In this regard, the joint use of remote sensing data and machine learning algorithms can assist in producing accurate mangrove ecosystem maps. This study investigated the potential of artificial neural networks (ANNs) with different topologies and specifications for mangrove classification in Iran. To this end, multi-temporal synthetic aperture radar (SAR) and multi-spectral remote sensing data from Sentinel-1 and Sentinel-2 were processed in the Google Earth Engine (GEE) cloud computing platform. Afterward, the ANN topologies and specifications considering the number of layers and neurons, learning algorithm, type of activation function, and learning rate were examined for mangrove ecosystem mapping. The results indicated that an ANN model with four hidden layers, 36 neurons in each layer, adaptive moment estimation (Adam) learning algorithm, rectified linear unit (Relu) activation function, and the learning rate of 0.001 produced the most accurate mangrove ecosystem map (F-score = 0.97). Further analysis revealed that although ANN models were subjected to accuracy decline when a limited number of training samples were used, they still resulted in satisfactory results. Additionally, it was observed that ANN models had a high resistance when training samples included wrong labels, and only the ANN model with the Adam learning algorithm produced an accurate mangrove ecosystem map when no data standardization was performed. Moreover, further investigations showed the higher potential of multi-temporal and multi-source remote sensing data compared to single-source and mono-temporal (e.g., single season) for accurate mangrove ecosystem mapping. Overall, the high potential of the proposed method, along with utilizing open-access satellite images and big-geo data processing platforms (i.e., GEE, Google Colab, and scikit-learn), made the proposed approach efficient and applicable over other study areas for all interested users.

Incremental Learning with Neural Network Algorithm for the Monitoring Pre-Convective Environments Using Geostationary Imager

Early warning of severe weather caused by intense convective weather systems is challenging. To help such activities, meteorological satellites with high temporal and spatial resolution have been utilized for the monitoring of instability trends along with water vapor variation. The current study proposes a retrieval algorithm based on an artificial neural network (ANN) model to quickly and efficiently derive total precipitable water (TPW) and convective available potential energy (CAPE) from Korea’s second geostationary satellite imagery measurements (GEO-KOMPSAT-2A/Advanced Meteorological Imager (AMI)). To overcome the limitations of the traditional static (ST) learning method such as exhaustive learning, impractical, and not matching in a sequence data, we applied an ANN model with incremental (INC) learning. The INC ANN uses a dynamic dataset that begins with the existing weight information transferred from a previously learned model when new samples emerge. To prevent sudden changes in the distribution of learning data, this method uses a sliding window that moves along the data with a window of a fixed size. Through an empirical test, the update cycle and the window size of the model are set to be one day and ten days, respectively. For the preparation of learning datasets, nine infrared brightness temperatures of AMI, six dual channel differences, temporal and geographic information, and a satellite zenith angle are used as input variables, and the TPW and CAPE from ECMWF model reanalysis (ERA5) data are used as the corresponding target values over the clear-sky conditions in the Northeast Asia region for about one year. Through the accuracy tests with radiosonde observation for one year, the INC NN results demonstrate improved performance (the accuracy of TPW and CAPE decreased by approximately 26% and 26% for bias and about 13% and 12% for RMSE, respectively) when compared to the ST learning. Evaluation results using ERA5 data also reveal more stable error statistics over time and overall reduced error distribution compared with ST ANN.

Proposing Several Model Techniques Including ANN and M5P-tree to Forecast the Stress at the Failure of Geopolymer Concrete Mixtures Incorporated Nano-silica

Geopolymers are innovative cementitious materials that can completely replace traditional Portland cement composites and have a lower carbon footprint than Portland cement. Recent efforts have been made to incorporate various nanomaterials, most notably nano-silica (nS), into geopolymer concrete (GPC) to improve the composite's properties and performance. Compression strength (CS) is one of the essential properties of all types of concrete composites, including geopolymer concrete. As a result, creating a credible model for forecasting concrete CS is critical for saving time, energy, and money, as well as providing guidance for scheduling the construction process and removing formworks. This paper presents a large amount of mixed design data correlated to mechanical strength using empirical correlations and neural networks. Several models, including artificial neural network, M5P-tree, linear regression, nonlinear regression, and multilogistic regression models were utilized to create models for forecasting the CS of GPC incorporated nS. In this case, about 207 tested CS values were collected from literature studies and then analyzed to promote the models. For the first time, eleven effective variables were employed as input model parameters during the modeling process, including the alkaline solution to binder ratio, binder content, fine and coarse aggregate content, NaOH and Na2SiO3 content, Na2SiO3/NaOH ratio, molarity, nS content, curing temperatures, and ages. The developed models were assessed using different statistical tools such as RMSE, MAE, SI, OBJ value, and R2. Results revealed that the ANN model estimated the CS of GPC incorporated nS more accurately than the other models. On the other hand, the alkaline solution to binder ratio, molarity, NaOH content, curing temperature, and ages were those parameters that have significant influences on the CS of GPC incorporated nS.

A Novel Artificial Neural Network-Based Correlation for Evaluating the Rate of Penetration in a Natural Gas Bearing Sandstone Formation: A Case Study in a Middle East Oil Field

This study presented an empirical correlation to estimate the drilling rate of penetration (ROP) while drilling into a sandstone formation. The equation developed in this study was based on the artificial neural networks (ANN) which was learned to assess the ROP from the drilling mechanical parameters. The ANN model was trained on 630 datapoints collected from five different wells; the suggested equation was then tested on 270 datapoints from the same training wells and then validated on three other wells. The results showed that, for the training data, the learned ANN model predicted the ROP with an AAPE of 7.5%. The extracted equation was tested on data gathered from the same training wells where it estimated the ROP with AAPE of 8.1%. The equation was then validated on three wells, and it determined the ROP with AAPEs of 9.0%, 10.7%, and 8.9% in Well-A, Well-B, and Well-D, respectively. Compared with the available empirical equations, the equation developed in this study was most accurate in estimating the ROP.