Synthetic generation of streamflow data facilitates the planning and operation of water resource projects. Significance of streamflow forecasting for intermittent river increases many fold in order to use available water yearlong for multipurpose water resources project. In the present study, monthly streamflow data has been used for intermittent river Goi in Narmada riverbasin. The performance of stochastic stream flow generation models– seasonal autoregressive integrated moving average (SARIMA) and Thomas-Fiering model are compared with Artificial Neural Network (ANN) approach. The performance of these models is evaluated on the basis of root mean square error (RMSE) and coefficient of determination (R 2 ). The study reveals that SARIMA performs better than Thomas-Fiering and ANNmodels. Thomas Fiering model is least reliable model among other two models. However Thomas-Fiering model performed well in case of high flow prediction whereas SARIMA and ANN performed well for lower and moderate flow. The predicted data can be used for the small hydropower projects development.
In this study, a simplified ANN based decomposed streamflow model is developed. The proposed data de- composition does not require any prior knowledge or understanding of physical processes. In this study the data is divided into two states namely rise and fall, based on the current state. The performance of the proposed SD-ANN model is compared to that of the feed-forward ANN model in terms of statistical indices such as coeffi- cient of correlation, coefficient of efficiency and root means square error. The exercise was carried out for the hourly data in Kolar riverbasin, India. It is observed that the proposed SD-ANN model and the FF-ANN model show similar results in terms of statistical indices except the case of RMSE where the former outperforms the lat- ter. Further, the SD-ANN model outperforms the FF- ANN model in prediction of high flows. The results show the significance of the streamflow decomposition when compared to single hydrograph. The performance of the SD-ANNmodels has to be tested on various time scales. Further extensions of this model can be examined to improve the forecasting accuracy .
In the present work, a novel and the robust computational investigation is carried out to estimate solubility of different acids in supercritical carbon dioxide. Four different algorithms such as radial basis function artificial neural network, Multi-layer Perceptron artificial neural network, Least squares support vector machine and adaptive neuro-fuzzy inference system are developed to predict the solubility of different acids in carbon dioxide based on the temperature, pressure, hydrogen number, carbon number, molecular weight, and acid dissociation constant of acid. In the purpose of best evaluation of proposed models, different graphical and statistical analyses and also a novel sensitivity analysis are carried out. The present study proposed the great manners for best acid solubility estimation in supercritical carbon dioxide, which can be helpful for engineers and chemists to predict operational conditions in industries.
σ = ∫ (6) As the groundwater movement is much slower than the surface water and near surface water system movements and the understandings about the bedrock is little, groundwater flow is simplified as a lumped linear reservoir in small GIS derived subwatershed scale. With considering the river damping effect for all flow components, overland flow and interflow are directed firstly from each grid cell to the main channel, and are joined with groundwater flow at the outlet of the subwatershed. Then the total hydrograph is routed to the outlet of the basin by the channel response function derived from Equation (4). The amount of to- tal discharge is sum of the overland flow, interflow, and groundwater flow, and is obtained by convolution of the flow responses of all grid cells. One advantage of this approach is allowing to the spatially distributed runoff and hydrological parameters of the basin for using as inputs for the model. Inputs of the model consist of digital elevation data, soil type, land use data, and measured climato- logical data. Stream discharge data are optional for model calibration. All hy- drological processes are simulated within a GIS framework. Because a large part of the annual precipitation is in the form of snow, snow melt simulating is done by a model based on hourly temperature data. The conceptual temperature in- dex or degree-day method is used in this study because of its simplicity but it has not a strong physical foundation. The method replaces the full energy balance with a term linked to air temperature. It is physically sound in the absence of shortwave radiation when much of the energy supplied to the snowpack is at- mospheric long wave radiation  . The equation is as follow:
The issue of prediction of various acids solubility in supercritical carbon dioxide and phase equilibrium investigation of supercritical carbon dioxide and different materials are the important topics in chemical engineering research. According to the hardships of experimental studies such as special tools and procedure which are needed, in the present work, the mathematical investigation is considered as a great solution for these problems (Anitescu, Atroshchenko et al. 2019, Guo, Zhuang et al. 2019, Rabczuk, Ren et al. 2019, Zarei, Razavi et al. 2019). In this paper four different algorithms, Radial basis function artificial neural network (RBF-ANN), Multi-layer Perceptron artificial neural network (MLP-ANN), Least squares support vector machine (LSSVM) and Adaptive neuro-fuzzy inference system (ANFIS) are developed to predict the solubility of different types of acid in supercritical carbon dioxide based on the various parameters such as structure of acid, pressure and temperature.
From Tables 4–6 and Figs. 6–9 , it is clear that the predicted outputs by the proposed ANFIS and ANNmodels are close to the experimental results, which shows the applicability of ANN and ANFIS as the accurate and reliable models for the modeling of the Squirrel-cage type induction motors. Also, from the obtained results it can be seen that the proposed ANFISmodels are more accurate than the proposed ANNmodels. The proposed ANFIS model is more accurate than the proposed ANN model. The ANFIS model could signifi- cantly reduce the average (the training and testing in the Single-cage motor) MRE% to less than 1.97% in comparison with the ANN model, which has the average (training and test- ing) MRE% less than 14.37%. In this case, the average CF for both training and testing data is greater than 0.99623 and 0.7795 for the proposed ANFIS and ANNmodels, respec- tively. For the Double-cage motor, the ANFIS model could significantly reduce the average (the training and testing) MRE% to less than 2.82% in comparison with the ANN model, which has the average (training and testing) MRE% less than 9.68%. In this case, the average CF for the both training and testing data is greater than 0.97000 and 0.95380 for the proposed ANFIS and ANNmodels, respectively. Also, the training process of the ANFIS model to reach the conver- gence is faster. In comparison with the proposed ANNmodels, the time needed for the training the ANFISmodels was less than the required time to design the proposed ANNmodels. The obtained set of parameters for the used motors can be effectively used to calculate the torque, power and power fac- tor accurately. To show it, one of the used induction motors (30 kW) is selected to calculate the maximum torque T M , the
An Adaptive Neuro-Fuzzy Inference System (ANFIS) refers, in general, to an adaptive network which performs the function of a fuzzy inference system. The most commonly used fuzzy system in ANFIS architectures is the Sugeno model since it is less computationally exhaustive and more transparent than other models. A consequent membership function (MF) of the Sugeno model could be any arbitrary parameterized function of the crisp inputs, most likely a polynomial. Zero and first order polynomials are used as consequent MF in constant and linear Sugeno models, respectively. In addition, the defuzzification process in Sugeno fuzzy models is a simple weighted average calculation. The fuzzy space is divided via grid partitioning according to the number of antecedent MF, and each fuzzy region is covered with a fuzzy rule. On the other hand, each fixed and adaptive node of the network performs one function or sub-function of the Sugeno model, as shown in Fig. 3, such that the overall performance of the network is functionally the same as that of the fuzzy model (Pothiya and Tantaswadi, 2006).
Usually, in Bioinformatics a challenge exists in endeavors seeking to predict and identify the location of promoter regions in the DNA molecule. In such quest to investigate on the same, this paper finds the ANN and ANFIS techniques as being potential eliminators of such challenge. Through the computational ANN, a model based on the structure and functions of biological neural networks is established. In this, information is expected to flow through the network structure of the ANN because a neural network changes - or learns, in a sense - based on that input and output. It will involve a hybrid learning procedure used to form an input-output mapping based on the training data pairs, in addition, employing a fuzzy inference system in the framework of adaptive networks. Thus, the writer finds it of significance that employing ANN for promoter prediction leads to promising results and leads to improvements, increasing the accuracy of the results obtained.
Use of glacier volume change in the calibration procedure ef- fectively reduced parameter uncertainty and helped to ensure that the model was accurately predicting glacier mass bal- ance as well as streamflow. This approach should be widely useful for quantifying glacier contributions to streamflow in glacier-fed catchments where mass balance observations are lacking. One drawback to the approach is that the calibration period must span the interval between glacier maps, in this case 15 yr. Because glacier cover can change significantly over decadal and longer time periods – historically and in future projections – approaches should be developed to ac- commodate glacier cover changes during model simulations. Although glaciers only cover 5 % of the Mica basin, they contributed up to 25 % of mean August flow and 35 % of mean September flow in the historic period. These contri- butions are particularly important during periods of warm, dry weather following winters with low accumulation and early snowpack depletion. Glacier retreat over the twenty- first century could therefore have significant implications for streamflow during critical late-summer periods.
For each of the 54 forested grid cells, if a grid cell contained land cover classification #1-6, then the value of the fractional coverage (Cv) for that particular land cover type was decreased by the specified amount for whichever of the five simulations were currently being performed. If a grid cell contained multiple types of forested vegetation, which occurred in most of the grid cells, then the percent decrease was divided by the total number of forest cover type classes #1-6 and then evenly distributed amongst the total number of forested land cover types within the grid cell. Once this was completed, then the total percent decrease that was applied for that particular scenario was then added to a less influential hydraulic land cover class, which for this analysis was assumed to be grasslands. In order to try to rationalize what was occurring in the NPRB, if a forest was initially in place and then was affected by beetle kill, in the process of modeling in VIC, it was concluded that the forest would basically transition from forest to grassland. In order to simulate decreasing forest cover in VIC, because all the total fractional coverage’s for each land cover type in a grid cell had to sum to one, the above mentioned procedure was carried out for each of the 54 forested grid cells. In essence, for modeling purposes, it was assumed that a forest affected by beetle kill would become similar to that of grassland, thus enabling a greater potential for surface water runoff to occur and potentially increased streamflow. The above rationale was applied to each of the five previously mentioned simulations by use of VIC and modeled streamflow results produced.
The present study evaluates the impacts of four precipitation datasets on stream ﬂ ow simulation at the Irrawaddy RiverBasin in Myanmar. For stream ﬂ ow simulation, we used the most widely used hydrological model SWAT (Soil Water Assessment Tool) (Arnold et al., 1998; Neitsch et al., 2011). Out of the four precipitation datasets used, two sets are derived from the available in-situ mea- surements (with and without interpolation), while the other two are satellite-based products. The reason for using two datasets from the same in-situ gauge measurements is governed by the way precipitation input is handled in SWAT. In its default setting, SWAT assigns one precipitation gauging station to each sub-basin which is closest to basin’s centroid, irrespective of the number of gauge stations available. This means that the same gauging stations may be assigned to more than one sub-basin while some gauging stations may not be assigned to any. As a result, the precipitation input to the model varies when the number of sub-basins varies. One way to avoid this issue is to spatially interpolate the gauge observations over the basinusing a ﬁxed grid and then to assign inter- polated values to each sub-basin. The two remote sensing based products used in this study are PERSIANN-CDR (Ashouri et al., 2015) and CHIRPS (Funk et al., 2015). PERSIANN-CDR is an abbreviation for Precipitation Estimation from Remote Sensing Information using Arti ﬁ cial Neural Network – Climate Data Record and CHIRPS stands for Climate Hazard Group Infrared Precipitation with Station data. The main reason for choosing these two products was their successes reported in recent studies (Ashouri et al., 2016; Casse and Gosset, 2015; Ceccherini et al., 2015; Guo et al., 2015; Le and Pricope, 2017; Shrestha et al., 2017; Zhu et al., 2017). Simulations with PERSIANN-CDR datasets have shown reasonably good performance in stream ﬂ ow prediction at two river basins (the Yangtze and the Upper Yellow River Basins) in the Tibetan Plateau (Liu et al., 2017). Based on diﬀerent extreme precipitation indices, PERSIANN-CDR data is capable in reproducing daily precipitation extremes of similar spatial and temporal patterns to those produced by East Asia (EA) ground-based gridded daily precipitation datasets (Miao et al., 2015). The CHIRPS product has shown good agreement with ground-based rain gauge data all over Cyprus (Katsanos et al., 2016). Hassels (2015) compared several open access satellite-based precipitation products for the Nile RiverBasin and recommends CHIRPS as one of the best products available for hydrological studies in that region. Stream ﬂ ow simulations with CHIRPS data have provided satisfactory results for Alpine catchments as well (Tuo et al., 2016).
The FRB is one of the largest watersheds draining the western slopes of the North American Cordillera, and home to both densely populated urban centres and a diversity of ecosystems closely linked to the economic prosperity of the region (Fraser Basin Council, 2010). The basin is also a fo- cal point for salmon migration and spawning, constituting a key component of the highly valued commercial and subsis- tence fisheries of the region. The FRB lies between the Coast Mountains and the Continental Divide, with the headwaters of the Fraser in the northeast ( ∼ 53 ◦ N, 118 ◦ W) and its out- let at the Pacific Ocean in the southwest (Fig. 1). Its vast 240 000 km 2 area encompasses a range of climatic zones, from the snowy mountains of the eastern Rockies to dry interior plateaus and wet fertile valleys nearest the Pacific west coast. The hydrological diversity of the basin is well described in Eaton and Moore (2010), who reviewed sea- sonal streamflow regimes at the catchment scale across BC, and also in Moore (1991), who focused specifically on the FRB. For the most part, streamflows are unregulated in the FRB, with the exception of Kenney Dam on the Nechako River in the northwestern sector of the FRB (operational since the early 1950s) and several dams associated with the Bridge River power project in the Interior Plateau (completed in 1960). Regulated subbasins represent less than 10 % of the total area of the FRB (Bawden et al., 2015).
There are three methods of predicting runoff hydrographs in ungauged basins (Parajka et al., 2013). The first is to estimate model parameters from basin characteristics a priori. Although some parameters of completely physics-based models can be well constrained, their performance is generally worse than that of cal- ibrated models (Duan et al., 2006). The second method is to cali- brate model parameters for well gauged basins and then apply them to an ungauged basin of interest by extrapolation of the model parameters along with basin characteristics while assuming hydrological similarity between basins. Extrapolations tend to be more reliable if process realisms are held (Tetzlaff et al., 2013); however, it is not easy to confirm the realisms without observa- tion. The third method is to calibrate modelsusing observed proxy data other than (but intrinsically associated with) discharge. For
emotions as well as the pure emotions of the basic layer. The next layer is further organized as collection of pure and primary emotion nodes along with the secondary emotion nodes. Strictly speaking, each stage can be derived directly from the first, but the stages are retained for the purpose of displaying the results of analysis at different granularities (Fig. 1).This layered network of nodes has two objectives: ease of construction and modification, and the ability to distil the data in the first layer into later layers. Thus a current layer node can derive values as a function of the nodes of the previous layer it is connected to and its link weight. As distillation is an operation meaningful only when at least two layers are present, it is layers, not nodes, which contain a reference to the previous layer. The basic structure (GE) is extended into a larger graph during the ANN process and analysis steps. The extracted images of facial expression from the video input and the textual document are checked for emotion content and then linked to the predefined set of Nodes
It is widely reported that watershed models are cost-effective and time-efficient measures for the assessment of pollutant loads and simulation of watershed processes and management practices in an effort to address non-point source pollution (Baginska et al., 2003; Shrestha et al., 2006; Li et al., 2015). Several watershed-scale hydrological and water quality models reviewed in Borah and Bera (2003), such as AnnAGNPS (Annualized Agricultural Non-Point Source) and SWAT (Soil and Water Assessment Tool) have been developed to evaluate the hydrologic and water quality responses of a watershed to alternative management practices and to understand hydrologic systems and pollutant loadings in recent years (Bingner et al., 2014; Arnold and Allen, 1999; Hua et al., 2012; Yuan et al., 2003). Also, these models can be used to predict the amount and effect of NPS as well as erosion from watersheds. These pollutants have a major impact on water quality especially in rural environments. In fact, watershed models can help to select suitable land uses and the best management practices to reduce the damaging effects of agricultural practices on the environment (Chahor et al., 2014). In general, there is no single best model for all applications but the most appropriate model will depend on the intended use and the characteristics of the watershed under study (Shamshad et al., 2008).
ABSTRACT: Stock Market is the market for security where organized trading of Stocks takes place either through exchange or over the counter in electrical or physical form. It plays an important role in canalizing capital from the investors to the business houses. As the stock prices get fluctuated every second it becomes very difficult to predict the stock price. Many have tried predicting the stock market, but very few have succeeded. It becomes difficult to guess or predict what would be stock price next second.We are implementing the predictionmodels of Artificial Neural Network (ANN) to predict the future stock price. The accuracy of models while predicting the stock price is tested with the real price which will be helpful for shareholder to predict the stock price of stocks.
it is critical to carefully screen and check their quality be- fore using them for analyses. Hence, the raw data were vi- sually inspected and screened for typos and outliers. Each station was carefully checked for data consistency by com- parison to nearby, upstream and downstream stations. Re- lationships between neighbouring stations can give prelim- inary evidence on the reliability of time series data, provided that there is no man-made water storage above the station (Hong et al., 2009). Identified unreliable data were fixed af- ter comparing its upper and lower boundary limits. Further- more, heterogeneity of the time series data was also detected using the double mass curve and residual mass plot methods. The monthly hydrological flow data were aggregated from the daily data and the seasonal and annual data were calcu- lated from the monthly data. Unreliable data were removed, whilst more stations were included to increase spatial cover- age. Missing data (> 2 years) were excluded from the analy- ses. However, during the peak rainy season, missing data for more than 2 weeks were excluded from the analyses. The rea- son to exclude only 2 weeks was to minimize untrustworthy data as more than 80 % of the river flow is generated during only two months (July and August).
The partial-pooling hierarchical Bayesian regression model provided a useful way to model spatial co-variability in seasonal hydrological predictions, while considering the potentially common effects of the predictors on regional hy- drologic response. An advantage of the approach is that it al- lows appropriate grouping of information in the region, and explicit modeling of the covariance of the model errors and the regression coefficients to better represent the uncertainty in both the model parameters and the final streamflow and rainfall forecasts. Cross-validated model results show good predictive skill, and the common effects as well as the at site effects of each predictor are identified well even un- der leave-10-out of 50 cross-validation. Comparison of the partial-pooling model with a no-pooling model in the same estimation framework (equivalent to the traditional regres- sion based modeling framework) showed that the Bayesian model is competitive or superior in terms of the validation statistics. An aspect of Bayesian modeling that is often cited is the ability to provide a systematic approach to the propa- gation and modeling of model parameter uncertainty. In this context, we argue that while there could be a potential up- per limit in predictability in seasonal rainfall using anoma- lous SST conditions as shown in Westra and Sharma (2010),
Watershed models contain many parameters; these parame- ters are classified into two groups: physical and process pa- rameters. A physical parameter represents physically mea- surable properties of the watershed (e.g. areas of the catch- ment, fraction of impervious area and surface area of wa- ter bodies, surface slope etc) while process parameters rep- resents properties of the watershed which are not directly measurable e.g. average or effective depth of surface soil moisture storage, the effective lateral inflow rate, the coef- ficient of non-linearity controlling the rate of percolation to the groundwater (Sorooshian and Gupta, 1995). Thus, cali- brations against available streamflow observations are often conducted to tune the model. Because automatic calibration relies heavily on the optimization algorithm and the spec- ified objective function we follow the recommendations of Gan (1998) to use both manual and automatic calibration procedures. We first conducted manual calibration of daily stream using the procedure developed by Santhi et al. (2001). Parameters identified from the sensitivity analysis were var- ied in sequence of their relative sensitivity within their ranges (Table 4) until the volume is adjusted to the required quan- tity (Zeray et al., 2007).This process continued till the vol- ume simulated is within ± 15 % of the gauged volume.The surface runoff adjustment was then followed by that of the baseflow. Here,the same apporach was followed being the adjustment made to the most sensitivity parameters affecting the baseflow. Each time the baseflow calibration is finalized, the surface runoff volume was also checked as adjustment of the baseflow parameters can also affect the surface runoff volume. The same procedure was followed to calibrate the water balance of the monthly flows. After each calibartion, the coefficient of determintation (proportion of the variance in the observations explained by the model, R 2 ) and Nash- Sutcliffe effficeny value (ENS) were checked (R 2 > 0.6 and