Trend Analysis of Water Inflow Into the Dam Reservoirs Under Future Conditions Predicted By Dynamic NAR and NARX Models

Nowadays, the use of artificial intelligence is extended to various scientific and engineering fields including water management and planning. This study investigates the performance of dynamic artificial neural network (ANN) models in prediction of water inflow into the Sefidruod dam reservoir (Iran). For this purpose, first, the discharge time series of tributaries of the Sefidruod dam were analyzed for trends for a 47 year time period (1967 to 2014) using parametric regression and non-parametric Mann–Kendall tests considering independence, short-term, and long-term persistence assumptions. Also, the homogeneity of the data was investigated using three statistical tests including Cumulative Deviations, Worsley's Likelihood Ratio, and Bayesian inference. Then, the inflow discharges into the reservoir of Sefidruod dam from GhezelOzan and Shahroud tributaries were simulated using dynamic Nonlinear Auto-Regressive (NAR) and Nonlinear Auto-Regressive with exogenous input (NARX) models. Further, water inflow values of both rivers were predicted for the next 5 years in future using dynamic NAR and NARX models. Finally, the simulated results were tested for trends. Obtained results showed a significant decreasing trend in both rivers. Results also showed a continuous downward trend for the following 5-year period predicted by NAR and NARX models. In addition, it was found that the results obtained by the NARX model were less accurate compared to those by the NAR model.

     

Improving ANFIS Based Model for Long-term Dam Inflow Prediction by Incorporating Monthly Rainfall Forecasts

The necessity of long-term dam inflow forecast has been recognized for many years. Despite numerous studies, the accurate long-term dam inflow prediction is still a challenging task. This paper presents an adaptive neuro-fuzzy inference system (ANFIS) based model and evaluates the applicability of categorical rainfall forecast for improvement of monthly dam inflow prediction. In order to obtain appropriate ANFIS model configuration for dam inflow prediction, several models were trained and tested using various numbers of input variables i.e. monthly observed rainfall, relative humidity, temperature, dam inflow and categorical monthly rainfall forecast. The ANFIS based models were configured and evaluated for six major dams of South Korea i.e. Andong, Chungju, Daecheong, Guesan, Soyang and Sumjin having high, medium and low reservoir capacity. The results showed significant improvement in dam inflow prediction for all the selected dams using the ANFIS based model with categorical rainfall forecast compared to the ANFIS based model with only preceding month’s dam inflow and weather data.     

Reservoir Inflow Prediction by Ensembling Wavelet and Bootstrap Techniques to Multiple Linear Regression Model

In this study, a new hybrid model, bootstrap multiple linear regression (BMLR) is suggested to investigate the potential of bootstrap resampling technique for daily reservoir inflow prediction. The proposed model compares with three other models: Multiple linear regression (MLR), wavelet multiple linear regression (WMLR) and wavelet bootstrap multiple linear regression (WBMLR). River stage data of monsoon season (1st July 2010 to 30 September 2010) from three gauging stations of Chenab river basin are used. In wavelet transformation, input vectors are decomposed using discrete wavelet transformation (DWT) into discrete wavelet components (DWCs). Then suitable DWCs are used to provide input to MLR model to develop WMLR model. Bootstrap technique coupled with MLR model to build up BMLR model. While WBMLR model is the conjunction of suitable DWCs and bootstrap technique to MLR model. Performance indices namely root mean square error (RMSE), mean absolute error (MAE), Nash-Sutcliffe coefficient of efficiency (NSC), and persistence index (CP) are used in study to evaluate the performance of model. Results showed that hybrid model BMLR produce significantly better results on performance indices than other models MLR, WMLR and WBMLR.

     

Development of Nonlinear Model Based on Wavelet-ANFIS for Rainfall Forecasting at Klang Gates Dam

Rainfall is one of the most complicated effective hydrologic processes in runoff prediction and water management. Artificial neural networks (ANN) have been found efficient, particularly in problems where characteristics of the processes are stochastic and difficult to describe using explicit mathematical models. However, time series prediction based on ANN algorithms is fundamentally difficult and faces some other problems. For this purpose, one method that has been identified as a possible alternative for ANN in hydrology and water resources problems is the adaptive neuro-fuzzy inference system (ANFIS). Nevertheless, the data arising from the monitoring stations and experiment might be corrupted by noise signals owing to systematic and non-systematic errors. This noisy data often made the prediction task relatively difficult. Thus, in order to compensate for this augmented noise, the primary objective of this paper is to develop a technique that could enhance the accuracy of rainfall prediction. Therefore, the wavelet decomposition method is proposed to link to ANFIS and ANN models. In this paper, two scenarios are employed; in the first scenario, monthly rainfall value is imposed solely as an input in different time delays from the time (t) to the time (t-4) into ANN and ANFIS, second scenario uses the wavelet transform to eliminate the error and prepares sub-series as inputs in different time delays to the ANN and ANFIS. The four criteria as Root Mean Square Error (RMSE), Correlation Coefficient (R ²), Gamma coefficient (G), and Spearman Correlation Coefficient (ρ) are used to evaluate the proposed models. The results showed that the model based on wavelet decomposition conjoined with ANFIS could perform better than the ANN and ANFIS models individually.     

Daily Forecasting of Dam Water Levels: Comparing a Support Vector Machine (SVM) Model With Adaptive Neuro Fuzzy Inference System (ANFIS)

Reservoir planning and management are critical to the development of the hydrological field and necessary to Integrated Water Resources Management. The growth of forecasting models has resulted in an excellent model known as the Support Vector Machine (SVM). This model uses linearly separable patterns based on an optimal hyperplane, which are extended to non-linearly separable patterns by transforming the raw data to map into a new space. SVM can find a global optimal solution equipped with Kernel functions. These Kernel functions have high flexibility in the forecasting computation, enabling data to be mapped at a higher and infinite-dimensional space in an implicit manner. This paper presents a new solution to the expert system, using SVM to forecast the daily dam water level of the Klang gate. Four categories are identified to determine the best model: the input scenario, the type of SVM regression, the number of V-fold cross-validation and the time lag. The best input scenario employs both the rainfall R(t-i) and the dam water level L(t-i). Type 2 SVM regression is selected as the best regression type, and 5-fold cross-validation produces the most accurate results. The results are compared with those obtained using ANFIS: all the RMSE, MAE and MAPE values prove that SVM is a superior model to ANFIS. Finally, all the results are combined to determine the best time lag, resulting in R(t-2) L(t-2) for the best model with only 1.64 % error.     

基于LSTM-Arima的大坝变形组合预测模型及其应用

大坝变形是水压、温度等多种因素综合作用的结果,变形监测数据是非平稳非线性的时间序列,并且在时间维度上具有关联性。为充分挖掘变形监测数据在长短时间跨度上的关联性,提出了应用长短期记忆网络(LSTM)预测大坝变形的方法。为进一步提升预测精度,利用自回归差分移动平均模型(Arima)对预测残差进行误差修正,从而建立基于LSTM-Arima的大坝变形组合预测模型。以某混凝土重力坝为例,将组合模型的预测结果与Arima模型、支持向量机(SVM)的预测结果进行对比分析。结果表明LSTM-Arima的预测结果优于Arima模型和SVM的预测结果,LSTM-Arima的均方根误差(RMSE) 比Arima模型和SVM分别降低了40.65%和59.00%,平均绝对误差(MAE)分别降低了35.49%和55.60%,表明LSTM-Arima模型具有较高的预测精度。研究成果对于更精确地开展大坝变形预测有一定参考价值。     

基于改进GWO和SVM的大坝变形预测

建立准确可靠的大坝变形预测模型是大坝安全评价的重要内容,为此,将差分进化算法的交叉和变异算子引入灰狼优化算法(GWO),提出一种基于改进灰狼算法(MGWO)优化支持向量机(SVM)的大坝变形预测方法。通过差分进化算法丰富初始种群,提出改进灰狼优化算法(MGWO),并采用MGWO算法优化SVM的惩罚因子和核函数,建立基于MGWO-SVM算法的大坝变形预测模型。以锦屏一级特高拱坝实测数据为例,将MGWO-SVM模型与SVM、GWO-SVM模型的预测结果进行比较。结果表明,MGWO-SVM模型可以有效提高大坝变形预测精度。     

A Multi-Objective Risk-Based Game Theoretic Approach to Reservoir Operation Policy in Potential Future Drought Condition

In this paper, by using the concept of Conditional Value at Risk (CVaR), a Leader-Follower game (LFG) based multi-objective optimization model is developed to determine the optimum 12-month operation policy of a reservoir in potential future dry periods. The minimization of CVaRs of storage loss and agricultural and environmental deficits along with maximization of planned allocation to agricultural sector are considered as leader’s objectives, while the followers try to maximize their share of water rights using Nash bargaining (NB) method. This framework is then used to model the operation policy of Dorudzan basin in Fars province, southwestern Iran. Water demand and daily climate data in the period of 2003 to 2015 for this basin, as well as future projections from fifteen IPCC-AR4 global circulation models (GCMs) for 2018–2030 under A2, B1 and A1B emission scenarios are considered to evaluate future dam operation policies. Future projections are downscaled using the LARS-WG model, which then feeds the HMETS watershed model to simulate the corresponding reservoir inflow time-series. Thereafter, three-hundred 12-month rainfall, evaporation and inflow time series with least inflow volume are used as input for the optimization model, which is solved using NSGA-II and GA algorithms. The results show while the model can determine the operation policy that keeps the associated risks in the acceptable range, it can satisfy the followers demands with respect to the available resources. The results also show that the agricultural sector of the study area can be hugely affected by potential future droughts.

     

Two-step daily reservoir inflow prediction using ARIMA-machine learning and ensemble models

The reservoirs play a crucial role in the development of civilisation as they facilitate the storage of water for multiple purposes like hydroelectric power generation, flood control, irrigation, and drinking water etc. In order to effectively meet these multiple purposes, the knowledge of the inflow in the reservoir is essential. Apart from the historical data, future prediction of the inflows is also necessary specially in context of climate change. A two-step algorithm for the prediction of reservoir inflow to enable meticulous planning and execution of daily reservoir operation keeping the historical variation of inflow in account has been proposed. The developed algorithm takes into account the patterns in the historic inflow data using the time series analysis along with the variability in the climatic patterns using the different predictors in the machine learning model. The first step uses time series model, ARIMA method to forecast the monthly inflows, which are then used as the targets in the second step for the month-wise daily forecasting of the inflows using the two types of ensemble models, namely, averaging and boosting models in machine learning. The test results show that for both the monthly models and daily models the NRMSE and NMAE values were low for the monsoon periods compared to the non-monsoon periods. The averaging ensemble models were found to perform better than the boosting ensemble models for maximum number of months. The yearly results show an error of less than 5% between actual and predicted values for all the test cases, showing the precision in the developed algorithm. Further, the uncertainty analysis shows that the prediction done using the weighted average of the different inflow scenarios performs better than the prediction against the single inflow scenario.     

基于PSO-SVM的碾压混凝土坝变形预测模型

为提升大坝变形预测能力,提出了一种基于粒子群算法(PSO)优化支持向量机(SVM)的混凝土重力坝变形预测模型.通过粒子群算法对支持向量机惩罚函数C与核函数σ进行寻优,避免了拟合过程中易陷入局部最优解的问题,提高了模型的拟合精度.以新疆北疆某碾压混凝土坝2014年～ 2019年变形监测数据为例,建立了逐步回归、SVM、P...     

考虑黏弹性滞后效应的拱坝位移监控组合模型

位移监控模型需要对拱坝变形性态兼具良好的解释和预测能力。水压-滞后-周期性温度-时效四因子HHST(Hydraulic,Hysteretic,Seasonal and Time)模型能够合理地解释锦屏一级拱坝的黏弹性滞后变形性态。为进一步提升该模型的预测精度,使用支持向量机(SVM)建立有限元计算所得拱坝黏弹性滞后位移与其因果因子之间的隐式关系,再将其融入到HHST模型中,进而基于多元线性回归建立拱坝位移的组合监控模型。以锦屏一级拱坝为例,减少输入因子数的组合模型的预测精度明显高于直接以HHST模型中18个因子作为输入的单一模型;SVM对滞后水压位移分量的预测精度明显高于基于约束最小二乘法的线性回归模型,采用2种滞后水压分量所建组合模型对拱坝变形性态具有相近的解释能力,而采用SVM滞后水压分量建立的组合模型可有效地提高拱坝位移的预测精度,多测点均方误差(MSE)平均降低21.67%,决定系数R²整体提高0.07%。     

Real-Time Operation of Reservoir System by Genetic Programming

Reservoir operation policy depends on specific values of deterministic variables and predictable actions as well as stochastic variables, in which small differences affect water release and reservoir operation efficiency. Operational rule curves of reservoir are policies which relate water release to the deterministic and stochastic variables such as storage volume and inflow. To operate a reservoir system in real time, a prediction model may be coupled with rule curves to estimate inflow as a stochastic variable. Inappropriate selection of this prediction model increases calculations and impacts the reservoir operation efficiency. Thus, extraction of an operational policy simultaneously with inflow prediction helps the operator to make an appropriate decision to calculate how much water to release from the reservoir without employing a prediction model. This paper addresses the use of genetic programming (GP) to develop a reservoir operation policy simultaneously with inflow prediction. To determine a water release policy, two operational rule curves are considered in each period by using (1) inflow and storage volume at the beginning of each period and (2) inflow of the 1^st, 2^nd, 12^th previous periods and storage volume at the beginning of each period. The obtained objective functions of those rules have only 4.86 and 0.44?% difference in the training and testing data sets. These results indicate that the proposed rule based on deterministic variables is effective in determining optimal rule curves simultaneously with inflow prediction for reservoirs.     

A neuro-fuzzy model for inflow forecasting of the Nile river at Aswan high dam

River flow forecasting is an essential procedure that is necessary for proper reservoir operation. Accurate forecasting results in good control of water availability, refined operation of reservoirs and improved hydropower generation. Therefore, it becomes crucial to develop forecasting models for river inflow. Several approaches have been proposed over the past few years based on stochastic modeling or artificial intelligence (AI) techniques. In this article, an adaptive neuro-fuzzy inference system (ANFIS) model is proposed to forecast the inflow for the Nile River at Aswan High Dam (AHD) on monthly basis. A major advantage of the fuzzy system is its ability to deal with imprecision and vagueness in inflow database. The ANFIS model divides the input space into fuzzy sub-spaces and maps the output using a set of linear functions. A historical database of monthly inflows at AHD recorded over the past 130 years is used to train the ANFIS model and test its performance. The performance of the ANFIS model is compared to a recently developed artificial neural networks (ANN) model. The results show that the ANFIS model was capable of providing higher inflow forecasting accuracy specially at extreme inflow events compared with that of the ANN model. It is concluded that the ANFIS model can be quite beneficial in water management of Lake Nasser reservoir at AHD.     

Application of ANN, Fuzzy Logic and Decision Tree Algorithms for the Development of Reservoir Operating Rules

Optimal use of scarce water resources is the prime objective for water resources development projects in the developing country like India. Optimal releases have been generally expressed as a function of reservoir state variables and hydrologic inputs by a relationship which ultimately allows the policy/water managers to determine the water to be released as a function of available information. Optimal releases were obtained by using optimal control theory with inflow series and revised reservoir characteristics such as elevation area capacity table, zero elevation level as input in this study. Operating rules for reservoir were developed as a function of demand, water level and inflow. Artificial Neural Network (ANN) with back propagation algorithm, Fuzzy Logic and decision tree algorithms such as M5 and REPTree were used for deriving the operating rules using the optimal releases for an irrigation and power supply reservoir, located in northern India. It was found that fuzzy logic model performed well compared to other soft computing techniques such as ANN, M5P and REPTree investigated in this study.     

Monitoring models for base flow effect and daily variation of dam seepage elements considering time lag effect

Affected by external environmental factors and evolution of dam performance, dam seepage behavior shows nonlinear time-varying characteristics. In this study, to predict and evaluate the long-term development trend and short-term fluctuation of the dam seepage behavior, two monitoring models were developed, one for the base flow effect and one for daily variation of dam seepage elements. In the first model, to avoid the influence of the time lag effect on the evaluation of seepage variation with the time effect component of seepage elements, the base values of the seepage element and the reservoir water level were extracted using the wavelet multi-resolution analysis method, and the time effect component was separated by the established base flow effect monitoring model. For the development of the daily variation monitoring model for dam seepage elements, all the previous factors, of which the measured time series prior to the dam seepage element monitoring time may have certain influence on the monitored results, were considered. Those factors that were positively correlated with the analyzed seepage element were initially considered to be the support vector machine (SVM) model input factors, and then the SVM kernel function-based sensitivity analysis was performed to optimize the input factor set and establish the optimized daily variation SVM model. The efficiency and rationality of the two models were verified by case studies of the water level of two piezometric tubes buried under the slope of a concrete gravity dam. Sensitivity analysis of the optimized SVM model shows that the influences of the daily variation of the upstream reservoir water level and rainfall on the daily variation of piezometric tube water level are processes subject to normal distribution.     

Drought Forecasting using Markov Chain Model and Artificial Neural Networks

Water resources management is a complex task. It requires accurate prediction of inflow to reservoirs for the optimal management of surface resources, especially in arid and semi-arid regions. It is in particular complicated by droughts. Markov chain models have provided valuable information on drought or moisture conditions. A complementary method, however, is required that can both evaluate the accuracy of the Markov chain models for predicted drought conditions, and forecast the values for ensuing months. To that end, this study draws on Artificial Neural Networks (ANNs) as a data-driven model. The employed ANNs were trained and tested by means of a statistically-based input selection procedure to accurately predict reservoir inflow and consequently drought conditions. Thirty three years’ data of inflow volume on a monthly time resolution were selected to enable calculation of the standardized streamflow index (SSI) for the Markov chain model. Availability of hydro-climatic data from the Doroodzan reservoir in the Fars province, Iran, allowed us to develop a reservoir specific ANN model. Results demonstrated that both models accurately predicted drought conditions, by employing a randomization procedure that facilitated the selection of the required data for the ANN to forecast reservoir inflow close to the observed values over a validation period. The results confirmed that combining the two models improved short-term prediction reliability. This was in contrast to single model applications that resulted into substantial uncertainty. This research emphasized the importance of the correct selection of data or data mining, prior to entering a specific modeling routine.     

Reservoir Inflow Modeling Using Temporal Neural Networks with Forgetting Factor Approach

In this paper, a recursive training procedure with forgetting factor is proposed for on-line calibration of temporal neural networks. The forgetting factor discounts old measurements through an on-line model calibration. The forgetting factor approach enables the recursive algorithm to reduce the effect of the older error data by multiplying the error data by a discounting factor. The proposed procedure is used to calibrate a temporal neural network for reservoir inflow modeling. The mean monthly inflow of the Karoon-III reservoir dam in the south-western part of Iran is used to test the performance of the proposed approach. An autoregressive moving average (ARMA) model is also applied to the same data. The temporal neural network, which is trained with the proposed approach, has shown a significant improvement in the forecast accuracy in comparison with the network trained by the conventional method. It is also demonstrated that the neural network trained with forgetting factor results in better forecasts compared to the statistical ARMA model, which has been calibrated through this approach.     

Robust Methods for Identifying Optimal Reservoir Operation Strategies Using Deterministic and Stochastic Formulations

Water allocation in a competing environment is a major social and economic challenge especially in water stressed semi-arid regions. In developing countries the end users are represented by the water sectors in most parts and conflict over water is resolved at the agency level. In this paper, two reservoir operation optimization models for water allocation to different users are presented. The objective functions of both models are based on the Nash Bargaining Theory which can incorporate the utility functions of the water users and the stakeholders as well as their relative authorities on the water allocation process. The first model is called GA–KNN (Genetic Algorithm–K Nearest Neighborhood) optimization model. In this model, in order to expedite the convergence process of GA, a KNN scheme for estimating initial solutions is used. Also KNN is utilized to develop the operating rules in each month based on the derived optimization results. The second model is called the Bayesian Stochastic GA (BSGA) optimization model. This model considers the joint probability distribution of inflow and its forecast to the reservoir. In this way, the intrinsic and forecast uncertainties of inflow to the reservoir are incorporated. In order to test the proposed models, they are applied to the Satarkhan reservoir system in the north-western part of Iran. The models have unique features in incorporating uncertainties, facilitating the convergence process of GA, and handling finer state variable discretization and utilizing reliability based utility functions for water user sectors. They are compared with the alternative models. Comparisons show the significant value of the proposed models in reservoir operation and supplying the demands of different water users.     

三峡水库实时调度水文气象预报应用风险及控制

为量化水文气象预报误差对水库实时预报调度的影响,以三峡水库为例,基于其历年水文气象预报信息的误差分析成果,假定入库过程,并叠加对应不同概率的预报误差,以此作为水库调洪演算模型的输入。分析了不同预见期、不同保证率水文气象误差条件下,不超过风险控制点的库水位指标,提出了三峡水库实时预报调度风险控制策略,可为水库科学调度提供参考。     

水库坝前水温统计分析

水库坝前水温是确定混凝土坝体温度场的一个最重要的边界条件，由于受坝址区的气候条件，水库的特征（库容、水深等）、水库运行方式等因素影响。水库坝前水温的分布非常复杂，本文以国内外２６个已建工程的坝址气温和坝前水温的多年实测资料为基础，建立了水库水温统计分析的数学模型。坝前水温随时间的变化用余弦函数表示，水温沿水深的分布用指数曲线表示，根据对各工程实测资料的统计分析结果，提出了一套简明实用的计算水库坝前