Browsing by Author "Li, Pei-Hao"

Now showing 1 - 4 of 4

A modified support vector machine based prediction model on streamflow at the Shihmen Reservoir, Taiwan
(2010-06-30)
Li, Pei-Hao
;
Kwon, Hyun-Han
;
Sun, Liqiang
;
Lall, Upmanu
;
Kao, Jehng-Jung
The uncertainty of the availability of water resources during the boreal winter has led to significant economic losses in recent years in Taiwan. A modified support vector machine (SVM) based prediction framework is thus proposed to improve the predictability of the inflow to Shihmen reservoir in December and January, using climate data from the prior period. Highly correlated climate precursors are first identified and adopted to predict water availability in North Taiwan. A genetic algorithm based parameter determination procedure is implemented to the SVM parameters to learn the non-linear pattern underlying climate systems more flexibly. Bagging is then applied to construct various SVM models to reduce the variance in the prediction by the median of forecasts from the constructed models. The enhanced prediction ability of the proposed modified SVM-based model with respect to a bagged multiple linear regression (MLR), simple SVM, and simple MLR model is also demonstrated. The results show that the proposed modified SVM-based model outperforms the prediction ability of the other models in all of the adopted evaluation scores. Copyright (C) 2009 Royal Meteorological Society
A segment-based optimization model for water pipeline replacement
(2007-07-01)
Kao, Jehng-Jung
;
Li, Pei-Hao
;
環境工程研究所
;
Institute of Environmental Engineering
By looking at valve-enclosed segments of a pipe rather than at each pipe as an independent unit, this research examined the extent and degree of water shortage resulting from pipeline damage. Two computer programs were developed to identify all segments of a pipe network that can be isolated with valve closure and to determine which segments are critical, i.e., pipe failures within them will result in a complete shutdown of water supply in any connected downstream segments. A third program was written using the EPANET 2 toolkit to estimate water shortages that would result because of pipe failures within noncritical segments. Results from these programs were used within a segment-based pipeline replacement optimization model to determine which pipes within a system should be replaced first in order to achieve the highest improvement in system reliability. Two case studies were implemented to demonstrate the effectiveness of the strategies derived from the proposed model for improving water supply reliability.
Optimization models for siting water quality monitoring stations in a catchment
(2012-01-01)
Kao, Jehng-Jung
;
Li, Pei-Hao
;
Hu, Wen-Shin
;
環境工程研究所
;
Institute of Environmental Engineering
A water quality monitoring network (WQMN) must be designed so as to adequately protect the water quality in a catchment. Although a simulated annealing (SA) method was previously applied to design a WQMN, the SA method cannot ensure the solution it obtained is the global optimum. Therefore, two new linear optimization models are proposed in this study to minimize the deviation of the cost values expected to identify the possible pollution sources based on uniform cost (UC) and coverage elimination uniform cost (CEUC) schemes. The UC model determines the expected cost values by considering each sub-catchment being covered by which station, while the CEUC model determines the coverage of each station by eliminating the area covered by any upstream station. The proposed models are applied to the Derchi reservoir catchment in Taiwan. Results show that the global optimal WQMN can be effectively determined by using the UC or CEUC model, for which both results are better than those from the SA method, especially when the number of stations becomes large.
Siting analyses for water quality sampling in a catchment
(2008-04-01)
Kao, Jehng-Jung
;
Li, Pei-Hao
;
Lin, Chin-Lien
;
Hu, Wen-Hsin
;
環境工程研究所
Pollution loads discharged from upstream development or human activities significantly degrade the water quality of a reservoir. The design of an appropriate water quality sampling network is therefore important for detecting potential pollution events and monitoring pollution trends. However, under a limited budgetary constraint, how to site an appropriate number of sampling stations is a challenging task. A previous study proposed a method applying the simulated annealing algorithm to design the sampling network based on three cost factors including the number of reaches, bank length, and subcatchment area. However, these factors are not directly related to the distribution of possible pollution. Thus, this study modified the method by considering three additional factors, i.e. total phosphorus, nitrogen, and sediment loads. The larger the possible load, the higher the probability of a pollution event may occur. The study area was the Derchi reservoir catchment in Taiwan. Pollution loads were derived from the AGNPS model with rainfall intensity estimated using the Thiessen method. Analyses for a network with various numbers of sampling sites were implemented. The results obtained based on varied cost factors were compared and discussed. With the three additional factors, the chosen sampling network is expected to properly detect pollution events and monitor pollution distribution and temporal trends.