Accuracy improvement of various short-term load forecasting models by a novel and unified statistical data-filtering method

ABSTRACT

Time-series and machine-learning methods are being strongly exploited to improve the accuracy of short-term load forecasting (STLF) results. In developing countries, power consumption behaviors could be suddenly changed by different customers, e.g. industrial customers, residential customers, so the load-demand dataset is often unstable. Therefore, reliability assessment of the load-demand dataset is obviously necessary for STLF models. Hence, this paper proposes a novel and unified statistical data-filtering method with the best confidence interval to eliminate unexpected noises/outliers of the input dataset before performing various short-term load forecasting models. This proposed novel data-filtering method, so-called the data pre-processing method, is also compared to other existing data-filtering methods (e.g. Kalman filter, Density-Based Spatial Clustering of Applications with Noise, Wavelet transform, and Singular Spectrum Analysis). By using an SCADA system?-based database of a typical 22kV distribution network in Vietnam, NYISO database, and PJM-RTO database, case studies of short-term load forecasting have been conducted with a conventional ARIMA model, an ANN forecasting model, an LSTM-RNN model, an LSTM-CNN combined model, a deep auto-encoder (DAE) network, a Wavenet-based model, a Wavenet and LSTM hybrid model, and a Wavelet Neural Network (WNN) model, which are to validate the novel and unified statistical data-filtering method proposed. The achieved numerical results demonstrate which the accuracy of the aforementioned STLF models can be significantly improved due to the proposed statistical data-filtering method with the best confidence interval of the input load dataset. The proposed statistical data-filtering method can considerably outperform the existing data-filtering methods.     

Salivary cortisol and cortisone levels,and breast milk dioxin concentrations in Vietnamese primiparas

There is a great deal of concern regarding the adverse effects of polychlorinated dibenzodioxins (PCDD) and polychlorinated dibenzofurans (PCDF) present in Agent Orange and other herbicides on Vietnam's population and ecosystems. The purpose of this study was to determine the effect of dioxin exposure on adrenal steroids in saliva, and dioxin levels in breast milk, of primiparas in an Agent Orange/dioxin hot-spot and a non-exposed area in Vietnam on the basis of epidemiological research. The subjects were 35 lactating women who had recently given birth to their first or second child. A further sub-study involved eight primiparas from each area. All subjects were aged between 20 and 30 years with infants aged between 4 and 16 weeks. The mean concentration of PCDD, PCDF, and PCDD?+?PCDF toxic equivalents (TEQ) in breast milk in the hot-spot area was significantly higher than in the non-exposed area. Cortisol and cortisone levels in the saliva of primiparas in the hot-spot area were also significantly higher than those in the non-exposed area. There was a significant negative correlation between cortisol and the cortisol/cortisone ratio and PCDD?+?PCDF and PCDF TEQ levels in the hot-spot area. Furthermore, the correlation between cortisol and cortisone and the PCDD + PCDF, PCDD, and PCDF TEQ in the combination of hot spot + non-exposed area was significant according to the curve (bell style). Our results suggest that Agent Orange/dioxin exposure still exerts a major influence on the salivary hormones of the Vietnamese population.     

Dioxin levels in the breast milk and estradiol and androgen levels in the saliva of Vietnamese primiparae

Human exposure to polychlorinated dibenzodioxins (PCDD) and dibenzofurans (PCDF), especially 2,3,7,8 tetrachlorodibenzo-p-dioxin (TCDD), was investigated in Vietnam since initial severe adverse health effects were reported in the late 1970s. The purpose of this study was to determine the effects of dioxin exposure on steroid hormones of primiparae in an Agent Orange/dioxin hot-spot and a non-exposed area in Vietnam. Sixteen primiparae (8 at each site), all of whom were aged between 20 and 30 years with infants aged between 4 and 16 weeks, agreed to participate in this study. The mean dioxin levels in breast milk of primiparae from the hot-spot area, in terms of PCDD, PCDF, and PCDD?+?PCDF toxic equivalents (TEQ), were significantly higher than those for the non-exposed area. PCDD TEQ, PCDF TEQ, and PCDD?+?PCDF TEQ levels showed a significant correlation with dehydroepiandrosterone (DHEA), androstenedione (A-dione), and estradiol (E2) in the saliva of primiparae in a combination of hot-spot and non-exposed areas in Vietnam. The dose–response curve between salivary E2 or A-dione levels and dioxin levels was U-shaped in humans. This study provides an overview of studies regarding dioxin hot-spots and effects on human health and steroid hormone levels in particular, with a focus on the toxicity attributed to dioxins and furans. Furthermore, causal evidence regarding the effects of dioxins on endocrine disruption in humans is provided.     

Influence of complexation on lead(II) and cadmium(II) percolation through a clay liner

Wastes are usually stocked in waste disposals without any pre-treatment. Runoff waters lead to the formation of leachates containing mineral and organic pollutants. In order to prevent groundwater contamination, a double barrier is placed on the landfill bottom. The first one is an active barrier made up of a draining system and a geomembrane, and the second one, consisting of a clay liner allowing pollutant retention, is considered as a passive system. The aim of this work is to evaluate the complexation impact on the organic and inorganic retention. Percolations on columns filled with clay are performed, which allow a better knowledge of the transfer of metallic pollutants, complexed or not, through a model porous media, considering charge and conditional stability of complexes.     

Numerical Modeling of Flow Fields and Dispersion of Passive Pollutants in the Vicinity of the Temelín Nuclear Power Plant

This study deals with numerical simulation of flow fields and dispersion of model passive admixtures in the planetary boundary layer in a 10 × 10 km square area, centered on the Czech Republic’s Temelín Nuclear Power Plant. Numerical calculations of three-dimensional flow fields with eight basic wind directions given for the inlet boundary of the computational domain are performed using the FLUENT CFD code with the standard k – ε turbulence model. The resultant modeling of the flow fields provides information as to probable local directions and velocities of flow vectors on a horizontal scale of 100 m, which are consistent with the data given for the boundaries of the calculation area in the framework of a scale of 10 km. The modeled flow fields generate the input data for related Lagrangian simulation of the ground concentration and deposition fields of passive particles assumedly emitted at the site of the Temelín plant. Simulated plumes describe eight cases of potential ground-level distribution of model passive admixtures in the area.     

Human ecological effects of tropical storms in the coastal area of Ky Anh (Ha Tinh,Vietnam)

Vietnam is prone to tropical storms. Climate change effects contribute to sea level rise, floods, progression of the low water line and coastal erosion. This paper inventories the perception of local people, assesses and values main aspects of the livelihood damage caused by the tropical storms of the period 2008–2013 in three coastal communes of the Ky Anh District of the Ha Tinh Province in Central Vietnam. The communes were selected because the location of their coastal line is perpendicular to the storm itself, which made them prone to damage. The effects of increasingly extreme weather conditions on three communities in an area most affected by storms and floods on the local residents and their responses to these changing environmental conditions are analyzed and assessed. The results of questionnaires completed by randomly selected local inhabitants of these communes show that storms and related hazards such as flood, sea level rise and heavy rain are perceived as the most impacting climate change intensified phenomena on agriculture and aquaculture, livestock, household property and income. Opinions and measured data provided by the commune and district authorities allow estimating the total direct cost of the tropical storm at 1.56 million $US (The used conversion rate VND/$US is 21,730 when the research was conducted in 2014) during the period 2008–2013. The long-term costs of adaptation and social impact measures will be significantly higher. Details of the monetary figures allow identifying the physical and natural capital of the area as being most affected by the storm. Trend and cost analysis show that the total financial support for hazard prevention and management during 2014–2019 is estimated at 1.19 up to 1.32 million $US. Local stakeholders indicate that climate change adaptation should not be limited to technical measures such as strengthening dikes, but also should target planting protection forests and mangroves and land use planning. Financial support for the relocation policy, stakeholder involvement and integrating climate change adaptation in both the socioeconomic development master plan and local land use planning are also of importance.     

Mercury capture into biogenic amorphous selenium nanospheres produced by mercury resistant Shewanella putrefaciens 200

Shewanella putrefaciens 200, resistant to high concentration of Hg(II), was selected for co-removal of mercury and selenium from aqueous medium. Biogenic Hg(0) reduced from Hg(II) by S. putrefaciens 200 was captured into extracellular amorphous selenium nanospheres, resulting in the formation of stable HgSe nanoparticles. This bacterial reduction could be a new strategy for mercury removal from aquatic environments without secondary pollution of mercury methylation or Hg(0) volatilization.     

Residual congener pattern of dioxins in human breast milk in southern Vietnam

This study evaluated residual congener patterns of dioxin/furan (=PCDD/DF) related to tactical herbicides aerially sprayed over the regions of southern Vietnam through Operation Ranch Hand. The study focused on Cam Chinh (CC) commune, Quang Tri province (an area sprayed with tactical herbicides), and the Cam Phuc (CP) commune, Ha Tinh province (a non-sprayed area). Breast milk samples for analysis were collected in September 2002 and July 2003 from lactating primiparous and multiparous mothers born after the war (<31 years old). We found the levels of each congener in the CC commune were higher than in the CP commune, and determined specificity in the PCDD/DF congener pattern in CC commune samples by cluster analysis. The congener pattern is characterized by higher (the hexa-, the hepta-, and the octa-) chlorinated PCDD/DFs; this appears to be the same profile as that presented by pentachlorophenol (PCP), rather than 2,4,5-trichlorophenoxy acid (2,4,5-T) contaminated with 2,3,7,8-TCDD. A GC/MS study in the 1970s detected the chlorophenols 2,4-di and 2,4,6-trichlorophenol in some Agent Orange samples, which contained, like PCP, a wide variety of PCDD/DF congeners. In this context, it may be expected that certain tactical herbicides contaminated with various chlorophenol impurities, have a unique congener pattern when compared with pure 2,4,5-T formulations.     

Long-term Performance of a Permeable Reactive Barrier in Acid Sulphate Soil Terrain

Deep drainage technique utilised for flood mitigation in low-land coastal areas of Australia during the late 1960s has resulted in the generation of sulphuric acid in soil by the oxidation of pyritic materials. Further degradation of the subsurface environment with widespread contamination of the underlying soil and groundwater presents a major and challenging environmental issue in acid sulphate soil (ASS) terrains. Although several ASS remediation techniques recently implemented in the floodplain of Southeast Australia including operation of gates, tidal buffering and lime injections could significantly control the pyrite oxidation, they could not improve the long-term water quality. More recently, permeable reactive barriers (PRBs) filled with waste concrete aggregates have received considerable attention as an innovative, cost-effective technology for passive in situ clean up of groundwater contamination. However, long-term efficiency of these PRBs for treating acidic groundwater has not been established. This study analyses and evaluates the performance of a field PRB for treating the acidic water over 2.5 years. The pilot-scale alkaline PRB consisting of recycled concrete was installed in October 2006 at a farm of southeast New South Wales for treating ASS-impacted groundwater. Monitoring data of groundwater quality over a 30 month period were assessed to evaluate the long-term performance of the PRB. Higher pH value (~pH 7) of the groundwater immediately downstream of the PRB and higher rates of iron (Fe) and aluminium (Al) removal efficiency (>95%) over this study period indicates that recycled concrete could successfully treat acidic groundwater. However, the overall pH neutralising capacity of the materials within the barrier declined with time from an initial pH 10.2 to pH 7.3. The decline in the performance with time was possibly due to the armouring of the reactive material surface by the mineral precipitates in the form of iron and aluminium hydroxides and oxyhydroxides as indicated by geochemical modelling.     

Persistence of white spot syndrome virus in shrimp ponds and surrounding areas after an outbreak

White Spot Syndrome Virus (WSSV) is the major and most serious pathogen in shrimp aquaculture industries. By using a sensitive PCR-based detection technique followed by sequencing multiple PCR products for confirmation, we address to the question of whether WSSV can persist in shrimp ponds and surrounding areas after an outbreak. The seawater samples were taken from two shrimp ponds and surrounding canals in a coastal area in northern Vietnam, Quang Ninh Province where the shrimps cultured in the two ponds had been killed due to a WSSV outbreak in April 2001 and the ponds were thereafter abandoned. A total of 480 seawater samples (30 samples each for July and December of 2001 and 2002) were subjected to WSSV genome detection. Although the detection rates of WSSV genome were generally gradually declined in seawater environments of both diseased shrimp ponds and surrounding canals, WSSV was still detected with rates of more than 10% in the diseased ponds and lower in surrounding canals in December 2002, 20 months after the WSSV outbreak.