Uncertainty handling using neural network-based prediction intervals for electrical load forecasting

The complexity and level of uncertainty present in operation of power systems have significantly grown due to penetration of renewable resources. These complexities warrant the need for advanced methods for load forecasting and quantifying uncertainties associated with forecasts. The objective of this study is to develop a framework for probabilistic forecasting of electricity load demands. The proposed probabilistic framework allows the analyst to construct PIs (prediction intervals) for uncertainty quantification. A newly introduced method, called LUBE (lower upper bound estimation), is applied and extended to develop PIs using NN (neural network) models. The primary problem for construction of intervals is firstly formulated as a constrained single-objective problem. The sharpness of PIs is treated as the key objective and their calibration is considered as the constraint. PSO (particle swarm optimization) enhanced by the mutation operator is then used to optimally tune NN parameters subject to constraints set on the quality of PIs. Historical load datasets from Singapore, Ottawa (Canada) and Texas (USA) are used to examine performance of the proposed PSO-based LUBE method. According to obtained results, the proposed probabilistic forecasting method generates well-calibrated and informative PIs. Furthermore, comparative results demonstrate that the proposed PI construction method greatly outperforms three widely used benchmark methods. © 2014 Elsevier Ltd.

Demo abstract: wind measurements for water quality studies in urban reservoirs

Water quality monitoring and prediction are critical for ensuring the sustainability of water resources which are essential for social security, especially for countries with limited land like Singapore. For example, the Singapore government identified water as a new growth sector and committed in 2006 to invest S$ 330 million over the following five years for water research and development [1]. To investigate the water quality evolution numerically, some key water quality parameters at several discrete locations in the reservoir (e.g., dissolved oxygen, chlorophyll, and temperature) and some environmental parameters (e.g., the wind distribution above water surface, air temperature and precipitation) are used as inputs to a three-dimensional hydrodynamics-ecological model, Estuary Lake and Coastal Ocean Model - Computational Aquatic Ecosystem Dynamics Model (ELCOM-CAEDYM) [2]. Based on the calculation in the model, we can obtain the distribution of water quality in the whole reservoir. We can also study the effect of different environmental parameters on the water quality evolution, and finally predict the water quality of the reservoir with a time step of 30 seconds. In this demo, we introduce our data collection system which enables water quality studies with real-time sensor data.

Characterization and prediction of issue-related risks in software projects

Identifying risks relevant to a software project and planning measures to deal with them are critical to the success of the project. Current practices in risk assessment mostly rely on high-level, generic guidance or the subjective judgements of experts. In this paper, we propose a novel approach to risk assessment using historical data associated with a software project. Specifically, our approach identifies patterns of past events that caused project delays, and uses this knowledge to identify risks in the current state of the project. A set of risk factors characterizing “risky” software tasks (in the form of issues) were extracted from five open source projects: Apache, Duraspace, JBoss, Moodle, and Spring. In addition, we performed feature selection using a sparse logistic regression model to select risk factors with good discriminative power. Based on these risk factors, we built predictive models to predict if an issue will cause a project delay. Our predictive models are able to predict both the risk impact (i.e. the extend of the delay) and the likelihood of a risk occurring. The evaluation results demonstrate the effectiveness of our predictive models, achieving on average 48%-81% precision, 23%-90% recall, 29%-71% F-measure, and 70%-92% Area Under the ROC Curve. Our predictive models also have low error rates: 0.39-0.75 for Macro-averaged Mean Cost-Error and 0.7-1.2 for Macro-averaged Mean Absolute Error.

A computational framework for uncertainty integration in stochastic unit commitment with intermittent renewable energy sources

The penetration of intermittent renewable energy sources (IRESs) into power grids has increased in the last decade. Integration of wind farms and solar systems as the major IRESs have significantly boosted the level of uncertainty in operation of power systems. This paper proposes a comprehensive computational framework for quantification and integration of uncertainties in distributed power systems (DPSs) with IRESs. Different sources of uncertainties in DPSs such as electrical load, wind and solar power forecasts and generator outages are covered by the proposed framework. Load forecast uncertainty is assumed to follow a normal distribution. Wind and solar forecast are implemented by a list of prediction intervals (PIs) ranging from 5% to 95%. Their uncertainties are further represented as scenarios using a scenario generation method. Generator outage uncertainty is modeled as discrete scenarios. The integrated uncertainties are further incorporated into a stochastic security-constrained unit commitment (SCUC) problem and a heuristic genetic algorithm is utilized to solve this stochastic SCUC problem. To demonstrate the effectiveness of the proposed method, five deterministic and four stochastic case studies are implemented. Generation costs as well as different reserve strategies are discussed from the perspectives of system economics and reliability. Comparative results indicate that the planned generation costs and reserves are different from the realized ones. The stochastic models show better robustness than deterministic ones. Power systems run a higher level of risk during peak load hours.

Sensor placement and measurement of wind for water quality studies in urban reservoirs

We study the water quality in an urban district, where the surface wind distribution is an essential input but undergoes high spatial and temporal variations due to the impact of surrounding buildings. In this work, we develop an optimal sensor placement scheme to measure the wind distribution over a large urban reservoir using a limited number of wind sensors. Unlike existing solutions that assume Gaussian process of target phenomena, this study measures the wind that inherently exhibits strong non-Gaussian yearly distribution. By leveraging the local monsoon characteristics of wind, we segment a year into different monsoon seasons that follow a unique distribution respectively. We also use computational fluid dynamics to learn the spatial correlation of wind. The output of sensor placement is a set of the most informative locations to deploy the wind sensors, based on the readings of which we can accurately predict the wind over the entire reservoir in real time. Ten wind sensors are deployed. The in-field measurement results of more than 3 months suggest that the proposed sensor placement and spatial prediction scheme provides accurate wind measurement that outperforms the state-of-the-art Gaussian model based on interpolation-based approaches.

Migrating swans profit from favourable changes in wind conditions at low altitude

Because energy reserves limit flight range, wind assistance may be of crucial importance for migratory birds. We tracked eight Bewick's swans Cygnus columbianus bewickii, using 95-g satellite transmitters with altimeters and activity sensors, during their spring migration from Denmark to northern Russia in 1996. During the 82 occasions where a swan's location was recorded in flight, average flight altitude was 165 m a.s.1. with a maximum of 759 m a.s.1., despite winds often being more favourable at higher altitudes. We also counted Bewick's swans departing from the Gulf of Finland and subsequently passing an observatory in the next major stop-over area 800 km further north in the White Sea, northern Russia, during the springs of 1994, 1995 and 1996. A comparison of these counts with wind data provided evidence for Bewick's swans using favourable changes in wind conditions to embark on migration. Changes in the numbers of birds arriving in the White Sea correlated best with favourable changes in winds in the Gulf of Finland 1 day earlier. Again, migratory volume showed a correlation with winds at low altitudes only, despite wind conditions for the swans being more favourable at high altitudes. We conclude that the relatively large Bewick's swan tends to gear its migration to wind conditions at low altitude only. We argue that Bewick's swans do not climb to high altitudes because of mechanical and physiological limitations with respect to the generation of power for flight and to avoid rapid dehydration.

Skipping swans: fuelling rates and wind conditions determine differential use of migratory stopover sites of Bewick's Swans Cygnus bewickii

Some migratory birds refuel at stopover sites that they by-pass on the return trip. In theory, this skipping behaviour is only expected in time-selected migrants when the overflown site is of a lower quality than the departure site. We provide empirical evidence that quality differences in stopover sites are the cause for skipping in Bewick's Swans Cygnus bewickii tracked by satellite telemetry. Two and five complete tracks were recorded in spring and autunm, respectively, showing that the White Sea was visited for c. 2 weeks in spring, but by-passed (or visited for a few days at the most) in autumn. Skipping of the White Sea in autumn was predicted by a dynamic programming model which was based on calculated gain rates during stopover in the Pechora Delta and the White Sea. This prediction was not sensitive to plausible variations in gain rates. Relative to the Pechora Delta the White Sea is a poor site because a large tidal amplitude precludes foraging on the beds of the submerged macrophyte Fennel Pondweed Potamogeton pectinatus during high tide. The dynamic programming model predicted a fast autunm migration. However, the phenology of autunm arrival dates of Bewick's Swans on the wintering grounds revealed that only in three out of ten years a significant number of birds was able to reach the wintering grounds without refuelling. In the other years, unfavourable wind conditions along the Russian/Baltic part of the route prevented such non-stop migration.

Flight costs and fuel composition of a bird migrating in a wind tunnel

We studied the energy and protein balance of a Thrush Nightingale Luscinia luscinia, a small long-distance migrant, during repeated 12-hr long flights in a wind tunnel and during subsequent two-day fueling periods. From the energy budgets we estimated the power requirements for migratory flight in this 26 g bird at 1.91 Watts. This is low compared to flight cost estimates in birds of similar mass and with similar wing shape. This suggests that power requirements for migratory flight are lower than the power requirements for nonmigratory flight. From excreta production during flight, and nitrogen and energy balance during subsequent fueling, the dry protein proportion of stores was estimated to be around 10%. A net catabolism of protein during migratory flight along with that of fat may reflect a physiologically inevitable process, a means of providing extra water to counteract dehydration, a production of uric acid for anti-oxidative purposes, and adaptive changes in the size of flight muscles and digestive organs in the exercising animal.

Energy expenditure in relation to flight speed: what is the power of mass loss rate estimates?

The relationship between mass loss rate and chemical power in flying birds is analysed with regard to water and heat balance. Two models are presented: the first model is applicable to situations where heat loads are moderate. i.e. when heat balance can be achieved by regulating non-evaporative heat loss, and evaporative water loss is minimised. The second model is applicable when heat loads are high, non-evaporative heat loss is maximised. and heat balance has to be achieved by regulating evaporative heat loss. The rates of mass loss of two Thrush Nightingales Luscinia luscinia and one Teal Anas crecca were measured at various flight speeds in a wind tunnel. Estimates of metabolic water production indicate that the Thrush Nightingales did not dehydrate during experimental flights. Probably, the Thrush Nightingales maintained heat balance without actively increasing evaporative cooling. The Teal, however, most likely had to resort to evaporative cooling, although it may not have dehydrated. Chemical power was estimated from our mass loss rate data using the minimum evaporation model for the Thrush Nightingales and the evaporative heat regulation model for the Teal. For both Thrush Nightingales and the Teal, the chemical power calculated from our mass loss rate data showed a greater change with speed (more 'U-shaped' curve) than the theoretically predicted chemical power curves based on aerodynamic theory. The minimum power speeds calculated from our data differed little from theoretical predictions but maximum range speeds were drastically different. Mass loss rate could potentially be used to estimate chemical power in flying birds under laboratory conditions where temperature and humidity are controlled. However, the assumptions made in the models and the model predictions need further testing.

Wingbeat frequency and the body drag anomaly: wind-tunnel observations on a thrush nightingale (Luscinia Luscinia) and a teal (Anas crecca)

A teal (Anas crecca) and a thrush nightingale (Luscinia luscinia) were trained to fly in the Lund wind tunnel for periods of up to 3 and 16 h respectively. Both birds flew in steady flapping flight, with such regularity that their wingbeat frequencies could be determined by viewing them through a shutter stroboscope. When flying at a constant air speed, the teal's wingbeat frequency varied with the 0.364 power of the body mass and the thrush nightingale's varied with the 0.430 power. Both exponents differed from zero, but neither differed from the predicted value (0.5) at the 1 % level of significance. The teal continued to flap steadily as the tunnel tilt angle was varied from -1° (climb) to +6° (descent), while the wingbeat frequency declined progressively by about 11%. In both birds, the plot of wingbeat frequency against air speed in level flight was U-shaped, with small but statistically significant curvature. We identified the minima of these curves with the minimum power speed (Vmp) and found that the values predicted for Vmp, using previously published default values for the required variables, were only about two-thirds of the observed minimum-frequency speeds. The discrepancy could be resolved if the body drag coefficients (CDb) of both birds were near 0.08, rather than near 0.40 as previously assumed. The previously published high values for body drag coefficients were derived from wind-tunnel measurements on frozen bird bodies, from which the wings had been removed, and had long been regarded as anomalous, as values below 0.01 are given in the engineering literature for streamlined bodies. We suggest that birds of any size that have well-streamlined bodies can achieve minimum body drag coefficients of around 0.05 if the feet can be fully retracted under the flank feathers. In such birds, field observations of flight speeds may need to be reinterpreted in the light of higher estimates of Vmp. Estimates of the effective lift:drag ratio and range can also be revised upwards. Birds that have large feet or trailing legs may have higher body drag coefficients. The original estimates of around CDb=0.4 could be correct for species, such as pelicans and large herons, that also have prominent heads. We see no evidence for any progressive reduction of body drag coefficient in the Reynolds number range covered by our experiments, that is 21600-215 000 on the basis of body cross-sectional diameter.

Integration of renewable generation uncertainties into stochastic unit commitment considering reserve and risk: a comparative study

The uncertainties of renewable energy have brought great challenges to power system commitment, dispatches and reserve requirement. This paper presents a comparative study on integration of renewable generation uncertainties into SCUC (stochastic security-constrained unit commitment) considering reserve and risk. Renewable forecast uncertainties are captured by a list of PIs (prediction intervals). A new scenario generation method is proposed to generate scenarios from these PIs. Different system uncertainties are considered as scenarios in the stochastic SCUC problem formulation. Two comparative simulations with single (E1: wind only) and multiple sources of uncertainty (E2: load, wind, solar and generation outages) are investigated. Five deterministic and four stochastic case studies are performed. Different generation costs, reserve strategies and associated risks are compared under various scenarios. Demonstrated results indicate the overall costs of E2 is lower than E1 due to penetration of solar power and the associated risk in deterministic cases of E2 is higher than E1. It implies the superimposed effect of uncertainties during uncertainty integration. The results also demonstrate that power systems run a higher level of risk during peak load hours, and that stochastic models are more robust than deterministic ones.

Optimal coordination of G2V and V2G to support power grids with high penetration of renewable energy

Electric vehicles (EVs) have recently gained much popularity as a green alternative to fossil-fuel cars and a feasible solution to reduce air pollution in big cities. The use of EVs can also be extended as a demand response tool to support high penetration of renewable energy (RE) sources in future smart grid. Based on the certainty equivalent adaptive control (CECA) principle and a customer participation program, this paper presents a novel control strategy using optimization technique to coordinate not only the charging but also the discharging of EV batteries to deal with the intermittency in RE production. In addition, customer charging requirements and schedules are incorporated into the optimization algorithm to ensure customer satisfaction, and further improve the control performance. The merits of this scheme are its simplicity, efficiency, robustness and readiness for practical applications. The effectiveness of the proposed control algorithm is demonstrated by computer simulations of a power system with high level of wind energy integration.

Modelling and analysis of type 4 wind turbine generator system for utilization in frequency regulation studies

Green energy targets for coming decades advocates high penetration of wind energy in main energy matrix which also pose incendiary threat to stability and reliability of modern electric grid if their dynamic performance aspects are not assessed beforehand. Considering increasing interest in dynamic performance along with ancillary service assessment related to frequency regulation, development of suitable generic modeling has gained high priority. This paper presents modeling of type 4 full converter wind turbine generator system suitable for frequency regulation focusing on active power control. Complete model is a modification of WECC generic model with additional aerodynamic and pitch control model. Descriptions of individual sub models are presented and performance results are compared manufacturer specific GE type 4 WTG generic model by means of simulations in the MATLAB ® Power System Block set.

Role of energy storage in the power system network

Today’s power system network is more complex with enhanced responsibility to maintain reliable, stable and quality supply of power at transmission and distribution level. Maintaining grid balance is a bigger issue, in case of any unexpected generation shortage or grid disturbance or any participation of an intermittent nature of renewable energy sources like wind and solar power in the energy mix. In order to compensate such imbalance and improve reliability, and stability of power system, an energy storage system (ESS) can be considered as a vital solution. Also ESS can be used to mitigate associated issues of renewable energy sources while integration into the power system network. Thus ESS supports to get a reduction in greenhouse gas (GHG) emissions by means of integrating more renewable energy sources to the grid effectively. There are various types of Energy Storage (ES) technologies which are being used in power systems network for large scale (MW) to small scale (KW) level. Based on the type and characteristics, each storage technology is suitable for a particular role of applications. This paper presents an extensive review study on various types of ES technologies in characteristics and applications point of view. It also demonstrates various applications of ESS in detail. Finally, with the aid of ES-selectTM tool software, a feasibility analysis has been carried out to identify a suitable ES technology for appropriate applications at different grid locations and also helps to develop a smart hybrid storage system for grid applications in future.

Full converter based wind turbine generator system generic modelling: Variations and applicability

Wind energy system integration can lead to adverse effects on modern electric grid so it is imperative toassess their dynamic performance before actual plant startup. Transmission system operators all over theworld stress the need for a proper wind turbine generator model for dynamic performance as well asancillary service assessments. Due to the bulk power system assessment requirements, developmentof suitable generic modeling has gained high priority. Generic modeling of type 4 full converter wind turbinegenerator system for application in frequency ancillary service investigations under varying windspeed and varying reference power has been presented in this study. Prevalent generic model, manufacturerspecific proprietary generic model along with detailed wind turbine model with synchronous generatoris also provided to highlight various modelling framework difference. Descriptions of individualsub models of proposed generic model are presented in detail and performance results are comparedand validated with GE’s proprietary generic model and detailed WTG model by means of simulationsin the MATLAB Power System Block set.