Bayesian Deep Learning-Based Probabilistic Load Forecasting in Smart Grids

GEFCom2014 probabilistic electric load forecasting: An integrated solution with forecast combination and residual simulation

Probabilistic load forecasts provide comprehensive information about future load uncertainties. In recent years, many methodologies and techniques have been proposed for probabilistic load forecasting. Forecast combination, a widely recognized best practice in point forecasting literature, has never been formally adopted to combine probabilistic load forecasts. This paper proposes a constrained quantile regression averaging (CQRA) method to create an improved ensemble from several individual probabilistic forecasts. We formulate the CQRA parameter estimation problem as a linear program with the objective of minimizing the pinball loss, with the constraints that the parameters are nonnegative and summing up to one. We demonstrate the effectiveness of the proposed method using two publicly available datasets, the ISO New England data and Irish smart meter data. Comparing with the best individual probabilistic forecast, the ensemble can reduce the pinball score by 4.39% on average. The proposed ensemble also demonstrates superior performance over nine other benchmark ensembles.

Advanced Metering Infrastructure (AMI) enables two-way communication between smart devices and utility control centers. This involves remote monitoring and control of energy consumption as well as other parameters in the electrical power network in real time. However, increasing technologies in AMI due to huge deployment of smart meters, integration of devices and application of sensors, demand a strong architectural model with the best network topology to guarantee efficient usage of network resources with minimal latency. In this work, a resource efficient multi-hop network architecture is proposed using hybrid media access protocols. The architecture combines queuing and random-access protocol to achieve optimal network performance. Numerical results show that the probability of delay incurred by an arbitrary smart meter depends on the mean and distribution of the queue switch over a period. It is also observed that for a single queued system, the throughput performance is equal to the existing hybrid method. As the number of smart meters increases to 500, the throughput of the proposed method improves by 10% compared to the existing method. Likewise, as the number of smart meters increases to 500, the delay reduced by 15% compared to the existing method. Keywords: Advanced Metering Infrastructure; hybrid media access protocols; Smart Meter; Smart Grid; Power Network.

Load probability density forecasting by transforming and combining quantile forecasts

Cold load pick-up (CLPU) has been a critical concern to utilities. Researchers and industry practitioners have underlined the impact of CLPU on distribution system design and service restoration. The recent large-scale deployment of smart meters has provided the industry with a huge amount of data that is highly granular, both temporally and spatially. In this paper, a data-driven framework is proposed for assessing CLPU demand of residential customers using smart meter data. The proposed framework consists of two interconnected layers: 1) At the feeder level, a nonlinear auto-regression model is applied to estimate the diversified demand during the system restoration and calculate the CLPU demand ratio. 2) At the customer level, Gaussian Mixture Models (GMM) and probabilistic reasoning are used to quantify the CLPU demand increase. The proposed methodology has been verified using real smart meter data and outage cases.

Smart electricity meters are currently deployed in millions of households to collect detailed individual electricity consumption data. Compared with traditional electricity data based on aggregated consumption, smart meter data are much more volatile and less predictable. There is a need within the energy industry for probabilistic forecasts of household electricity consumption to quantify the uncertainty of future electricity demand in order to undertake appropriate planning of generation and distribution. We propose to estimate an additive quantile regression model for a set of quantiles of the future distribution using a boosting procedure. By doing so, we can benefit from flexible and interpretable models, which include an automatic variable selection. We compare our approach with three benchmark methods on both aggregated and disaggregated scales using a smart meter data set collected from 3639 households in Ireland at 30-min intervals over a period of 1.5 years. The empirical results demonstrate that our approach based on quantile regression provides better forecast accuracy for disaggregated demand, while the traditional approach based on a normality assumption (possibly after an appropriate Box-Cox transformation) is a better approximation for aggregated demand. These results are particularly useful since more energy data will become available at the disaggregated level in the future.

This Paper examines, in detail, all aspects related to rollout of Advanced Metering Infrastructure (AMI) in India. It covers key issues such as smart meter standards and specification, testing of smart meters, retrofitting of existing meters, communication technology options, procurement strategy and business models for AMI rollout. Under the UDAY program, Government of India plans to deploy smart meters for all customers with monthly consumption above 200 kWh by December 2019. Distribution Companies (DISCOMs) have the responsibility of choosing the most optimum deployment strategy. While the traditional AMI approach proposes deployment of smart meters for all customers on a feeder, another option is to deploy only for customers having monthly electricity consumption greater than 200 units. Part A of this Paper presents a detailed cost-benefit analysis which strongly advocates deployment of smart meters for all customers on a feeder primarily because of the potential to reduce Aggregate Technical and Commercial (AT&C) losses and the substantially lower cost for the last mile communication network when the full feeder is covered. Part B of this Paper covers a set of recommendations for AMI roll out in India.

Deep ensemble learning based probabilistic load forecasting in smart grids

The deployment of smart meters is a promising innovation that comes to enhance the energy efficiency measures in the smart grid. The smart meter enables distributors to better understand the electrical network and reduce complexity of the management operations. It offers to households monitoring and control possibilities to their everyday energy consumption through the distribution of detailed information on household consumption and its evolution. This involves disaggregation of individual household loads in term of their individual energy consumption known as Non intrusive loads monitoring. In this paper, we present a supervised NILM approach based on dynamic fuzzy c-means events clustering and KNN label matching. First, a filtering method is involved to enhance the edge/events detection step. Then we perform a dynamic Fuzzy c-means clustering procedures to build appliances signature data based on active and reactive power measurements taking into account the time of day usage. The data base is further refined to map potential clusters centers that best identify the different appliances. A performance evaluation of the proposed approach is conducted showing a recognition rate over 90% for high consumption loads and promising results for low consumption loads.

To mitigate climate change, renewable energy sources are increasingly integrated into the energy system. Due to their volatility and uncertainty, smart grid applications are required to deal with this uncertainty and maintain grid stability. To operate effectively and in an automated manner, each of these applications requires a quantification of the uncertainty in future demand and generation, which probabilistic forecasts can provide. Furthermore, these applications often require specific probabilistic forecast properties, such as coverage rate and sharpness. However, existing probabilistic forecasts cannot be easily customised to exhibit these required properties. Therefore, we present a novel approach that creates loss-customised probabilistic forecasts using automated hyperparameter optimisation based on custom loss metrics. We combine a deterministic base forecaster and a conditional Invertible Neural Network to include specified uncertainty in a deterministic forecast. This uncertainty is defined by automated hyperparameter optimisation based on flexible and adaptable loss metrics, enabling the generation of loss-customised probabilistic forecasts with different properties without computationally expensive retraining. We evaluate our approach on four real-world data sets and compare the generated loss-customised forecasts with three state-of-the-art probabilistic forecasting benchmarks. We show that our approach generates probabilistic forecasts that can be customised to achieve state-of-the-art performance in either Continuous Ranked Probability Score, Pinball Loss, or Coverage Rate Error, depending on the selected customised loss metric.

Probabilistic load forecasting (PLF) is able to present the uncertainty information of the future loads. It is the basis of stochastic power system planning and operation. Recent works on PLF mainly focus on how to develop and combine forecasting models, while the feature selection issue has not been thoroughly investigated for PLF. This paper fills the gap by proposing a feature selection method for PLF via sparse L_1-norm penalized quantile regression. It can be viewed as an extension from point forecasting-based feature selection to probabilistic forecasting-based feature selection. Since both the number of training samples and the number of features to be selected are very large, the feature selection process is casted as a large-scale convex optimization problem. The alternating direction method of multipliers is applied to solve the problem in an efficient manner. We conduct case studies on the open datasets of ten areas. Numerical results show that the proposed feature selection method can improve the performance of the probabilistic forecasting and outperforms traditional least absolute shrinkage and selection operator method.

With a significant deployment of smart meters across end-user platforms, the dynamic visibility of energy flow among the end-users has been increased significantly. The granular information of smart meters can be used to improve the load forecast accuracy and to influence energy consumption patterns with demand side management (DSM) schemes. This paper addresses the challenges of smart meter data size, complexity, variability and volatility for efficient use in load forecast and DSM. A novel clustering-based approach for analysis of smart meter data, aimed at more accurate and detailed load profiling, reduced profile complexity, improved load forecast accuracy and providing optimal DSM solutions is proposed. The proposed approach utilizes an advanced clustering algorithm to reduce the data size. The approach addresses data complexity, variability and volatility by linearizing the load profiles and minimizing the errors. The validity of the approach is demonstrated on an Irish smart meter dataset and on a simulated solar photovoltaic (PV) data and showed an improved load forecast accuracy, improved DSM solutions, and reduced computational burden. The improvements in the DSM solution are evidenced by a higher cost saving with a higher peak load reduction at the lower level of demand flexibility.

Nonparametric probabilistic load forecasting based on quantile combination in electrical power systems

The electricity market has been significantly changing in the last decade. The deployment of smart meters is enabling the logging of huge amounts of data relating to the operations of utilities with the potential of being translated into knowledge on consumers and enable personalized energy efficiency programs. This paper proposes an approach for mining characteristics of a residential consumers (income, education and having children) from high-resolution smart meter data using transparent fuzzy models. The system consists in: (1) extraction of comprehensive consumption features from smart meter data, (2) use of fuzzy models in order to estimate the characteristics of consumers, and (3) knowledge extraction from the fuzzy models rules. Accurate estimates of consumer income and education level were not achieved (60 % accuracy), for the presence of children accuracies of over 70 % were achieved. Performance is comparable to the state of the art with the addition of model interpretability and transparency.

Smart Grid is an advanced power grid system that integrates digital technology, two-way communications, smart sensors and process automation to streamline the production, transmission, distribution and consumption of electricity. The implementation of smart grid components in the Oltenia region has been analysed by tackling network improvement projects on both the transmission and distribution sides. Through the SMART TRANSFORMATION program of the distribution operator in the region, smart grid components such as: Smart Metering Deployment, Meter Data Management and Metering Data Collection Systems, Advanced Distribution Management System and Asset Management System are being implemented. Among the renewable energy sources, the category of prosumers was analysed, as their number has been continuously increasing in recent years in the Oltenia region. To understand the degree of acceptance of smart grid components (smart meters, prosumer applications, smart devices) by consumers in the Oltenia region, we applied the Technology Acceptance Model (TAM). The assessment of the results highlighted a positive perception of the use of technology, as it was considered to be both beneficial and user friendly.