Big Data Is So 2016. We Need Smart Data.
Big Data Is So 2016. We Need Smart Data. ===> https://tiurll.com/2t02d5
Data analytics are now playing a more important role in the modern industrial systems. Driven by the development of information and communication technology, an information layer is now added to the conventional electricity transmission and distribution network for data collection, storage and analysis with the help of wide installation of smart meters and sensors. This paper introduces the big data analytics and corresponding applications in smart grids. The characterizations of big data, smart grids as well as huge amount of data collection are firstly discussed as a prelude to illustrating the motivation and potential advantages of implementing advanced data analytics in smart grids. Basic concepts and the procedures of the typical data analytics for general problems are also discussed. The advanced applications of different data analytics in smart grids are addressed as the main part of this paper. By dealing with huge amount of data from electricity network, meteorological information system, geographical information system etc., many benefits can be brought to the existing power system and improve the customer service as well as the social welfare in the era of big data. However, to advance the applications of the big data analytics in real smart grids, many issues such as techniques, awareness, synergies, etc., have to be overcome.
In power grid, the traditional fossil fuels are facing the problem of depletion and the de-carbonization demands the power system to reduce the carbon emission. Smart grid and super grid are effective solutions to accelerate the pace for electrification of human society with high penetration of renewable energy sources (Ak et al., 2016). Although the rising awareness of sustainable development have become the impetus to the utilization of renewable energy sources, the intermittent characteristics of wind and photovoltaic energies bring huge challenges to the safe and stable operation in a low inertia power system (Wenbin & Peng, 2017; Ye et al., 2016). The data analytics based renewable energy forecasting methods are a hot research topic for a better regulation and dispatch planning in such cases. Traditional electricity meters in distribution systems only produce a small amount of data which can be manually collected and analyzed for billing purpose. While the huge volume of data collected from two-way communication smart grids at different time resolutions in nowadays need advanced data analytics to extract valuable information not only for billing information but also the status of the electricity network. For example, the high-resolution user consumption data can also be used for customer behavior analysis, demand forecasting and energy generation optimization. Predictive maintenance and fault detection based on the data analytics with advanced metering infrastructure are more crucial to the security of power system (Chunming et al., 2017).
Thus, the great progress of information and communication technology (ICT) provides a new vision for engineers to perceive and control the traditional electrical system and makes it smart. An embedded information layer into the energy network produces huge volume of data, including measurements and control instructions in the grid for collection, transmission, storage and analysis in a fast and comprehensive way. It also brings a lot of opportunities and challenges to the data analysis platform. This paper is to discuss the concepts of data analysis and their applications in smart grids. The intent of this paper is three-fold. First the potential data collected with advanced metering infrastructure in smart grid are discussed. Next, the paper briefly reviews the concepts of data analytics and the popular techniques. Finally, the paper illustrates the detailed applications of data analytics in smart grid.
As an intelligent system of both energy and information, smart grid is the abundant source of information, which covers the data from process of electricity generation, transmission, distribution and consumption. These data include the electrical information from distribution stations, distribution switch stations, electricity meters, and non-electrical information like marketing, meteorological as well as reginal economic data as shown in Fig. 1 (Keyan et al., 2015). Collection and analysis of them provide essential help in scheduling of power plants, operation of subsystems, maintenance for vital power equipment and business behavior in marketing.
In smart grid, the data are collected and transmitted with help of smart meters which provide energy related information to both the utility company (or DSO) and customers. For the energy consumption of residential customers, the number of smart meter readings for a large utility company is expected to rise from 24 million a year to 220 million per day (SAGIROGLU et al., 2016). As an emerging component in electricity market and smart grid, electric vehicles (EVs) and plug-in hybrid EVs (PHEVs) have seen a growing popularity with the movement of electrification in transportation sector and progress of artificial intelligence. To control the normal operation status of the distribution system, DSO traditionally relies on the measurements in the primary substation, at the beginning of each MV feeder, where the protection systems are normally installed. The current magnitude information is also needed for the automatic on-load tap changer in HV/MV transformers for voltage regulation. The measurements of a typical smart meter include the node voltage, feeder current, power factor, active and reactive power, energy over a period, total harmonic distortion as well as load demand, etc. The intelligent devices for data collection in smart grid are listed as Table 2.
Basic types of communication technologies for smart meters include wired and wireless infrastructures. The wireless communication technology allows the data center to gather measurement information from smart meters with low costs and simple connections while it may face the electromagnetic problem. Power line communication (PLC) is a wired communication technology by add a modulated carrier signal to the power cables and already successfully implemented in power system. The existing communication technology include ZigBee, WALN, cellular communication, WiMAX, PLC, etc. (Baimel et al., 2016).
As shown in Fig. 4, the main procedure of data analytics in smart grid is to extract valuable information from historical data for guiding the operation and maintenance with the comparison to real-time data (Siryani et al., 2017). The huge amount of data collected from smart meters and sensors are arranged and stored with data management techniques. After preparation, the mathematical model can be established through data mining techniques based on the clean data. With the input of real-time measurements, the state status can be evaluated in the derived model, which provides the possible schemes to guide practical actions and solve potential problems.
As a real-time social sensor for the smart grid, social media like Twitter or Facebook could contain potential information indicating the occurrence and location of power outages (Bauman et al., 2017). A probabilistic framework is devised in (Sun et al., 2016) for detecting a targeted event from the fragmented and noisy tweets. The method shows a good performance in locating accrual outage areas in experiment, which could be integrated to a social data-driven outage management.
Distribution automation (DA) is a concept of smart grid which focuses on the operation and system reliability at the distribution level. A successful DA has the capability to localize and isolate the faults in distribution system with a reduced restoration time and improved customer satisfaction. Under the concept of DA, increasing volume of operational data have been collected from supervisory control and data acquisition (SCADA) or advanced metering infrastructure (AMI) for state monitoring and fault diagnosis.
Under the concept of smart grid, a large amount of data collected via AMI are involved in the state assessment of power systems to support the energy management, system operation and decision making. Therefore, efficient summarization techniques are required for extracting useful patterns and discovering valuable information from redundant measurements in power system. A DT-based framework is proposed in (Liu et al., 2014) (Vittal, 2013) for the dynamic security assessment (DSA) in power system with high penetration of DGs. Two contingency-oriented DTs are trained based on the databases generated from real-time simulations. One of the well-trained DT is fed with real-time wide-area measurements to identify potential security issues, and the other DT provides the online corresponding preventive control strategies to deal with the problems. In (He et al., 2016) the dominant instability generation group (DIGG) in power system is identified without time domain simulation since the features adopted for TSA are extracted from steady-state variables. Reference (Parate et al., 2016) proposed an approach to classify the collected data from smart grid into two classes called vulnerable and non-vulnerable data sets with the data analytics such as multichannel singular spectrum analysis (MSSA), principal component analysis (PCA) and SVM. A framework for online contingency screening is presented in (Dimitrovska et al., 2017) with respect to first swing transient stability. The large spectrum of pre-fault operating state variables and critical clearing times of several contingencies are collected to compose a dataset for pattern recognition methods. The metric which can be used for operating condition evaluation is developed through PCA.
In addition to the renewable energy micro-sources distributed in smart grid (SG), the grid-connected high capacity wind farms are also widely accepted and applied for an effective utilization of pollution free and abundant nature resources. The improvement of technologies for large wind turbine generators and high capacity power converters accelerate the amount of wind energy integration into power system. To address the potential deterioration and stability problem caused by the large integration of wind energy to power grid, reference (Andalib-Bin-Karim et al., 2017) proposed data-driven analytics to determine the Q-V characteristic curve at the point of interconnection of the wind farm with valuable information for voltage stability extracted. Without prior knowledge of the system configuration and parameters, different curve-fitting techniques are adopted in a real case study in Canada. 2b1af7f3a8