OMICRON Electronics, USA
The reliability of the electric power system is one of the key requirements for the Smart Grid. The transition of the protection and control systems from electromechanical and solid-state relays with hardwired interfaces to microprocessor devices with IEC 61850 based process bus interfaces creates an environment for the development of new and innovative condition monitoring technologies.
One of the main characteristics of the Smart Grids of the 21st century is the digitization of the electric power system and the rapid evolution of communications and computer technologies. The development and widespread implementation of IEC 61850 based protection, automation, monitoring and control systems generates a huge amount of real-time data that can only be processed using artificial intelligence technologies.
Managing a portfolio of electric power system assets presents significant challenges related to the need for good data  and the changes of the electric power grid into a Smart Grid, including the integration of distributed renewable energy resource. The traditional methods for asset management do not provide the efficiency required by today’s electric power systems and may have an impact on their security and reliability. That is why it is necessary to take advantage of the new technology available as a result of the digitization of the grid. This can be accomplished based on the analysis of the assets in the Smart Grid, the available data sources and the requirements for the asset management systems.
Electric Power Systems Assets
The electric power system in the age of the Smart Grid has a wide range of assets that impose very different requirements on an asset management system. The following is a list of the groups of assets that are typically available:
- Transmission line equipment
- Distribution feeder equipment
- Primary electric power plant equipment
- Primary substation equipment
- Renewable distributed energy resources
- Protection, automation, control, monitoring and recording devices
- Communication devices
- Time-synchronization devices
One of the main challenges for asset management systems is their integration in a single system. This is where IEC 61850 again demonstrates why it is established as a cornerstone technology for the smart grid. The standard is not just a communication protocol. Its modeling principles and the System Configuration Language (SCL)  create an environment supporting the development of integrated utility asset management systems.
The reason is that SCL describes not only the individual components of the system, but also their connectivity and associations. The object model in SCL has three basic parts:
Substation / Line / Process describes the primary process related functions and devices like switch yard, respectively any primary process in the functional view according to IEC 81346-1, electrical connections on single line level (topology), and the designation of equipment and functions
Product this stands for all SA product-related objects such as IEDs and logical node implementations
Communication contains communication-related object types such as subnetworks and communication access points and describes the communication connections between IEDs as a base for communication paths between logical nodes as clients and servers. The process / substation part and the product part form hierarchies, which are used for naming and can be mapped to the functional and product structures.
The communication model part just contains the communication connection relations of IEDs to subnetworks, between subnetworks by means of routers at an IED, and the placement of master clocks at the subnetworks for time synchronization.
Considering that the primary substation equipment is not a data source itself, this model helps an artificial intelligence-based asset management system associate the large amount of data from the process interface devices and IEDs with the primary equipment they interface with.
Data Sources in Electric Power Systems
The IEC 61850 based digital substations use various sensors connected to the primary system equipment. All function elements in such systems are represented in the IEC 61850 model by Logical nodes that belong to different groups according to their role in the system. The sensors in a protection, automation, control, monitoring and recording system belong to the group T. While in Edition 1 of the standard they were only for currents (TCTR) and voltages (TVTR), Edition 2 [3-5] added many new sensor logical nodes – for temperature, vibration, etc. that are required by condition monitoring functions of the asset management system. Figure 2 shows an abstract block diagram of the functional decomposition in the IEC 61850 model with the sensor T logical nodes at the bottom sending sampled analog values (SAV) to the function elements that need them – protection (P group), measurements (M group), monitoring (S group). Each logical node represents a function element in the system that is a data source providing data using the services defined in the standard.
Figure 3 shows the IEC 61850 data model and the control blocks used for the various communication services.
The data model hierarchy includes Logical Devices (LD) representing the different functions – protection, control, measurements, monitoring, recording – in the device. The logical device may contain child logical devices representing sub-functions in a hierarchical structure. Logical devices contain logical nodes representing function elements – the building blocks in the model.
The logical nodes contain data objects belonging to various common data classes. Each data object contains attributes. Some of the data objects and attributes are mandatory, but most of them are optional.
As can be seen from Figure 3 the communication services represented by the control blocks are associated to data sets containing data objects and data attributes. This is the data that is available over the substation communications network and becomes an input into the artificial intelligence-based components of the asset management system.
The amount of available data in digital substations can be huge, especially when the merging units are publishing streams of current and voltage samples for both protection and power quality and disturbance recording purposes. Today this is based on the implementation agreement known as IEC 61850-9-2 LE. It defines for protection and control applications a sampling rate of 80 samples/cycle at the nominal system frequency. The digital output publishing rate is 4000 frames per second at 50 Hz and 4800 frames/sec at 60Hz with one ASDU (Application Service Data Unit) per frame. For power quality monitoring and disturbance recording the sampling rate is 256 samples/cycle at the nominal system frequency with 8 ASDUs per frame. It is not difficult to imagine the huge amount of data that this represents.
Since most of today’s advanced IEDs also have embedded phasor measurement units (PMU) that calculate M and P class synchrophasors that may be published 4 times per cycle, the amount of data to be processed by the asset management system further increases.
Any operation of a function in an IED also can be configured to send an event report or to log the data, further adding to the data to be processed by the asset management system. Depending on the type of asset management function in some cases we may need to process sampled values streaming in real time, while in others we will be working with recorded data. It is impossible all this data to be processed by humans, that is why we need the help of artificial intelligence platforms to solve the big data asset management problem.
Artificial Intelligence-Based Asset Management System (AIBAMS)
The large amount of data available to the asset management system requires careful consideration of the system hierarchy that will ensure its most efficient operation. That is why the data processing should be performed as close as possible to the data source to extract useful information that can be used as an input to the upper levels of the system hierarchy. Considering the substation topology and the allocation of primary and secondary equipment, the bottom level is the bay level. In a typical substation shown in Figure 4 we can identify three types of bays:
- Transmission line
- Power transformer
- Distribution feeder
Each of these bays contains primary equipment (breakers and disconnectors) and secondary equipment (process interface devices and multifunctional IEDs).
The data from the bay process interface devices and IEDs is processed by the Bay level AIBAMS which is then forwarding the related information to the upper levels of the hiearachy. The Voltage level of AIBAMS processes selected raw data and information from all bays connected to the busbar at this level. It takes into consideration in the analysis also the real topology of the busbar based on information for the status of the breakers and the switches.
The Substation level of AIBAMS processes selected raw data and information from all bays and voltage levels depending on the asset management functions that are executed at this level. It may take into consideration in the analysis also information regarding the status of adjacent substations available through the Gateway to SCADA or other communication interfaces.
Considering that asset management decisions are typically made at the regional or grid level, the information from the substations AIBAMS is communicated over secured wide area communications network to the regional or grid AIBAMS. This can be especially important in areas with large numbers of distributed renewable energy resources where AI based predictions of their status play important role in the reliable operation of the grid.
With the advancement of communications and cloud technologies we can envision in the future Asset Managements as a Service (AMaaS) running in a private utility cloud.
Machine Learning in Artificial Intelligence Systems
Artificial intelligence (AI) is already part of many aspects of our everyday life. The AI concepts and methods have been related to electric power applications for more than half a century.
There are several reasons that now electric power utilities are turning their attention to artificial intelligence to address many of the challenges they are facing:
- Data availability: The transition of our industry towards a digital grid with merging units and phasor measurement units streaming sampled values and synchrophasor measurements, together with the GOOSE messages and reports from protection and control IEDs generates a large amount of data which, combined with decreasing costs of data storage, is easily available for use. Machine learning can use this as training data for learning algorithms, developing new rules to perform increasingly complex tasks.
- Computing power: Powerful computers and the ability to connect remote processing power through the Internet make it possible the development and implementation of machine-learning techniques that process enormous amounts of data
- Algorithmic innovation: New machine learning techniques, specifically in layered neural networks (also known as “deep learning”) are enabling innovation in different domains of the electric power industry
Today there is a significant, government sponsored development of new algorithms and models in a field of computer science referred to as machine learning. Instead of programming the computer every step of the way, this approach gives the computer instructions that allow it to learn from data without new step-by-step instructions by the programmer. This means computers can be used for new, complicated tasks that could not be manually programmed, as is sometimes the case with protection and control applications in systems with inverter-based DER interfaces. The basic process of machine learning is to give training data to a learning algorithm. A machine learning model may apply a mix of different techniques, but the methods for learning can typically be categorized as three general types:
- Supervised learning: The learning algorithm is given labeled data and the desired output
- Unsupervised learning: The data given to the learning algorithm is unlabeled, and the algorithm is asked to identify patterns in the input data
- Reinforcement learning: The algorithm interacts with a dynamic environment that provides feedback in terms of rewards and punishments
Deep learning is a subset of machine learning. It refers to the number of layers in an artificial neural network (ANN). A shallow network has one so-called hidden layer, and a deep network has more than one (Figure 8). Multiple hidden layers allow deep neural networks to learn features of the data in a so-called feature hierarchy. Nets with many layers achieve some amazing results but are more computationally intensive to train because they pass input data (features) through more mathematical operations than nets with few layers.
The selection of machine learning methods or deep learning will depend on the specific asset management tasks to be implemented by the AIBAMS at the different levels of the system hierarchy.
The digitization of the electric power grid based on the developments of the IEC 61850 standard creates an environment supporting the next generation of asset management systems. The enormous amounts of data produced by streams of sampled values from merging units and synchrophasor measurements from phasor measurement units, combined with GOOSE messages and reports from IEDs requires the processing of the data by artificial intelligence-based computer applications at the different levels of a hierarchical artificial intelligence-based asset management system (AIBAMS).
Modern machine learning and deep learning AI platforms can be used in the development of various asset management applications.