Energy Efficient Agent Function Block: A semantic agent approach to IEC 61499 function blocks in energy efficient building automation systems

Highlights

• An agent-based and knowledge-driven control software entity for BAS is proposed.

• To enhance IEC 61499 function block to maintain energy efficiency in buildings.

• We compared energy efficiency of proposed and traditional control strategies.

• In predefined situations, proposed model reduced considerable amount of energy consumption.

• Ontology & agent radically enhance function block concerning energy efficiency.

Abstract

Automation plays a vital role in improving energy efficiency in buildings. Building Automation Systems (BASs) are drastically growing in size and complexity. The traditional centralized automation systems cannot suitably address this growth due to insufficient flexibility and scalability. IEC 61499 Function Block (FB) is a promising software paradigm for industrial automation that addresses these complexities because of its object-oriented, event-driven and distributed paradigm. However, efficient energy management in a heterogeneous system comprising of many energy consuming components, functioning in highly dynamic and partially predictable environments, requires a higher level of intelligence that FB currently does not possess. To cope with this issue, a software entity called 2eA-FB has been introduced by combining FB, intelligent software agent and semantic web technologies. Being capable of field-level control, communication and reasoning over semantically enriched domain knowledge, 2eA-FB bridges the gap between management and automation layers of BAS and controls the energy consumption for individual energy consuming components, separately and intelligently. This in turn results in more efficient overall energy consumption for the whole system.

Introduction

According to published statistics from 2006 to 2009, the amount of energy consumption by buildings in a global scale has reached a critical level. In the US, for instance, 40% of energy nationwide is consumed by buildings [8]. This amount has been reported up to 37% in European Union and 35% in UK [25] while buildings consume 25% of total energy consumption in China [26]. It is obvious that even a little reduction in buildings' energy consumption will bring very substantial economy and great positive ecological impact for the society.

Modern computer technologies can help reducing energy consumption in buildings by monitoring, controlling and managing energy consuming devices. Nowadays large residential and office buildings are equipped with, building automation system (BAS) [24] that is a system that controls electrical devices in a building [23] such as heating, cooling, ventilation and lighting and it can be used to automatically manage and control the efficient utilization of these devices. For instance, as claimed by SIEMENS, their BAS solutions can increase the energy efficiency of heating, ventilation and air conditioning in commercial buildings by up to 30%.

The potential of further improvements is in making BAS more adaptable to various disturbances, such as ever-changing patterns of energy usage by the building inhabitants, or ongoing changes in the buildings' layout and machinery, which requires more flexibility and more intelligence required in BAS. Nevertheless, current BASs are not intelligent and flexible enough yet, due to their centralized architecture and the so-called “mechanical logic” [24], implemented in monolithic code.

From one side, a traditional automation system, in general, is divided into four levels, namely; Plant Level, Process Level, Device Level and Component Level. These levels together build a structure known as the automation ICT pyramid [28]. From the other side, the architecture of a BAS is divided into three functional layers, namely: Management Layer, Automation Layer and Field Layer. As depicted in Fig. 1, each one of these layers contains one or more levels of automation functionalities and their underpinning devices and used technologies.

One way to increase flexibility in automation systems is via applying more modular, object-oriented design of software, so that it could be easily reconfigured in response to changes in the system. Another improvement can be sought by making the control more intelligent. In this work a combination of these two approaches is explored.

The modularity of design and distributed code deployment are addressed by using the notation of Function Block (FB), which is, from software development point of view, an abstraction that hides the complexity of the automation layer comprising of process level as well as device level machineries. In the IEC 61499 standard a novel FB concept has been introduced, targeting the design and implementation of distributed industrial automation software in a hardware independent way, without taking into account at initial design stage, the specification of controllers such as PLCs, which are going to execute the software [29], [30]. Furthermore, software components can be efficiently reused so that their internals are hidden from the developers. Using this technology, an appropriate FB can be provided for each field device by its vendor, which eases the process of managing the automation complexity.

Furthermore, examining the usefulness of utilizing the FB paradigm in BAS is easier, in comparison with other industries as the most industrial installations of IEC 61499 have been reported in BAS sector. The IEC 61499 architecture facilitates modular design of BAS software following the modular structure of buildings and of the equipment used in the various subsystems, such as lighting, heating and cooling, security, office equipment, etc. In addition, the function blocks encapsulate functions or services to be performed by the corresponding equipment units, or related to some parts of the building (geographically or functionally separated).

For instance, as described in [36], FB has been successfully utilized in developing a virtual smart metering in automated energy efficient lighting systems for BAS. By means of visualization, this smart metering can assist consumers to develop reasonable energy consuming habits. In addition, using this smart meter, different automation strategies can be simulated, evaluated and selected pertaining to the amount of their required energy consumption. Another example is a powerful design framework for decentralized control systems in BAS described in [37] that utilizes simulation in the loop. Thanks to that, different control system configurations can be simulated before actual development takes place, which in turn, reduces costs and, more importantly, the energy usage that an inadequate configuration can impose on a BAS.

Although FB is successful in managing automation layer complexity and despite the fact that FB architecture complies with object-oriented and event-driven software development paradigms, there are still some sorts of complexities that they are not able to address and manage. Most importantly, complexity of processes at the management layer is not manageable by FB. First of all, processes at the management layer implement the so called “business logic”. Secondly, they are subject to unexpected changes, which make the whole system dynamic, only partially predictable and risky. As an example, consider a meeting scheduling system, which is running at the management layer to keep track of all business meetings within an office building. One may also imagine a part of BAS installed in meeting rooms capable of doing necessary preparation of the meeting rooms such as turning on/off heating, cooling, lighting systems and some projectors, pulling some screens down or up, etc. Each one of these two systems can perform perfectly on their own but connecting them together may be challenging, as meeting scheduling is subject to unexpected changes and disruptions caused by human factors, organizational policies and so on. It is obvious that cancellation of a prescheduled meeting after completion of room preparation process by BAS implies wasting of energy which in the long run reduces the efficient use of energy in the building. One should note the generic and illustrative nature of this example. Very similar problems can be encountered in scheduling of any energy consuming processes, for example, in manufacturing systems.

In such environments, intelligent software agents empowered by communication capability and capable of reasoning over domain knowledge happened to be a practical and effective solution [38], [39]. Therefore, combining benefits of modular design with function blocks and the power of multi-agent systems with reasoning capabilities can fill the gap between management layer and automation layer. In this paper, such composite software entity called Agent-FB has been proposed. An abstract view of Agent-FB is depicted in Fig. 1. In addition, Fig. 1 shows the role and place of Agent-FB in BAS.

Another reason of introducing Agent-FB is to distribute intelligence amongst software components responsible for controlling field-level devices. It is reasonable to assume that an energy management system will be more efficient if each one of its components can exhibit certain level of intelligence. Agent-FB is an enhanced function block, which exhibits intelligence, via communication to its external world (outside of the boundary of automated control system) and reasoning over domain knowledge. Having these capabilities, they can be effectively used for decentralization of control in energy efficient BAS. In the following sections energy efficient Agent Function Block (2eA-FB), which is a specific type of Agent-FB is modeled and exemplified.

The structure of this paper is as follows. In Section 2, a background study on utilizing intelligent software agent, ontology and semantic web, and FB in energy efficient BAS is discussed. Section 3 describes the limitations of the existing systems and proposes a solution to overcome these shortcomings. In Section 4 a case study is described and the results of applying the proposed technique in saving energy are demonstrated. Section 5 summarizes the paper and includes conclusion and suggestions for future works.