Distributed Control Applications

Guidelines, Design Patterns, and Application Examples with the IEC 61499

Alois Zoitl, Thomas Strasser, Alois Zoitl, Thomas Strasser

• Publisher: CRC Press
• Year: 2017
• ISBN:
  9781351830935
• Edition: 1
• Topic: Technology & Engineering
• Subtopic: Electrical Engineering & Telecommunications
• Index: Technology & Engineering

Distributed Control Applications: Guidelines, Design Patterns, and Application Examples with the IEC 61499 discusses the IEC 61499 reference architecture for distributed and reconfigurable control and its adoption by industry. The book provides design patterns, application guidelines, and rules for designing distributed control applications based on the IEC 61499 reference model. Moreover, examples from various industrial domains and laboratory environments are introduced and explored.

Part I

IEC 61499 Basics

1

Challenges and Demands for Distributed Automation in Industrial Environments

Thomas Strasser

AIT Austrian Institute of Technology GmbH

Alois Zoitl

fortiss GmbH

CONTENTS

1. 1.1 Trends in Industrial Automation
2. 1.2 Requirements for Future Automation Architecture
3. 1.3 Outlook
4. Bibliography

1.1 Trends in Industrial Automation

The industrial automation domain is a key driver and supporter for other industries. With its innovations it can be seen as the backbone for many industrial sectors, like manufacturing, process technology, or the power and energy systems domain. Since production processes are more and more performed by automated machines, factories and plants, an increased level of automation is observable today. The major trend and driving factor for industrial automation in recent years is the growing need for customized and individualized products and goods. There is a need to evolve corresponding automation solutions in response to the rapidly changing demands due to new production processes and technologies. Also recent advancements in hardware and software solutions influence this trend. This requires that production lines, machines, and components have to be constructed and adapted to new products and production processes quickly [3, 10].

Such highly automated systems are mainly controlled by a vast set of embedded hardware and software components which are heterogeneous in nature. Typically, corresponding specifications and architectures lead to a tremendous increase in the design complexity. As a result of this trend, an increase of the software engineering costs from about 50% of the overall system engineering costs up to 80% in the next 15 years can be expected [11]. However, more recent studies also show that the engineering effort can be reduced up to 70% by optimizing the overall engineering process using proper methods, architectures, and corresponding tools [4]. Such an optimization result cannot really be achieved for the control software development, which is still a key problem. A further specialty of the manufacturing and also of the power and energy systems domain is the fact that each plant or system can be seen as a prototype. Hence, engineering costs are still key cost factors for the above mentioned processes and systems.

Applied design methods and approaches strongly depend on the specific application field. Well known control approaches in industrial automation are computerized numerical control (CNC), robot control, programmable logic controller (PLC), distributed control system (DCS), and supervisory control and data acquisition (SCADA). Also other approaches like field programmable gate array (FPGA)-based controllers becoming popular. These approaches are often used together to perform the automation and control tasks of a specific application [1]. The design and engineering of such heterogeneous systems is elaborate and needs the knowledge from different domains during all phases of development. For the engineering of industrial automation and control solutions, the system complexity, the domain or platform dependence, and also the design time are the most critical factors. In order to keep the complexity under control, the following key requirements and needs have to be satisfied [9]:

• The system specifications should be prepared at an easily understandable level of abstraction. Specification means should support designers of industrial automation and control systems at each level of the plant architecture to define the desired functionality and avoid to implement it on code level.
• Different domain-specific techniques and approaches should be combinable in a flexible way, overcoming potential limitations of each procedure. This allows modeling, analysis and implementation of complex, heterogeneous systems forming industrial automation and control systems.
• Methods and tools for specifying systems architectures should not depend on general languages used for embedded system design.

To cope with the challenges and demands of industrial automation and control systems a key design trend is to put components—blocks made of hardware, software and intellectual property (like algorithms and data structures)—together [15, 16] into one reusable component. This requires common, language-independent models for representing, saving and reusing the engineering artifacts of such components. Today, state-of-the art approaches, methods and corresponding tools for designing and engineering industrial automation systems are not fully capable of providing applicable solutions to the above mentioned issues [1, 15, 16]. Solutions developed for other domains like automotive or aerospace focus on the specific needs and requirements of their own sectors. They are different from the demands of industrial automation. Moreover, most of these solutions require a deep knowledge in software engineering. In the industrial automation sector, however, electrical and control engineers usally design the control systems. These domain experts may have only a basic education in software design. As software development becomes more and more important for industrial automation, methods and approaches have to be developed that allow also non-software engineering experts to effectively and efficiently develop control software.

1.2 Requirements for Future Automation Architecture

The above identified trends show the need for a more flexible and adaptable future automation system architecture and corresponding tools fulfilling requirements of different products and customers. As identified by [14] distributed architectures and dynamic reconfiguration are key drivers supporting the adaptivity of automation environments. In order to make this concept applicable, a future automation architecture has to fulfill the following key requirements:

Distributed intelligence has the potential to provide an opportunity to achieve manageability and robustness of large and high complex industrial processes and machines.

Usability is an important requirement for automation architecture. It needs to be simple to use, understandable, and maintainable as current PLC-based systems are. At the same time, design should allow a more expressive way to develop and maintain them.

Domain-specific modeling languages are useful tools allowing domain experts with little software engineering knowledge to describe automation tasks on a higher level of abstraction. This ensures reusability, reduces the engineering effort, and increases the software quality.

Easy communication configuration is needed for defining the interaction as well as the data exchange of a large number of distributed (embedded) control devices.

Predictability of distributed applications (in terms of timing behavior and its flow of execution) should be as easy to understand as the well known cyclic execution behavior typically implemented in PLC systems.

Heterogeneous execution is required to cope with the different needs of application tasks. Time-triggered execution with jitter below a defined border (especially for the execution of closed loop control algorithms), and event-based execution within a specified timing window must be supported within the same execution context or environment. Hence, the functions don't have to disturb each other and they may have to interact in a deterministic way.

Real-time execution with guarantees are needed as industrial automation and control systems interact with real-world processes, machines, and components. In case of dynamic reconfiguration, timing constraints may also have to be met deterministically.

Real-time reconfiguration with guarantees allows changing the functionality of control applications and systems during full operation. This requires that the reconfiguration process does not result in observable disturbances of the automation application. The switching or relocation of full control applications is required along with a fine grained adaption of application parts in a deterministic. The execution environment has to support reconfiguration on specific events or states of the industrial automation system as well as the switching at fixed points in time. Guarantees have to be given that the reconfiguration and switching processes do not violate the execution constraints of the running application. Typically, those processes have to be performed within the available resources (e.g., memory, processing power, network bandwidth) as well.

Durability of industrial automation and control systems needs to support system lifespans up to decades. Even if the produced products are more short-lived, the manufacturing system may also be used for a vast range of other products. This requires long-term upward and downward compatibility for replacement of components and devices.

Standards compliance is required to provide i...