Sei sulla pagina 1di 14

PLC Formal Control Methodologies: Does Academia Supply What Industry Demands?

Houshang Darabi Rupa Sampath David Naylor


MIE Department (M/C 251) MIE Department (M/C 251) Starthis, Inc.
University of Illinois at Chicago University of Illinois at Chicago 10600 W. Higgins Road
2039 ERF, 842 W. Taylor St 2039 ERF, 842 W. Taylor St Suite 414
Chicago, Illinois 60607 Chicago, Illinois 60607 Rosemont, Il 6061

Keywords

Logic Control, Programmable Logic Controller, Formal Methods, Academia, Industry

Abstracts

Discrete part manufacturing systems consist of numerous machines working together in a


coordinated and sequential fashion. Programmable Logic Controllers (PLCs) are widely used in
the implementation of control algorithms for these machines. The PLCs have been modeled as
Discrete Event Systems (DESs) using formal control theories like Petri Nets and Finite
Automata. Though such methods have received wide attention within academia, their industrial
implementations are scarce. The work reported in this paper aims identify the reasons for the gap
between academia and industry in implementing formal control methodologies for PLCs in
Industrial manufacturing and to provide solutions to reduce the identified gap.

I. Introduction

A logic controller is a discrete event system whose purpose is to control the behavior of a
process which is itself a discrete-event system, considering the state of this process, and
information coming from other systems [1]. A PLC is a micro-processor based controller that
continuously reads inputs and in real-time determines new outputs [6]. They are used for
implementing various control systems based on Timing and Logic. A model PLC is depicted in
Figure 1. Most existing programs for PLCs are written in Relay Ladder Logic (RLL) [12]. A
typical ladder logic diagram is shown in Figure 2. PLCs are the workhorse for controlling event-
driven industrial automated systems and are used in the Instrumentation and control systems of
process industries as well as in automation environments due to their low cost. The industrial
importance for PLCs has created the need for research in the field of Control engineering.
Much of todays developments in control systems have primarily been due to the efforts of
academia and industry. Academia, comprising of students, faculty and research associates,
approaches problem solving based on formal methods [2]. Formal methods are mathematics-
based techniques, which specify and model the behavior of a system and verify that the system
design and implementation satisfy functional and safety properties. Industry, comprising of
workers, control engineers, practitioners, and programmers use methods that are easy to use and
understand, and support quick delivery of product to market. It prefers non-formal methods
because majority of its work-force comprises of workers at the shop-floor level who might not
have the education needed for practicing formal methods. Formal methods, while receiving wide
attention within academia, are scarcely implemented in industry. Years of research by academia
has created little impact on the mainstream industry. The two groups look at control engineering
as different concepts and with different needs. The objective of this paper is to identify the
reasons for the gap between academia and industry in the implementation of PLC formal control
methodologies and propose solutions to solve the identified problems. This paper is structured in
four sections. Section II gives an overview of existing works on PLCs using formal methods.
Section III presents the basic reasons for the gap between academia and industry. Section IV
provides solutions to the problems identified in section III. Section V is the conclusion.

II. An Overview of Existing Works on PLCs

The control life cycle gives the sequence of events associated with any control problem. A
description of the various stages in the cycle is given.

The problem statement is a description of the uncontrolled process and requirements for the
controlled system.

Modeling is the specification of the requirements using various approaches. It results in a


process model, which is needed to check different properties of a system.

The model is then tested and verified to check whether the implementation runs as the
specification.

The general approach to get the realization from the problem statement is the direct
implementation of controller using a PLC program. This includes both hardware and
software.

Program Verification is the proof that the internal semantics of a model is correct. Validation
determines if the model agrees with the designers purpose. The same techniques are used for
model and program verification since both check the correctness between specification and
implementation.

Implementation consists of deriving the realization from the formal specification.

Documentation consists of defining the terminologies used in industrial environments.


Maintenance is an important segment in process control and is defined as the continuous
assurance of product quality and safety after commissioning.

A thorough study of over eight hundred documents (papers, books, reports etc.,) on control
engineering practices, various interviews and surveys conducted with relevant personnel has
given us a clear picture of the state of the art in industrial and academia practice of PLC formal
methodologies. The information presented in Table 1 is a quantitative measure of the works done
on various aspects of PLCs. Note that there might be some overlaps between the works in
different classes of formal methods presented in the Table.

TABLE 1 STATISTICS FOR APPLICATION OF FORMAL METHODS

CLASS INDUSTRY ACADEMIA COMBINED

Modeling using 12 42 2
Automata

Modeling using Petri 46 103 18


Nets (Ordinary PN,
Real-time PNs and
Interpreted PNs)

Modeling using 5 24
Timed Models

Modeling using Fault 3 12


tree

Modeling using 22 40 5
Grafcet

Modeling Other 10 19 11
tools

V&V Using Formal 12 78 14


Methods

V&V Simulation 36 12 7

V&V Emulation 17 8

V&V Others 4 9 2

PLC Hardware 66 18 12
Programming using 50 48 24
RLL

Programming using 64 40 13
IEC 1131-3

Programming 18 20 6
Others

Implementation 26 65 10

Testing 31 17 2

Diagnostics 58 20 14

Maintenance 24 10 2

PLC Networking 16 16 6

Applications 79 47 30
* The authors can be contacted directly to obtain an electronic copy of the list of references referred to in this Table.

2.1 System Modeling

Academia has developed new tools for system specification because of the following needs:

1. To understand the change in output behavior of a control device as a result of the change of
its process or environment (input).

2. To describe the sequence of states of discrete event systems.

3. Generally, given a state of the system, only a few inputs affect the state, and only a few
outputs change. Thus there was a need to describe only the behavior corresponding to these
input changes.

A brief description of the various tools and techniques presented in Table 1 is given.

A Petri Net (PN) [11,15] is a graph where states are represented by places and events by
transitions. The successful use of PNs for modeling purposes is clear from Table 1 which
has a high value of 45%. Extensions of PNs are used based on the nature of the control
problem. Real-time Petri Nets (RTPNs) are a class of PNs for sequential control design,
obtained by associating timing information and I/O sensory information to the untimed PNs.
Interpreted Petri Nets (IPNs), also an extension to ordinary PNs, uses input and output
languages to capture the control and output languages of a DES so as to capture the
relationship between the PLC and its environment.

Another formal way to study the behavior of logic control systems is based on the Finite
Automaton (FA) method [10,13]. In FA method, we distinguish between the system total
behavior (what can happen) and system desired behavior (what is desired to happen). This
approach is different from the PNs which requires the development of only one behavior set.
From Table 1, it is found that, next to Petri Nets, FA is the most widely used tool for system
modeling.

Grafcet [1], a modeling tool that draws its inspiration from Petri Nets. Grafcet provided the
basis for Sequential Function Chart (SFC), an international standard for control specification.

Timed models are used to model systems, which change their states by time events. The
modeling tools discussed prior to the timed models considered only specified event and state
sequences and did not include timing information. Table 1 shows that less than 5% of the
work done by industry on modeling are based on timed models. The scarce usage is because
of the complex nature of these models and their huge number of states.

Fault Tree (FT) is a modeling tool used to access the safety requirements in critical systems.
If the system to be modeled is highly redundant and contains similar components, the use of
FT enables the replicated units to be folded. Thus, only a single element for each class is
explicitly included in the model. This tool overcomes the state space explosion problem but
is not widely used because of the limitations and assumptions in its approach.

2.2 System Verification

Verification & Validation is a class of formal methods for PLCs. From Table 1, it is found that
16% of academia work has been related to the development of V & V procedures. The methods
commonly used are Simulation, Reachability Analysis, Model Checking and Theorem Proving
[14]. The Table shows that almost 73% of the V & V models developed by academia involve
formal methods. Methods on reachability analysis build the complete state-space of the modeled
system and check properties by investigating the components of this state-space. However, the
problem with this method is state-space explosion. Hence this method is not ideal for system
verification. In Model-checking, specifications of the system are checked on a model of the
system. The modeling tool used is the PNs or the FA. This technique also does not overcome
the state space problem. In theorem proving [7], the system and its expected properties are
formalized using mathematical logic. The property formulas are then proved from the axioms of
the system description using some interference rules. A great advantage of theorem proving is
the avoidance of the state-explosion problem. In spite of this, it is not widely used because this
method is applicable only to a specific class of control problems.

Simulation [4] is a verification tool widely used for system analysis. This statement is supported
by Table 1 which shows that more than 50% of the work done by industry for system analysis is
using simulation. Simulation refers to the task of simulating the dynamic action of the field
devices. Using simulation, we can simulate the whole system or just the controller interacting
with a device that simulates the target real world system. It is widely used because

1. When industry has a lot of similar projects, it could develop standard libraries of simulation
functions.

2. Developing a simulation model does not need expertise in formal verification & validation
methods Another technique for verifying a PLC based system is Emulation [3]. Emulation allows
the PLC program to be run on the PC without the PLC and test the ladder logic in whole or in
part. It consists of writing a response program to provide feedback to the PLC program. It can be
stored for possible reuse on a future project. However, it is less widely used compared to
simulation.

2.3 System Programming

A control system is composed of a hardware part (PLCs, PCs) and a software part (programs).
Most of the existing programming languages depend on the PLC manufacturers. RLL is the only
technique that is currently common to most types of PLCs and yet each product appears to have
its own ladder programming dialect. IEC [5,9] is a language standard developed by reviewing
common programming languages and techniques. IEC provides a consistent approach to
programming and reduces the learning curve for people who deal with products from different
vendors. The acceptance of IEC as a programming standard is evident from Table 1 which shows
that about 48% of the work done by industry on is based on the IEC standards. EC 1131-3
standard defines the syntax for four programming languages namely Relay Ladder Logic (RLL),
Function Block Diagram (FBD), Structured Text (ST) and Instruction List (IL) and one
structured language namely SFC which enables the partitioning of a PLC program into a set of
steps and transitions inter-connected by directed links. Programming using IEC standards
provides more quality and safety.

2.4 System Diagnosis

In an Industry, the dynamic nature of systems controlled in real-time makes situation assessment
difficult. However, detecting and predicting faults is essential to maintain the availability of
equipments. Hence Industry emphasises on diagnostic procedures for systems. Diagnosis is the
framework of PLCs concerned with surveying failures that occur on a physical system [8].
Diagnostic systems have been proposed by academia using Colored Timed Petri Nets, Neural
networks, and Expert systems among others.

PLC Hardware is the class of PLCs concerned with the equipments that comprise the control
system. PLC Networking refers to the inter-connection of PLCs for executing control programs.
Here each PLC takes a portion of the work. PLC Hardware and networking do not involve the
use of formal control methodologies and hence are not discussed in detail in this section. From
this discussion, it is inferred that the formal tools developed by academia are not implemented in
industry. For example, academia has proposed formal methods for system analysis. On the other
hand, industry, for its system verification, uses a non-formal method like simulation. Similarly,
Industry has placed much emphasize on providing support for fault tolerance and handling.
Again, academia has also developed diagnostic procedures for systems. However the models
developed by academia for diagnostics have considered only the normal operating cycle of a
system (not considering the effects of exceptional fault handling). But in actual practice, 90% of
the control is associated in dealing with fault detection and diagnosis. This helps us conclude that
academia and industry have different priorities and working methodologies to follow while
developing a solution. Tables 2 and 3 summarize the information discussed in this Section.

Table 2*-Statistics (As A % of Specific Class)

CLASS INDUSTRY ACADEMIA COMBINED

Modeling using 11* 18 6


Automata

Modeling using PNs 43 45 50

Modeling using 30 13 14
Grafcet

V & V Formal 17 73 61
methods

V & V Simulation 52 11 30

Programming using 48 39 30
IEC
*11% of the work done by industry on modeling uses FA as the modeling tool.

Table 3-Statistics (As a % of Total Work)

CLASS INDUSTRY ACADEMIA COMBINED

Modeling 18* 35 21

Verification 11 16 14

PLC Hardware 11 3 7

Programming 22 19 23

Implementation 4 10 5

Testing 5 3 1
Diagnostics 9 3 9

Maintenance 4 2 1

PLC Networking 3 2 3

Applications 13 7 16
*18 % of the work done by academia focuses on modeling of PLCs

*Table 2 includes the classes of PLCs from Table 1 that are supported by formal methods.

III. Reasons for Gap Between Industry and Academia

In this section, we consider the various actions performed by industry and academia. We aim at
identifying how these actions lead to the gap between academia and Industry. This analysis will
help us to propose solutions to the identified problems in Section IV.

3.1 Academia Views

The primary objectives of academia are to publish papers, attract research funds, and provide
education & research issues to students. By publishing papers, academicians gain recognition as
researchers. The University associated with the researcher is also benefited because the ranking
of a university, while relying on various factors, is also dependent on the amount of research
performed. Since academia is not time-bound to produce results because it has students to do
research, it can choose problems that are time-consuming and research labor intensive.
Moreover, to attract research funds from most government agencies, academia has to explore
high risk intellectually challenging issues, even though they might not have practical value. They
have to choose a problem with potential for future research. Above all, current use of advanced
control devices necessitates the development of tools for modeling, programming, verification
and implementation of control systems. The above factors and the unavailability of problems
from industry makes academia choose problems that use formal theories. Hence they are less
concerned about the relevance of the problem to industry. Academia has developed techniques to
support systems of varying complexities. Table 3 shows that about 35% of academia works have
focused on modeling. Newer modeling techniques have been developed because of the following
reasons. a. Problems with existing modeling tools and practices b. Need to develop modeling
tools that provides better support for system analysis c. Need for a smooth transition from
electromagnetic relays to electronic control devices. Academia has developed numerous V & V
procedures using formal methods. They have not focused on non-formal methods because of the
following reasons:

1. Non-formal works has little to offer in terms of theoretical content.


2. Developing non-formal simulation tools do not provide scope for future research because the
tools are application specific and do not support generic analysis.

3. It does not prefer to use its resources and time in the development of easy to create simulation
tools.

The above discussion is supported by our study of existing practices in section II and the data
presented in Table 2. From Table 2 it is also found that only 11% of academia work has been on
simulation compared to a 52% contribution by industry.

3.2 Industry Views

Since the primary aim of an industry is the quick delivery of product to the market, it tries to
have a flexible action scheme to meet changes in demand. To remain competitive, industry needs
systems that can be handled easily and personnel skillful to handle them. In case of system
failure, its design should be such that fault identification as well as repair is easy. This has made
industry spend a huge sum of money to hire quality professionals. They also set up R & D teams
to ensure a consistent improvement in the quality of its products. For an Industry, constraints in
the market place demand that any new technique adopted by it fits within the development
process, and does not delay that process any more than the current way of doing it. Industry
prefers easier methods like tables and charts that capture the essence of the phenomena of
interest because its backbone is the personnel at the shop-floor level. These highly compact
forms permit accurate modeling and analysis without explicitly resorting to axioms and
theorems. Similarly, by using simulation, the total behavior of the system can be predicted

by studying certain scenarios. In contrast, formal procedures like reachability analysis and
modelchecking require detailed knowledge of underlying mathematical theory. These techniques
are also time-consuming. For the industry, earlier the job done- lesser the money spent and more
the profit made.

Summarizing, at a time when the future of PLC technology is debated, both technologically and
from a business perspective, industry cannot afford to spend its time on time-consuming
approaches. Moreover, companies are reluctant to be the first to enter new territory: Any
approach developed by academia is tested theoretically and not using real-world systems. Hence,
without sufficiently large evidence of success, formal methods are considered too immature for
adoption by industries. Also, many industries have put much effort in their development process.
Therefore, introducing a new technology into the existing process is difficult, as it requires
fundamental change. At the same time, academia also cannot stop itself from pursuing
theoretical issues. For them, publishing papers having research content is more important than
developing industry-oriented tools. Only theoretical issues can generate funds for them. From the
industry perspective, diagnostics should be a key feature of todays equipments. Industry
requires a monitoring technology that can perform diagnosis without human intervention. For an
online diagnostic model to be implemented in a system, all the knowledge required to replace the
operator must be put in the system. Industry can make use of the limited size control strategies
like the supervisory-control methodology that can be efficiently generated on-line. These would
actually lead a person through a procedure to trouble-shoot and repair equipment. This would
eliminate the need for hardcopy instruction manuals. Academia has provided diagnosis tools to
the industry. However, as explained in the previous Section, those models consider only the
normal operating mode of a system and do not consider exceptions.

From this study, we understand that, academia and industry work differently to achieve a
common goal. This helps us conclude that part of the gap between academia and industry is not a
gap at all. This is because both academia and industry have individual objectives to achieve. And
most of their actions are influenced by their objectives. The main reason for industry not
choosing formal methods is because the use of formal methods hampers the progress and the
primary aim of the industry making profit by quick delivery of product. Academia does not
prefer non-formal methods because those methods do not have the theoretical content for
publishing papers. The discussion of this section is summarized as follows:

Industrys fear that formal methods may slow down development unacceptably.

Industry requiring techniques that are easily understood by shop-floor personnel.

Industrys inability to provide training in formal methods to its personnel due to lack of time.

Difficultly in the use and practice of formal methods.

Unavailability of experts from academia to create awareness on usage of formal methods.

Industrys reluctance to discuss problems with academia.

Academia concentrating on industrially irrelevant problems.

Academia does not communicate with industry to find out real application problems.

Academia developing approaches that require strong theoretical backgrounds.

Academia developing new tools rather than improving existing tools.

With the information provided by sections II, III, we proceed to section IV where we aim to
propose solutions to the problems identified in section III.

IV Solution to Identified Problems

Here we aim to provide solutions to overcome the barriers which cause the gap between
academia and industry and suggestions to further strengthen those factors where academia and
Industry already have commonality. Based on a detailed study and analysis, the following
solutions are proposed to reduce the gap.

Industry overcoming myths about formal methods


There is a myth that the use of formal methods delays the development process. It is true that
several formal-methods projects have run over schedule. This does not imply that this problem is
inherent in formal methods. Estimating development time is very difficult even with traditional
development methods. Historical information about formal development techniques is also
scarce as formal methods have not yet been applied to a sufficient number of projects. Moreover
many of the publicized formal methods projects have been in specialized domains, producing
data of limited use. Hence, the Industry should make a thorough survey of formal development
processes and the highlights of successes, failures, and hindrances for getting the required
information before deciding its credibility.

Academia and Industry improving educational levels

The main reason for the unsuccessful use of formal methods by industry is the lack of education
in formal methods. A pre-requisite to introduce formal methods in an existing setup is to
investigate how they can be combined with methods that are in use. This does not mean that
everyone should be trained to know finite state machines. The fact is that the principles offered
by formal methods are different from the principles about which the industry engineers are
concerned. There is no easy way of applying a formal method to engineering problems. Hence,
one way of overcoming this problem is to have the needed education in formal methods. The
following is a result of a survey conducted with a group of students at the University of Illinois at
Chicago.

A course titled Advanced Manufacturing Information systems, based on PLC programming and
control, was offered to graduate students and accompanied by lab sessions at the Discrete Event
system (DES Lab) at the University of Illinois, Chicago. The students had prior knowledge
related to the use of Ladder logic techniques using RSLogix 5000 and Allen Bradley PLCs. A
survey was given based on the curriculum. The survey questions included

1. What is your level of comfort while using formal theories to solve simple manufacturing
problems?

2. How useful are the formal theories to solve real-world problems?

3. Can the theoretical knowledge provided by these theories make a good Control Engineer?

From the survey responses, it was clear that formal methods were difficult to apply to real-world
problems. It was realized that formal methods should no more remain merely an academic
exercise. The formal methods provide support for a formal specification. But they fail to support
many aspects of the more traditional structured-development methods. In the context of control
engineering, a method consists of an underlying development model and guidance for applying
these in a coherent manner. However formal methods, while supporting support the design
principles of traditional methods like stepwise refinement and top-down design, place little
emphasis on the underlying development model and provide little guidance as to how the
development must proceed. The students felt that the formal approaches were time-consuming
and difficult. It becomes more complicated for complex systems. This hints the possibility of
academia spending too much time in solving the wrong problems, thus not providing the solution
to solve industry problems. For too long, formal methods researchers have worked only on
intellectually interesting problems. Although academic-style work is essential to the
advancement of the field, it cannot be the only work that is done. Because so few researchers
have concentrated on industrially relevant problems, most of the examples and models that have
been developed have borne little resemblance to the examples and models needed in industry.

Academia solving industrially relevant problems

Industry cannot be expected to develop all its models alone. It requires a contribution from
academia also. So, more formal methods researchers must become knowledgeable about the
problems relevant to industry, and develop examples and tools appropriate to those domains.
Academia must be willing to tackle problems that are not intellectually interesting but that
industry needs solved. It also means that it must be willing to forgo powerful and sophisticated
methods, when simple ones will suffice.

Academia focusing more on existing tools and curbing the desire to create new tools

An impediment to the industrial use of formal methods is the expectation that if one builds a new
tool, it will be used. This viewpoint overlooks the gulf that exists between research world and the
industrial world. Industry rarely has the time and resources to keep up with the vast number of
ideas and emerging from the research community and the research community is often ignorant
of the challenges that industry faces. Tool developers could curb their desires to create more
powerful tools. When a tool is available to industry, academia experts must be available to
provide guidance on the tools use and ways to overcome its shortcomings. Thus, there is a
growing need for training courses that are tailored for industrial users.

Creation of a forum for academia and industry to meet

One solution to prevent the academia from solving the wrong problems is to have a forum for
academicians and people from industry to meet. This forum should be used by industry to
present its problems to academia. The forum could act as a channel through which industry and
academia interact. This provides an opportunity for academia to get real problems from industry
and solve them. Academia can continue to solve its own research issues. However, since models
developed by them do not find real applications, it will be advantageous to them to solve real
problems also.

Industry academia interaction through students

Another way in which academia and industry can interact is by industries offering projects and
summer internships to students. These programs, started with the intention of providing industry
experience to students, while giving them hands-on experience, helps them identify problems of
industrial relevance that they could work on in their universities.

Academia offering courses on current trends in industries


Majority of the control engineering positions in manufacturing today involves the use of rugged
PLCs. The university has been slow to recognize this trend. Only a few universities have entire
courses devoted to PLCs. In our opinion, it is necessary that every university that teach control
engineering has one course related to PLCs. The curriculum provided at universities should
keep the students abreast of the latest trends in industry. The formal aspects of courses should be
integrated with the syllabus. The gap between application domains and mathematics should be
bridged by providing better education for the control-engineering practitioners, so that they are
comfortable with formal methods. Many of the solutions proposed here, if implemented, could
lay the foundation for a good interaction between academia and industry and pave way for
reducing the gap.

V- Conclusion

In this paper, the clues for the gap between industry and academia in implementing formal
control methodologies were identified. The solutions suggested were based on a thorough study
of various documents. Common and conflicting areas of work between industry and academia
were identified. Solutions were provided to reduce the gap; in some cases it was found that the
gap between academia and industry was not because of the difference in the approaches of
academia and industry but only because of miscommunication and misunderstanding between
them.

References

[1] R. David, Grafcet: a powerful tool for specification of logic controllers, IEEE Transactions
on Control Systems Technology, 3(3) (1995) 253-268.

[2] G. Frey, L. Litz, Formal methods in PLC programming, IEEE International Conference on
Systems, Man, and Cybernetics, 4(4) (2000), 2431-2436.

[3] F.G. Gonzalez, W. J. Davis, Developing a physical emulator for a flexible manufacturing
system, IEEE International Conference on Systems, Man, and Cybernetics, 1, 1998, 710-715.

[4] A. M. Graham, M. Etezadi-Amoli, Design, implementation, and simulation of a PLC based


speed controller using fuzzy logic, IEEE Power Engineering Society Summer Meeting, 4, 2000,
2475 - 2480.

[5] F. Jimenez-Fraustro, E. Rutten, A synchronous model of IEC 61131 PLC languages in


SIGNAL, Euromicro Conference on Real-Time Systems, 2001, 135 -142.

[6] K. Johnsson, Performance of logic controllers in industrial environment, IEEE Conference


Record of the Industry Applications Society Annual Meeting, 2, 1991, 1699-1702.

[7] S. Lamperiere- Couffin, O. Rossi, J._M. Roussel, J.-J. Lesage, Formal validation of PLC
Programs: A survey, Proceedings of European Control Conference, 1999,
http://www.lurpa.enscachan.fr/publications/col_act.html.
[8] K. Lemmer, B. Ober, E. Schnieder, Model-based programming and diagnosis for
programmable logical controllers, IEEE International Conference on Systems, Man and
Cybernetics, 5, 1995, 4474-4479.

[9] R.W. Lewis, What are the benefits from using IEC 1131-3 PLC languages as the basis for
training future control engineers? IEE Colloquium on Mechatronics in Education: Delivery of a
New Engineering Discipline into the Workplace, 1996, 711-717.

[10] J. Liu, H. Darabi, Ramadge-Wonham Supervisory Control of Mobile Robots: Lesson from
Practice, IEEE International Conference on Robotics and Automation, 1, 2002, 670-675.

[11] S.S. Peng, M.C. Zhou, Conversion between ladder diagrams and PNs in discrete-event
control design-a survey, IEEE International Conference on Systems, Man, and Cybernetics, 4
(2001) 2682- 2687.

[12] J. R. Pollard, Ladder Logic Remains the PLC Language of Choice, Control Engineering,
1994, 77- 79.

[13] P.J.G.Ramadge, W.M. Wonham, The control of discrete event systems, Proceedings of the
IEEE conference, 77 (1) (1989), 81 -98

[14] M. Rausch, B.H. Krogh, Formal verification of PLC programs, Proceedings of the
American Control Conference, 1, 1998, 234-238.

[15] K. Venkatesh, M.C. Zhou, R.J. Caudill, Comparing ladder logic diagrams and Petri nets for
sequence controller design through a discrete manufacturing system, IEEE Transactions on
Industrial Electronics, 41 (6) (1994) 611-619.

Potrebbero piacerti anche