Volume 9, Issue 7, July – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24JUL664

Artificial Intelligence Based Embedded Security

Solution Model for PID Controller
Bitrus Haruna1; Mathew Ehikhamenle2
Centre for Information and Telecommunication Engineering
University of Port Harcourt Choba, Rivers State, Nigeria

Abstract:- This study investigated the impact of artificial I. INTRODUCTION

intelligent based embedded security solution model for
PID controller. Recent studies have revealed that the In general, we are living in the era of digital warfare
conventional mitigation techniques like zoning, (Kim, Countdown to Zero Day, 2021). According to the
demilitarization, firewalls, company’s policies to mention American National Standard Institute and the International
but a few are no longer enough to match the sophisticated Society of Automation (ANSI/ISA), it has moved up from the
intelligence being deployed by attackers hence the urgent internet domain, which is level 5, to the enterprise domain,
need for the adoption of artificial intelligent based level 4, and finally to the Process Control Domain, which is
embedded security solution model. Furthermore, levels 0, 1, and 2 (ISA, ANSI/ISA-TR99.00.01-2021 Security
conventional IT security solutions cannot be deployed Technologies for Industrial Automation and Control
directly in process controllers given the real time Systems, 2021). According to Kotenko & Ulanov's Packet
availability requirement of industrial control systems. Level Simulation of Cooperative Distributed Defense against
This has limited the security solutions to firewalls, Internet Attacks (2018), it appears that conventional
network segmentation and some laid down policies which mitigation methods like zoning, demilitarization, firewalls,
the control system operators are not expected to violate. and company policies, to name just a few, are no longer
This study will ultimately help deliver algorithms that will sufficient to match the sophisticated intelligence being
be implemented in industrial field device controllers deployed by attackers.
making them resilient to cyber-attacks whose
consequences have far reaching implications. With In light of these prevalent threats, it is necessary to
regards to the methodology of the work, four sets of data investigate the control system loop, the fundamental
were collected from the test bed. The first data, Data_1, component of a control system, in order to identify any
shown in appendix D is the plant’s response to the existing security gaps and offer appropriate solutions to those gaps.
PID algorithm. This was done by taking the temperatures The developed solution can also be used in the
of the plant at interval of 3 seconds after loading the pharmaceutical, automotive, and robotics industries, among
controller with the existing (insecure) PID algorithm other automation-related fields.
shown in appendix A. A total of 100 samples were taken.
From the methodology employed, it was discovered that II. PROBLEM STATEMENT
it is seen that the existing PID algorithm was able to
achieve the control objective of maintaining the A distributed control system's process controller is
temperature of the process plant within the temperature similar to a football goalkeeper. Despite the fact that it is the
range of 38 º C and 43 º C with 40 º C as the optimal or last "man" to be defeated when it comes to attacking
ideal performance. From table I in appendix I, the mean industrial processes, it lacks any form of security other than
steady state error (MSSE) of the existing PID algorithm security by obscurity, which cannot be completely relied
considering the 27th to 101th temperature data is upon given the open network architecture of contemporary
0.062631579 which is approximately 0.06. It means the distributed control systems. Tragically, the customary IT
accuracy of the existing PID algorithm is ((40 – security arrangements can't be sent straightforwardly in
0.06)/𝟒𝟎) ∗ 𝟏𝟎𝟎 = 𝟗𝟗. 𝟖𝟓 %. This showed that the process regulators given the continuous accessibility
existing PID control algorithm’s performance is necessity of modern control frameworks. As a result, security
acceptable under normal condition since the minimum measures have been restricted to firewalls, network
control accuracy required to achieve the control objective segmentation, and a few predetermined policies that control
in the considered process plant is (40 – (3-2)/2)/𝟒𝟎) ∗ system administrators are not expected to violate (Shell group
𝟏𝟎𝟎 = 𝟗𝟖. 𝟕𝟓 %. of Companies, 2021). Field controllers, which have a
recovery time objective (RTO) of zero, can be argued to be
Keywords: Artificial Intelligence, Security Solution more important than other components of a distributed
Model, PID Controller. control system, so it has become important to develop
enhanced and appropriate security solutions for them. This
study sought to critically examine the impact of an embedded
security solution model for PID controller based on artificial
intelligence on the foregoing.

IJISRT24JUL664 www.ijisrt.com 889

Volume 9, Issue 7, July – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24JUL664

III. SIGNIFICANCE OF THE STUDY Figure 1 above is the sequence of the approach adopted
to realize the aim of this work.
 The Significance of this Study cannot be Over-
Emphasized. These Include but not Limited to: V. RESULTS AND DISCUSSIONS

 It will help deliver algorithms that will be implemented in The plant's response to the existing PID algorithm under
industrial field device controllers making them resilient to normal circumstances is depicted in figure 2, while the plant's
cyber-attacks whose consequences have far reaching response to the secured PID algorithm is depicted in figure 3.
implications. Under normal conditions, the responses of the process plant
 This study is also important based on the fact that without to the secured and existing PID algorithms are compared in
field device controllers, industries would lack required Figure 4.
efficiency in their business operations. Consequently, it
can be said that the device controllers do not only control The process plant's response to the existing PID
field devices, but also control economy. algorithm under threat or attack is depicted in Figure 5, while
the process plant's response to the secured PID algorithm is
IV. METHODOLOGY depicted in Figure 6 under the same threat or attack
conditions.
This work employed a top-down and bottom-up design
approach in combination. The mathematical models of the The plant's response to the existing and secured
secured PID algorithm, or SPIDA, an artificial intelligence algorithms under the same threat conditions is compared in
algorithm security solution for industrial process control, Figure 7.
were initially developed using a top-down approach. The
software codes and flowcharts for putting the secured PID
algorithm into action were also created using the same
method. The product codes were created utilizing inserted c
language with Arduino uno as the objective gadget. Granular
perspective was then used to plan and execute a model
intensity control framework which was utilized to approve
the got PID calculation.

Fig 1: Methodology Flow Chart

IJISRT24JUL664 www.ijisrt.com 890

Volume 9, Issue 7, July – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24JUL664

Fig 2: Plant’s Response to the Existing PID Algorithm Under Normal Condition

Fig 3: Plant’s Response to the Secured PID Algorithm Under Normal Condition

Fig 4: Response of the Process Plant to the Secured and the Existing PID Algorithms Under Normal Condition

IJISRT24JUL664 www.ijisrt.com 891

Volume 9, Issue 7, July – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24JUL664

Fig 5: Response of the Process Plant with the Existing PID Algorithm Under Threat Conditions

Fig 6: Response of the Process Plant with the Secured PID Algorithm Under Threat Conditions

Fig 7: Response of the Process Plant to the Secured and the Existing PID Algorithms Under the Same Threat Conditions

IJISRT24JUL664 www.ijisrt.com 892

Volume 9, Issue 7, July – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24JUL664

A. Discussion on the Plant’s Response to the Existing PID C. Contribution of this Article to the Body of Knowledge
Algorithm Under Normal Condition The following are the contributions that this work made
From figure 2, it is seen that the current PID calculation to the existing body of knowledge:
had the option to accomplish the control objective of keeping
up with the temperature of the interaction plant inside the  It created and incorporated proactive security into the
temperature scope of 38 º C and 43 º C with 40 º C as the ideal widely used back-end control algorithm used in the
or optimal execution. The current PID algorithm's mean automation industry, positioning the sector for cloud-
steady state error (MSSE) for the 27th to 101st temperature based automation without posing a threat from within or
data is 0.062631579, or approximately 0.06, according to outside the organization.
table I in appendix I. This indicates that the current PID  The cycle plant worked to approve the got PID calculation
algorithm is accurate to ((40 – 0.06)/40)*100, or 99.85 can be taken on in colleges and polytechnics for effective
percent. Since the minimum control accuracy required to educating and learning of PID control standards
achieve the control objective in the considered process plant particularly in the creating and immature nations of the
is ((40 – (3-2)/2)/40)*100=98.75 percent, this demonstrated existence where the investigation of control designing is
that the performance of the existing PID control algorithm is a greater amount of numerical speculations than
acceptable under normal conditions. reasonable. Students will be better prepared for successful
careers in the automation industry as a result of this.
Figures 3 and 4 show that the secured PID algorithm
was also able to achieve the control objective of maintaining REFERENCES
the process plant's temperature within the temperature range
of 38 o C to 43 o C, with 40 o C representing the optimal or [1]. bdel-geliel, M., Qaud, F., & Ashour, H. (2020).
ideal performance Realization of adaptable PID controller within an
industrial automated system. IEEE 11th international
B. Discussion on the Plant's Response to the Secured PID conference on control & automation (pp. 965-970).
Algorithm Taichung: IEEE.
Under Normal Conditions The secured PID algorithm's [2]. Anthony, K. (2018). Fundamental of PID control.
mean steady state error (MSSE) for the 27th to 101st PDH Center.
temperature data is 0.288421, or approximately 0.29, [3]. Eric, D. K. (2011). Industrial network security:
according to table I in appendix I. This indicates that the securing critical infrastructure networks for smart
secured PID algorithm is accurate to ((40 – 0.29)/40)*100, or grid, SCADA, and other industrial control systems.
99.28 percent. This demonstrated that, despite the fact that the USA: Elsevier publisher.
secured PID control algorithm has an effect on the availability [4]. Fairchild, S. (2017, August 29). Datasheet catalog.
and integrity of the process plant, the impact is still within an Retrieved August 29, 2017, from Datasheetcatalog:
acceptable range. http://www.datasheetcatalog.com/datasheets_pdf/4/N
/3/5/4N35.shtml
It is seen from figures 6 and 7 that the got PID [5]. Fawzi, H., Tabuada, P., & Diggavi, S. (2012). Security
calculation (SPIDA) had the option to accomplish the control for control systems under sensor and actuator attacks.
objective even within the sight of the mimicked danger IEEE international conference on Decision and
specialists. Taking into account temperature data from the Contro (pp. 3412 – 3417). Maui HI : IEEE.
27th to the 126th, the MSSE of SPIDA was determined to be [6]. Gupta, D., Mohapatra, P., & Chuah, C. (2008).
0.372475248, or approximately 0.37, according to table J2 in Efficient monitoring in wireless mesh networks:
appendix J. ((40 – 0.37)/40)*100 = 99.08 percent accuracy. overheads and accuracy trade-offs. IEEE 5th
Still, this is perfectly acceptable! Figure 5 makes it international conference on Mobile Ad Hoc and
abundantly clear that the current PID algorithm (EPIDA) was Sensor systems (pp. 13-23). Atlanta: IEEE hjkml.
unable to accomplish its control goal. Based on the 27th to (2017). gghhl. ghjk (p. 67). bbb: bb
126th temperature data in appendix J's table J1, the MSSE of [7]. Hong, R. (2011). Research and application of TCP/IP
EPIDA is 9.50029703. ((40 – 9.5)/40)*100=76.25 % is the protocol in embedded system. IEEE 3rd international
accuracy. In the process plant under consideration, the conference on communication software and networks
minimum control accuracy required to achieve the control (pp. 584-587). Xi'an: IEEE.
objective is currently 98.75 percent. The EPIDA is not [8]. Hongji, W., & Zhenqiu, H. (2011). PID parameter
suitable for remote applications where threat agents may have optimization of pressure and phear testing controller
access to process control parameters because 76.25 percent is based on CLPSO algorithm. International conference
less than 98.75 percent. The fact that the finished product on computer-aided design, manufacturing, modelling
does not meet standard requirements will have a practical and simulation (pp. 105-109). Switzerland: Trans
effect on the acid gas removal procedure from natural gas. Tech Publications.
The company (SNEPCo) will face sanctions as a direct result.
In addition to making customers unhappy, this will give
competitors an advantage in business.

IJISRT24JUL664 www.ijisrt.com 893

Volume 9, Issue 7, July – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24JUL664

[9]. ICS-CERT. (2019, December 12). ICS-CERT. [23]. Smith, D. J., & Kenneth, G. S. (2011). Safety Critical
Retrieved September 24, 2015, from https://ics- Systems Handbook: a straightforward guide to
cert.us-cert.gov/Training-Available-Through-ICS- functional Safety, IEC 61508 and related standards.
CERT#need: USA: Elsevier pulishers.
https://d36spl5w3z9i0o.cloudfront.net/dcd/scormapi_ [24]. Stan, Ż. (2013). An Introduction to Proportional-
v60/launcher.html?host=ics-cert- Integral-Derivative (PID) Controllers. Lafayette:
training.inl.gov&id_user=22129&id_reference=203 Purdue University.
&scorm_version=1.3&id_resource=24203&id_item= [25]. Stuart, M. (2000). Operational aspects of oil and gas
111&idscorm_organization=111&id_package=111&i well testing. USA: Elsevier publisher.
d_course=17&launch_type=popup&auth_ ICS- [26]. Texas, I. (2015, June 4). DatasheetCatalog.com.
CERT. (2021, June 30). ICS-CERT. Retrieved November 10, 2015, from LM35 Precision
[10]. Ikhlef, A., Kihel, M., Boukhezzar, B., Mansouri, N., Centigrade Temperature Sensors:
& Hobar, F. (2015). Remote PID control of tank level http://www.datasheetcatalog.com/datasheets_pdf/L/
system. International Conference on Interactive M/3/5/LM35.shtml
Collaborative Learning (pp. 20-24). Italy: IEEE. [27]. Timberlake. (2015). Chemistry: An Introduction to
[11]. NIST. (2021, January 30). National Institute of General, Organic, and Biological Chemistry. Los
Standards and Technology. Retrieved 2 3, 2015, from Angeles: Pearson .
www.csrc.nist.gov: [28]. Torres, G. (2014, November 13). Hardwaresecrets.
http://csrc.nist.gov/publications/nistpubs/800- Retrieved March 25, 2015, from
82/SP800-82-final.pdf www.hardwaresecrets.com:
[12]. Okeke, P. N., & Ndupu, B. L. (1983). Ordinary level www.hardwaresecrets.com/article/433
physics. New Jersey: Prentice Hall Press. [29]. Vicomsoft. (2014, September 12). Vicomsoft Leraning
[13]. Peng, Z. (2010). Advanced Industrial Control Centre. Retrieved April 4, 2015, from
Technology. Amsterdam: Elsevier. www.vicomsoft.com: www.vicomsoft.com/learning-
[14]. Peter, L. (2013, April 23). TechTarget. Retrieved center/firewalls/ Weli, W. (2014). Bonga South West
September 25, 2015, from Instrument Operation Summary. Lagos: SNEPCo.
http://searchsecurity.techtarget.com: [30]. Wikipedia. (2015, January 3). Wikipedia
http://searchsecurity.techtarget.com/resources/Hacker encyclopedia. Retrieved February 12, 2015, from
-Tools-and-Techniques-Underground-Sites-and- http://en.wikipedia.org:
Hacking-Groups http://en.wikipedia.org/wiki/Systems_analysis
[15]. Philips. (2021, August 29). Datasheet Catalog. [31]. WWittrisch, C., & Chole, H. (2012). Progressing
Retrieved August 29, 2017, from DataSheetcatalog: cavity pumps: oil well production artificial lift.
http://www.datasheetcatalog.com/datasheets_pdf/B/T Energies Nouvelles.
/1/3/BT139.shtml [32]. Yadong, L., Wenqiang, C., Danlan, L., & Ru, i. Z.
[16]. Reddy, Y. J. (2015). Industrial process automation (2011). Research based on OSI model. IEEE
systems: design and implementation. Amsterdan: international conference on communication software
Elsevier. and Networks (pp. 554-557). Xi'an: IEEE.
[17]. Schneidar, E. (2013, July 23). The Engineering
Toolbox. Retrieved October 23, 2015, from
www.engineeringtoolbox.com:
http://www.engineeringtoolbox.com/air-properties-
d_156.html
[18]. Schneider, E. (2021, September 25). PID Control
made easy. Retrieved August 28, 2017, from
Eurotherm: http://www.eurotherm.com/pid-control-
made-easy
[19]. Shahrukh, A., & Sandip, G. (2014). Microcontroller
implementation of digital pid . Rourkela: National
Institute of Technology.
[20]. Shell Global Solutions, I. (2015). DESIGN AND
ENGINEERING PRACTICE. Netherlands: Shell
Group of Companies.
[21]. Shell group of Companies. (2021, May 12). DEP-
00.00.05.05-GEN. Retrieved February 03, 2015, from
www.scribd.com: Shell group of Companies, “Design
Engineering Practice Specification:
prowww.scribd.com/archive/plans?doc=238070612&
escape=false&
[22]. Silva, V., Carvalho, V., Vasconcelos, R., & Soares, F.
(2016). Remote PID Control of a DC Motor.
International Journal of Online Engineering.

IJISRT24JUL664 www.ijisrt.com 894