Development of mobile surface water filtration system through simulation using coloured petri net (cpn)

IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163
__________________________________________________________________________________________
Volume: 02 Issue: 03 | Mar-2013, Available @ http://www.ijret.org 367
DEVELOPMENT OF MOBILE SURFACE WATER FILTRATION
SYSTEM THROUGH SIMULATION USING COLOURED PETRI NET
(CPN)
Aashish Vatsyayan1
, B. Hemalatha2
, A Vimala Juliet3
, Ishwarya Venkatram4
AP(Sr.G), HOD, SRM University, SRM University, SRM University, SRM University
Aim
Testing the working of the mobile surface water filtration system through the design and simulation of a CPN generated schema
Methodology
For understanding the required action for different conditions the programming is developed using CPN tools for analysis of different
conditions and obtaining a holistic picture of the system.
Result
The simulation model has demonstrated how the working of the mobile system is intended and has the actions of a microcontroller
clearly cut out for the programmer to understand and write the code. Each module was looked upon in the different scenarios and the
changes in the system after simulation were shown, providing a holistic understanding of the entire system. The sensors,
microcontrollers and actuators have been discussed to bridge the gap between completely theoretical concepts to an implementable
entity.
Conclusion
Impending release of tool for direct conversion of Coloured Petri Net simulation into hexadecimal format opens up infinite avenues
for system modelling in Coloured Petri Net environment. The model design can be improved upon by adding
time dependency property of CPN to simulate more dynamic and time dependant processes, this aspect has not been explored as a
holistic illustration, but the system for mobile surface water filtration device has been thoroughly and holistically explained and
explored.
Key words: Coloured Petri Nets, simulation, modelling, filtration system, mobile, navigation, electro coagulation.
--------------------------------------------------------------------***-----------------------------------------------------------------------
1. INTRODUCTION
The simulation of CPN is based on triggering of inputs and
demonstrating the resultant action impact on the system, it is
parallel process simulation software where the triggering of
one input has response from different sub systems, all these
processes can be observed to code a microcontroller for
performing similar actions in real time situations.
Before actual fabrication of an electrical system involving
microcontrollers, sensors, actuators it is necessary to develop a
logic that dictates when which actuator should react to given
input, here CPN is used to display a blueprint of the logic that
will be implemented into an actual functioning hardware
prototype.
C P nets have various characteristics and some of the major
characteristics touched upon while modeling the design were
the liveness, boundedness and stochastic nature of the net. The
design is that of a discrete, continuous and hybrid net. We also
have introduced a dimension of high-level Petri nets by
modeling using a CPN interface. It allows us to generate high -
level nets with token values and hierarchy that allow
the transitions on various levels to fire. The problem of non-
determinism has been tackled and the net has been
incorporated with concurrent behavior in selected simulations.
The Bayesian nature of the Petri nets has been explored and
incorporated in the net schema.
The entire paper revolves around creating a simulation model
to understand the working of the developed system and
display the interconnectivity amongst the different parameters.
Coloured Petri Nets is chosen as the platform for creating a
simulation model as the token transfer in CPN is clearly
understandable along with it treating all values in 0 or 1,
which creates a direct colliery with the working of a
microcontroller. CPN adds the concept of coding, graphical
interface and Petri nets to produce a system which has
powerful primitives and has a vast potential. The obtaining of
a hexadecimal code from a CPN simulation is in its path of a
commercial release and it will improve current modelling
schemes as CPN will greatly reduce the processing time in
microcontrollers. The simulation displays the working of the
system under different situations and what response will be
triggered in each scenario.

__________________________________________________________________________________________
The case by case analysis is done to explain the working of the
mobile surface water filtration device. The modules discussed
are: navigation, electro coagulation and an overall working
and output of the system. Before and after screen shots are
used to explain the reaction of the output actuators on the basis
of triggering of the input sensors.
MATERIALS AND METHODS
The CPN model was designed using the CPN tools interface
window. A new page was created in the workspace and was
saved using .cpn extension. The places, transitions required for
the model were created and arcs connecting the places and
transitions were designed. The places and the transitions were
labelled with unique names and assigned data types. The
places then were designated with a set number of tokens (i.e. 0
or 1). Next the variable declarations were made by creating an
empty declaration and then modified according to the variable
names and their respective data types. The function
declarations were defined according to the various input
parameters and an optimal code for generation of suitable
output was written, for the respective functions. The variables
and functions were then labelled onto the arc inscriptions, thus
developing a proper place -arc-transition schema defining the
structure of the autonomous mobile surface water filtration
system. Subsequently, hierarchy and other parameters were
added to increase the parallelism and complexity in the net and
the net was colour coded. The non-determinism was verified
and the schema was completed. The model generated was
simulated with various input parameters and the optimal
output was generated by the net.
The sensors are hydrocarbon sensors which can differentiate
between crude oil and water, depending upon the range of
their output a suitable analogue to digital converter can be
used to give a high (i.e. 1) value when oil is detected and a low
(i.e. 0) to the microcontroller. The sensors are connected on
the body of the mobile device at 5 locations in a star format, 5
sensors are considered only to show how the discrimination of
direction is to be made however more sensors can be added to
increase the accuracy of system, it would result in more data to
process though. One hydrocarbon sensor is connected on the
input pipe to determine is oil is present in immediate
proximity of the input site; it plays a nominal role in
navigation.
Level sensors are used inside processing chamber to
determine the level of the process liquid, floating switches
type can be used.
Microcontroller of choice is 8051 is due to it being familiar
and easy to use with it and as the mobile device is still in
development stage 8051 provides with the processing
requirements at this stage.
Figure 1 A pin diagram shows the number of available inputs
and outputs.
Output actuators
The output actuators in this model are primarily valves and
motors and propellers
Valves: for open and close operation at the input site. A simple
solenoid valve can be used for this operation.
Motors: the motor being used is a simple DC motor having
high torque and low rpm to facilitate the movement of input
pipe. The motor is enclosed inside the processing tank and is
replaceable in case the motor burns out; a high torque has been
chosen to move the input pipe despite the viscosity offered by
water.
Propellers: the propellers are high torque and high rpm to
facilitate the movement of the mobile device.
The battery backup is not discussed in this paper, plausible
avenues for research into supply of electricity or energy are:
crude oil internal combustion engines, solar cells, microbial
fuel cells and wind energy.
The process of electro coagulation mentioned in the paper can
be looked up in relevance in the paper ‘mobile surface water
filtration system’.
Operation:
The navigation and electro coagulation units work in 6
different modes. The before and after simulation screen shots
are explained.

__________________________________________________________________________________________
CASE 1: When all the sensors test positive for oil:
1.1 BEFORE SIMULATION:
1.1.1 WHOLE MODEL SCHEMA:
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 1
Sensor 5 : 1
Navigation(Propeller main,Propeller left, Propeller right) : 0
Electrocoagulation(Input valve,Level sensor1,Levelsensor2) :
0
Output : 0
Figure 1.1.1 Model before simulation when all sensors are 1
1.1.2 NAVIGATION SCHEMA:
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 1
Sensor 5 : 1
Figure 1.1.2 Navigation model before simulation when all
sensors are 1 and propellers are 0
1.1.3 ELECTRO COAGULATION SCHEMA:
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 1
Sensor 5 : 1
0
Output : 0

__________________________________________________________________________________________
Figure 1.1.3 Electro coagulation before simulation when all
sensors are 1, level sensors are 0 and output is 0
1.2. AFTER SIMULATION:
Electrocoagulation(Input valve,Level
sensor1,Levelsensor2) : 0
Output : 1
Figure 1.2.1 Model after simulation yields output 1
Propeller main : 0
Propeller left : 0
Propeller right : 0
Figure 1.2.2 Navigation model after simulation yields
propeller outputs as 0
INPUT VALVE : 1
OUTPUT : 1
Figure1.2.3 Electro coagulation after simulation yields output
as 1

__________________________________________________________________________________________
CASE 2: When only three sensors test positive for
oil(i.e. sensor1,sensor2,sensor3):
2.1 BEFORE SIMULATION :
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 0
Sensor 5 : 0
0
Figure 2.1.1 Model before simulation when all sensor1,
sensor2, sensor3 are 1
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 0
Sensor 5 : 0
sensor1, sensor2, sensor3 are 1 and propellers are 0
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 0
Sensor 5 : 0
0
Output : 0

__________________________________________________________________________________________
sensor1, sensor2, sensor3 are 1, level-sensors are 0 and output
is 0
2.2 AFTER SIMULATION :
Output : 0
Figure 2.2.1 The simulation yields no output
Propeller main : 500
Propeller left : 0
Propeller right : 500
Inference: Left turn is made

__________________________________________________________________________________________
Figure 2.2.2 The simulation yields the system to turn left
INPUT VALVE : 0
OUTPUT : 0
Figure 2.2.3 Electro coagulation after simulation yields no
output
CASE 3: When only three sensors test positive for
oil(i.e. sensor3,sensor4,sensor5):
Sensor 1 : 0
Sensor 2 : 0
Sensor 3 : 1
Sensor 4 : 1

__________________________________________________________________________________________
Sensor 5 : 1
0
Figure 3.1.1 Model before simulation when all sensor3,
sensor4, sensor5 are 1
Sensor 1 : 0
Sensor 2 : 0
Sensor 3 : 1
Sensor 4 : 1
Sensor 5 : 1
sensor3, sensor4, sensor5 are 1 and propellers are 0
Sensor 1 : 0
Sensor 2 : 0
Sensor 3 : 1
Sensor 4 : 1
Sensor 5 : 1
0
Output : 0

__________________________________________________________________________________________
sensor3, sensor4, sensor5 are 1, level-sensors are 0 and output
is 0
Output : 0
Figure 3.2.1 The simulation yields no output
Propeller main : 500
Propeller left : 500
Propeller right : 0
Inference: Right turn is made

__________________________________________________________________________________________
Figure 3.2.2 The simulation yields the system to turn right
INPUT VALVE : 0
OUTPUT : 0
Figure 3.2.3 Electro coagulation after simulation yields no
output
CASE 4: When all 5 primary sensors and 1 shaft
sensor test positive for oil:
4.1.1 MODEL SCHEMA:
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 1
Sensor 5 : 1
Sensor 6 : 1
Shaft position : 0
0

__________________________________________________________________________________________
Figure 4.1.1 All 6 sensors test positive for oil and shaft
position is 0
4.2.1 MODEL SCHEMA:
The inlet shaft is connected to a stepper motor of step angle
900 and hence it has 4 postions in its circular path at 00, 900,
1800, 2700 and then back to 00.
Shaft Position : 0
Inference: Shaft position is unchanged and input valve is
opened and electro coagulation happens
Figure 4.2.1 The simulation yields no change in position of
the shaft
CASE 5: When all 5 primary sensors except 1 shaft
sensor test positive for oil:
5.1.1 MODEL SCHEMA:
Sensor 1 : 1
Sensor 2 : 1
Sensor 3 : 1
Sensor 4 : 1
Sensor 5 : 1
Sensor 6 : 0
Shaft position : 0

__________________________________________________________________________________________
0
Output : 0
Figure 5.1.1 All 5 sensors test positive for oil except sensor 6
and shaft position is 0
5.2.1 MODEL SCHEMA:
Shaft Position : 1
Inference: Shaft position changes from 0 to 1 and new position
is assigned value 0 again. Though the shaft position changes to
1, till sensor 6 sends a new input as positive the input valve
does not open.
Figure 5.2.1 The simulation yields the shaft position to
change
CASE 6: When no oil sensors detect oil :
In this case there is no oil in the immediate vicinity of the
mobile device. There are 2 approaches to aid the device in
such a situation.
One approach is to make the mobile device to trace circles of
increasing diameters until it interacts with oil after which the
movement via sensors is guided. The movement will be
achieved by switching off one propeller (taking the left
propeller) and keeping the main propeller ON. This mode can
also be understood as ‘scouting’ for oil.
An alternate method is the use of thermal imaging to
determine the concentrated oil puddles and directing the
mobile device in that direction. Thermal imaging can be
utilized due to the unique property of oil that it heats up
rapidly during the day and cools off in the night. Hence at all
points of time oil will have a different heat signature from
water. By setting up a threshold algorithm to differentiate the
pixels in greyscale the position of oil can be determined. The

__________________________________________________________________________________________
thermal camera can be installed at a height at an inclination so
that it is facing the water body. The thermal camera will
operate within a range, the camera will be installed on a base
connected to a stepper motor having a step angle of 900, such
a step angle is kept due to the high amount of data that needs
to be processed, we can have a smaller step angle and operate
at a higher accuracy but it would need higher computation and
data storage. The implementation of thermal imaging with
relevant data on setting threshold has not been discussed in the
paper.
Arc inscription code
Navigation:
Propeller-main
fun propellermain(a,i,j,k,l) =
if(a > 0 andalso i > 0 andalso j > 0 andalso k > 0 andalso l >0 )
then 1`0
else if(a > 0 andalso i > 0 andalso j > 0 andalso k < 1 andalso l
< 1) then 1`500
else if (a > 0 andalso i < 1 andalso j < 1 andalso k > 0 andalso
l >0 ) then 1`500
else 1`1000
Propeller -left
fun propellerleft(a,i,j,k,l) =
then 1`0
else if(a > 0 andalso i > 0 andalso j > 0 andalso k < 1 andalso l
< 1) then 1`0
l >0 ) then 1`500
else 1`1000
Propeller- right
fun propellerright(a,i,j,k,l) =
then 1`0
else if(a > 0 andalso k > 0 andalso l > 0 andalso i < 1 andalso j
< 1) then 1`0
l >0 ) then 1`500
else 1`1000
In the ever changing trend of modern day society, there
develops a need to bridge the gap between user interface and
direct implementation into hard drive. CPN thus offers a much
simpler, more dynamic interface with powerful primitives that
enable parallel processing and high throughputness. The
system that has been modelled is aimed at cleaning oil-spills
but can also be used to clean still water bodies against
eutrophication and other pollutants. The paper describes the
structure and concepts of mobile surface water filtration
system but not that of process detail and power management,
as they are not necessary for simulation. The impending
release of a compiler that can generate hexadecimal code from
CPN xml code can facilitate easier code generation from a
dynamic and user friendly tool such as CPN. Also the future
design of the paper could incorporate time dependency to
make the simulation more real-time oriented.
ACKNOWLEDGEMENTS
The authors would like to thank the departments of
Instrumentation and Control Engineering and Bioinformatics
Engineering of SRM University, Chennai for their support and
aid in making this paper possible.
REFERENCES
[1]. ‘Mobile surface water filtration system’- Aashish
Vatsyayan, B. Hemalatha, IJET
[2]. OIL SPILL SENSOR USING MULTISPECTRAL
INFRARED IMAGING VIA `1 MINIMIZATION, Yingying
Liy_, Wei-Chuan Shihy, Zhu Hany, and Wotao
Yin_yElectrical & Computer Engineering Department,
University of Houston_Computational and Applied
Mathematics, Rice University
[3]. An Introduction to the Practical Use of Coloured Petri
Nets, Kurt Jensen, Department of Computer Science,
University of Aarhus, Ny Munkegade, Bldg. 540, DK-8000
Aarhus C, Denmark

Development of mobile surface water filtration system through simulation using coloured petri net (cpn)

More Related Content

Viewers also liked (11)

Similar to Development of mobile surface water filtration system through simulation using coloured petri net (cpn) (20)

More from eSAT Journals (20)

Recently uploaded (20)

Development of mobile surface water filtration system through simulation using coloured petri net (cpn)