Implementation of a Smart AC Automation System with Room Temperature Prediction

Abstract

Room Temperature prediction in Air Conditioners is highly challenged and ambiguous in today's life. To develop the system, some hardware like raspberry pi zero, thermal sensor, microcontroller, IR sensor and the room’s AC with existing remote are used. The proposed system is implemented through an embedded system by using the Python programming language. Based on the data, a mathematical formula can be derived and an algorithm of the proposed system has been designed and developed for the predicted temperature data with the values of the two sensors. The sensors are used to detect temperature and the AC performs automatically turn on or turn off. This system can be implemented in any smart AC room where anyone can utilize the AC system automatically switched on/off with the predicted temperature. This also can used in all over the places including for disabled peoples, personal room, conference room, hall room, classroom and transports, where manually control of AC is not feasible

Full text

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
23
Implementation of a Smart AC Automation System with Room
Temperature Prediction
F.M. Javed Mehedi Shamrat1, Shaikh Muhammad Allayear2 and
Md. Ismail Jabiullah3
1. 2Dept. of Multimedia and Creative Technology, Daffodil International University
3Dept. of Computer Science and Engineering, Daffdil International University
E-mail: 1shamrat777@diu.edu.bd, 2drallayear.swe@diu.edu.bd
3drismail.cse@diu.edu.bd
Abstract
Room Temperature prediction in Air Conditioners is highly challenged and ambiguous in today's life. To
develop the system, some hardware like raspberry pi zero, thermal sensor, microcontroller, IR sensor and the
room’s AC with existing remote are used. The proposed system is implemented through an embedded system by
using the Python programming language. Based on the data, a mathematical formula can be derived and an
algorithm of the proposed system has been designed and developed for the predicted temperature data with the
values of the two sensors. The sensors are used to detect temperature and the AC performs automatically turn on
or turn off. This system can be implemented in any smart AC room where anyone can utilize the AC system
automatically switched on/off with the predicted temperature. This also can used in all over the places including
for disabled peoples, personal room, conference room, hall room, classroom and transports, where manually
control of AC is not feasible.
Keywords: smart AC, Microcontroller, Automation, temperature prediction, thermal sensor, raspberry pi zero,
IR sensor and IR remote.
1. INTRODUCTION
Temperature prediction in Air Conditioners is exceedingly challenged in the present life. Automation
in air condition is as of now vital themes in the territory of IoT. It's to a really hard to change the
temperature erratically by physically for debilitated people groups, in sleeping hours, meeting time
and in for the most part where countless people exist there must be issue for comfort temperature for
all. A large number of the researchers work with room environment not with the object temperature,
so the outcome what is found is might be more productive if temperature prediction depends on object
temperature. If there are one or two peoples in a place, there can be possible to give an comfort
temperature for all but if there are huge peoples there is turmoil to give comfort temperature depend
on room temperature.
2. LITERATURE REVIEW
In recent years some research paper has been published, where researchers have where researchers
have demonstrated to lessen the utilization of power amid the utilization of air condition. But they
have a few works on Smart AC system by predicting in AC temperature.
The System can distinguish the surface temperature of tenants by a non-contact recognition at the
limit of 6 meters far. Separating human from other moving as well as static item by warmth variable is
almost inconceivable since human, creatures and electrical apparatuses produce heat. The wild
warmth properties which can change and exchange will add to the discovery issue. Coordinating the
ease MEMS based warm sensor can comprehend the first of human detecting issue by its capacity to
distinguish human in stationary [1].

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
24
Fig. 1: Human-detecting from Nearness Identification to the Recognizing Character. Just Two Detecting
Degree will be Required in this Undertaking which is Presence and Area
The system consequently controls cooling by methods for changing temperature settings in forced air
systems. Inside gadgets of climate control systems accordingly, don't need to be supplanted. A
versatile neuro fuzzy deduction framework and a molecule swarm calculation are received for
unraveling a nonlinear multivariable backward PMV model in order to decide warm solace
temperatures [2].
Fig. 2: Signal Transmission among Gadgets in the Remote Control Arranges
Most of the specialists worked for a Smart AC framework that can be constrained by remote gadgets.
They considered a remote sensor sent in the objective zone for detecting the encompassing
temperature. The remote sensor issues control directions to a remote cooling framework when the
privately detected surrounding temperature surpasses a specific attractive temperature extend. There
are a few issues considered in our sensor model [3].

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
25
Fig. 3: A Representation of the ON/OFF Cycle of Cooling
The creators of [4] utilized a model-prescient control system to learn and adjust for the measure of
warmth because of tenants and hardware. They utilized measurable techniques together with a
numerical model of warm elements of the space to evaluate warming burdens because of occupants
and hardware and control the AC likewise.
On the other hand, this paper presents a technique by methods for transmitting the temperature
directions by means of a remote sensor organize [5] to control forced air system tasks for tenants'
warm solace. The remote system is likewise used to acquire condition data including the temperature,
moistness, and airspeed at spots around inhabitants. Along these lines, utilizing the proposed control
setup does not need to change inside gadgets of existing climate control systems.
Fig. 4: System Built on the Berkeley Grounds that Permits Testing of Various Control Methodologies for
Controlling an AC so as to Investigate Tradeoffs between Vitality Utilization and following a Temperature Set
Point.
Warm comfort and cooling essentialness usage are two basic issues in spots of business. This paper
fundamentally separates the association between cooling essentialness use and warm comfort in
different urban networks. Directly off the bat, we present the significance of in-passage cooling
burden, and figure the ordinary office's indoor cooling load in thirteen urban regions; Secondly, in
perspective on a fundamental condition for Predicted Mean Vote (PMV), a tweaked control
enhancements method for cooling control structures has been proposed in the paper; At last, we
reproduce the cooling imperativeness use under different warm comfort expand. The results show that
cooling essentialness saving adequacy is solidly related with the tasteful solace range and city's
geographic region [6].

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
26
Fig. 5: Traditional Temperature Control Method and Thermal Comfort Control Method
According to use research gap, a methodology for recognizing human temperature from a situation
and anticipate an ideal temperature is proposed. For building up our examination, we have utilized
python programming with its default library, develop algorithm, Raspberry pi zero, Thermal sensors,
IR sensors, IR remote. We have built up an implanted system by utilizing these types of gear, which
works with object temperature and programmed on-off rely upon either object is exist in room or not.
At long last it deliver consequence of temperature depend of the criteria of number of object, object
value, environment temperature, object is in environment or not.
3. PROPOSED METHODOLOGY
The ideal value of the room temperature is considered as 18°C for the proposed research work. It is
noted that the World Health Organization recommends a minimum indoor temperature of 18ºC, with a
2-3ºC warmer minimum temperature for rooms occupied by sedentary elderly, young children and the
handicapped [6]. Below 16ºC temperature, there is an increased risk of respiratory diseases, while
below 12ºC temperature the risk is of increased problem of the cardiovascular strain [10]. The main
objective of this research is to set an ideal temperature that predicts temperature with automation of
the AC system so that one can avoid manually operating the room AC. A set of hardware components
is used here to develop the proposed embedded system. In Fig. 6 we shown our proposed model and
Fig. 7 represents the flowchart of the entire system.
Fig. 6: Process of the Proposed Model.

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
27
Fig. 7: Flowchart Diagram of the Entire System
Mathematical Exploration: To develop the algorithm of proposed entire embedded system, first
temperature data and object data are gathered by utilizing thermal sensors1 and sensor2 to set up an
anticipated temperature. The sensors play the role of distinguishable sorts of momentous worth.
Sensor 1 is utilized to detect the temperature upper than 18 (temp>18) and sensor 2 is utilized to
identify the temperature below (temp<18). Here, the ideal temperature is deliberated as 18°C [6].
Let,
Sensor1 = s1.
Sensor2 = s2.
Object = x.
Temperature = t.
No of object in environment = n.
Read the value from s1 (t.value>18 ).
Read the value from s2 (t.value<18).
S1.value = x1.t, x2.t, x3.t, x4.t, x5.t, x6.t,…,xn.t
S2.value = x1.t, x2.t, x3.t, x4.t, x5.t, x6.t,…,xn.t
∑s1= x1.t + x2.t + x3.t + x4.t + x5.t + x6.t+…+xn.t /n
∑s2= x1.t + x2.t + x3.t + x4.t + x5.t + x6.t+…+xn.t /n
Predict Temperature (PT) = (∑s1 + ∑s2) /2.
Compare Temperature (CT) = PT – 18°C.
#If CT value is equal or more then 20 (CT>=20)
PT = CT – 5.
#If CT value more then 10 and less than 20 (10<CT<20)
PT = CT – 2.
# If CT value more then 1 and less than 10 (1<CT<10)
PT = CT – 1.
# If CT value is equal or more then -5 (CT>= -5)
PT = CT + 3.
# If CT value is equal or more then -5 (-1 <CT< -5)
PT = CT + 1.
# If CT value is 0
PT = 18 °C
# After a couple of sec.
Repeat [process ()]

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
28
Description of used component: Building up the proposed embedded smart AC system, following
equipment segments are used. Here, the segments with their figures and portrayal are introduced.
Thermal Sensor: This sensor is an 8x8 exhibit of IR warm sensors from Panasonic. When it is
associated with the microcontroller (or Raspberry Pi) it will restore a variety of 64 singular infrared
temperature readings over I2C. It is as like as those of extravagant warm cameras, sufficiently
straightforward for simple reconciliation. This part will quantify temperatures going from 0°C to 80°C
(32°F to 176°F) with an exactness of ± 2.5°C (4.5°F). It can distinguish a human article present in the
room from a separation of up to 7 meters or 23 feet. Program codes are created for utilizing this
breakout on an Adriano or a perfect or on a Raspberry Pi with Python. On the Pi, with a touch of
picture handling help from the SciPy python library by which it can introduce the 8x8 networks and
get some truly decent outcomes. The AMG8833 is the up and coming age of 8x8 warm IR sensors
from Panasonic and offers higher execution than its forerunner the AMG8831 [18]. Fig. 8 is a warm
sensor that is utilized for the framework.
Fig. 8: Thermal Sensor
Raspberry pi zero: The Raspberry Pi is ease, MasterCard estimated PC that connects to a PC screen
or TV, and utilization a standard console and mouse. It is a competent little gadget that empowers
individuals of any age to investigate registering, and to figure out how to program in dialects like
Scratch and Python. The Raspberry Pi Zero is a large portion of the span of a Model A+, with double
the utility. A modest Raspberry Pi that is moderate enough for any task. It contains 1GHz single-
center CPU, 512MB RAM, Mini HDMI port, Micro USB OTG port, Micro USB control, HAT-
perfect 40-stick header, Composite video and reset headers, CSI camera connector [19]. Fig. 9 is an
image of Raspberry pi zero.
Fig. 9: Raspberry pi zero
IR Sensor: An infrared sensor is an electronic instrument that is used to identify certain
characteristics of its condition. It does this by either creating or recognizing infrared radiation.
Infrared sensors are in like manner prepared for assessing the glow being released by an inquiry and
perceiving development [20]. In Fig. 10 we demonstrated the IR sensors which pass the esteem and
get the incentive for the direction the air conditioner.

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
29
Fig. 10: Infered Sensor
IR Remote: Infrared remote control a handheld, a remote gadget used to work sound, video and other
electronic hardware inside a room utilizing light flags in the infrared (IR) run. Infrared light requires a
viewable pathway to its goal. Low-end remotes utilize just a single transmitter toward the finish of the
unit and must be pointed legitimately at the hardware. Excellent remotes have three or four ground-
breaking IR transmitters set at various edges to give the room signals [21]. In Fig. 11 we showed the
image of an IR remote.
Fig. 11: IR Remote
Microcontroller (NodeMCU V-3 Development Kit): The NodeMCU is an open-source firmware
and advancement pack that causes us to Prototype our IOT item inside a couple of Lua content lines.
Open-source, Interactive, Programmable, Low cost, Simple Smart, and WI-FI empowered. The
Development Kit dependent on ESP8266 incorporates GPIO, PWM, IIC, 1-Wire, and ADC across the
board. Power your improvement in the quickest manner mixes with NodeMCU Firmware! USB-TTL
included, plug and play, 10 GPIO, each GPIO can be PWM, I2C, 1-wire, FCC CERTIFIED WI-FI
module, PCB radio wire [22]. In Fig. 12 we demonstrated the image of the microcontroller.
Fig. 12: Microcontroller
By assembling all these segments, proposed embedded system has been structured and created. Fig.13
represents the proposed diagram of entire smart AC system.

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
30
Fig. 13: Diagram of the Entire Syst
4. RESULT AND ANALYSIS
The proposed strategy has been structured and created in default IDE and furthermore can develop
Anaconda application. A small conference room is selected for test the system, where 12 persons were
available in the room. They all are of different temperatures, and the temperatures are detected by two
thermal sensors (Sensor1, Sensor2). Then a data table is prepared by what exactly the sensor has
provided. The fracture values like 28.5 to 29, 28.3 to 28 are skipped in this case. Several experiments
have been observed to test the algorithm of the proposed system. But we have shown one experiment
there one of them. Compare these results of the experiments with another, the result of the proposed
system is found in the expected range.
Total 12 objects are measured for the temperatures that are presented in Table 1 and the
corresponding graphical representations are presented as line chart and are shown in Fig. 14.
TABLE1: Object with Temperature by Implemented Sensors.
Utilizing the estimation of the table distinguishes by the sensors and the output gets by utilizing of our
calculation. Table 1. Object Temperature and Predicted Temperature
No of Object
Object Temp In °C
Predict Temp
1
19
2
19
3
21
4
22
5
18
6
20
18°C
7
18
8
20
9
19
10
18
11
19
12
21
In Fig. 14 we showed the Line chart of the Tabular 3 data and output.

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
31
Fig. 14: Diagram of Table 1 Data
In several experiments, temperature data is found from objects by using sensors from a small
conference room and save it in a table and deploy our proposed algorithm1 on it, we have found a
predict temperature 20°C from experiment, then after 10 sec we again collected data and stored raw
data in the another table and deploy algorithm1 on it and get a predict temperature 19°C. If we
concentrate on the past predict temperature, we can easily judge that the temperature of object is
changing dynamically. As not unlike that experiment we again collect temperature from object and
store it to table and this one is last and final experiment. Deploy algorithm on the table that have
shown in table1 and get predict temperature 18°C. Generally what temperature predict by the system
using algorithm1 that is comfort temperature for all we can easily see it from fig.14. Every predict
temperature is near to close of comfort temperature 18°C.
5. CONCLUSION
Automation and temperature prediction of a room is very much demanding issues in the present
period for some reasons. A few methodologies are working in this purpose. A smart AC system room
by mechanization and temperature prediction has been designed, developed and implemented by the
python programming language. Several experiments have been performed to analyses the approach
and found at satisfactory dimension. It can be applied in any manual AC system room to convert it
into smart AC system environment by utilizing machine learning, artificial intelligence and expert
system.
REFERENCES
[1] S. Parnin and M. M. Rahman,” Human Location Detection System Using MicroElectromechanical
Sensor for Intelligent Fan,” International Conference on Mechanical, Automotive and Aerospace
Engineering 2016.
[2] K. L. Ku, J. S. Liaw, M. Y. Tsai, and T. S. Liu,” Automatic Control System for thermal Comfort
Based on Predicted Mean Vote and Energy Saving ” IEEE TRANSACTIONS ON AUTOMATION
SCIENCE AND ENGINEERING ,VOL . 12, NO.1,JANUARY2015.
[3] Muhammad Aftab, Chi-Kin Chau, and Peter Armstrong,” Smart Air-Conditioning Control by Wireless
Sensors: An Online Optimization Approach”, Masdar Institute of Science and Technology Abu Dhabi,
UAE.
[4] D. V. Babu, D. K. Yadav, N. K. Yadav, and V. K. Verma 2015 MEMS Based Smart & Secure Home
Automation System with Multi-Way Control & Monitoring Facility using Smart Phone J. Chem.
Pharm. Sci. 9357–359
[5] Aswani, N. Master, J. Taneja, D. Culler, and C. Tomlin. Reducing transient and steady state electricity
consumption in hvac using learning-based model-predictive control. Proceedings of the IEEE
100(1):240 –253, jan. 2012.
[6] J. Y. H. Wang, Y. K. Chuah, S. W. Chou, and T. H. Lo, “Noninvasive Zigbee wireless controller for
air conditioner energy saving,” in Proc.IEEE 7th Conf. WiCOM, Wuhan, 2011, pp. 1– 4.

J. Bangladesh Electron. 18 (1–3); 01–05, 2018
32
[7] Nigel P. Isaacs, “Poverty and Comfort”, Presented at the Fourth National Food Bank Conference,
Wellington, November 13, 1998
[8] M.M.Rahman and F. Radzi, 2015 Sensory System to locate Human for Smart Fan. 1st International
Workshop on Mechatronics Education
[9] S. Kaur and H. C. Vashist 2013 Automation of Wheel Chair Using MEMS 3 227–232.
[10] T. Ayesha. S. Smitha, and M, Rumana, 2015 Hand Gesture Based Home Automation for Visually
Challenged Int. J. Innov. Eng. Research Technol. 21–7
[11] N. Pranathi and S. M. Ahmed, 2013 Tri-Axis Motion Detection using MEMS for Unwired Mouse
Navigation System in the Future Generation Machines Int. J. Adv. Res. Comput. Commun.
Eng.23672–3675.
[12] B. Sun, P. B. Luh, Q. S. Jia, Z. Jiang, F. Wang, and C. Song, “Building energy management:
Integrated control of active and passive heating, cooling, lighting, shading, and ventilation systems,”
IEEE Trans. Autom. Sci. Eng., vol. 10, no. 3, pp. 588–602, Jul. 2013.
[13] Collins, K.J 1993 Cold- and heat-related illness in the indoor environment in Burridge R & Ormandy
D (eds) Unhealthy Housing: Research Remedies and Reform London: E & FN Spon.
[14] W. Kong, T. Chai, J. Ding, and S. Yang, “Multifurnace optimization in electric smelting plants via load
scheduling and control,” IEEE Trans. Autom. Sci. Eng., vol. 11, no. 3, pp. 850– 862, Jul. 2014.
[15] J. S. R. Jang, “ANFIS: Adaptive-network-based fuzzy inference system,”IEEE Trans. Syst, Man,
Cybern., vol. 23, no. 3, pp. 665–685,May/Jun. 1993.
[16] V. Bobál, J. Böhm, and R. Prokop, “Practical aspects of selftuning controllers,” Int. J. Adapt. Control
Signal Process., vol. 13, pp.671–690, 1999.
[17] S. Senthilkumar and R. Vinothraj 2012 Design and study of ultrasound-based automatic patient
movement monitoring device for quantifying the infraction motion during teletheraphy treatment.
Journal of Applied Clinical Medical Physics136 .
[18] Thermal camera, https://learn.adafruit.com/adafruit-amg88338x8-thermal-camera-sensor/overview,
28.09.2018
[19] Raspberry pi zero, https://learn.adafruit.com/introducing-theraspberry-pi-zero?view=all, 28.09.2018
[20] IR sensor, https://learn.adafruit.com/ir-sensor?view=all, 28.09.2018
[21] IR remote, https://learn.adafruit.com/using-an-ir-remote-with-araspberry-pi-media-center?view=all,
28.09.2018
[22] Microcontroller,https://learn.adafruit.com/programmingmicrocontrollers-using-openocd-on-raspberry-
pi?view=all, 28.09.2018