ZigBee-Based UV-IR Flame Detection for Security System

M.Naveen [1], M.T.S.Nivash [2], S.Venkatesh Prasath [3], Mr.P.Aravind M.E (Asst Prof.) [4] Department of ECE, Kamaraj college of Engineering and Technology, Virudhunagar, Tamil Nadu, India.
reachnaveenece@gmail.com[1], nivash.nivash.mts@gmail.com[2], s.venki21@gmail.com[3], mcetarv@gmail.com[4] AbstractThis paper describes a ZigBee based ultraviolet/infrared detection of flame, security & surveillance systems. The sensor node is composed of a current-voltage convertor circuit with an UV filter, IR and PIR sensor, PIC microcontroller, LCD module & a ZigBee transceiver. This transceiver module is operated at a 2.45GHz industrial, scientific, and medical band. A passive phototransistor is designed to convert the UV emission of flame into mill amperes. IR, PIR and temperature sensor is used to ensure the presence of flame and for security purpose The sensor node consumes only an average of 2.3 mW from a 3.3-V power supply. Including mixed signal processing and ZigBee transmission, in order to provide wireless communication within nodes and centralised system. Keywords Flame analysis, ultraviolet (UV) detector, Wireless sensor network (WSN), Zig-Bee.

INTRODUCTION FIRE is considered as one of the serious threats in our daily life with fast spreading speed and totality of property destruction. Low cost, reliable, and wide coverage fire alarm systems are indispensable in industry to protect the equipment and assets. The most commonly used fire detector in the fire safety sector is the smoke detector even if they always have false alarms. Some estimates are as high as 11 to 1 for the ratios of false to actual alarms. Their applications are also limited in a confined area as the smoke concentration may be diluted in a large monitoring area, which delays or even worst desensitizes the alarm triggering. Being able to compensate the aforementioned disadvantages of smoke detectors, optical flame detectors offer higher reliability, good long-term stability and prompt response to accidents. Flame detectors that use optical sensors working at specific spectral ranges to record the incoming electromagnetic emission at the selected wavelengths include IR [1][3], visible [2] and ultraviolet (UV) [1], [2], [4][6]. UV-only flame detectors work with wavelengths shorter than 400 nm. They detect flames at a high speed of 34 ms due to the UV high-energy radiation emitted by fires and explosion at the instant of their ignition. UV/IR detectors compare the threshold signal in two spectral ranges and their ratio to each other to confirm the reliability of the fire signal, minimizing false alarms. However, available UV flame detectors [7] are based on the GeigerMuller counter, a quartz tube filled with an inert gas that conducts electricity when a photon of wavelength between 185 and 260 nm of a flame temporarily makes the gas conductive. This type of flame detector is indeed bulky and expensive, operates at high voltage, has short lifetime, and optically interferes with others in close proximity. Therefore, the low-voltage and compact semiconductor UV sensor is always the welcoming device.

Aside from the flame detection, the installation of fire alarm systems may be constrained. For example, in ancient and historic buildings, the construction works of the cable layout for wired fire alarm systems may damage the building architecture and decoration, which are usually unrecoverable. A wireless sensor network (WSN) with flame detectors as WSN nodes [8]-[10] can alleviate this problem with no wiring network installation. Furthermore, its ad hoc infrastructure offers high flexibility in sensor placement and needs no particular human control and intervention [11]-[12]. The sensor nodes are self organized to gather in real time the environmental data and detect abnormalities in the monitoring region. A large number of sensor nodes can also be organized in piconet so as to extend the coverage area. In a wireless sensing network for fire safety, there is a tacit assumption of a tiny WSN node with low power consumption. In this context, this paper reports a WSN node for fire safety, and we first characterize the phototransistor to detect the flame in UV range using colour filter. Finally, the aforementioned UV photo detector is integrated with IR detector and a ZigBee transceiver so as to implement a compact WSN node for fire safety. I. FLAME CHARACTERISTICS A Flame is the visible, gaseous part of a fire. It is caused by a highly exothermic reaction taking place in a thin zone. If a fire is hot enough to ionize the gaseous components, it can become plasma. Colour and temperature of a flame are dependent on the type of fuel involved in the combustion, for example when a lighter is held to a candle. The applied heat causes the fuel molecules in the candle wax to vaporize. In this state they can then readily react with oxygen in the air, which gives off enough heat in the subsequent exothermic reaction to vaporize yet more fuel, thus sustaining a consistent flame. The high temperature of the flame causes the vaporized fuel molecules to decompose, forming various incomplete combustion products and free radicals, and these products then react with each other and with the oxidizer involved in the reaction. Flame colour depends on several factors, the most important typically being blackbody radiation and spectral band emission, with both spectral line emission and spectral line absorption playing smaller roles. In the most common type of flame, hydrocarbon flames, the most important factor determining colour is oxygen supply and the extent of fuel-oxygen pre-mixing, which determines the rate of combustion and thus the temperature and reaction paths, thereby producing different color hues. Flame temperatures of common items include a blow torch which can burn usually up to around 1,600 C (2,900 F), a candle at 1,400 C (2,600 F),a propane torch at 1,995 C (3,620 F), or a much hotter oxyacetylene combustion at 3,000 C (5,400 F).

IV. SENSOR NODE

II. UV DETECTION
The optical flame detector is a detector that uses optical sensors to detect flames. As shown in the figure 1 there are several types to detect the flame namely ultraviolet, infrared, ultraviolet/infrared, infrared/infrared, and infrared/infrared/infrared.

The sensor node consists of phototransistor (BPW77N), low power dual operational amplifier (LM358), NPN transistor (BC547). A. BPW77N is a very high sensitive silicon NPN epitaxial planar phototransistor in a standard TO18 hermetically sealed metal case. A base terminal is available to enable biasing and sensitivity control. It has high photosensitivity. The turn-on time and turn-off time of BPW77N is 6 microseconds and 5 microseconds respectively. The cut-off frequency of the phototransistor is 110 KHz. B. The BC547 is a NPN transistor which is used to amplify the small output current and it can operate high current device. It is also used to convert current to voltage. It has the following characteristics: Ic(max)=100mA,Vce(max)=45V,hfe(min)=200,P(max)=5 00mW. C. LM358 is used for linear operation of the sensor (ie.)it must not reach within 3V of supply. The output range is similarly limited to within 3V to 5V and to eliminate the negative supply its operating temperature is -40 deg to +85 deg C. In our circuit design, R1=1K Ohm connected to the collector terminal of the phototransistor, R2=10K ohm connected to the collector of NPN transistor, R3=1K ohm, R4=1K ohm and capacitor 1 mega farad is used to reduce the noise present in the circuit (figure 3). D. IR323 sensor has a peak wavelength of 940 nm.It has low forward current (20 mA). It made of GaAlAs material with blue colour lens. The voltage follower is used which follows the voltage that is sent in. It has the ability to buffer a high impedance signal as shown in figure 2. The output directly back to the negative (inverting) input, it sends out the exact same thing that came in.

Fig: 1 Infrared flame detectors work within the infrared spectral band. The disadvantages of infrared are subjected to false alarms in the presence flickering infrared sources and poor detection performance for stable flames. The Dual Infrared detectors compare the threshold signal in two infrared ranges. One sensor looks at the 4.4 micrometer range and other sensor at a reference frequency. It has the disadvantage that it affected by infrared sources. The ultraviolet detectors work with the wavelength shorter than 400 nm. These detectors detect fire and explosions within 3 to 4 milliseconds. The temperature dependent emissions are strongest in the early stages of combustion. There are several advantages in ultraviolet detectors they are unaffected by solar radiation, unaffected by hot objects, quick response and rapid sensitivity. The ultraviolet radiation is invisible to the eye and it is non-ionizing form of radiation in the 100 nm to 400 nm wavelength region of electromagnetic spectrum. III. IR DETECTION Infrared (IR) light is electromagnetic radiation with a wavelength longer than that of visible light, measured from the nominal edge of visible red light at 0.74 micrometers (m), and extending conventionally to 300 m. These wavelengths correspond to a frequency range of approximately 1 to 400 THz, and include most of the thermal radiation emitted by objects near room temperature. Microscopically, IR light is typically emitted or absorbed by molecules when they change their rotational-vibrational movements. Sunlight at zenith provides an irradiance of just over 1 kilowatt per square meter at sea level. Of this energy, 527 watts is infrared radiation, 445 watts is visible light, and 32 watts is ultraviolet radiation. The balance between absorbed and emitted infrared radiation has a critical effect on the Earth's climate.

Fig: 2 E. The PIR (Passive Infra-Red) Sensor is a pyroelectric device that detects motion by measuring changes in the

infrared levels emitted by surrounding objects. This motion can be detected by checking for a high Signal on a single I/O pin. It has a features of single bit output and can work with 3.3V & 5V operation with <100 microampere current draw. The device contains a special filter called a Fresnel lens, which focuses the infrared signals onto the element. The PIR Sensor has a sensitivity range of approximately 20 feet (Refer figure 2).

the peer-to-peer topology is selected instead of the point-to-multipoint topology. The decentralized nature of Peer to Peer networks increases robustness because it removes the single point of failure that can be inherent in a client-server based system. For fire safety, as shown in Fig: 1.a WSN configured in an ad hoc infrastructure is designed for the UV detection of flame. It includes the WSN node and a control center. The low-power consumption short-range communication technology ZigBee is used, and it has 16 channels of data rate 250 kb/s in the license free industrial, scientific, and medical band of 2.42.4835 GHz.

Fig: 3 PIR configurations

F. The LM35 are precision integrated-circuit temperature sensors, whose output voltage is linearly proportional to the Celsius (Centigrade) temperature. The LM35s low output impedance (0.1ohm for 1 mA load), linear output (+ 10.0 mV/C scale factor), and precise inherent calibration make interfacing to readout or control circuitry especially easy. The value of resistor Ra range from 80K ohm to 600K ohm. The general equation used to convert output voltage to temperature is: Temperature ( oC) = Vout * (100 oC/V), so if Vout is 1V, then, Temperature = 100 oC.

Fig: 4 VI. BLOCK DIAGRAM The Fig 4.a shows the UV sensor consists of photo-transistor and a UV filter. The output of the UV sensor is of few milliamperes. Hence current is converted to voltage and amplified. The analog output of the voltage is given as the input for the PIC microcontroller. The in-built A-D converter used for the conversion and the output digital voltage is given to the LCD and to the ZigBee transceiver. At the output we obtain the result from HyperTerminal or by using Visual Basic we can set a threshold value to make a warning in the PC.

Fig: 3. UV sensor Circuit G. Programmable Intelligent Computer (PIC 16F877A) is a microcontroller used to convert analog to digital and used for interfacing the devices. It is High Performance RISC CPU. All single cycle instructions except for program branches, which are two-cycle. Operating speed: DC - 20 MHz clock input DC - 200 ns instruction cycle. The PIC16F877A has an 8192 (8k) 14bit instruction program memory. H. LCD-016M002B is used to display the output which is interfaced to the PIC microcontroller. It is a 16 x 2 Character LCD which works with + 5V power supply and with 1/16 duty cycle. It has 8 data bus lines. V. NETWORK CONFIGURATION In a wireless sensing network for fire safety, lots of nodes can be connected. So, that fire can be detected anywhere in an infrastructure. In order to increase the coverage of the flame detection,

Fig: 4.a VII. RESULTS AND MEASUREMENTS FOR UV ALONE

The results obtained by the project are shown as per the categories below Line Of Sight (at 00): 1: When flame not detected: DISTANCE (cm) 5 10 25 50 75 100 125 150 FLAME (o/p in LCD) 0 1 2 2 3 3 4 4

2: When flame is detected:

ANGLE DIST (cm) 5 10 25 50 100 140 5 10 25 50 75 100 125 150 5 25 50 75 100 5 25 50 75 5 25 DEG (clockwise & anticlockwise) 10 10 10 10 10 10 30 30 30 30 30 30 30 30 50 50 50 50 50 70 70 70 70 80 80 FLAME 0 0 2 7 5 5 0 0 2 4 6 5 6 6 0 3 4 6 8 0 3 4 6 0 4 3: When any interruption is detected:

4: When any interruption is not detected:

5: Temperature sensor reading:

S. No 1 2 3 4 5 6 7 8

Distance (feet)

LIGHTER IR 7 88 103 107 115 110 115 115 103 UV 9 7 9 11 14 12 27 56 35

CANDLE IR 35 95 110 111 113 115 120 125 100 UV 10 8 10 12 22 45 80 90 25

0.5 1 1.5 2 2.5 3 3.5 3 AVERAGE

6: Result obtained using hyper terminal: S.No Distance (feet) CAMPHOR IR 0 2 7 30 50 88 90 110 120 120 60 FIRE UV 5 6 7 9 7 10 10 12 14 12 10

VIII.RESULTS USING UV & IR The results obtained by the project are shown as per the categories below LIGHT SOURCE

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 10 10 12 AVERAGE

ANGLE (clockwise & anti-clockwise) S.No 1 2 3 Degree LIGHTER IR 110 125 110 115 UV 20 82 32 44

20 40 70 AVERAGE Degree 20 40 70 AVERAGE

i)

LIGHTER

S.No 1 2

CAMPHOR FIRE IR 65 60 70 65 UV 8 22 35 20

ii) CANDLE

THERSHOLD THERSHOLD S.No iii) CAMPHOR Line Of Sight (at 00): 1 2 Degree 0 20 IR 88 86 UV 23 14

3 4

40 70 AVERAGE

92 90 90

52 33 30

threshold for fire detection. Also it has been proven that this type of surveillance systems can be achieved up to this amount of low power consumption model.

i)

Reading at room light & temperature X.REFERENCES

[1]D. Starikov, C. Boney, R. Pillai, and A. Bensaoula, Dualband UV/IR optical sensors for fire and flame detection and target recognition , in Proc. ISA Sensors Ind. Conf., 2004, pp. 3640. [2] L. Xu and Y. Yan, A new flame monitor with triple photovoltaic cells, IEEE Trans. Instrum. Meas., vol. 55, no. 4, pp. 14161421, Aug. 2006. [3] B. U. Toreyin, E. B. Soyer, O. Urfalioglu, and A. E. Cetin, Flame detection using PIR sensors, in Proc. 16th IEEE Signal Process., Commun. Appl. Conf., 2008, pp. 14. [4] A. R. Pauchard, D. Manic, A. Flanagan, P. A. Besse, and R. S. Popovic,A method for spark rejection in ultraviolet flame detectors, IEEE Trans. Ind. Electron., vol. 47, no. 1, pp. 168 174, Feb. 2000. [5] Z. Djuric, K. Radulovic, N. Trbojevic, and . Lazic, Silicon resonant cavity enhanced UV flame detector, in Proc. 23rd Int. Conf. Microelectron., 2002, pp. 239242. [6] R. C. Luo and K. L. Su, Autonomous fire-detection system using adaptive sensory fusion for intelligent security robot, IEEE/ASME Trans. Mechatronics, vol. 12, no. 3, pp. 274281, Jun. 2007. [7] Hamamatsu Photonics K.K., Electron Tube Division, 2010 Photonic Devices: Electron Tube Devices and Applied Products, p. 24, Iwata City, Japan, 2001. [8] L. Ferrigno, A. Pierosanto, and V. Paciello, Low-cost visual sensor node for bluetooth-based measurement networks, IEEE Trans. Instrum.Meas., vol. 55, no. 2, pp. 521 527, Apr. 2006. [9] G. Boggia, P. Camarda, L. A. Grieco, and M. R. Palattella, Fire detection using wireless sensor networks: An approach based on statistical data modeling, in Proc. New Technol. Mobility Security, 2008, pp. 15. [10] S. Liu, D. Tu, and Y. Zhang, Multiparameter fire detection based on wireless sensor network, in Proc. Int. Intell. Comput. Intell. Syst. Conf., 2009, pp. 203206. [11] X.-H. Cao, J.-M. Chen, Y. Xiao, and Y.-X. Sun, Buildingenvironment control with wireless sensor and actuator networks: Centralized versus distributed, IEEE Trans. Ind. Electron., vol. 57, no. 11, pp. 35963605, Nov. 2010. [12] V. Gautam and A. K. Gupta, Spectroscopic analysis of fuel lean flames for propulsion applications, in Proc. ASME Power, 2004, pp. 495504.

ii) Readings when flame detected

IX.CONCLUSION This paper has presented a low-cost and lowpower ZigBee-based WSN node for the detection of flame, contributing to the fire safety protection. This wireless flame detector infrastructure offers high flexibility in sensor placement and needs no human control and intervention. This paper gives a complete solution for the fire safety and surveillance systems for the society. In this paper the false alarms has also be reduced because of using the UV & IR equivalent voltage as a cut-off