Lifestyle

4th International Conference on Computers and Management (ICCM) 2018 Detection of Forest Fires at Early Stages using Wireless Sensor Network and Graphical User Interfacing

Description
4th International Conference on Computers and Management (ICCM) 2018 Detection of Forest Fires at Early Stages using Wireless Sensor Network and Graphical User Interfacing
Categories
Published
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  4th International Conference on Computers and Management (ICCM) 2018   CHANDAN ET. AL. 1 Detection of Forest Fires at Early Stages using Wireless Sensor Network and Graphical User Interfacing “ Chandan a , Jaskaran Singh b , Shivani Sharma a , Kailash Chand a , Pankaj Mukhija c , Paramita Guha a,* , K. Chathuranga d , S. Lalitharathne d , A.L. Kulasekera d , R. Gopura d ”   a CSIR-Central Scientific Instruments Organization, Delhi Centre b Guru Tegh Bhahadur Institute of Technology, Delhi c  National Insitute of Technology, Delhi d   Mortuwa University, Sri Lanka Abstract:  Forest fires are a menace to the environment and disturb the ecological balance in the environment. There is a need to detect forest fire in early stages, so as to take appropriate preventive measures. Wireless Sensor Networks have proved to be useful for the early detection of Forest fires. There are many methods proposed for early detection of forest fires, from which monitoring the environmental conditions and gasses have proven to be most useful. In this paper, we discuss the early developments of one such wireless sensor network systems for early and efficient detection of forest fire. 1.   Introduction Fire is characterized as fast oxidation of a substance, thereby producing gasses and chemical productions, which can be read by sensors to provide an insight about the type and place of the fire. Forest fires prove to be a great menace to ecologically healthy grown forests and protection of the environment causing the loss of several acres of forest areas. This has  been a matter of interest for researchers around the world and many studies have been conducted in the path of efficient detection of forest fire (Ahmad, 2014; Antonio, 2016; Ahmad, 2013).This raises the need for early detection of forest fires and early steps to extinguish the forest fires in early stages (Krüll, 2012). The detection systems for forest fires are divided into three categories: ground based, Ariel based and space bound  based. Many works have been carried out for space bound based detection of forest fires (Aslan, 2010) but these methods owe their own disadvantages. Ariel based fire detection techniques have been discussed in (Krüll, 2012). As a part of ground based forest fire detection systems, Wireless Sensor Network based systems have been proven to be more effective in detecting in forest fire in early stages. Wireless Sensor Network (WSN) is a system consisting of several small subsystems (referred as sensor node), which collect the environment data  based on the sensors incorporated in the sensor node , and transfer this data wirelessly to the computer system for effective processing of the collected data( Murat, 2015; Pokhrel, 2018). Our approach is also based on the WSN paradigm has been designed and developed in the context of a research project that included all the key actors in the creation of a forest fire (Chowdhary, 2018). The important design and functional requirements of the WSN for Early Forest Fire detection included contribution to the forest fire fighting strategy,  providing critical data to improve their safety and efficiency; providing of valuable forest and climate parameters input to the fire propagation models (Ameh, 2018). This included temperature, humidity, rainfall, wind, smoke, and solar radiation; early detection of forest fires, alarm management and real time reporting about the fire evolution (Liu, 2018). They provide alarms by integrating information from several nodes; following of a “deploy and forget” policy. The Sensor nodes will not need any kind of maintenance or battery recharging. Energy harvesting techniques and an optimized low power consumption will ensure uninterrupted functioning (Antonio, 2016; Kechar, 2013; Vivek, 2014). This paper describes the development of a sensor node for early forest fire detection by monitoring the environmental conditions simultaneously with the timeline of initial stages of forest fire to full flaming fire. This work is  proposed to be carried out in a number of stages with system improvements’ taking place at each stage. In this paper, we discuss the initial works and research carried out towards the development of the system with a limited number of sensor nodes and their testing in a controlled environment. Fig. 1  –   WSN Overview.  4th International Conference on Computers and Management (ICCM) 2018   CHANDAN ET. AL. 2 The structure of paper is described as follows; Section 2 gives a brief idea about the previous works carried out regarding the development of sensor nodes for early forest fire detection. Section 3 explains the proposed work and methodology by the author to carry out the development of the sensor network. Section 4 discusses the results obtained by real time simulation of the network, while the possibilities of future improvements have been discussed in Section 5. 2.   Related Work Lot of researchers has worked previously for development of forest fire detection system (Murat, 2015) and a wireless sensor node for this  purpose (Vijaylakshmi, 2016). A brief background about the forest fire and the decision making steps for fire detection has been provided in (Bayo, 2010; Sarde, 2018; Deve, 2016). The systems developed commonly employed the use of sensors such as temperature /humidity sensor, smoke sensor and even thermal camera (Bosch, 2013) for detection of forest fire. For efficient data transfer from the sensor node to the base station, many wireless platforms are available, which are discussed in (Islam, 2013; Suneel, 2011), as a result it has been concluded that Zigbee is efficient platform for data transfer as used in (Suneel, 2011; Wang, 2010; Arnoldo, 2012). In (Guozhu, 2010), General Packet Radio Services  ( GPRS) has also been used along with Zigbee for transfer of data from central node to the base station; the advantages of GPRS for this task have also been discussed. It has also been proposed in (Bahrepour, 2009), regarding the use of Artificial Intelligence (AI) techniques for efficient detection of forest fire. Apart from this various algorithms for efficient data interpretation from the sensors have been employed in (Arnolodo, 2012; Krull, 2012). Algorithms have not only been developed for data interpretation but also for efficient node positioning in (Jiawen, 2016). .Any wireless data transfer system faces security issues which need to be addressed; these issues have been addressed in detail in (Harish, 2017). 3.   Proposed Work and Methodology For the development of the sensor node, the general methodology of typical sensor node was studied and observed for crucial components as it is intended to develop the sensor node in economical and efficient method. It has been commonly observed that any typical sensor node consists of a processing unit to control functioning of the node. This node activates the sensor node when measurement is required, controls the data transfer among the various sensor nodes and the central node, applies data transfer and energy conservation protocols within the sensor node. It has been observed that during the whole series of events of a forest fire, various gasses are emitted in the pyrolysis phase, flaming phase and cool down phase as shown in Table 2. The increased or decreased quantity of any of these gasses during a forest fire depends on the fuel matter present in the forest in the occurrence of forest fire. Table 1  –  Gasses emitted during various phases of forest fire. Stages of Forest Fire Gasses Emitted Pyrolysis/ Smoldering Phase Carbon Monoxide (CO) Flaming Phase Carbon Dioxide (CO 2 ) Cool Down Phase Methane (CH 4 ) and Smoke For each gas to be observed in order to successfully detect a forest fire in its early stages, a deep literature review of the possible sensors available and their ranges were carried out. The detailed methodology followed for the development of the sensor node and sensor node testing is explained as under.  3.1.    Selection of Sensors As it was evident from the literature survey carried out that we need to monitor the existence and concentration of selected gasses to successfully  predict the possibility of a forest fire and detection of forest fire as well. The gasses needed to be monitored for this research development purpose included Carbon Monoxide (CO), Carbon Dioxide (CO 2 ) , Methane (CH 4 ) and Smoke. After the vigorous sensor search in the research papers and internet sources, some sensors for each gas was selected for screening to select the sensors to be used for the development of the sensor node. As in the case of Temperature and Humidity Sensor, three sensors were selected after the research study which included HYT221 by Innovative Sensor Technology, SHT15 by Sensiron and DHT11 by Digi-key. These three sensors were simulated together to check their accuracy and resolution when compared to a highly accurate temperature and humidity sensor. The results of this simulation have been discussed in the late section.  3.2.   Development of Sensor Nodes Early development of the sensor node started with the deep study of each sensors selected and its behaviors. Most important part of any sensor node is the wireless data transfer. Owing to the numerous advantages of Zigbee  protocol among the various available data transfer protocols, Zigbee S2C module was chosen to be the transceiver for the sensor node. The Zigbee S2S module contains an inbuilt 10 bit Analog to Digital Convertor (ADC), because of which the early development of sensors nodes was carried out in the absence of any processor or controller for the node. As the development of the node was carried out, it was observed that many sensors need different type of configuration and communication protocols for efficiently giving the output. Due to this factor, Arduino Uno was later added in each sensor node for efficient interfacing of the sensors. Zigbee S2C is compatible with Arduino Uno board because of which the decision of using Zigbee S2C module for data transfer was retained.  4th International Conference on Computers and Management (ICCM) 2018   CHANDAN ET. AL. 3  3.3.   Development of Simulation Chamber After the sensor node was prepared, the basic experiments were needed to  be carried out. It was required to create a forest like environment in the laboratory. Owing to this requirement, a chamber was created using waste cardboard. The purpose of creating this chamber is to concentrate the gasses emitted by the fire towards the sensors, so as to check the behavior of these sensors and the ability of sensor node to efficiently detect a fire in its early stages.  3.4.   Experimental Setup The experimental setup to test the system prepared was created in laboratory as shown in Figure 2. Some leaves from the nearest forest were collected for burning. The purpose of using the leaves from the forest is to create a forest fire like environment for proper simulation of the sensor node created. The leaves were burnt with an external fire, and the  behavior of the system was observed. The results obtained for this setup and its outcomes are discussed in the next section. Fig. 2  –   Experimental Setup.  3.5.   Development of Python based Graphical User Interface (GUI)  for data display and alert generation The final phase of the proposed project was the need for display of the real time environment data at the coordinator node (base station) in the form of a GUI for alert generation. For this a GUI has been developed in Python based software commonly called as PyCharm. This software allows the development of application based GUI in an easy and sophisticated manner. PyCharm has inbuilt Python interpreter. The environmental parameters data collected by the sensor node will be fed to this GUI which has thresholds stored of each parameter. This GUI decides the creation of alert by comparing the environmental data with predefined thresholds. If the parameters exceed the threshold values, the alert is generated, otherwise not. The GUI created is one of the important parts of the development of the system as it allows the user to visualize the environment of the forest, and can predict the chances of forest fire. The created GUI picks up the sensor data and processes it internally by comparing the output from each sensor with a predefined threshold limit. This threshold limit can be altered with respect to the type of environment in which the system has to be installed. The created first prompts the user for providing the identification, which when verified, shows the data received from the sensor node. The fire alert is produced using visual modifications in GUI and a buzzer sound embedded in GUI software. Fig. 3  –   User authentication provided in GUI. 4.   Results and Discussion The earlier results observed from the continuous monitoring of the temperature sensors selected included the decision of incorporating which sensor into the sensor node. This experimental setup consisted of the three sensors, HYT221 by Innovative Sensor Technology, SHT15 by Sensiron and DHT11 by Digi-key, an external Hygro-thermometer clock by Extech and a highly accurate and precise Hygrometer 1620 by FLUKE. The outputs of all these sensors were monitored simultaneously to check the accuracy of these sensors for inclusion in the development of the node. The results from this monitoring are shown in Figure 5 and 6. From the results it was evident that SHT15 showed output much closer to the Hygrometer 1620 by Fluke both in terms of temperature and humidity, so SHT15 was considered to be the best choice for sensing temperature  4th International Conference on Computers and Management (ICCM) 2018   CHANDAN ET. AL. 4 and humidity in the sensor node. In the second part of experiment analysis, the sensor node was tested in the simulation chamber and artificial environment of forest fire was created by setting up fire from the dried leaves, small dried tree branches and paper bits. Here the parameters monitored included Temperature, Humidity, Smoke, Methane, and Carbon Monoxide as shown in Figure 7. Fig. 4  –   Front View of GUI created in Python. The obtained sensor readings were plotted in an excel file for future interpretations. It was observed from the readings that as the fire build up the temperature rises and humidity falls in the initial stages of the fire. As the fire builds up, the levels of smoke and carbon monoxide rises along with small level rise in methane. During the flaming fire, the temperature continues to rise while humidity continues to fall, and smoke and carbon monoxide rises continuously exponentially. As the fire begins to die out, the temperature falls slowly and humidity rises slowly, while smoke and carbon monoxide falls rapidly and methane levels falls slowly. From the above data collected (Figure 7), it was observed that there are some thresholds for every parameter to correctly detect forest fire in early stages. These thresholds are then fed to the developed GUI for data display and alert generation. From the observed data, we have selected the threshold for every parameter which acts as deciding factor , wether an alert should be generated or not. The thresholds selected for each  parameter are discussed in Table 2. Table 2  –   Thresholds selected for various parameters to be monitored. Parameter Threshold Temperature > 40 o C Humidity < 30 % RH Smoke > 1 PPM Methane > 2.5 PPM Carbon Monoxide (CO) > 1 PPM Fig. 5  –   Temperature data from Continuous Simulation of Temperate/Humidity Sensors. Fig. 6  –   Humidity data from Continuous Simulation of Temperate/Humidity Sensors.  4th International Conference on Computers and Management (ICCM) 2018   CHANDAN ET. AL. 5 Fig. 7  –   The collected data of the sensors from the experimental setup for Temperature and Humidity. Fig. 8  –   The collected data of the sensors from the experimental setup for Smoke, Methane, and Carbon Monoxide. The GUI then catches the collected data and shows alert if there is a  probability of f  ire in the forest in terms of two states namely, “Fire” and “No Fire” as shown in Figure 9 and 10.   Fig.9  –   GUI displaying the collected data of the sensors and displaying the state of the system as ‘FIRE’.  Fig. 10  –   GUI displaying the collected data of the sensors and displaying the state of the system as ‘NO   FIRE’.   5.   Future Improvements and Conclusion Forest fires are a very serious problem in many countries, and global warming may contribute to make this problem worse. Experts agree that, in order to prevent these tragedies from happening, it is necessary to invest in new technologies and equipment that enable a multifaceted
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x