BRIDGE SAFETY MONITORING USING SVM

Transcription

1 International Journal of Electrical Engineering & Technology (IJEET) Volume 9, Issue 5, September- October 2018, pp , Article ID: IJEET_09_05_005 Available online at ISSN Print: and ISSN Online: Journal Impact Factor (2016): (Calculated by GISI) IAEME Publication BRIDGE SAFETY MONITORING USING SVM Poonam T Sathe VLSI & Embedded, Electronics and Telecommunication, MITCOE Pune, India Dr. Virendra V Shete VLSI & Embedded, Electronics and Telecommunication, MITCOE Pune, India ABSTRACT Bridges are a key infrastructure particularly in the worlds cities and towns where either waterways or land barriers prevent transport to the other side. It also impacts life quality and safety of the civilian. In spite of bridge s own weight, structural elements of the bridge are also significantly affected by disaster and environmental conditions. Thus it is decisive to have proper inspection and detection during both regular and unsafe incident. A Bridge safety monitoring system using SVM described in this paper in this system various sensors are used for sensing the various parameters of bridge and environmental signals. This signals processed by the SVM to provide accurate result i.e bridge is safe or not for travelling, to the end user through the Thingtweet application. This system works 24/7 and it gives the alert signal to the authority center in case of any emergency. Keywords: Bridge safety monitoring system, IoT, WSN, Sensors, SVM, Thingtweet. Cite this Article: Poonam T Sathe and. Dr. Virendra V Shete, Bridge Safety Monitoring Using SVM: International Journal of Electrical Engineering & Technology, 9(5), 2018, pp INTRODUCTION A structure constructed to cross physical obstacles such as river,valley orroad without closing the route below is called as bridge[1]. The damage of bridges can cause the severe cost of human life and property[2]. Severe damages to the bridges cause the complete closure of the route[3]. the structural members of the bridge are also affected by live load and the impacts generated by environment and disaster such as flood, wind and earthquake[4]. Other parameters like bridge age, heavy weight vehicles and structural deficiencies also plays crucial role in it, because of this natural disasters, and other parameters like bridge age and structural deficienecies lots of bridge collapses happened in India and in foreign also. So to protect human lives there is need to have more reliable and efficient bridge safety monitoring system[5]. The appropriate work condition of bridge structure can not be obtained in time because of the earlier bridge test method which is manual method, with which it is very difficult to make an objective assessment for the basic safety of the bridge structure. If we want to know the operation state of bridge correctly, the real-time monitoring of each work 42 editor@iaeme.com

2 Bridge Safety Monitoring Using SVM factors (includes temperature, humidity, air speed, water level, vehicle load) must be realized first, and the operation performance of bridge should be evaluated by using of bridge evaluation expert system[6]. There are few drawbacks in previous methods and to overcome this drawbacks a method which includes WSN & IoT network is presented in this paper. IoT is defined as a network of inter-connected devices, which is accessed through Internet. Wireless sensors networks (WSN) are considered as eyes and ears of IoT, WSN is the link, which is able to connect real worlds to digital objects Traditional methods There are a lot of traditional methods for bridge safety monitoring [7]. But bridge safety monitoring system using traditional methods, is labor intensive and time consuming. [8]. Experts are expected to make physical visits periodically, set up the equipment, collect relevant data, perform off-line analysis and inform authorities accordingly [9]. This conventional approach cannot meet the requirements of modern infrastructure management 1.2. Disadvantages of traditional method 1. The nature of labor-intensive tasks, on-site data sampling cannot be performed in a comprehensive manner and thus do not render a complete picture of the inspected infrastructure. The missing information, sometimes, leads to major mistakes and incorrect decisions[10]. 2. Conventional approaches present a gap between data collection and data interpretation. Such gaps range from hours to even days. Monitoring of bridges, therefore, cannot be done in real-time and thus its efficiency is greatly reduced, resulting in potential safety risks. However, up till now, virtue diagnosis techniques for existed bridges have mainly been conducted by individual visual inspections. Therefore, there must be a strong demand to establish objective and effective bridge safety monitoring techniques for existing bridges[11]. 2. TECHNOLOGY OVERVIEW The concept of Internet of things was developed using the network of radio frequency identification (RFID) system by the automatic identification center which was established in Massachusetts institute of technology (MIT) in In this system, all the items can be connected to the Internet by radio frequency identification information such as sensing devices.[12]. It is used to realize intelligent recognition and management. Its main function includes: Information acquisition Information transmission Information processing Information apply-effect The network of physical devices, the vehicles, home appliances and other items Embedded with electronics, software, sensors, actuators and connectivity which enables these objects to connect and exchange data is called Internet of Things(IoT).The Internet of things makes us in a world of information and communication technology. The connection described in Fig.1 shows that we can connect any devices to anyone, at any time anyplace within the existing Internet infrastructure editor@iaeme.com

3 Poonam T Sathe and. Dr. Virendra V Shete Figure 1 The connection dimension of Internet of things. 3. PROPOSED SYSTEM The proposed system mainly consists of following main units as shown in block diagram. It includes sensor unit, power supply unit, wireless communication system unit, server unit, decision making unit. Figure 2 Block diagram of proposed system 3.1. Sensor Unit Sensor units main task is to monitor the realibility state of the bridge. It consists of various types of sensors for sensing the different signals it includes Wind speed sensor, Air pressure, Water level sensor, humidity sensor, temperature sensor, Water flow sensor and Accelerometer sensor. Wind speed sensors are mainly deployed on the bridge to measure the wind speed in real time. Water level sensors deployed on the bridge pier for monitoring the water level in real time when the water level crosses the threshold it gives the alert signal. Humidity sensors are deployed on the bridge near water or the culvert when the humidity is too high, the bridge surface and the internal steel stent are prone to corrosion, results in a damage to the bridge. Temperature sensors are deployed on the lower bridge and inside of the surrounding soil to monitor the change of soil in extrem weather, when temperature is too low, there is lead to change of soil and results in change of the position of the bridge. Accelerometer sensor used 44 editor@iaeme.com

4 Bridge Safety Monitoring Using SVM to measure the tilt, roll and pitch of the bridge. Water flow sensors are deployed on the bridge pier to measure the flow or speed of the water Power unit Power supply unit gives the power supply to the individual sensor, WiFi module and controller Wireless communication system unit The major function of the wireless communication system is to connect all the components in the bridge safety monitoring system, including the sensors, computing system and signal receptors. It consists of Wi-Fi module. WiFi exchange data amongst things within 100 m range by using radio waves. Smart devices communicate and exchange information using WiFi module Server unit The data collected by the above sensor units are transmitted using the Wi-Fi technology to this server system for further computing and decision making. The decisions made by the system, related analysis contents and alert messages are all transmitted by the server system via the internet to the management center and mobile devices of management staff for them to have real-time and comprehensive understanding of the bridges surrounding environment and keep records of the data for proper responses when a disaster occurs Decision making unit Decision making unit uses SVM classifier for the decision making. SVM classifier used on MATLAB platform. In it when parameters crosses their threshold value it gives alert signal to the authority center and people through thingtweet application. 4. METHODOLOGY Figure 3 Flow chart of the system editor@iaeme.com

5 Poonam T Sathe and. Dr. Virendra V Shete Above figure.3. shows the step by step flow of the system. 5. EXPERIMENTAL SETUP AND RESULTS Simulation result of the proposed system is shown in below figure. Figure 4 Simulation result. The DHT22 sensor is connected to the arduino board and output is obtained on the virtual terminal and on LCD as shown in the above figure. Following figure shows the hardware setup of the system. In it various sensors are deployed on bridge and there data is transferred on thingspeak server using WiFi(ESP8266). Figure 5 Hardware setup of the system Hardware connection Below section shows the module wise hardware connection of various sensors. As shown in the below Figure.6. hardware connection of DHT22 sensor and BMP180 sensor with arduino is shown which are used to measure temperature + humidity and atmospheric air pressure respectively editor@iaeme.com

6 Bridge Safety Monitoring Using SVM Figure 6 Hardware connection of various sensor Hardware results Following section shows the module wise result of the system. Below Figure shows the hardware connection of BMP180 and Rotary encoder sensor. And results are shown on LCD with WiFi module for transfering sensor data. Figure 7 Air Speed and Air pressure values on LCD. Hardware connection of DHT22 sensor with arduino is shown in below Figure.8. and Temperature and humidity values is displayed on the LCD screen. Figure 8 Hardware Connection of DHT22 sensor showing result on LCD editor@iaeme.com

7 Poonam T Sathe and. Dr. Virendra V Shete Below Figure.9. shows the result of accelerometer sensor deployed on the bridge, which gives the measurement in terms of pitch and roll of the bridge. Figure 9 Pitch and Roll measurement of the bridge. Following Figure.10. shows the hardware connection and result of the water module sensors which are water flow and water level sensor Software results Figure 10 Water module sensors result ThingSpeak ThingSpeakis an open source Internet of Things(IoT) application for storing and collecting data from things using thehttp protocol over the Internet or via a Local Area Network. Using ThingSpeak we can create sensor logging, location tracking applications and a social network of things with status updates. ThingSpeak provides support for the numerical computing software like MATLABby using MATLAB math work ThingSpeak users analyze and visualize uploaded data. Below figures shows the ThingSpeak results in terms of various sensor value chart. This chart values varies according to sensors data on regular time basis editor@iaeme.com

8 Bridge Safety Monitoring Using SVM Figure 11 Air pressure, Wind speed, Humidity and Temperature chart of respective sensors on ThingSpeak server. Figure 12 Water flow, Water height, Pitch and Roll chart of respective sensors on ThingSpeak server. Figure 13 Alert message through the ThingTweet application in the form of tweet. Above figure shows the alert message given to the end user as well as to the authority center through the ThingTweet application in the form of tweet MATLAB Matlab software is used for further processing of the data which is retrived from the thingspeak server. in it SVM classifier is used for accurate result. Below figures shows the SVM analysis results

9 Poonam T Sathe and. Dr. Virendra V Shete Figure 14 Linear kernel result of SVM algorithm. Above figure shows the result using Linear kernel of SVM algorithm which gives the 96% accurate result. Figure 15 Quadratic kernel result of SVM algorithm. Above figure shows the result of the system using Quadratic kernel of SVM algorithm with 99% accuracy. 6. ADVANTAGES 1. The system generates the alert if water level, and the load on the bridge are increased. 2. The system saves the human lives. 3. The system provides real time data of the temperature, humidity, load, water level, and pressure. 4. The system takes quick action and responses by reducing resources. 7. CONCLUSION Bridge safety Monitoring system using IoT and SVM classifier, is unique in its ability to monitor the bridge environment, transmit the environmental data through wireless communication and send alerts to the bridge management staff in real time for prompt reactions. And it gives the alert message if sensor value goes above certain threshold on thingspeak server through thingtweet application. This system will help to reduce major accident in future. We can save many lives using this system. This system will be enabled 24x7 and also all the collected environmental data sent to the server in the system can be used for big data analysis or follow-up research editor@iaeme.com

10 Bridge Safety Monitoring Using SVM FUTURE SCOPE Further enhancement can be done by addition solar cells to power up the hardware required for sensing environmental data and bridge parameters. The algorithm used in the project requires huge amount of structured data for training of the network architecture. So it becomes necessary to harness big data, analytics, social media and mobility that allows emergency & disaster management organizations to accommodate massive amounts of incoming public safety data. Moreover, giving mobile access to the information helps rest responders to better anticipate and respond to rapidly evolving critical situations. Critical situation data can also be sent using android application. REFERENCES [1] Pinkesh Machhi, Rishikesh Nandavadekar et.al, Structural Audit and Redevelopment of Shivaji bridge. Imperial Journal of interdisciplinary Research [2] Chih-Chyau Yang, Ssu-Ying Chen, Yi-Jie Hsieh, Fu-Chen Cheng, et.al, A Rugged Sensor System for Real-time Bridge Safety Monitoring in Taiwan, IEEE [3] Harutoshi Ogai, et.al, Damage Detection of Bridge using Wireless Sensors, International Federation of Automatic Control (IFAC) [4] Ittipong Khemapech, et.al, A Real-time Health Monitoring and Warning System for Bridge Structures, IEEE [5] Amro Al-Radaideh1, et.al, A Wireless Sensor Network Monitoring System for Highway Bridges, 1st International Conference on Electrical and Information Technologies ICEIT2015. [6] CHEN Shuli, SU Mubiao and LIU Yuhong, WANG Qingmin. Vibration Remote Monitoring System of Continuous Steel Truss girder The 14th World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China. [7] Ying Sun Research on the Railroad Bridge Monitoring Platform Based on the Internet of Things, International Journal of Control and Automation [8] Andrew Gastineau, Tyler Johnson et.al, Bridge Health Monitoring and Inspections Systems - A Survey of Methods, IRRB.org [9] Tyler Harms, Sahra Sedigh, and Filippo Bastianini Structural Health Monitoring of Bridges Using Wireless Sensor Networks IEEE In strumentation & Measurement Magazine December [10] Jin-Lian Lee2, Yaw-Yauan Tyan3, Ming-Hui Wen1, Yun-Wu Wu2 Development of an IoT based Bridge Safety Monitoring System, IEEE International Conference on Applied System Innovation IEEE-ICASI [11] Jivesh Kumar,Ramansh Bajpai et.al, Application of MEMS in bridge structures health Monitoring, International Journal of Engineering and Innovative Technology (IJEIT) [12] KuBo, The Research of IoT Based on RFID Technology,7th International Conference on Intelligent Computation Technology and Automation editor@iaeme.com