|
|
| Description: |
Santhosh is invoved in a study which
aims to automate the online remote interrogation of a Sensor network inside
the pipes using the Internet and to check for it’s data consistency using
Artificial Neural Networks. "Suppose you’re sitting in front of your computer screen and you want to monitor the behavior of a pipeline (or any other structure). You can accomplish this by constructing an Internet web site and transferring data from that remote location to this website which you are viewing. My research accomplishes the same task for the Trench less Technology center, Louisiana Tech University. In addition this data is checked for its consistency using artificial Neural Networks. The remote data is captured by the network of sensors embedded in the pipelining network. These sensor’s can record any data type of data e.g. Temperature, stress, humidity etc. My research describes an innovative way to collect and transmit data using Internet Protocols. A servlet program running on the web server accepts transmitted data and incorporates it into a project database. Bi-directional interaction between the web client and web server is accomplished through the use of a servlet program and a Java applet running inside a browser located on the web client’s computer. The second part deals with the predicting the data transmitted by the sensors based on the previous values and raising a warning flag for the engineer to check if any abnormality is detected. A suitable neural network has to be selected based on the values seen. Since a Back propagation algorithm can remember a continuous data function or approximate a continuous function, I decided to use this for prediction. This designed neural network should be trained on the old sets of data and should predict future sets. If the amount of variation of the actual output is beyond a threshold percentage of the actual value a flag is raised show an abnormality at that point. In future this work can be extended to use wireless Internet data Transfer protocols." View Santhosh's PowerPoint presentation . |