College of Engineering, Pune

About: College of Engineering, Pune is a based out in . It is known for research contribution in the topics: Computer science & Sliding mode control. The organization has 4264 authors who have published 3492 publications receiving 19371 citations. The organization is also known as: COEP.

Automated test bench for an induction motor using LabVIEW

Abstract: The testing of electrical machines is necessary for validating the design of a machine. Induction Motor (IM) is widely used in the industrial applications. More than 90% motors used are of squirrel cage type in the industry. With the development of sensors and various automation components, the testing process can be automated using software such as LabVIEW, Programmable Logic Controller (PLC), Supervisory Control and Data Acquisition (SCADA) etc. LabVIEW is a graphical programming language which is widely used throughout the industry, as well as research and academic labs, for data acquisition and as an instrument control software. LabVIEW is compatible with Windows, Mac OS X, and Linux. The traditional motor testing method possesses many disadvantages such as more time requirement, lack of reliability, and stability of measuring instruments in the production environment. This paper aims at development of an automated test bench for Induction Motor using data acquisition and controlling of the motor by developing suitable Graphic User Interface (GUI) and integration of sensing components into the system.

Image Classification Using Advanced Block Truncation Coding with Ternary Image Maps

Abstract: Incredible escalation of Information Technology leads to generation, storage and transfer of enormous information. Easy and round the clock access of data has been made possible by virtue of world wide web. The high capacity storage devices and communication links facilitates the archiving of information in the form of multimedia. This type of information comprises of images in majority and is growing in number by leaps and bounds. But the usefulness of this information will be at stake if maximum information is not retrieved in minimum time. The huge database of information comprising of multiple number of image data is diversified mix in nature. Proper Classification of Image data based on their content is highly applicable in these databases to form limited number of major categories. The novel ternary block truncation coding (Ternary BTC) is proposed in the paper, also the comparison of Binary block truncation coding (Binary BTC) and Ternary Block Truncation Coding is done for image classification. Here two image databases are considered for experimentation. The proposed ternary BTC is found to be better than Binary BTC for image classification as indicated by higher average success rate.

Object recognition using horizontal array of ultrasonic sensors

Abstract: The aim is to design a low cost and reliable Object Recognition System. In the system, horizontal array of ultrasonic sensors is used for object recognition. The advantage to use ultrasonic sensor is the ease with which distances can be obtained from immediate objects without intensive processing. Recognition methods with an ultrasonic sensor are often utilized in situations where optical sensors cannot be used or in objects which are hard to be identified by the approach based on the light. Also ultrasonic sensors have low cost compared to other devices. Thus ultrasound sensor seems to be a good solution to detect and recognize several objects. In the paper, new approach is proposed in which response of ultrasonic sensor from rough surface objects is used for object recognition. When object surface has roughness of specific degree, incident ultrasonic waves scatter more and part of ultrasonic may return to the sensor receiver. Thus recognition of objects having critical orientation with respect to ultrasonic sensors array or rough surfaces has become possible.

Application of neural networks in resolution of disputes for escalation clause using neuro-solutions

Abstract: Increasing scope and complexity of construction industry, results into clashes of interests among the parties to the contract. Clashes lead to claims which may culminate into disputes. Project duration being considerable, escalation forms one of the reasons for occurrence of disputes. Dispute resolution consumes valuable time and money of the parties involved. Hence a faster, convenient and cheaper method needs to be evolved. This study throws light on the possibility of implementing neural networks in resolving disputes arising out of escalation claims. For this, factors influencing the decisions for escalation claims are identified and Neuro-solutions is implemented for network building. Various iterations help to assess the effect of change in network parameters. Comparison of their results is made in the study giving an optimum combination of parameters for effective resolution of escalation disputes using neural network. The approach is demonstrated to be a feasible alternative and the network is able to give a very high prediction rate.