Embedded system for high performance digital temperature sensor on distributed parallel computer system

Abstract

This paper presents the design and development of the digital temperature sensor (DTS) for data collecting of temperature behavior and heat flow profile. The transition of collected temperature is implemented by reading the analogue signals from temperature sensor and convert into digital signal. Then, the High Performance Computing software (HPC) is developed to organize the process of digital data collecting, to observe the temperature behavior and to predict over the monitored area. Some numerical analysis and parallel performance measures are used to validate the automated temperature measurement sensor and to process the visualization of digital data collecting. The mathematical modeling based on partial differential equation (PDE) with multidimensional parabolic type are used to predict the temperature behavior. The parallel performance measures are based on run time, convergence rate, accuracy, root means square error, speedup, efficiency and effectiveness. Distributed parallel computer system (DPCS) with DPCS is developed based on MATLAB Distributed Computing Software version R2011a with 8 processors of Intel® Core™ Duo Processor Architecture. Local area network is generated to communicate via Ethernet network on LINUX operating system. The communication paradigm of MATLAB Distributed Computing Engine allows the communication activities among the neighboring processors. As a conclusion, the embedded system of DTS is efficient on DPCS for visualizing the digital data.