Date of Publication :10th August 2017

Abstract: Fractional calculus a field dealing with mathematical analysis has its applications in various domains such as power transmission units, image processing, financial system design, automobiles and various control system. There are many advantages of fractional calculus in analytical world. But, the computational cost accompanied with it has prevented software implementations to achieve real-time performance for large and complex computations. This paper exhibits the parallel computing power of the Graphics Processing Unit (GPU) in the area of fractional-order integration. Numerical methods for implementing different fractional-order derivatives and integrations are available. By using MATLAB Parallel Computing Toolbox, GPU computational power can be easily accessed with minimum knowledge of GPU architecture and MATLAB code can be executed on the GPU. The fractional-order integration by Trapezoidal formula using NVIDIA GPU with support of MATLAB Parallel Computing Toolbox is implemented in order to achieve faster execution. Performance comparison of the algorithm for sequential implementation on CPU and parallel implementation on GPU is carried out. This new algorithm produces significant speedup in the computations of fractional-order integration and provide required result in much less time as compared to execution on CPU.

Reference :