A parallelization of instance methods of a remote object of a distributed .NET application
Igor Košťál · 2012
An effective parallelization of a application code can significantly improve its the execution efficiency. This is an important parameter of the control application of mechatronic system, too. The subject of this paper is to find more efficient parallelized instance methods of a remote object of the distributed .NET application and comparing the results of execution effectiveness to serial methods of the same remote object of the same application. Parallelization of these methods is carried out by the method Parallel. For from the Task Parallel Library (TPL) and by data parallelism. Compared instance methods calculate the values of the definite integrals of the functions f 1 (x) = x2e−x^2, f 2 (x) = cos(x)/x, f 3 (x) = sin(x2) for which can not be expressed antiderivates as any algebraic combinations of elementary functions. One way to compute the values of definite integrals of these functions is to use numerical approximation methods. This is shown in the sample distributed .NET application developed using the .NET Remoting technology in the Microsoft Visual Basic 2010 and in the Microsoft Visual Studio 2010 development environment. Using the values of definite integrals can be calculated for example, centers of gravity and moments of inertia of homogeneous solids of revolution created by the revolution of curves of just such or similar functions as we presented above around the x axis. These may be important static and dynamic parameters of mechatronic system components, which need to be the most effectively determined by the control software of the system. The sample distributed .NET application can be a part of the control software of the mechatronic system. It can efficiently compute the values of definite integrals of mentioned or similar functions and to provide them to the control software that can compute by them values of these static and dynamic parameters of mechatronic system components.