Multiprocessor task scheduling method based on harmony search and simulated annealing

Abstract

The invention relates to a multiprocessor task scheduling algorithm based on harmony search and simulated annealing. The algorithm lies in fusion of harmony search and simulated annealing and includes: firstly, initializing harmony algorithm related parameters and a harmony base; secondly, directly performing annealing process to a task sequence corresponding to an optimal harmony solution in the harmony base so as to enhance search speed of the algorithm. The optimal solution is stored, probabilistic leap of SA (simulated annealing) algorithm is given to full play, a local minimum point is got rid of, and accordingly algorithmic performance is improved effectively. In addition, the modified algorithm is successfully applied to task scheduling of a multiprocessor. Verification shows that the algorithm has a scheduling strategy better than the traditional algorithm in operating efficiency.

Description

technical field [0001] The invention relates to a method for improving the processing speed of a computer, in particular to a multiprocessor task scheduling method based on harmony search and simulated annealing. Background technique [0002] The problem of multiprocessor scheduling arises with the development of computers. With the rapid enhancement of the performance of microprocessors, new fast networks appear constantly, resulting in high-performance computer systems, namely distributed systems. The efficiency of distributed system processing tasks largely depends on the task scheduling of processors. Therefore, the emergence of distributed systems promotes researchers to study the scheduling of multiprocessors more deeply. [0003] Multiprocessor Task Scheduling Problem (MTS) is a kind of combinatorial optimization problem, it is NP-hard problem, and it is also an integral part of parallel and distributed computing. The multiprocessor scheduling problem is primarily a...

AI Technical Summary

Problems solved by technology

They proposed it based on the phenomenon that musicians rely on their own memory to repeatedly adjust the pitch of each instrument in the band to achieve a state of beautiful harmony. Although HS was proposed later, it is simple, robust, and The advantages of fast convergence speed have been applied and developed in many fields. The harmony search algorithm belongs to the heuristic group intelligence optimization algorithm. It has a continuous nature and has been widely used in many continuous optimization problems. The simulated annealing algorithm is the earliest It is proposed for combinatorial optimization. Its characteristic is that it can jump out of the local minimum during the search process and eventually tend to the global optimum. Therefore, it has the advantages of high quality, strong initial value robustness, and easy implementation in general. For example, Yu Liang et al. used the harmony search algorithm to solve the water distribution optimization system design, and achieved better results, but it has not been widely used in the field of multiprocessor scheduling

Images