Multi-dimensional function optimization acceleration method based on artificial bee colony algorithm

Abstract

The invention relates to the technical field of agricultural data processing, in particular to a multi-dimensional function optimization acceleration method based on an artificial bee colony algorithm, which comprises the following steps of: in order to evaluate the performance of an ABC algorithm, testing the performance of the ABC algorithm by adopting four different multi-dimensional functions for solving extreme values, and utilizing a heterogeneous architecture mode of CPU + DCU (Central Processing Unit + Data Control Unit) to optimize the performance of the ABC algorithm; a multi-thread technology of an HIP heterogeneous language is adopted for parallel optimization, and on a DCU platform, calculation tasks are divided into different threads; according to the method, the efficiency of solving the optimal solution of the multi-dimensional function by the ABC algorithm is improved, compared with a serial ABC algorithm and a GPU-ABC algorithm, the method can obtain better performance, the operation efficiency of the multi-dimensional function is improved, mass data brought by agricultural rural development can be further efficiently processed, then the data is expanded to a plurality of operation nodes, and meanwhile, the calculation efficiency of the multi-dimensional function is improved. And a vectorization optimization method can be carried out, so that the parallelism in the algorithm can be found more fully, and the execution efficiency of the algorithm is improved to the greatest extent.

Description

technical field [0001] The invention relates to the technical field of agricultural data processing, in particular to a multi-dimensional function optimization acceleration method based on an artificial bee colony algorithm. Background technique [0002] Agricultural and rural data has a long history, large quantity, and many types. However, there have been problems such as lack of core data, low quality data processing, and insufficient development and utilization for a long time, which cannot meet the needs of agricultural and rural development in the new era. Relying on the emergence of various advanced computing applications and analysis methods such as supercomputing and cloud computing, it has effectively accumulated experience in massive data processing for agricultural and rural development, and provided an effective way to solve the difficulties and problems faced by agricultural and rural big data development. [0003] Traditional optimization algorithms for functi...

AI Technical Summary

Problems solved by technology

[0003] Traditional optimization algorithms for function optimization problems, such as evolutionary calculation, golden section method and polynomial approximation method, can no longer meet the needs of agricultural and rural big data processing at the present stage, because the results of their local optimal solutions have large Limitations, and the dependence between the values ​​is strong, and an optimization algorithm with strong parallelism and randomness is required to optimize the data. The artificial bee colony algorithm is a swarm intelligence algorithm, which simulates the bees in nature to find the best honey source The process to find the optimal solution of the problem has the advantages of less preset parameters, simple implementation, and strong parallelism. However, for the multi-dimensional function solution problem, the algorithm takes too long to solve and still cannot meet the needs of big data processing in the industry. need

Images