Research On Selection Method Of Mine Technology Evaluation Index Based On Improved BGWO Algorithm

In recent years,with the continuous development of swarm intelligence algorithms,many excellent swarm intelligence algorithms have been proposed,and achieved good results in different fields.In the establishment of mine technology evaluation index system,because in the past,the indexes need to be selected by experts,the index system often has great subjectivity,resulting in a series of problems such as too many indexes and overlapping among indexes.Therefore,this paper studies the wrapped feature selection method based on swarm intelligence algorithm,selects binary grey wolf optimization(BGWO)as the feature subset search method,and puts forward a variety of different improvement strategies to improve the convergence speed and calculation accuracy of the algorithm.The main work includes the following aspects:Firstly,aiming at the slow convergence speed of binary gray wolf optimization algorithm,a binary gray wolf optimization algorithm based on Golden sine(BGWO-GS)is proposed.By introducing the golden section number into the traditional binary gray wolf optimization algorithm,BGWO-GS algorithm is more inclined to search the area where the optimal solution is more likely to appear,so as to effectively reduce the search space and improve the search efficiency.Compared with other algorithms,BGWO-GS algorithm can obtain higher accuracy and lower average fitness value than BGWO algorithm,The running time and feature selection rate of the algorithm are reduced,and experiments are carried out on eighteen standard data sets of UCI to prove its effectiveness.Secondly,aiming at the problem that the binary gray wolf optimization algorithm is easy to fall into local optimization,a binary gray wolf optimization algorithm based on Levy flight and spiral encirclement(BGWO-L-SE)is proposed.By introducing Levi flight mechanism into BGWO algorithm,the algorithm can jump out of local optimization with higher probability in the later stage of iteration,strengthen the global search ability of the algorithm,and achieve a good balance between exploration and development.At the same time,the spiral encirclement mechanism of whale optimization algorithm is introduced into gray wolf optimization algorithm to strengthen the search ability of the algorithm and further improve the performance of the algorithm.In order to evaluate BGWO-L-SE algorithm,eighteen standard data sets of UCI are selected and compared with BGWO,PSOGWO and BWOA algorithms.The results show that BGWO-L-SE algorithm performs better in accuracy,average running time,average fitness value and standard deviation compared with the above algorithms.