Attribute Reduction Of Neighborhood Rough Set Based On Improved Fish Swarm Algorithm And Its Application

Attribute reduction is the key technology of neighborhood rough set.Its main idea is to delete redundant and irrelevant condition attributes without affecting the classification ability of decision system.Since the attribute reduction proves to be an NP-hard problem,the traditional attribute reduction algorithm has a small search space,and often cannot obtain a smaller reduction set.The attribute reduction algorithm combined with the group intelligent algorithm has a larger search space.It is possible to obtain a smaller reduction set,so the study of a more efficient and faster group intelligent attribute reduction algorithm becomes one of the main research topics in the field of attribute reduction.In this paper,a neighborhood rough set attribute reduction algorithm based on improved fish swarm algorithm is proposed.The effectiveness of the algorithm is verified based on the UCI test set.The algorithm is applied to the evaluation of the fatigue life influencing factors of aluminum alloy welded joints.The evaluation system of fatigue life influencing factors of aluminum alloy welded joints is obtained by quantitative evaluation and weight calculation of fatigue factors of aluminum alloy welded joints,which provides a basis for fatigue life prediction of aluminum alloy welded joints.The research work of the thesis is mainly reflected in the following three aspects:First,an improved artificial fish swarm algorithm is proposed.Based on the artificial fish swarm algorithm,the algorithm introduces the piecewise adaptive function to control the field of view and the moving step size of the artificial fish.The artificial fish moving strategy in the cluster and rear-end behavior is improved,and the crowding factor is ignored.The default foraging behavior was abolished;the foraging behavior was improved.After a round of trials,if the better solution was not found,the artificial fish made another round of trials with new horizons and steps;The rebirth mechanism,after each iteration,eliminates the worst artificial fish and regenerates it into a more adaptive artificial fish,ensuring that the overall fitness is at a higher level.Secondly,combined with improved artificial fish swarm algorithm and neighborhood rough set attribute reduction theory,attribute dependency degree is used as fitness function,and a neighborhood rough set attribute reduction algorithm based on improved fish group is proposed.By comparing the algorithm with the neighborhood rough set forward greedy search attribute reduction algorithm and the artificial fish group based neighborhood rough set attribute reduction algorithm,the experimental results are verified from the previous time.The feasibility of improved fish school algorithm is proved that the algorithm is a fast and effective neighborhood rough set attribute reduction method from the reduction rate and the classification accuracy.Finally,by referring to the literature and the fatigue test of welded joints,the data sets of fatigue life influencing factors of aluminum alloy welded joints are obtained,and the influence factors of fatigue life of aluminum alloy welded joints are established based on the neighborhood rough set attribute reduction algorithm based on improved fish stocks.The evaluation model completed the research and development of the evaluation system for the fatigue life of aluminum alloy welded joints,determined the set of key factors affecting the fatigue life of aluminum alloy welded joints,and quantitatively calculated the weight values of each key influencing factors,and accurately predicted the aluminum alloy welded joints for subsequent follow-up.The fatigue life provides the basis.