Parallel industrial internet of things big data clustering method based on meta-heuristic algorithm

Abstract

The invention provides a military dog-based cognitive industrial Internet of Things big data clustering parallel algorithm, which specifically comprises the following steps: (1) preparing clustering data, (2) distributing tasks to different machines in an MR-MHBC-Map stage to simulate a search process of a military dog on suspicious targets to perform clustering and update a clustering center, and (3) on each machine, performing clustering on the suspicious targets in the MR-MHBC-Map stage. The optimal clustering center of each data point is solved during each iteration, and (4) the decomposed tasks are combined in the MR-MHBC-Reduce stage, and whether the termination condition of the algorithm is met is judged. According to the method, the advantages of MapReduce are utilized, and a new clustering method based on meta-heuristic is provided to solve the problem of big data. According to the method, the potential of searching a suspicious target by a military dog is fully utilized, and a big data set is processed by adopting a MapReduce structure. The MR-MHBC algorithm is superior to other existing algorithms in the aspect of clustering the big data set, and has important practical significance.

Description

technical field [0001] The present invention relates to the field of big data clustering, in particular to a big data clustering method for the parallel industrial Internet of Things based on a meta-heuristic algorithm. The optimal clustering center of the data, and the computing framework MapReduce oriented to the parallel processing of big data is used to process large data sets. Background technique [0002] With the development of wireless communication, Internet of Things and big data, using high-performance data analysis tools and algorithms to provide people with intelligent services has become a research hotspot in recent years. In the Industrial Internet of Things, a large number of sensors are usually deployed to collect data, and the data generated by these sensors is usually high-capacity, manifold, and unstructured. Analyzing such a large volume of streaming data is a challenging problem. Furthermore, streaming data acquired from sensors is time-dependent, whi...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to propose a parallel method for industrial Internet of Things big data clustering under the inspiration of military dogs, aiming at the problems of dense calculation objective function and high calculation cost faced by the traditional meta-heuristic algorithm in the current Industrial Internet of Things. Using a new meta-heuristic algorithm, by simulating the search process of trained military dogs for suspicious targets and using the MapReduce structure to perform parallel calculations on large data sets, to find the optimal clustering of large data clusters generated by the Industrial Internet of Things Class center, and effectively solve the data processing and computational complexity problems faced by the clustering of large industrial IoT data sets

Images