Research And Development Of Industrial Internet Platform For Chair Industry Based On Artificial Intelligence

The rise of the Industrial Internet has brought huge changes to the traditional manufacturing industry,which has promoted the development of the manufacturing industry from the digital stage to the network stage.The Industrial Internet has been serving the high-end manufacturing industry for a long time,mainly used for manufacturing process optimization and equipment operation and maintenance management,but it is rarely used in the production of small and medium-sized enterprises.This paper analyzes the problems of low level of information interaction,loose data value,and low degree of data application between the production lines of the office chair production line,establishes an industrial Internet platform for office chair production services,and manages production data in a unified manner.For the queuing problem of the workpiece buffer,this paper proposes an improved locust optimization algorithm to solve the optimal capacity configuration of the buffer.For the production line balance optimization problem,this paper selects the production takt and the balance index as the optimization target values,and uses the improved genetic algorithm.The minimum production takt time and balance index are solved.The main contents of this article are as follows:Through the discussion of the production process of office chairs,the problems existing in the current production line are analyzed,and the importance of building an industrial Internet platform is explained.The data involved in the production line is classified,and the corresponding data acquisition scheme is selected according to the data type,and the data acquisition and storage architecture is designed,which consists of the perception layer,the network acquisition layer,and the application service layer.The historical data can be obtained through the local database,and the real-time data needs to build a wireless sensor framework for data collection,and the composition of the real-time data collection architecture is described in detail.For historical offline data,this paper designs an offline data warehouse based on Hadoop ecology for data stream processing and batch processing.Compared with the mature big data industry,the office chair industry does not have standardized standards for data processing,so this paper A data standard set for office chairs is designed,and a hierarchical data warehouse is constructed.The warehouse is mainly used for data storage and data batch processing,and the data is cleaned to obtain the required number of office seats,the number of defective products and other target parameters,which are mainly implemented by Map/Reduce coding.For the application of data processing in the production line,this paper mainly solves the problems of queue congestion in the buffer and unreasonable production planning.From the perspective of buffer queue congestion,in order to solve the optimal buffer capacity configuration,this paper uses an improved locust optimization algorithm for data stream processing.From the perspective of unbalanced production planning,this paper selects two important parameters of production takt and balance index as the optimization target value,and establishes a multi-objective optimization mathematical model.Considering the defects of the traditional genetic algorithm in solving the production line problem,this paper improves the genetic algorithm from the perspectives of coding,inheritance and crossover operation,and obtains the optimal target parameter value in the process of multiple iterations.Finally,the design basis and functional interface of the industrial Internet platform are described in detail.