Research And Implementation Of Quality Information Management System Based Statistical Process Control

In the era of economic globalization,enterprises play a strong role in promoting the rapid development of national economy,and quality has become the core competitiveness of enterprises.In the complex market environment,there have been phenomena such as the lack of information accuracy,timeliness and management requirements,insufficient process control capabilities,differences in regional quality standards,and inaccurate decision-making.Facing the new situation of quality,it is an important research topic to make use of rapid development of computer science and artificial intelligence technology in the research of quality science to actively promote enterprise information management.Aiming at the problems of quality management,this dissertation puts forward the standard based on sampling inspection and statistical process control,and improves the product quality from three aspects of planning,inspection and control.Aiming at the limitation of statistical control chart,a control chart recognition algorithm based on transfer learning and convolution network is proposed.According to actual investigation and demand analysis,a quality information management system based on statistical process control is designed and implemented.The main tasks are as follows:Firstly,aiming at the problem that eight kinds of abnormal patterns in the control chart were fixed,and could not completely identify all process patterns,a control chart recognition algorithm based on transfer learning and convolution neural network was proposed.Firstly,according to the characteristics of abnormal patterns,six kinds of typically abnormal patterns were abstracted.At the same time,the numerical data was converted into image data as sample data.Then,according to the feature based transfer learning in the isomorphic space,VGG16 convolution neural network was transferred to complete the feature extraction,which could improve the generalization ability of the network.The classifier was trained on the target dataset.Finally,the output of the feature extraction was used as the input of the classifier,the network was fine-tuned according to the recognition results during the process of training on the target dataset,and the optimal control chart recognition model was obtained.The total recognition rate is above 98%.Secondly,in view of the differences in the description and understanding of the same problem or item for the staff,it causes the obstacles to information exchange.In this dissertation,the concept of data dictionary is introduced,and a series of strict and consistent definitions are established to explain the entry of system application,which is beneficial to the effective communication between the staff.Thirdly,in order to solve the problem of manufacturing process,national standards are introduced in the process of production,mainly including sampling inspection procedures and conventional control chart,which standardize production processes and are conducive to scientific and standardized data collection and quality control in the production process,and to ensure the unification of process standards.Thirdly,aiming at the problem that inspection of batch products is non-standard and the control ability of production process is insufficient,this dissertation introduces national standards,mainly including sampling inspection procedures and conventional control chart,which is conducive to inspect quality and control process scientifically as well as normatively,so that production of high quality products continuously.Fourthly,due to the large coupling between internal systems and the poor adaptability of heterogeneous data environments,the overall system adopts web service integration technology,which is platform-independent and integrates functions between heterogeneous platforms,providing enterprises with more flexible and effective solutions.Fifthly,applying the above research results according to the actual needs,the quality planning,quality inspection and quality control modules are designed and implemented.From the quality objectives of products and services,the scientific and standardized inspection plan and more standard and reasonable inspection process are formulated,and the quality information management system is effectively integrated with the target of the enterprise.