Research And Development Of Visualization Pipeline Supporting Full Process Analysis Of Industrial Control Data

In the historical data of industrial control fields such as thermal power generation,there are valuable control laws that can be used to guide equipment operation optimization.However,due to the complex structure of the control system,it is difficult to analyze and optimize the control parameters,therefore,the visual analysis technology can be introduced into the data analysis and modeling process of complex industrial control processes.Visual analysis technology can provide flexible view interaction to support complex data analysis processes,but visual analysis of data in industrial control fields has problems such as difficult requirements for refining,difficult process design,and complex technical implementation.Therefore,a visualization pipeline is proposed to support the whole process analysis of industrial control data,and comprehensively guides the design and implementation of visual analysis of industrial control data from the aspects of overall architecture,analysis process and technology implementation.The pipeline contains three components:(1)Conceptual design framework,which extracts the key elements data,visualization,modeling,knowledge and control optimization in the industrial control field,and abstractly describes the main line of visual analytics.(2)Process division architecture,which designs a visual analytics process structure for the full process analysis of industrial control data,and refines the visual analytics tasks and design aims in each process.(3)Technique implementation classification,which combs the main technique challenges in industrial data processing,visual mapping and interactive analysis for visual analytics process,and develops the corresponding technique classification system and solutions.The real application study cases in the field of thermal power control show that the proposed pipeline can effectively support visualization designer and domain experts to characterize visual analytics problems in industrial control field,and rapidly develop more feasible analytical processes and more usable technical implementations.