Monitoring And Optimizing For Carbon Fiber Spinning Process Based On Intelligent Algorithms

Carbon fiber spinning is a large-scale production system with multiple processes and complex conditions. The implementation of monitoring and controlling the production process requires not only mastering the production equipments and technologies, but also learning to design and optimize production process. This paper, as a sub-topic of the carbon fiber spinning forming process monitoring, process optimization and systems integration projects, undertakes a research on the optimized operation of the polyester filament's spinning process. The main research is as follows:Firstly we studied the carbon fiber spinning process, comfirmed the impact parameters of every process and the fiber's precursor structure and quality indicators. Then we analyzed the relationship between the quality of the carbon fibre and the spinning process parameters with gray relational analysis.Second, we established the carbon fiber spinning process monitoring system based on the Netcon system, including the chossening of parameters to be monitored and building the system's hardware and software platform, and building database for monitoring data. In this platform, we proposed carbon fibre quality prediction model which based on the support vector regression machine and the genetic algorithm, so we can predict the quality of carbon fibre according to online parameters, this act a role in early warning.Then, we established a carbon fibre spinning process optimization expert system. Expert system used the collaborative framework, which divided into the main system and some subsystems. The implementation of expert system included knowledge representation and inference engine designing. The knowledge base divided into global knowledge base and sub-domain knowledge base. The inference engine used RBF neural network for backward reasoning, obtarined the performace parameters after optimization to online optimization purposes.At last we propose the implementation of the carbon fibre spinning process monitoring and optimization system software based on the C/S structure, and show the overall design framework of the system, the practical function of modules and the form designs of the system database, which provide a good software platform for the monitoring and intelligent optimization of the carbon fiber spinning process in the future.