Study On Self-organizing Fuzzy Control System Of Dyeing Machine

Textile industry is one of traditional competitive industries in China, while dyeing machineis one of large essential equipment. Reliability and accuracy of temperature control are the mostimportant factors for quality of fabric. That is to say, textile dyeing process should strictly obeytemperature curve. In the dyeing temperature control process, overshoot, disturbance, responsetime and deviation are to be minimized to improve quality and productivity. Proper selection forfuzzy control rules is the key to design a successful fuzzy controller. Rules of conventional fuzzycontroller are chosen with trial-and-error method which is pre-established based on experts’experience and fixed in the control process. The fuzzy controller does not achieve in abilityadapting to constantly changing process, which results in excellent control effect can not beguaranteed. Self-organizing fuzzy control and optimization are proposed to solve these problems.Simulation and experiment in temperature control of dyeing machine is implemented.Main work and contributions are shown as following:(1) Review for background and development of dyeing equipment and fuzzy controlresearch. Dyeing process for dying machine is described, temperature characteristic is analyzedin detail, and desired temperature object model is established.(2) Dyeing machine is a system with nonlinear, large delay, uncertainty model structure andother striking features. Rules of conventional fuzzy control can not be self-tuned, so thatdifficulty in adapting sustainable change of plant. A new self-organizing fuzzy control methodbased on knowledge association is proposed. Original rules based on expertise and correctionrules based on performance evaluation are pre-established, relativity of which is studied.Knowledge association function is built for updating rules. Simulation results show that thecontrol algorithm is effective.(3) A novel strategy is brought forward which is based on modified artificial fish swarmalgorithm (MAFSA) for self-organizing fuzzy logic controller design (SOFLC). By addinginformation of global best artificial fish (AF) in behaviors of AF, considerable advantages, suchas high convergence speed, flexibility and high accuracy, are achieved. SOFLC is designed and optimized by seven parameters, including three correlation factors of membership functions, onemodifying factor of fuzzy logic rules, two quantitative factors and one scaling factor. The fitnessfunction is considered as optimized objective defined by comprehensive performance index ofthe controller. Finally, simulation results show that MAFSA based SOFLC can effectively avoidpremature and achieve good dynamic and static performances.(4) On the basis of MCU STM32F107, intelligent control system including hardware andsoftware are accomplished, and proposed method is applied in temperature control of dyingmachine. The control system shows good temperature real-time tracking performance, highcontrol accuracy and promotional value.