Research Of A Kanban System For Glass Tempering Guidance Based On FNN Control Algorithm

Today, the requirement of the tempered glass is continuously growing. How toimprove the yield and the rate of the tempered glass whose quality is good in a unitof time has become the focus point in the enterprises’competition.There are many problems in the process of glass tempering, such as complexoperation procedure, unstable control of production parameters, limitations inprocess information monitoring, bad retroaction of quality etc. Facing theseproblems, a KanBan system that working in a glass tempering furnace which has adouble room for convection has been designed. This system has four modules,including the prescheduling module，the guidance module for controlling parameteroptimization，the process information monitoring module and the quality tracingmodule. In this paper, researches have been done from three aspects of the controlparameter optimization of the tempering furnace, process visualization of temperingand tempered glass’s quality. The main work is shown as follows:(1) Method for control parameters’ optimization based on experience values hasbeen analyzed to find its advantages and disadvantages. Basing on the fuzzy BP neuralnetwork control algorithm, according to the classification look-up tablemethod,using feedback information of the tempered glass’s defect detection，theoptimal control parameter database can be established. This system can update thecontrol parameters itself and has the adaptive function, to control the parameteroptimization effect.(2)Series of lattice sensor is used to record the temperature sets of valid locationin the glass tempering furnace for based on optimal coverage theory. The display modeis various like numerical matrix, curve etc. The system can monitor process parametersonline and achieve the effect of glass tempering process visualization.(3) Basing on the RFID anti collision algorithm, using radio frequencyidentification technology, the tempered glass quality traceability function can be realized bycollecting the tracking information from glass’s order, production process and transportation.The running results show that, this system has a series of characteristics such ashigh reliability, good control stability, weak dependence on technical workers. Also,the efficiency and finished product rate of toughened glass could be effectivelyimproved.