Design Of Temperature Control System For Automatic Packaging Machine Based On Interpolation And Neural Network

With the continuous improvement of the Chinese people's economic level,the demand for puffed food is increasing rapidly,which makes the market of domestic food packaging machines develop continuously.However,compared with developed countries,there is a big gap between the domestic control technology,especially in the temperature control technology.With the growing maturity of the intelligent control theory,automatic packaging machine is developing towards the direction of intelligent control in the world.This paper discusses and studies the temperature intelligent control technology of heat-seal temperature control system of automatic packaging machines.The hardware circuit of the heat-seal temperature control system and the related peripherals are designed which was base on the STM32F103ZET6 microcontroller of the Cortex-M3 kernel.SPI protocol is used to read and write nonvolatile memory FLASH.FSMC protocol is used to expand memory.ModBus communication protocol is used to realize communication between controller and thermal seal speed actuator.The data processing ability and integration degree of the heat-seal temperature control system are improved.In order to solve the influence of the heat-seal parameters of the packaging film in the traditional heat-seal process.This paper established a mathematical model which can be used in the heat-seal process of automatic packaging machines by studying the nonlinear relationship between heat-seal parameters in the heat-seal process of plastic filmand.The BP neural network was trained by the additional momentum method under the sample data which was collected from experimental data to establish the nonlinear mathematical model between the heat-seal parameters.Using neural network to predict the heat-seal speed,and interpolation algorithm is used to establish the polynomial mathematical model between heat-seal temperature and heat-seal time.The results show that through the combination of interpolation algorithm and neural network,the mathematical relationship between heat-seal temperature and heat-seal spped is described accurately.At the same time,the interpolation function realizes the simplification of the neural network model,and the error is within the allowable range.The intelligent level of automatic packaging machine is improved.In order to solve the difficult problem of the traditional PID control algorithm parameters,the paper studies neural network PID algorithms which was radial basis neural network combined with the traditional PID.The analog PID was converted for digital PID,and compares with neural network PID algorithm.The simulation results of system output response in Simulink simulation tools show that the RBF neural network PID algorithm are superior to traditional PID whatever in dynamic or static performance aspects.The application of neural network PID algorithm to automatic packaging machine will improve the accuracy of heat-seal temperature control,and at the same time,as the PID parameters are adjusted by the system,the intelligent level of automatic packaging machine is improved.The control system is realized and debugged on the designed hardware experiment platform.The output curve of RBF neural network PID and the result of serial port debugging are given.