Research On Wireless Monitoring System For The Storage Environment Of Electronic Components Based On IGA-RBF

In order to monitor temperature and humidity changes for the storage environment of electronic components,and realize the function of data processing,signal wireless fidelity(WIFI)technology is combined with predictive control method for radial basis function(RBF)neural network based on improved genetic algorithm(IGA),a wireless monitoring system for the storage environment of electronic components based on IGA-RBF is proposed and implemented.The system consists of sensor module,IGA-RBF neural network predictive monitoring module and the actuator module.The sensor module is used to collect temperature and humidity of storage environment.The IGA-RBF neural network predictive monitoring module mainly consists of microcontroller and peripheral circuits,WIFI serial communication circuits,IGA-RBF neural network predictive model and RBF neural network controller.Combined with the design of system hardware circuit and software,data transmission of storage environment can be realized through WIFI,which is used as the input of the IGA-RBF neural network prediction model.According to the IGA-RBF algorithm,nonlinear mapping relationship between input and output of the IGA-RBF model is learned,and mathematical model of storage environment is established.Finally,The RBF neural network controller is designed to correct the deviation between temperature and humidity preset values and predictive model output values,and then output the control volume of the actuator,which is sent to microcontroller through WIFI.The microcontroller drives the execution units of the execution module to perform corresponding actions,thereby realizing the control for the storage environment of electronic components.The function of data processing,display,alarm and WIFI serial communication is realized by the lower computer software.The upper computer monitoring software is designed under Windows system,mainly to manage IGA-RBF neural network predictive model,RBF neural network controller and storage environment data.The temperature and humidity output range for the storage environment of electronic components is set to 18~22 ? and 42~45 %RH respectively.The performance of IGA-RBF neural network predictive model is tested and experimental results show that root-mean-square error of predicted temperature and humidity is less than 0.16 ? and0.31 %RH,and fitting precision of fitted value and actual value are greater than 0.98 and 0.95.It can be seen that IGA-RBF neural network predictive model has the strong correlation andgood generalization ability,and fitting error more less.Therefore,IGA-RBF neural network predictive model as controlled object for the storage environment of electronic components,and combined with design of RBF neural network controller to form a predictive control structure for the storage environment of electronic components based on IGA-RBF.At the same time,WIFI serial communication and overall structure of predictive control system are validated and tested.Experimental results show that system can realize functions of data processing and wireless transmission,and the values of temperature and humidity can change quickly following set values.