The Design Of Wireless Data Collection System Based On Neural Network Temperature Compensation

In industry, the process of monitoring and controlling the parameters usually use pressure sensors.In order to improve the non-linearity of pressure sensor and the output error characteristics caused by the variation of temperatures,this paper has discussed tree methods about the temperature drift compensation, they are method which is composed by polynominal-match method and Newton insert numerical value method,BP neural network arithmetic and the ntural network arithmetic which is put forward by us. We discussed the third method specicaly.It consists of the regulation for four weights, which stand for the coefficients of one and two powers of output signals and temperatures, respectively. An optimum output formula is obtained after a lot of iterations. This formula satisfies both sample values and other values among them and can check the correctness of the final iteration results of this artificial neural network computing. Because the third method has the merits which are higher accuracy, faster speed and lower cost, we used this method.A wireless data acquisition system for smart pressure sensor based on wireless module PRT2000 is presented in this paper. In order to insure the reliable transmission of the data, a protocol must be made to adjust the transmission between the master and slave.The system can be used in some special situations for the signal collection, processing, and transmitting. It can also replace the complicated wiring in local field. Furthermore, this wireless data acquisition system for smart pressure sensor is low cost and reliable with high practicability.