Optimizing output performance of sensor based on wavelet neural networks

Sensors´ output performance is easily impacted by factors such as temperature, humidity and supply voltage fluctuations etc. To solve the issue, a mechanism that takes advantage of wavelet neural networks is proposed to optimize the sensors´ output performance. Mexican hat wavelet has been used as hidden layer function of the RBF neural networks, and pressure sensor has been used for simulation. The simulation results indicate that the proposed mechanism can effectively eliminate the impact of circumstantial factor to target sensor. The convergence rate and compensation accuracy with the proposed algorithm are better than with traditional methods and neural networks. The algorithm can be easily extended to other kinds of sensors.