Temperature and Humidity Sensor Location Optimization Based on BP Neural Network

A system for sensor location optimization based on BP neural network is described in this research, which consists of the digital temperature and humidity sensors, microcontrollers, data transmission unit and the PC. The system is tested by experiments and the results show that temperature standard deviation is ± 0.3°C and relative humidity standard deviation is ± 0.5%RH, which meets the accuracy requirements of the design. In the process of sensors distribution optimization, by building and solving the contact degree function, characteristic analysis method is used to pick up the best information collection point. The BP neural network model is built, and the temperature and humidity data and other factors (such as height, weight, etc.) are used as network input. The experimental conclusion is that the two evaluation methods have high consistency (the average accuracy rate > 81.1%, the trained sample n = 6, the test sample n = 4), which is sufficient to prove that objective office chairs comfort evaluation method has good evaluation effect using temperature and humidity acquisition system and BP neural network.