Research On Calibration Instrument Development And Error Compensation Method Of New Automatic Weather Station

In order to meet the requirements of China Meteorological Administration for the verification of automatic weather stations every two years,the development of a calibration system for the data acquisition unit of the new automatic weather station and the improvement of the sensor error compensation algorithm play an important role in improving the quality of meteorological observation data and the competitiveness of China's meteorological equipment industry.This paper analyzes the structure of the new automatic weather station,Aiming at the existing problems of JJQ1 automatic weather station signal simulator,the circuit modules such as temperature,humidity,pressure,wind direction and radiation of power supply are improved.Aiming at the problem of poor adaptive ability of current sensor compensation methods,an adaptive piecewise compensation algorithm is proposed.The main research contents are as follows:(1)The temperature measurement principle of the CAWS600 and ZQZ series automatic weather station data acquisition units is analyzed.The temperature channel of the existing automatic weather station signal simulator was tested using different types of data acquisition unit.Point out the reason why the test data has jitters.Proposed its improved circuit.Based on the principle of four-wire temperature measurement,a four-wire resistor network with high precision and very low temperature drift is constructed with relays,supplemented by software compensation.Develop temperature channels that accommodate different models of data acquisition unit.(2)Aiming at the problem that the output channels of analog quantities such as humidity,air pressure,evaporation and radiation of existing automatic weather station signal simulators are greatly affected by ambient temperature and humidity,a programmable weak standard source is proposed.The source is based on MSP430,with16-bit DAC device and precision conditioning circuit as the core.Accuracy is improved by measures such as power isolation,bus isolation,and software compensation.The temperature and humidity adaptability experiment and electromagnetic compatibility experiment were carried out on the circuit.The circuit can be programmed output of 0~20mV weak voltage and 4~20mA current,the maximum error is ±3uV and ±2uA respectively.The performance of the circuit remained normal after electro-static discharge and electrical fast transient/burst immunity experiments.The output signal provides a standard source for verification of analog channels such as humidity,air pressure,evaporation,and radiation from an automatic weather station collector.(3)An algorithm for adaptive segmentation compensation measurement error of humidity sensor is proposed.Through the experimental test,the error curve of the humidity sensor affected by temperature is determined.error curve.According to the segmentation results,the humidity measurement data of different intervals are compensated by the least squares method and BP neural network improved by the genetic simulated annealing algorithm(GSA-BP).The compensation function model is established,and the efficiency of the GSA-BP algorithm is tested.The measured data is used to verify the compensation effect.The experimental results show that the range and error of the improved automatic weather signal simulator,such as temperature,humidity,air pressure,evaporation and radiation,can meet the industry standard QX/T 346-2016 "Automatic Weather Station Signal Simulator",which has strong environmental temperature and humidity adaptability.Multiple linear regression least square method and GSA-BP are combined to compensate the error of humidity sensor.This method effectively reduces the influence of ambient temperature on the measurement results.These studies provide technical means and theoretical basis for improving the observation quality of automatic weather stations.