Designing Of Noxious Matter Data Processing System Estimating Missing Data In An Animal Barn

The livestock and poultry breeding of China has developed rapidly since reform and opening up.However,the production level of livestock and poultry breeding in our country is relatively backward compared with the developed countries.To increase the efficiency and output of livestock and poultry breeding,in recent years,animal industry has been growing as scale and factory in our country,resulting in various noxious gases assembling and volatilizing,which has seriously polluted our environment.Therefore,agriculture, environmental protection departments and academic community all over the world pay attention to research on emission of noxious matter in the animal barn.In order to calculate emission rate of noxious matter in the animal barn,it's necessary to collect large amount of raw data from the animal barn and process them,such as data analysis,statistic summary. Because there are large amount of data and many types of data in an animal barn,it takes us much time to complete calculation.When calculating or processing these data by hand,we easily make a mistake.So the design and implementation of noxious matter data processing system in an animal barn is of great importance.After the paper analyzes deeply all factors received means in environment monitoring system of livestock factory and data processing theory of nocuous matter,developed under Visual BASIC 6.0 development platform,the data processing system for noxious matter in an animal barn includes data input,data processing and computing,graphic drawing etc.It can calculate the emission rate of noxious matter automatically according to category and situation users choose,show environmental monitoring historical data and emission data from an animal barn by graph or table.And the invalid data is marked with different color points on the curve.The system makes it more convenient for farmers to observe environmental factors impacting on animal growth,and for technicians to check cause of environmental pollution.For the missing data due to the sensor faults,considering fully these factors such as time,space,temperature,humidity,wind speed etc,the author advances application of neural network based on BP algorithm to approximate missing data.A three-layer feedforward network is first built according to the characteristics of noxious gas,and a module based on BP algorithm to estimate noxious gas concentration is compiled with VB secondly.This module can accurately estimate missing data,run steadily,be used easily, train sample data,and map various related curves.After this module is tested with experimental data,the result shows that the biggest relative error is 5.99 percent,the smallest relative error is 0.07 percent and the average relative error is 3.17 percent. Compared with the traditional linear interpolation method,the method considers fully the nonlinear relation among noxious gas concentration and time,space,environmental factors, such as temperature,humidity,wind speed etc.Estimating effect is good and offers assurance precision in calculation of noxious gas emission rate,so the technicians can obtain accurately emission amount of noxious gas,then control the noxious gas beyond normal.