The Key Techniques For Dairy Farm Digital Management Based On Internet Of Things

Compared with the developed countries, China’s dairy industry is faced with precisionfeeding is not systematized, low body measuring efficiency, heat detection is not timely,and other issues, which greatly limit the domestic dairy industry’s survival anddevelopment space. This thesis aims to through research to analyze the current domesticdairy farm’s precision feeding series problems, explore the key technology of dairy digitalfarming Internet of things, solve the technical problem of the cow body measurementbased on digital image processing, achieved the intelligent automatic determination of cowconformation and body size, and realized the real-time and automatic feedback of dairyinformation, improved the dairy information monitoring efforts,. The research results ofthis paper have important application value and economic benefits for effectively improvecow’s digital fine breeding levels and promote the development of the dairy industry.This paper mainly completed the following research work:(1) Aiming at the problem of milk cow industry management level is low, introducethe “internet of things” concept to the dairy precision feeding, proposed a tree networkstructure based dairy internet of things system, build a ZigBee-based dairy internet ofthings system. According to the characteristics of the dairy precision feeding, designed theinformation databases related to dairy environmental parameters and cows individualparameters; applied LabWindows/CVI2012independently developed the monitoringcenter software for C/S system; using DataSocket communication mechanism based onTCP/IP protocol to achieve the real-time remote monitoring and transmitting of the data;applied the Z-Stack-CC2530-2.3.3-1.4.0protocol stack to developed the WSN networknodes software, designed the coordinator node, routing nodes and terminal nodesapplication software; comprehensive applied the radio frequency identification (RFID)technology, wireless sensor networks (WSN) technology, digital image processingtechnology, virtual instrument technology to achieve the all-weather real-time monitoringof the dairy farm information.(2) Using digital image processing technology, achieved the automatic measurementof body size cows, resolved the problem of low efficiency of traditional body size measurements and poor extraction accuracy of body size image processing. For the cowsblack and white image have the problems of unevenness of illumination, the background ismore complex and includes noise etc., using median filter, restriction contrast andhistogram equalization and other methods for image enhancement, using Prewitt operatorand Otsu method for image segmentation, the use of expansion, corrosion, small objectsremoved, external gradient computing and other morphological image processingtechniques to remove image interference and extract image edge, to achieve the effect ofoutstanding the cow’s body shape characteristics; proposed that using the target orientedparallel search algorithm to automatically identify cows body feature points and measurecows body size, compared with manual measurement the maximum absolute error ofmeasurement is less than0.8cm, and measuring time is shortened4-6times, effectivelyimprove the measurement accuracy and measurement efficiency.(3) Comprehensive application of the single-factor test, orthogonal, polynomial fitting,error compensation method, optimized the physical structure of the dairy internet of thingssystem. Applied the single factor test methods, determined the optimal mounting heightinterval node environment; applied the point-point communication packet loss rateorthogonal experiment to optimize installation height and installation distance between therouting and notes; applied the polynomial fitting method to establish an environmentalnode installation height and packet loss rate error model; conducted a hardware selectionand node settings from the perspective of interference, proposed that take the averagemethod and RMS method to compensation the measuring error, effectively inhibited themeasurement error, reducing the system loss rate.(4) Carried out the research about dispersion and credibility of the measurementresults of dairy internet of things system. According to the rules of the “Guide to theExpression of Uncertainty in Measurement”, analyzed the uncertainty source of dairy farminternet of the things system, obtained the main factors which restricted the measurementresults; Class B Uncertainty Evaluation Method was applied for the synthetic evaluationmethod of uncertainty of the dairy farm’s4system metrics namely temperature, humidity,the average content of ammonia and RMS value of exercise, achieve the quantitativecharacterization of the uncertainty degree of the system measurement result.