| Since the 1980 s,with the development of society,especially urbanization,grain food consumption per capita and animal-derived foods consumption have undergone a historic turn in China.The food structure has undergone a transformation of the times.The proportion of grains as the "staple food" in food has been declining,and the per capita consumption of animal-derived foods such as milk has continued to grow.In order to adapt to the changes in the diet structure of the people in China,the agricultural structure needs to be adjusted and transformed accordingly.In recent years,the CPC Central Committee and the State Council have issued a series of policies to support the transformation and optimization of China’s agricultural structure and the development of smart agriculture."The Agro-grassland is a new type of agricultural system that takes into account ecology and production,food and feed,adapts to the development of the times,meets the overall planning of urban and rural areas,and faces the integration of global economy.It is also a great transformation for China to solve the development problems and optimize the agricultural structure." Dairy industry,as an important part of animal production layer and post biological production layer in four production layers of the Agro-grassland,connects plant production layer and pre plant production layer at the same time,is a strategic industry of coordinated development of primary,secondary and tertiary industries.As a "locomotive" in the development of grassland agriculture,dairy industry plays a key role.In recent years,with the rapid development of information technology,big data has become a new strategic resource,which,like material and energy,has become an important driving force for the transformation and upgrading of modern agriculture.Under the guidance of Agro-grassland theory,this study takes dairy industry as the starting point to carry out the preliminary exploratory research on the construction of Agro-grassland big data platform,using the ETL(extract transform load)architecture of big data and Apache Flink,the big data processing engine,big data storage and analysis technologies have built a big data platform system for dairy industry,which has been deployed in 197 dairy farms in 21 provinces of China.Based on the big data platform system for dairy industry,the status of key production performance of dairy cows in China has been studied.The main achievements of the research are as follow:1)According to the construction requirements of dairy big data platform system,a dairy real-time data platform is designed and constructed based on the characteristics of dairy big data,which realizes the real-time and efficient collection and management of dairy data.2)Based on the biological characteristics and production management practice of dairy cows,the evaluation model of key production performance,the early warning model of key indicators and the analysis and decision-making model of dairy production management are constructed.The model is integrated between the big data platform systems in a fully integrated way.The integrated system of "evaluation-early warning-decision making" for dairy production is realized to support the application function of dairy big data platform system.3)Using Java SDK 1.8 and Apache Flink developed a big data platform system for dairy industry is constructed and developed,including cow production management system,precise feeding management system based on material flow,intelligent identification management system,intelligent analysis and decision-making system,and big data visualization system.It can be deployed in a variety of cluster servers and can realize real-time and fast calculation of different scale data,forming a comprehensive technology of dairy big data platform system construction.4)The dairy big data platform system has been deployed and applied in 197 dairy farms in 21 provinces of China.Based on this platform,it helps dairy farms to establish a standardized dynamic cow file and production management process,assists dairy farms managers to make scientific decisions,and improves the production efficiency(increased by 2-3 times)and sustainable profitability(more than 20%)of dairy farms.5)Based on the grassland comprehensive and sequential classification system(CSCS),115 dairy farms with accumulated data more than one year were selected from 197 dairy farms applied in the dairy big data platform system,with a total of 363062 cows,including 203341 mature cows(Including 173229 milking cows),157351 heifers.The key performance data of dairy cows in China were analyzed by descriptive statistical analysis,the key production performance parameters of dairy cows in China were obtained,including the quartile range of 21 day pregnant risk of adult cows,the mean value of which was 18.8%,the median value of which was 16~21%;the quartile range of cow death rate of mature cows,the mean value of which was 30.5%,the median value of which was 30.8%;the death rate of 60 days postnatal cows,the median value of which was 19.4%~39.7%,the quartile range of rate is 5.3%~10.8%,with an average and median of 7.9%;the quartile range of 21 day pregnant risk of heifers is 15.5%~27%,with an average of 21.7%,with a median of 20%;the quartile range of age ≤ 60 days Cull Rate by Age≤60 days is 4.6%~19.7%,with an average of 15.2%,with a median of 8.2%,Avg.Milk/Day of Mature Cow,the quartile range is 23~28.4kg,with an average of 25.4kg and a median of 25.1kg;the quartile range of Avg.Milk/Day of Milking Cow is 27.2~32.4kg,with an average of 30 kg and a median of 30 kg.In this study,the big data platform of dairy industry was constructed by using big data technology in the framework of Agro-grassland.Through the application research,the evaluation system and standard of key production performance of dairy were initially established in China,which will provide important data reference for the production management of dairy farms and the formulation and implementation of industrial policies in China,as well as for the research of biological dairy industry and the coupling of plant production layer and animal production layer interface by scientific basis. |
PDF Full Text Request