In the second phase of the project, in addition to establishing a risk monitoring system for salmonella and many other farms in the poultry supply chain, we also need to establish a multidimensional big data research system for poultry farms. Through the establishment of a quantitative research system for farms, the multi-dimensional feature analysis of the farms is realized, and the risk assessment and risk early warning of the farms are quantified by means of big data. From the original single disease warning to the breeding chain and processing chain, the source of the risk can be found in advance to ensure the safety of meat and poultry adaptation. Then establish a set of farm-based meat and poultry traceability system based on blockchain to improve food safety, traceability, early warning and rapid response in terms of transparency.
In order to establish a multi-dimensional big data research system for poultry farms, the blockchain and risk assessment team led by Professor Yuxing Han carried out a series of tasks and tasks. We have carried out research on big data analysis, established research teams, and analyzed and researched big data on poultry. The team went to Jiangfeng Poultry Farm to learn about the production and breeding conditions of the farm, to understand the quantitative description indicators of the farm, and to accumulate rich experience for establishing a big data research system.
Figure 1 Team Visit Jiangfeng Farm
The team also established a research model on the historical data in the past, further analyzed the quantitative analysis of the big data model on poultry farming, and verified the validity of the model. Among them, we carried out big data analysis on Jiangfeng breeding farm samples, established a data analysis model, and obtained preliminary correlation feature analysis results.
Figure 2 Data Analysis Model
Among them, in the meat and poultry treatment process, it was effectively analyzed that the proportion of positive results of infection in the anal test, shaving, and pre-cooling was relatively small, and the proportion of positive results of infection in the opening process was a certain proportion. The positive results of infection in the washing process were the largest Proportion, greater influence. It can be seen from the water sample detection stage that the disinfection stage has a small effect on the positive result of the infection. The 9 o'clock point is that the water sample rinsing stage has a certain effect on the positive result of the infection. The cleaning stage has the largest positive effect on the infection. High risk of infection. Anal swabs, shaving, and pre-cooling links have low correlation with other results, and have little impact on their subsequent processes. Rinse wash links have the greatest correlation with positive infection results. As the main influencing link, open routs have a correlation with bleach wash in their subsequent links. Larger, the results will affect subsequent processes. The rinsing process has a certain influence on the positive result of the infection in the cleaning process, and the cleaning process has the most correlation with the positive result on the infection, which is the main influence link. The correlation between the two stages of cleaning is relatively large, and a positive result at 9 o'clock will affect the water sample at 11 o'clock.
Figure 3 Quantitative Analysis