Quality and safety traceable information mining in wheat flour processing based on risk matrix and flow analysis method

Traceability information in wheat flour processing is complex because of various sources of raw and auxiliary materials. Collecting and confirmation of the information in existed method depends on developer's experience or information availability. The lack of in-depth mining and analysis has affected the efficiency of information collection and the continuous application of traceability. The study is based on the data of 15 518 batches of wheat flour sampled by the former State Food and Drug Administration from January 1, 2014 to March 31, 2018. The geographical distribution of the unqualified products and the causes of the unqualified products were analyzed. In the 74 batches of unqualified products, 45 batches were caused by deoxynivalenol. The result from benzoperoxide is 12 batches. The other substances such as aluminum, lead, cadmium, benzopyrone and dehydroacetic acid, is respectively 10 batches, 2 batches, 2 batches, 2 batches and 1 batch. Therefore, the ranking of seven main factors causing wheat flour quality and safety were deoxynivalenol, benzoperoxide, aluminum, lead/cadmium/benzopyrone and dehydroacetic acid. The ranking was acted as the probability of hazard occurrence. According to different international measurement indicators, the risk was assigned to 7 kinds of risks by using different hazard indicators such as LD50 value and carcinogenicity. The higher assessment value was, the higher risk harm degree. In the 7 substances, benzopyrone had the highest risk harm degree and benzoperoxide had the lowest risk harm degree. A two-dimensional risk matrix was constructed based on the probability of hazard occurrence and the severity of hazard occurrence. According to the risk identification, the risk of deoxynivalenol was the highest, which was level 4; the risk of benzopyrone was level 3; the risk of benzoperoxide, aluminum, lead and cadmium was the second level; the risk of dehydroacetic acid was the lowest. Through wheat processing flow analysis and hazard substances orientating, the key nodes of wheat flour traceability was identified, which included dynamic links such as raw material storage, raw grain storage and material addition, as well as static status of equipment and packaging materials feature. Based on the risk analysis result, the wheat flour processing management and traceability system was designed and developed with the batches coding and association rules of raw batches, processing batches and traceability batches, and the key traceable information. The system includes basic information management, raw material storage information collection, raw material storage information management, processing information collection, testing information collection, bar code generation and printing, query and statistics and system management functions. This paper proposes a key traceability information mining method based on risk matrix method and process analysis in wheat flour processing, and is applied to the development of actual traceability system. Compared with existing systems, the traceability of the processing process can be realized by correlating the different batches of raw materials, processing batches and finished products with this system. The traceability accuracy can be improved. When quality and safety problems occur, the system is more conducive to the realization of accurate recall. On the other hand, the system is improved by combining the traceability and hierarchical information of mining, which can improve the efficiency and pertinence of information acquisition in the process of wheat flour processing and facilitate the application of the system in wheat flour processing enterprises. © 2019, Editorial Department of the Transactions of the Chinese Society of Agricultural Engineering. All right reserved.