Research On Key Technologies In Food Safety Early Warning System

As food safety problem constantly occurs, the serious food safety accidents and contingencies of public hygiene security challenge the food safety monitoring and controlling capabilities of a state or a region. The food pollution derived from the new technology, new process, and new material used in agriculture and food industries is the main problem that developed countries are confronting with; developing countries are up against both the chemical pollution and biological pollution; what less developed countries faced with is the shortage of food supply. With a vast territory, our country is developing unevenly between regions. Meanwhile the issues of ecological environment and the sustainable development of resources also exist. The overexploitation of natural resources, together with the varying pollution of water, air, and soil caused by environmental deterioration, results in the hidden danger to food safety, which exposes the food safety problem.The usage of all kinds of food additives (leaching powder coloring agent、 sweetening agent、nitrite、preservative agent、flavoring agent, etc.) and expired food lays big hidden dangers to food safety. AQSIQ (General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China) has circulated the notices dozens of times, in which food additives, quality and microorganism are listed as the principle reasons to unqualified food. Nowadays food safety problem seems to be increasingly fierce. According to the analysis, it has following three features:(1) the unqualified food convers a wider range;(2) the unqualified food brings the harm to a deeper extent, which turns from food’s external health hazard to food’s internal safety hazard;(3) the methods of producing toxic matters are diversified,"deeper" and more obscure. The potential physical safety hazard is converted into chemical or biological one.At present, there are quite many researches on food safety controlling and early warning. Mostly they can be classified into three types, firstly the researches on management framework of food safety, then the researches on the tracking and tracing of food safety, and lastly the researches on the early warning system of food safety. Judging from the current researches at home and aboard, we find that so far the researches on food safety issue seldom refer to the study of food’s property itself. The majority of them is applied to relieve food safety problem from the aspects like sampling inspection, supervisory tracking, management and polices, while the analyses of the latent relevance between foods are lacking.Food system is a typical Grey System that partial information is clear and definite, but the other is not. This "incomplete information" is absolute. Most processed food has far more than one kind of additive, and the mutual reactions between different additives are complicated. Moreover there’s little known among the numerous factors result in food safety issue. Food safety early warning system is a sophisticated system, which not only has some intrinsic necessity, but also random contingency. The spot check of food is random and uncertain. Although the processing looks like a happenstance, some inherent laws that are unacknowledged in food safety problem itself still exist in it after the thorough analysis. This paper will focus on the two key technical issues that affect food safety. One is the additives during food productive process; the other is the assessing of microbial population in circulation link. The main contents are as follows.(1) Put forward the segmental prediction model for microorganism (ISPMM) to forecast the margin of shelf life.Microorganism is one of the dominant factors that cause food safety problem. Cold-chain logistics is the weakest link in the course of the whole logistics action, and it is the phase that most likely changes the amount of microorganism as well. In the light of the analysis of chief specific putrefying bacteria, we shall put forward the microorganism’s growth model at particular temperature. At the same time, we shall study the main environmental factors affecting the growth of microorganisms, construct the segmental prediction model for microorganism, and eventually realize the warning of ultimate shelf life of food against the changes of temperature, reducing the safety problems about food hygiene.(2) The automatic construction of food association model (FAM)Additive is another important factor of food safety issue. Based on the analysis of food intrinsic categories, food spot check categories, and food targets, we come up with the reticulate structure involving hierarchical features. Also we provide the corresponding formalized description model. This model stores both the explicit and implicit features of food resources. In order to realize the transfer of food problem degrees, we will quantify their severity degree via defining the attributes (PS and VA) of risk indicator on different nodes.(3) The simplification of search space There are hundreds and thousands of food resources, which makes it rather hard to analyze the relative correlations in this huge system. After analyzing the features of food relative correlation model, on the basis of Least Common Ancestors (LCA) concept, this paper proposes a simple but effective way to seek for the arithmetic of LCA. It fills the gap in seeking for arithmetic of LCA when several parent nodes are processed, so that the search space of relative correlations is effectively cut down.(4) The measurement algorithm of food relative correlation between different nodes.On the grounds of the explicit food attributes analysis and Grey Theory, we provide the related definitions and thermos, and put forward the measurement algorithm of food relative correlation between different nodes. According to the features of explicit food attributes, we construct a multi-attributes and multi-format sequence. Their property-format cover the format of ordered characters set, unordered characters set, interval number, mathematical set. We construct their calculation functions of correlation coefficient aimed at each one’s features. Taking the explicit attributes into consideration, we also construct the measurement algorithm of food relative correlation between different nodes by the integration of implicit food attributes.(5) The food safety early warning model.This paper studies the food safety issue from two aspects. They are the research on the assessing and early warning of microbial population in circulation link and the research about the warning to food relative correlation. Microbial population is applied to warning the margin of shelf life, and the food relative correlation is applied to warning the degree of risk about food categories. The confirmation of food relative correlation between different nodes cannot indicate the occurrence of food safety issue. In other words, even the relative correlation between two food nodes is quite high; food safety problem isn’t bound to happen unless the sequential severity degree of food problems is considered. Therefore, we need to forecast the issue by combining food relative correlation and the degree of risk about food categories in spot check.