【作者】 郗伟东；

【导师】 孙永海；

【作者基本信息】 吉林大学 ， 农业电气化与自动化， 2009， 博士

【摘要】 建立科学的生猪生产与消费链安全评价及预警系统对其预警决策和用户控制有毒有害物质所致风险具有重要意义。目前,我国以及欧美等国在宏观层面上广泛展开食品风险评估,但多是基于单一有毒有害物质的理化危害后果以及质量保障规制进行的。国内外对食品,尤其是对动物性食品生产过程隐含风险的微观评价研究较少。本文在理论上深入研究了生猪生产与消费链风险形成机理和风险结构,提出了风险系统的“2-3结构”模式。基于RFS理论并利用MATLAB提取了抽样数据所隐含的生猪生产、居民消费等决策知识规则,验证了风险的“2-3结构”模式;根据该模式构造了“生猪散养”、“猪肉消费”安全评价框架与指标系统,设计了风险与危险预警指标系统。基于最小二乘法提出了修正G1-法,并对生猪散养、猪肉消费系统构成模块在评价中的作用进行了分析,通过优化专家数据给出了相应的因子水平,对区域安全评价与预警提出了“隶属度频率法”。利用吉林地区的抽样数据验证了文中提出评价指标体系与方法的有效性。利用Java平台技术与MVC模式研制了生猪生产与消费链安全评价及预警的智能评价系统,为提高生猪生产与消费链安全、有效控制风险以及风险分析提供了支持工具。更多还原

【Abstract】 Pigs, on the one hand are important human Food-Producing Animals, on the other hand are important carriers of zoonose diseases and harmful substances. Assessing the risk status of pig industrial chain scientifically is significant for controlling and reducing the risk that meat posed to human, sending out risk warning timely, protecting the health of residents and social public health security. At present, Europe,the United States and other Western countries carry out risk assessment of food at the macro level extensively, but it is based on the consequences of physical and chemical hazards of a single toxic and harmful substance, as well as the quality assurance regulation; China’s Food Risk Assessment is similar to Europe and the United States. There are few studies on the formation of food-derived risks especially Food-Producing Animal at home and abroad, as well as the assessment that every main body of pig industrial chain cause risk status. Therefore, in this paper, it studies formation mechanism of the pig industrial chain’s risk in depth, puts forward a risk framework in theory, designs risk evaluation index system of pig industrial chain, combining with the actual situation of our country, accordingly, it develops a safety assessment and warning computer-aided system. This paper does exploration and research in depth as follows:1) It clarifies relevant existing confusion of the concept of risk assessment; clear the essential meaning of the risk of pig industrial chain. In theory, it studies formation mechanism of the pig industrial chain’s risk in detail, puts forward the risk’s three-dimensional structure and the‘2-3 structure’of its formation. Risk-response capacity, toxic and hazardous materials awareness, control credit quality of toxic and hazardous substances make up three-dimensional space of a pig industrial chain risk. Risk tree based on‘2-3 structure’indicates that, in theory, the lack of risk control, leads the possibility of risk overflow reaching 87.5%. Risk proliferation and the‘2-3 structure’of risk control, describe formation mechanism of the pig industrial chain’s risk completely.2) Study the formation of pig industrial chain in detail, and mine data from the process of live pig small-scale farming sample using FRS (Rough Fuzzy Sets) theory and MATLAB tools, then extract rules on decision-making knowledge of the process of live pig small-scale farming; The rules verify three-dimensional structure and its formation mechanism of pig industrial chain’s risk. The rules extracted from decision-making knowledge divide the attribution of live pigs’production process into decision-making dependent on (conditions) and non-reliance, both reveal the risk of live pig small-scale farming from different angles, while deduce risk of pig industrial chain based on system isomorphism. Based on the‘2-3 structure’of risk,“process– component– risk”framework on pig industrial chain is set up, in this framework model, 45 modules exist in model system of the risk of pig industrial chain in theory. Model based on risk behavior is put forward using FD (Fishbone Diagram) and AHP (Analytic Hierarchy Process) methods, which is divided into system level and individual level; Establishing the "component-dimensional - target-type" modular system for the individual level of the pig small-scale farming, among all of impact factors analyzed in every module, the ability to cope with in risk behavior has impact on the system, reaching 44.4%, equipment and facilities (including technology) in system components has a greater impact on the system, reaching 61.5%.3) Using the findings of the conceptual model of risk behavior, it puts forward four principles building a pig industrial chain security evaluation framework which should be followed, the safety assessment framework of the pig industrial chain and technical routes of safety evaluation; it puts forward the principle of safety evaluation index structure, combining with the structural of risk to design the safety evaluation index system of the pig small-scale farming. Evaluation index system is divided into risk assessment and danger assessment; The former includes the ability to cope with the risks, science and technology perceived risk, three sub-PC(H), PD(H), PM(H) of credit quality risk assessment,the various subsystems is divided into property layers, the basic layer, object layer etc.3 structures; Accordingly, combining with GB, HACCP, GMP, GAP, SSOP it designs the safety evaluation indicator system of the pig small-scale farming such asPCPI、Pthe system assesses risk of while reveals the state of risk distribution,making up for shortcomings of traditional safety evaluation.In allusion to the distinction between risk and danger combining with national standards on limits of toxic and hazardous substances, it puts forward deterministic type, fuzzy type as well as comprehensive assessment type of the region interpretation guideline of safety assessment of pig industrial chain.4) It studyies calibration of various indicators, according to the nature of safety assessment indicator, it puts forward fuzzy number calibration method;Indicators’synthesis selects fuzzy algorithm, and puts forward key indicator method as supplementary. It puts forward modify G1- method using least square method, and extract assessment indicator weight of pig breeding on expert sample data. Combining with safety assessment grading of every area, it puts forward 5 gradings for safety status of pig industrial chain giving triangular model algorithm of each grading. About state assessment of the regional risk,it puts forward stratified(group) machinery technology based on traditional stratified sampling, and puts forward“membership frequency method”to estimate overall safety status of region. Using this method it carries out an assessment of Jilin City, and verifies the validity of the method.5) For the end of pig industrial chain, according to sample data from consumer of pork, using Probabilistic Rough Sets theory it extracts rules on decision-making knowledge of pork’s consumption, based on this rule it designs safety assessment indicator system of pork consumption CCCI ,CI facing two-dimensional framework of risk,and analyzes the extent they impact on the system using expert sample data, showing that the impact factor of CI is 58.3%, while the impact of individual attributes difference module on I reaches 61.5%. 6) Construct warning model and technology of pig industrial chain with the use of safety assessment framework. Compared with traditional warning model this model makes up warning type, adds a time dimension to warning domain, proposes just-in-time warning, focus tracking warning and periodic warning. Propose warning technology of key indicator tracking, balanced warning technology and mixed warning technology firstly; in allusion to characteristics of pig industrial chain’s danger, import control charts, smooth movement and forecast passing rate as warning technology to reveal risk level. Corresponding to pig production safety assessment, it designates pig-producing chain four grades.7)In order to make producers, consumers and quality & safety regulators in the pig industrial chain use safety assessment and warning system designed in this paper conveniently, it develops a corresponding computer-aided intelligent system. In order to apply to Internet environment and enable all kinds of users access the system through network easily,it chooses MVC model and Java tools as system development,and carried out by MyEclipse+MySQL+Tomcat platform. In order to reduce development cost and system cost, the system uses B/S structure, using JSP+Servlet+JavaBean to achieve the whole system,users can achieve system call through any Browser. This system can be accessible to platform transplant.With the use of Java technology and MyEclipse+MySQL+Tomcat the system functional expansion convenient, and maintenance easy.In short, safety assessment of pig industrial chain is a very fuzzy, with large subject span and scientific job because of safety’s ambiguity, universality of risk, the importance of safety assessment. It designs safety assessment structure and warning system based on risk structure, and achieves intelligent safety assessment and warning through computer technology. The risk analysis model, risk structure framework, safety assessment system, in this paper, compared to the previous assessment model has the following advantages:1) Complete structured and systematic. The safety assessment and warning model proposed in this paper possesses structured risk analysis, structured risk model, structured safety assessment framework, and structured risk warning. Therefore, it can improve the system through the structural analysis of system, thus improves the scientificity of safety assessment and warning. Assessment model proposed is based on object’s level structure, so the safety assessment can embed into the different levels within object along the hierarchy structure to interpret systemic reasons of risk occurrence, and then makes warning point clear and specific.2) Strong objectivity. The design of safety assessment indicators attaches to each specific attributes of object thereby makes the indicators express various risk sources attached to object objectively.3) Adaptable. It selects fuzzy quantitative assessment indicators and methods, making the process and results of assessment fit for the relativity of safety assessment problem, and at the same time it uses the knowledge and experience of experts effectively. The risk analysis model, risk structure model, safety assessment model, risk warning model as well as the system realization proposed in this paper, can be applied to food production and consumption system, environment analysis and protection systems, chemical system, agriculture and animal husbandry and so on.更多还原