Pre-warning analysis and application in traceability systems for food production supply chains

Production quality in the food production supply chain is studied in this paper. The deficiencies of quality monitoring that exist in traceability systems are analyzed. An abnormality diagnosis algorithm, pre-warning method and structure of pre-warning system are presented. Four abnormal data types in supply chain are analyzed, they are substandard abnormality, over-range abnormality, abnormal distribution and abnormal tendency. All the detection data of the whole supply chain are monitored timely and pre-warned. The production abnormality of the logistics unit is diagnosed and automatically warned, and the decision support information is given. A standard hierarchy evaluation indicator system for abnormalities is developed in this paper. A mathematical model for abnormality detection is developed by combining radial base function (RBF) neural network, fuzzy control, and statistical analysis methods. This model is used in detecting and recognizing different types of abnormalities in the food production supply chain, especially hidden problems. The simulation results show that the proposed pre-warning system can effectively identify abnormal data types, and accurately determine whether a warning should be issued, depending on the warning level when an abnormality is detected by the system. The pre-warning system for food production supply chain performs well and effectively.