An Optimization Model for Process Traceability in Case-Based Reasoning Based on Ontology and the Genetic Algorithm

A traceability system can quickly and accurately query product supply chain circulation information. It is difficult to obtain processing information and achieve sharing due to the big data, long duration, large amount of equipment and scattered and easily lost processing information. In tracing the application of the case-based reasoning (CBR) to product processing, we can solve the problem of establishing an accurate and complete mathematical model between the processing information and product quality defects by finding a target record in a large amount of historical data. The traceability model of ontology and CBR uses the web ontology language (OWL) for case representation to simplify the CBR framework for big data, classifies process data and standardizes information storage. A joint optimization algorithm based on the genetic algorithm (GA) and CBR is proposed to establish an optimization model of process traceability, which is applied to case retrieval and reuse. This algorithm optimizes the genetic operator by combining the optimal preservation strategy and roulette selection method and uses an exponential-scale transformation method to stretch the fitness function. The experiments show that the optimized traceability model can infer information from garbled codes, wrong codes and missing messages to quickly determine the problematic products, thus effectively improving the traceability accuracy of product processing.