A study on ally searching in maintenance alliance based on relation web model

摘要

Aiming at the risk aversion of adverse selection and efficiency improvement in maintenance ally searching, relation web and Mobile Agent are introduced into the study of maintenance ally searching. A maintenance ally searching method based on relation web model is proposed. First, the method decomposes complicated plant maintenance tasks, and then by setting up ally Searching Sub-Agent, the relation web model is mapped as properties, plans, behaviours and interactions of the Searching Sub-Agent. Based on the relation categories, shortest path, reliability and other definitions in relation to web model, a quantified evaluation method for acquaintance reliability and capability in maintenance alliance and ally searching strategy are designed. A solution for recommendation deadlock and dynamic management of Searching Sub-Agent is discussed. The ally searching system in maintenance alliance based on the above method is effectively applied in a plant overhaul of a partner petrochemical company.