摘要
目前RBI软件因扩展性不足不能满足RBI标准不断更新的需求,它亦不能实现根据实际情况调整风险预测模型,并且在多用户并发进行RBI风险计算时效率低下.针对以上问题,设计并实现了基于C/S的RBI专家系统.该系统以RBI中通用的API581标准为依据构建知识库,采用XML存储知识,使其达到了RBI标准动态更新以及在RBI实施中根据实际情况及时调整风险预测模型的要求.采用层次式知识库结构、应用数据缓存和基于线程池的任务管理器,使系统达到了在多用户并发时具有快速响应能力的目的.
To satisfy the continuous updating RBI standards and the requirement of adjusting RBI model according to practical demands,the expert system in the RBI was applied. Using XML to store knowledge and updating the knowledge through user-interface,the system is extensible and flexible in adjusting RBI theoretical prediction model. Regarding the complexity and parallelism of RBI risk calculation,the RBI expert system uses stratified knowledge base structure,application data buffering and task manager based on threads pool to develop a highly efficient RBI expert system.
基金
科技部科技型中小企业技术创新基金(04C26223400621)资助
