National Field Office for Loran Data Support Hybrid Expert System/Artificial Neural System for Loran Area Monitors

The National Field Office for Loran Data Support (NFOI-DS) laid out a plan to establish and operate a system that collects. monitors, processes, and distributes data ucts P for the Federal Aviation Administration for Loran’s use as a tandard Instmment Approach Pmcedure (SIAP) navigational aid. The Early Implementation Project (EIP) collected data and demonstrated the feasibility of the system. As the NFOLDS S~S~CUI goes into effect, the FAA is installing 196 monitors across the nation which send data to central NFXXDS proce&ng in Oklahoma City, OK The EIP gram demonstrated that NKLDS pemonnel obtained reliable time dif erence (TD) data products when appropriate data filtering p” techniques were used. This iabor-intensive process, however, requires more experience analyzing Loran data and expert& pmcessing it than most operators will have acquired. This paper describes an approach using Artifiiial Intelligence (AI) techmques for developing an advisor program to help NFOLDS operators develop reliable TD corrections. The Expert System for Lmm Monitoring (EXSLAM) is a hybrid of AI techniques. It combines an Artificial Neural System (ANS) front-end processor with an expert system approach. The ANS will be used as an anomaly detector to identify known anomalies with the Loran Area Monitor (LAM) data as well as novel cases. The expert system will diagnose and recommend courses of action for NFOLDS operators pmcessmg TD data conecdons. The expert system approach preserves the knowledge of the Volpe National Transportation System Center (VNTSC) experts, commumcates that knowledge to less experienced operators, and allows them to apply that knowledge for more consistent and higher quality performance, assuring thequality of the Loran TD corrections