Inverse Neural-Network-Based Automated Antenna CAD Tool Navy SBIR FY2012.2 Sol No.: Navy SBIR FY2012.2 Topic No.: N122-119 Topic Title: Inverse Neural-Network-Based Automated Antenna CAD Tool Proposal No.: N122-119-0087 Firm: Physical Optics Corporation Products and Engineering Division 1845 West 205th Street Torrance, California 90501 Contact: Shahzad Khalid Phone: (310) 320-3088 Web Site: www.poc.com Abstract: To address the Navy's need for an antenna design tool that will rapidly produce CAD models of commercial off-the-shelf (COTS) aircraft antennas, Physical Optics Corporation (POC) proposes to develop a new Inverse Neural-Network-based Automated Antenna CAD Tool (INNCAD). The innovation in algorithms based on trained knowledge-based inverse neural networks (INN) and an innovative antenna knowledge database (ADB) produce rapid automated design CAD models of COTS aircraft antennas that are ready to use in the Navy's CEM simulation tools to evaluate performance when installed. This device offers a methodology that uses COTS antenna performance and physical data to yield complete antenna designs and their physical descriptions, directly addressing Navy requirements. In Phase I, POC will develop a complete system architecture with details of the graphical user interface (GUI) and demonstrate the ability to produce a valid antenna design from top-level specifications for at least two types of antennas. In Phase II, POC will develop INNCAD into a complete engineering tool, which will include a GUI, interface to the Navy's CEM codes, and a complete version of the ADB and INN to design a comprehensive set of aircraft antennas. Benefits: POC's INNCAD can support significant commercial applications, including radar, wireless communications, and navigation. The INNCAD can start with incomplete manufacturer data and obtain CEM-ready CAD models from INNCAD in minutes, representing a much more efficient product for both industry and government. Return