National Institute of Standards & Technology, Acquisition & Assistance Div., Bldg. 301, Rm B117, Gaithersburg, MD 20899

D -- DEVELOPMENT OF INTELLIGENT SYSTEM FOR HEAT EXCHANGER DESIGN SOL 53SBNB860066 DUE 060598 POC Contracting Officer, Diane Loeb, 301 975-6399 WEB: NIST Contracts Homepage, http://www.nist.gov/admin/od/contract/contract.htm. E-MAIL: NIST Contracts Office, Contract@nist.gov. The National Institute of Standards and Technology (NIST), Gaithersburg, MD, issues this request for quotes (RFQ) to develop an intelligent system for heat exchanger design (ISHED). ISHED shall optimize refrigerant circuitry of the finned- type heat exchanger to obtain the maximum capacity for specified general heat exchanger parameters. Usually, heat exchanger manufacturers produce established lines of heat exchangers with the same tube material, diameter, spacing and a preferred fin design. For a given application, the designer selects the number of tubes in a depth row, number of depth rows, and the length of the tube, which determine the external dimensions of the heat exchanger, according to the company's established practice or the heat exchanger application. The designer must define refrigerant circuitry to indicate how to connect tubes with each other to route the refrigerant through the coil assembly. The optimized refrigerant circuitry shall provide maximum coil capacity for all other parameters being fixed. This project shall develop an ISHED for optimizing refrigerant circuitry. The parties will determine coil capacity for this project employing a NIST tube-by-tube evaporator simulation model. The optimization of a heat exchanger presents complex problems because of a large number of possible arrangements of the tubes and their connections. NIST anticipates that the development of ISHED shall incorporate innovative ideas that integrate evolutionary learning and symbolic learning methods. The contractor shall perform various services to develop ISHED, including the following tasks: Line item number (LIN) 0001: Develop an overall design of ISHED and Evolutionary Learning Module (ELM). This task shall develop an overall design of ISHED, its modules, and the way they shall interact, develop and test the Evolutionary Learning Module for ISHED, and develop suitable representation of designs and evolutionary operators, one job. LIN 0002: Develop and test the Symbolic Learning Module (SLM) for ISHED. The SLM shall automatically invent new attributes that are more relevant to the problem at hand than original attributes, learn composite rules involving both logical and arithmetic operators, reason with multitype attributes, and optimize rules according to several user-specified criteria to reflect the needs of the application. Any learned rules shall be easy to understand and interpret, one job. LIN 0003: Develop the Control Module (CM), which shall coordinate execute and synergistically integrate the SLM, ELM, and the heat exchanger simulator. CM shall also support the conduct of experiments and the evaluation of results, one job. LIN 0004: Implement the prototype of ISHED, includes implementing, experimenting, and evaluating the prototype of ISHED. Integrate all developed modules into one system interacting with the evaporator simulation model to formulate an optimized circuitry design, one job. LIN 0005: Develop the final version of ISHED, involves preparing ISHED documentation, including a user manual, describing the experimental results, and ISHED applicability, one job. The successful contractor shall employ key personnel holding Ph.D.'s in relevant fields and displaying significant experience in evolutionary computation and symbolic machine learning and some experience in heat exchanger design. The successful contractor shall submit all deliverables and documents, FOB destination, Gaithersburg, MD. Deliverables include the executables and source code of custom-developed software operable on Windows 95 or higher version, with a general program description, program logic, and comment statements to facilitate future modifications. The contractor shall accomplish all services within 12 months after issuance of the purchase order. NIST anticipates issuance of a firm-fixed price purchase order on August 8, 1998. The cost range falls between $50,000 and $100,000. Significant evaluation factors include: A. PAST EXPERIENCE: Quoters shall submit evidence of past experience with similar research efforts. Such evidence shall be clear, factual, and include resumes of key personnel, titles of contracts, grants, or projects in closely related areas, and list all peer review articles and reports in the field of artificial intelligence; B. DEGREE TO WHICH QUOTE MEETS OR EXCEEDS REQUIREMENT: Quoters shall submit a well-organized, clear and concise quote with sufficient information indicating understanding of the requirement. Quoters shall address each element of the technical requirement found in the performance work statement; C. SCHEDULE COMPLIANCE: Quoters shall submit a schedule for completion for each task above described, indicating efforts to insure quality in service and products quoted. NIST will consider a quotation from all responsible sources. NIST will accept written requests for this solicitation through FAX 301.963.7732, ATTN: Diane Loeb; e-mail, diane.loeb@nist.gov., or regular US mail; or the NIST Contracts Group External Website: http://www.nist.gov/admin/od/contract/contract.htm. See Numbered Note 26. (0131)

Loren Data Corp. http://www.ld.com (SYN# 0017 19980513\D-0001.SOL)