Title:
LEARNING DATA COLLECTION METHOD FOR CONTROLLING AIR-CONDITIONING DEVICE, AND AIR-CONDITIONING SYSTEM
Document Type and Number:
WIPO Patent Application WO/2023/176152
Kind Code:
A1
Abstract:
Provided is a learning data collection method for controlling an air-conditioning device with which it is possible to collect learning data for a prediction model in a region to be controlled reliably in a short time without placing a burden on the device. This learning data collection method includes the steps of region setting, learning start point setting, state point movement, and data collection. In the region setting step, a region R1 in which a prediction model is produced is set on a psychrometric chart. In the learning start point setting step, a plurality of learning start points S1 to S9 are set in the region R1 as start points of random excitation. In the state point movement step, an air-conditioned air state point K1 at the same position as an outside air state point G1 is moved to the first learning start point S1. At this time, temperature and humidity control is performed by operating air-conditioning facilities 22 to 25. In the data collection step, random excitation for the air-conditioning facilities 22 to 25 is performed while the air-conditioned air state point K1 is moved between the learning start points, and learning data is collected.
Inventors:
MURAMATSU SATOSHI (JP)
FUKATSU TAKUYA (JP)
HISHINUMA TORU (JP)
FUKATSU TAKUYA (JP)
HISHINUMA TORU (JP)
Application Number:
PCT/JP2023/002014
Publication Date:
September 21, 2023
Filing Date:
January 24, 2023
Export Citation:
Assignee:
TRINITY IND CORP (JP)
PROXIMA TECH INC (JP)
PROXIMA TECH INC (JP)
International Classes:
B05D3/00; B05B16/40; F24F11/49; F24F11/64; F24F110/12; F24F110/22
Foreign References:
| JP3993358B2 | 2007-10-17 | |||
| JPH11132532A | 1999-05-21 | |||
| JPH09244708A | 1997-09-19 |
Other References:
ANONYMOUS: "Annual energy bill reduced by 3 million yen! Succeeded in significantly reducing the energy consumption of air conditioning equipment using our unique control AI "Smart MPC"! Since learning is possible with a small amount of data, it can be introduced in one day, and performance improves rapidly wit", PROXIMA TECHNOLOGY CO., LTD., 28 February 2022 (2022-02-28), pages 1 - 4, XP093092749, Retrieved from the Internet [retrieved on 20231018]
Attorney, Agent or Firm:
ATSUMI Hisahiko (JP)
Download PDF: