Information processing device, method for generating an inference model, method for generating training data, inference model generation program, and training data generation program

By concatenating and normalizing time-series data from multiple sources, the device creates a versatile inference model that enhances accuracy and applicability across different facilities, addressing the limitations of single-source models.

JP7875057B2Active Publication Date: 2026-06-17CANADEVIA CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
CANADEVIA CO LTD
Filing Date
2022-07-06
Publication Date
2026-06-17

AI Technical Summary

Technical Problem

Conventional inference models generated using plant data from a single source lack generality and have low accuracy when applied to data from another plant.

Method used

An information processing device that concatenates multiple time-series data from different sources to generate a single pseudo-time series data, applies standardization or normalization processes, and uses machine learning to create a highly versatile inference model.

Benefits of technology

Generates an inference model that is more versatile and accurate across multiple facilities, allowing for immediate application in new facilities without requiring extensive data collection and reducing the time needed to generate training data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007875057000001
    Figure 0007875057000001
  • Figure 0007875057000002
    Figure 0007875057000002
  • Figure 0007875057000003
    Figure 0007875057000003
Patent Text Reader

Abstract

To enable generation of an inference model with high versatility.SOLUTION: An information processing device (2) comprises: a data connection unit (204) which connects a plurality of pieces of time-series data based on data collected by each of a plurality of facilities to generate pseudo time series data; a teacher data generation unit (205) which applies standardization processing to the pseudo time series data to produce teacher data; and a learning unit (206) which generates an inference model by machine learning using the teacher data.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art