Search

CN-122021861-A - Inference system, inference method, and learning completion model

CN122021861ACN 122021861 ACN122021861 ACN 122021861ACN-122021861-A

Abstract

The invention relates to an inference system, an inference method and a learning completion model. The inference system according to the present invention inputs input data into the learning completion model. The learning completion model is a model obtained by machine learning a learning data set including 1 st data including failure diagnosis data, which is data for diagnosing each of a plurality of types of failures related to the equipment, and 2 nd data including history data indicating a repair history of the failure in natural language. The input data is 1 st data concerning a fault detected in the device, i.e., a detected fault. The inference system obtains, as output data from the learning completion model, data expressed in natural language concerning at least one of a cause of occurrence of a detected failure and a repair method expressed by the input data.

Inventors

  • KATO SEI

Assignees

  • 丰田自动车株式会社

Dates

Publication Date
20260512
Application Date
20251023
Priority Date
20241112

Claims (5)

  1. 1. An inference system, characterized by: Inputting 1 st data concerning a detected failure, which is a failure detected in an apparatus, as input data into a learning completion model which is a model obtained by machine learning a learning data set including 1 st data including failure diagnosis data which is data in which each of a plurality of the failures related to the apparatus is diagnosed, and 2 nd data including history data representing a repair history of the failure in natural language, and Data expressed in natural language concerning at least one of a cause of occurrence of the detected failure and a repair method expressed by the input data is obtained as output data from the learning completion model.
  2. 2. The inference system of claim 1, wherein, The 1 st data includes data representing a state of the apparatus during a period before an occurrence time point of the fault represented by the fault diagnosis data.
  3. 3. The inference system as claimed in claim 1 or 2, characterized in that, The 2 nd data includes inspection data described in natural language as a result of inspecting the device.
  4. 4. An inference method, characterized by causing a computer to perform: Inputting 1 st data concerning a detected failure, which is a failure detected in an apparatus, as input data into a learning completion model which is a model obtained by machine learning a learning data set including 1 st data including failure diagnosis data which is data in which each of a plurality of the failures related to the apparatus is diagnosed, and 2 nd data including history data representing a repair history of the failure in natural language, and Data expressed in natural language concerning at least one of a cause of occurrence of the detected failure and a repair method expressed by the input data is obtained as output data from the learning completion model.
  5. 5. A learning completion model for causing a computer to function as input data and output data, characterized in that, The learning completion model is a model obtained by machine learning using a learning data set including 1 st data including failure diagnosis data which is data for diagnosing each of a plurality of types of failures related to the apparatus, and 2 nd data including history data representing a repair history for the failures in natural language, wherein the learning completion model is capable of inputting the 1 st data regarding a detected failure, which is a failure detected in the apparatus, as the input data after machine learning, and obtaining data represented in natural language as the output data regarding at least one of a cause of occurrence of the detected failure and a repair method represented by the input data.

Description

Inference system, inference method, and learning completion model Technical Field The invention relates to an inference system, an inference method and a learning completion model. Background Japanese patent laying-open No. 2010-241263 describes a fault diagnosis device for supporting reproduction of a fault based on a fault code in fault diagnosis of a vehicle. Disclosure of Invention However, in the technique described in japanese patent application laid-open No. 2010-241263, a fault corresponding to the fault code can be reproduced, and even if the situation can be reproduced, the cause of the fault or the repair method cannot be determined. Therefore, it is desired to develop a technique for easily and rapidly determining the cause of a fault or a repair method from fault diagnosis data for diagnosing the fault. The inference system according to the present invention performs the following processing: Inputting 1 st data concerning a detected failure, which is a failure detected in an apparatus, as input data into a learning completion model which is a model obtained by machine learning a learning data set including 1 st data including failure diagnosis data which is data in which each of a plurality of the failures related to the apparatus is diagnosed, and 2 nd data including history data representing a repair history of the failure in natural language, and Data expressed in natural language concerning at least one of a cause of occurrence of the detected failure and a repair method expressed by the input data is obtained as output data from the learning completion model. In the estimation method according to the present invention, a computer is caused to perform the following processing: Inputting 1 st data concerning a detected failure, which is a failure detected in an apparatus, as input data into a learning completion model which is a model obtained by machine learning a learning data set including 1 st data including failure diagnosis data which is data in which each of a plurality of the failures related to the apparatus is diagnosed, and 2 nd data including history data representing a repair history of the failure in natural language, and Data expressed in natural language concerning at least one of a cause of occurrence of the detected failure and a repair method expressed by the input data is obtained as output data from the learning completion model. The learning completion model according to the present invention enables a computer to function to input data and output data, wherein, The learning completion model is a model obtained by machine learning using a learning data set including 1 st data including failure diagnosis data which is data for diagnosing each of a plurality of types of failures related to the apparatus, and 2 nd data including history data representing a repair history for the failures in natural language, wherein the learning completion model is capable of inputting the 1 st data regarding a detected failure, which is a failure detected in the apparatus, as the input data after machine learning, and obtaining data represented in natural language as the output data regarding at least one of a cause of occurrence of the detected failure and a repair method represented by the input data. According to the present invention, the cause of the failure or the repair method can be easily and quickly determined from the failure diagnosis data. Drawings Features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described below with reference to the accompanying drawings, in which like reference numerals denote like elements, and in which: fig. 1 is a functional block diagram showing a configuration example of a learning system for generating a learning completion model according to the embodiment. Fig. 2 is a schematic diagram showing an example of an architecture of a learning system for generating a learning completion model according to the embodiment. Fig. 3 is a flowchart for explaining an example of processing in the learning system for generating the learning completion model according to the embodiment. Fig. 4 is a functional block diagram showing an example of the configuration of the estimation system according to the embodiment. Fig. 5 is a flowchart for explaining a processing example in the inference system according to the embodiment. Detailed Description The present invention will be described below with reference to embodiments of the invention, but the invention according to the claims is not limited to the following embodiments. The configuration described in the embodiments is not necessarily required as a means for solving the problems. Description of the embodiments The estimation system according to the present embodiment is a system that uses a learning completion model to estimate at least one of the cause of occurrence of a fault detected in an apparatus and the repair method, and outputs the