SYSTEM FOR MONITORING A SET OF COMPONENTS OF A DEVICE

• US 20150287249A1
• Filed: 10/25/2013
• Published: 10/08/2015
• Est. Priority Date: 10/26/2012
• Status: Active Grant

• Current Assignee: Safran Aircraft Engine Company
• Original Assignee: Snecma (Safran SA)
• Inventors: Lacaille, Jerome, Rabenoro, Tsirizo
• Granted Patent: US 9,754,429 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: B64D 2045/0085 Devices for aircraft health monitoring, e.g. monitoring flutter or vibration
• CPC Class Codes: B64D 45/00 Aircraft indicators or protectors not otherwise provided for camouflage F41H3/00
• CPC Class Codes: G05B 23/0251 Abstraction hierarchy, e.g. "complex systems", i.e. system is divided in subsystems, subsystems are monitored and results are combined to decide on status of whole system
• CPC Class Codes: G07C 5/006 Indicating maintenance
• CPC Class Codes: G07C 5/0808 Diagnosing performance data testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005

Abstract

The invention relates to a system for monitoring a set of components of a device such as an engine, comprising:

• • means for detecting (2) comprising a set of agents for detecting anomalies (a₁, . . . , a_i, . . . , a_n), with each one of said agents being configured to receive measurements on physical parameters relating to a sub-set of components of a device such as an engine and to deliver an initial distribution of the probability of an anomaly,
  • means for acquiring (3) configured to receive feedback information (17) as well as the initial distributions of the probability of an anomaly coming from said set of agents,
  • means for emerging (5) configured to iteratively emerge current distributions of the probability of an anomaly in line with said feedback information, with said current distributions of the probability of an anomaly converging towards a set of optimum distributions of the probability of an anomaly in accordance with said feedback information, and
  • means for synthesis (7) configured to synthesise optimum distributions of the probability of an anomaly relating to at least one sub-set of interest of components of a device such as an engine in order to extract the risks of an anomaly specific to said sub-set of interest.

Claims

1. System for monitoring a set of components of a device, comprising:

means for detecting (2) comprising a set of agents for detecting anomalies (a₁, . . . , a_i, . . . , a_n), with each one of said agents being configured to receive measurements on physical parameters relating to a sub-set of components of said device and to deliver an initial distribution of the probability of an anomaly,

means for acquiring (3) configured to receive feedback information (17) as well as the initial distributions of the probability of an anomaly coming from said set of agents,

means for emerging (5) configured to iteratively emerge current distributions of the probability of an anomaly in line with said feedback information, with said current distributions of the probability of an anomaly converging towards a set of optimum distributions of the probability of an anomaly in accordance with said feedback information, and

means for synthesis (7) configured to synthesise optimal distributions of the probability of an anomaly relating to at least one sub-set of interest of components of said device in order to extract the risks of an anomaly specific to said sub-set of interest.

2. System according to claim 1, characterised in that the means for emerging comprise:

means for merging (51) configured to create each current or optimum distribution of the probability of an anomaly by merging preceding distributions of the probability of an anomaly according to predetermined rules for merging,

means for optimisation (53) configured to put into line each initial or current distribution of the probability of an anomaly with the feedback information that corresponds to it, in order to associate an efficiency value to it, and

means of management (55) configured to change the set of agents towards an optimum population of agents by creating new agents corresponding to current and optimum distributions of the probability of an anomaly and by selecting the agents that are the most compliant with the feedback information according to the efficiency values associated with said current and optimum distributions of the probability of an anomaly.

3. System according to claim 1, characterised in that the means for emerging (5) are furthermore configured to associate to each initial, current or optimum distribution of the probability of an anomaly relative to a sub-set of components of said device, an implication vector reflecting a contribution of each element of said sub-set on the corresponding distribution of the probability of an anomaly.

4. System according to claim 3, characterised in that the means for synthesis (7) are configured to synthesise the optimum distributions of the probability of an anomaly relating to said sub-set of interest by weighting the influence of each element of said sub-set of interest according to the corresponding implication vectors.

5. System according to claim 2, characterised in that the predefined rules for merging are chosen from among the following rules:

expert rules, reduction rules, voting rule, business rules, Bayesian rules, and fuzzy logic rules.

6. System according to claim 2, characterised in that the means for optimisation (53) are furthermore configured to adjust the merging rules according to the feedback information.

7. System according to claim 2, characterised in that the set of agents or the optimum population of agents is formed by instances of algorithms comprising different choices of measurements, or different processing methods, or different configurations.

8. System as claimed in claim 7, characterised in that the initial distributions of the probability of an anomaly are standardised with respect to the context of the measurements.

9. System as claimed in claim 8, characterised in that it comprises means for assisting in taking decisions (21) that make it possible to determine the operations of maintenance according to the risks of anomalies.

10. System as claimed in claim 9, characterised in that said device is an engine and in particular an aircraft engine.

11. Method for monitoring a set of components of a device, comprising the following steps:

acquiring feedback information as well as initial distributions of the probability of an anomaly coming from a set of agents for detecting anomalies, with each one of said agents being configured to receive measurements on physical parameters relating to a sub-set of components of said device before delivering an initial distribution of the probability of an anomaly,

iteratively emerging current distributions of the probability of an anomaly in line with said feedback information, with said current distributions of the probability of an anomaly converging towards a set of optimum distributions of the probability of an anomaly in accordance with said feedback information, and

synthesising optimum distributions of the probability of an anomaly relating to at least one sub-set of interest of components of said device in order to extract the risks of an anomaly specific to said sub-set of interest.