FDIC method for minimizing measuring failures in a measuring system comprising redundant sensors
First Claim
1. A method for minimizing a contributions of sensors affected by possible measuring failures detected by a plurality of redundant interconnected sensors forming a measuring system by detecting and isolating said sensors affected by failures comprising the steps of:
- a) mapping the sensor values measured by all of said sensors such that the sensor values are combined to form a measured vector by means of a linear transformation into a measured parity vector; and
b) determining a dimension of said vector space by redundance of the measuring system;
thenc) determining an absolute magnitude of said parity vector; and
d) comparing said magnitude with a first detection threshold; and
e) concluding a failure-free state when said magnitude does not exceed said first detection threshold; and
f) if said magnitude exceeds said first detection threshold, projecting said measured parity vector onto all subspaces characteristic of the possible failure states wherein dimensions of said subspaces are determined by the number of possible faulty sensors associated with the respective failure state; and
g) establishing which projection of the parity vector onto a subspace, respectively belonging to relevant failure determination level, providing the largest proportion of the measured parity vector, in order to isolate said failures;
thenh) testing whether a residual failure exceeds a second detection threshold; and
i) proceeding to step j) when said residual failure does not exceed said second detection threshold while carrying out steps f) and g) with the next failure determination level if said residual failure exceeds said second detection threshold; and
thenj) determining the best sensor combination by omitting a sensor value combination belonging to the respective subspaces with the largest proportion of the parity vector at the relevant failure determination level.
1 Assignment
0 Petitions
Accused Products
Abstract
A method for minimizing measured quantities determined from sensors affected by measuring failures and detected by a plurality of redundant sensors connected to form a measuring system by detecting and isolating the sensors affected by failures. Sensor values measured by all the sensors are mapped by a linear transformation into a vector in the parity space (parity vector). The dimension of the parity space is determined by the redundance of the measuring system (i.e., the number of sensors and the dimension of the quantity to be measured).
The subspaces at each failure level that contain the largest proportion of the measured parity vector are determined by projection of the measured parity vector onto all possible subspaces. The best sensor combination at each failure level can be determined by omitting the sensor combinations belonging to such subspaces.
37 Citations
5 Claims
-
1. A method for minimizing a contributions of sensors affected by possible measuring failures detected by a plurality of redundant interconnected sensors forming a measuring system by detecting and isolating said sensors affected by failures comprising the steps of:
-
a) mapping the sensor values measured by all of said sensors such that the sensor values are combined to form a measured vector by means of a linear transformation into a measured parity vector; and b) determining a dimension of said vector space by redundance of the measuring system;
thenc) determining an absolute magnitude of said parity vector; and d) comparing said magnitude with a first detection threshold; and e) concluding a failure-free state when said magnitude does not exceed said first detection threshold; and f) if said magnitude exceeds said first detection threshold, projecting said measured parity vector onto all subspaces characteristic of the possible failure states wherein dimensions of said subspaces are determined by the number of possible faulty sensors associated with the respective failure state; and g) establishing which projection of the parity vector onto a subspace, respectively belonging to relevant failure determination level, providing the largest proportion of the measured parity vector, in order to isolate said failures;
thenh) testing whether a residual failure exceeds a second detection threshold; and i) proceeding to step j) when said residual failure does not exceed said second detection threshold while carrying out steps f) and g) with the next failure determination level if said residual failure exceeds said second detection threshold; and
thenj) determining the best sensor combination by omitting a sensor value combination belonging to the respective subspaces with the largest proportion of the parity vector at the relevant failure determination level. - View Dependent Claims (2, 3, 4, 5)
-
Specification
-
- Resources
-
Current AssigneeLitec-LP GmbH
-
Original AssigneeLitec-LP GmbH
-
InventorsFischer, Harald
-
Primary Examiner(s)Beausoliel, Jr., Robert W.
-
Assistant Examiner(s)TU, CHRISTINE TRINH LE
-
Application NumberUS08/578,840Time in Patent Office609 DaysField of Search371/49.1, 371/49.4, 371/50.1, 371/57.1, 371/64, 371/67.1, 371/25.1, 371/28, 395/184.01, 395/185.01, 376/245, 376/215, 376/216, 376/217, 367/124, 367/125, 364/560, 364/561, 364/562, 364/571.02, 364/571.03, 364/571.04, 364/571.07, 318/563, 318/564US Class Current714/800CPC Class CodesG01D 3/08 with provision for safeguar...
