Critical system and associated monitoring method

• US 10,261,505 B2
• Filed: 11/19/2014
• Issued: 04/16/2019
• Est. Priority Date: 11/19/2013
• Status: Active Grant

• Current Assignee: Thales SA
• Original Assignee: Thales SA
• Inventors: Bobrinskoy, Alexandre, Guerineau, Olivier, Cazaurang, Franck, Gatti, Marc, Bluteau, Bruno
• Primary Examiner(s): Mustafa, Imran K
• Application Number:
  US15/036,168
• Publication Number:
  US 20160291586A1
• Time in Patent Office: 1,609 Days
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: G05B 9/03 with multiple-channel loop, i.e. redundant control systems

Abstract

A critical system comprises a control pathway comprising an actuation unit having a nonlinear behavior, that is able to be controlled by applying an input quantity and whose state is characterized by an output quantity. The behavior of the actuation unit being described by a flat nonlinear model, wherein the system comprises an estimation unit comprising a dynamic inversion linearization module calculating from the measured output quantity, the input quantity of the equivalent linear model; an estimation module estimating a flat output quantity from the input quantity of the equivalent linear model; and a computing module computing a residual based on the estimated flat output quantity and a measured flat output quantity, corresponding to the measured output quantity.

Claims

1. A critical system, in particular a critical flight system, including a control pathway comprising:

an actuation unit having a nonlinear behavior in the order of n and following the dynamic equation;

{dot over (x)}=f(x,u), x∈

ⁿ, u∈

^m

where x is a vector state of dimension n, the components x_i thereof representing an instantaneous state of the actuation unit, u is an input vector of dimension m, the components u_i thereof representing input quantities of the actuation unit, f is a sufficiently derivable function, and n and m are integers, and is controlled by applying the input quantities and whose instantaneous state is characterized by output quantities measured by at least one appropriate sensor,

wherein the behavior of the actuation unit is described by a flat nonlinear model, for which there exists an equivalent linear model establishing a relation between a flat input vector v of dimension m and a flat output vector z of dimension m, the flat output vector z depending on the state vector x and the input vector u and their time derivatives, according to the relationship;

z=ψ

_z(x,u,{dot over (u)}, . . . ,u^(p))

such that the state vector x and the input vector u can be expressed only from the flat output vector z and its derivatives, according to the relationships;

x=ψ

_x(z,ż

, . . . ,z^(q))

u=ψ

_u(z,ż

, . . . ,z^(q+1))

where g^(p) represents the p^th derivative of the function g relative to time,

wherein the system comprises a monitoring pathway for the control pathway, having an estimation unit comprising;

a passage module calculating, from the measured output quantities, the flat output vector of the equivalent linear model;

a dynamic inversion linearization module, calculating from the flat output vector calculated, the flat input vector of the equivalent linear model;

an estimation module estimating the flat output vector of the equivalent linear model from the flat input vector of the equivalent linear model, based on said equivalent linear model; and

a computing module computing a residual based on the flat output vector estimated at the output of the estimation module and the flat output vector computed at the output of the passage module.

2. The system according to claim 1, wherein the estimating module implements a SKF observer.

3. The system according to claim 1, wherein the actuation unit comprises an induction motor.

4. The system according to claim 3, wherein the induction motor is of the hybrid stepping.

5. The system according to claim 4, wherein the flat input vector v of the equivalent linear model is:

6. The system according to claim 5, wherein the mechanical and electromagnetic equations of the behavior of the induction motor defining the equivalent linear model according to:

7. The system according to claim 1, wherein the actuation unit further comprising a control module generating, from a setpoint quantity, the input vector of the actuation unit.

8. The system according to claim 1, wherein the monitoring pathway further having a diagnostic module using at least the residual as input and generating an inhibition signal of the control pathway.

9. The system according to claim 1, comprising:

a first control pathway associated with a first monitoring pathway, the latter generating first residuals;

a second control pathway associated with a second monitoring pathway, making the first control and monitoring pathways redundant, the second monitoring pathway generating second residuals,

wherein the system further comprising a shared diagnostic module inhibiting at least one of or the other of the first and second control pathways, once a fault is detected, based on the first and second residuals.