By Andreas Varga
This booklet addresses fault detection and isolation subject matters from a computational standpoint. not like so much latest literature, it bridges the distance among the present well-developed theoretical effects and the area of trustworthy computational synthesis techniques.
The model-based method of fault detection and analysis has been the topic of ongoing study for the prior few many years. whereas the theoretical elements of fault prognosis at the foundation of linear types are good understood, lots of the computational equipment proposed for the synthesis of fault detection and isolation filters should not passable from a numerical viewpoint. a number of positive factors make this booklet detailed within the fault detection literature:
- Solution of ordinary synthesis difficulties within the so much basic atmosphere, for either non-stop- and discrete-time structures, whether they're right or now not; accordingly, the proposed synthesis techniques can resolve a selected challenge every time an answer exists
- Emphasis at the most sensible numerical algorithms to unravel the synthesis difficulties for linear platforms in generalized state-space shape (also often called descriptor systems)
- Development of normal synthesis approaches hoping on new computational paradigms, resembling factorization-based layout in accordance with clear out updating ideas and nullspace-based synthesis
- Availability of a entire set of unfastened accompanying software program instruments for descriptor platforms, which permits readers to simply enforce all synthesis strategies offered within the booklet and guarantees that each one effects are reproducible
This ebook is essentially meant for researchers and complex graduate scholars within the parts of fault analysis and fault-tolerant keep watch over. it's going to additionally entice mathematicians with an curiosity in control-oriented numerics.
Read Online or Download Solving Fault Diagnosis Problems: Linear Synthesis Techniques PDF
Best linear books
Fresh years have noticeable an important upward thrust of curiosity in max-linear concept and methods. as well as supplying the linear-algebraic heritage within the box of tropical arithmetic, max-algebra offers mathematical thought and methods for fixing quite a few nonlinear difficulties coming up in parts similar to production, transportation, allocation of assets and knowledge processing expertise.
- Formation and Containment Control for High-order Linear Swarm Systems
- Fundamentals of the Theory of Operator Algebras: Special Topics Volume IV Advanced Theory—An Exercise Approach
- Lie Groups and Compact Groups
- Brauer Groups
- A Practical Approach to Linear Algebra
Additional resources for Solving Fault Diagnosis Problems: Linear Synthesis Techniques
Large variations of these parameters are considered parametric faults. 2. 12). The equivalent fault input is defined as f (t) := ΓA (ρ)x(t). For the fault vector f (t), we can use a more structured representation using the alternative affine representation of A(ρ) as A(ρ) = A(0) + A(1) ρ1 + A(2) ρ2 , which leads to A(ρ)x(t) = A(0) x(t) + A(1) x(t) A(2) x(t) ρ . 18). 19) where, for i = 1, . . , N, x (i) (t) ∈ Rn and y(i) (t) ∈ Rp are the state vector and output vector of the i-th system, respectively.
7) can be easily obtained for LPV models whose matrices depend rationally on the components of ρ. 11) S(ρ) = LFTu (M, Δ), M11 M12 is a certain constant matrix with M11 square and Δ = M21 M22 Δ(ρ) is a diagonal matrix depending on the components of ρ such that Δ(ρ0 ) = 0. Straightforward algorithms are available to obtain the above representation. The above LFT-based representation of S(ρ) allows to immediately obtain S (0) = M22 , ΔS = M21 and ΓS (ρ) = Δ(I − ΔM11 )−1 M12 . 5(1 + ρ1 ) C= 011 , 110 Du = 00 .
An important aspect of this approach is that the resulting models with additive faults can simultaneously cover several categories of actuator and sensor faults. 1 Flight actuators with faults are often modelled as continuous-time LTI models, whose transfer-function representation is y(s) = G u (s)u(s) + G f (s)f(s) , where u(t) and y(t) are respectively, the commanded and achieved surface positions and f (t) is a fault signal. For an input (actuator) fault we can take G f (s) = G u (s), while for an output (sensor) fault G f (s) = 1.