Richard Saeks

Fault diagnosis for linear systems via multifrequency measurements

The fault diagnosis problem for a linear system whose transfer function matrix is measured at a discrete set of frequencies is formalized. A measure of solvability for the resultant equations and a measure of testability for the unit under test is developed. These, in turn, are used as the basis of algorithms for choosing test points and test frequencies.

A search algorithm for the solution of the multifrequency fault diagnosis equations

A search algorithm for the solution of the fault diagnosis equations arising in linear time invariant analog circuits and systems is presented. By exploitation of Householders formula an efficient algorithm whose computational complexity is a function of the number of system failures rather than the number of system components is obtained.

On the implementation of an analog ATPG: the nonlinear case

A self-testing algorithm in which post-test simulation with failure bounds is employed, has been proposed. Based on this self-testing algorithm, an analog Automatic Test Program Generation (ATPG) for linear circuits or systems is being developed. The AATPG code is subdivided into off-line and on-line components while the actual test can be run in either a fully automatic mode or interactively.

Analog Fault Diagnosis

A simulation-after-test algorithm for the analog fault diagnosis problem is presented, in which a bound on the maximum number of simultaneous faults is used to minimize the number of test points required. Based on this self-testing algorithm, an analog automatic test program generation (AATPG) for both linear and nonlinear circuits and systems is being developed. The AATPG code is subdivided into off-line and on-line processes. The actual test can be run in either a fully automatic mode or interactively. In addition, the issues of parallel processing and testability for analog circuits and systems are also addressed.

On the encirclement condition and its generalization

A new sufficiency proof for the Nyquist criterion is obtained via the analyticity properties of the log function. The intuition derived therefrom is then used to formulate a generalized stability criterion applicable to feedback systems whose open loop gains are characterized by arbitrary Lipschitz continuous unbiased operators.

Neural networks and fault diagnosis in rotating machinery

A series of experiments in which neural networks are used to detect and classify faults in rotating machinery are described. Specific experimental testbeds included a helicopter tail rotor gearbox, fire pumps and condensation pumps.<<ETX>>

The Adaptive Dynamic Programming Theorem

The centerpiece of the theory of dynamic programming is the HamiltonJacobi-Bellman (HJB) equation, which can be used to solve for the optimal cost functional Vo for a nonlinear optimal control problem, while one can solve a second partial differential equation for the corresponding optimal control law ko.Although the direct solution of the HJB equation is computationally untenable, the HJB equation and the relationship between Vo and koserves as the basis for the adaptive dynamic programming algorithm. Here, one starts with an initial cost functional and stabilizing control law pair (Vo, k0) and constructs a sequence of cost functional/control law pairs (Vi, ki) in real time, which are stepwise stable and converge to the optimal cost functional/control law pair, for a prescribed nonlinear optimal control problem with unknown input affine state dynamics.

A neurocontroller for robotic applications

The neural network based robotic arm control concept covers three areas, decentralized adaptive joint control, an inverse kinematics, and path planning. Included are new results from the decentralized adaptive joint controller. This joint controller uses neural networks to adapt a proportional-integral-derivative (PID)/PVA controller. The results show that neural networks allow for fast, accurate control. The authors have tested the joint controller in a robotic testbed simulation software. The neural driven inverse kinematic system has produced accurate performance. The results show that the overall system outperforms conventional methods.<<ETX>>

Parallel system simulation

A relaxation algorithm composed of both a time-step parallel algorithm and a component-wise parallel algorithm is proposed for solving large-scale system simulation problems in parallel. The interconnected nature of the system, which is characterized by the component connection model, is fully exploited by this approach. A technique for finding an optimal number of the time steps is also described. Finally, this algorithm is illustrated via several examples in which the possible trade-offs between the speedup ratio, efficiency, and waiting time are analyzed.

Development and application of a Lyapunov synthesis based neural adaptive controller

The focus of this paper is to describe a neural adaptive control (NAC) technology derived using a Lyapunov synthesis technique. The NAC is a nonlinear adaptive controller which requires minimal plant information. It adapts its gains in real time to maintain the desired performance and to automatically compensate for changes in plant dynamics caused by system failures, environmental changes, or component damage. As such, the NAC control technology has the potential to enhance system performance by automatically optimizing its control laws for each operating regime. It can also increase reliability by automatically compensating for plant damage and system failures.