Mechanical Engineering - Research Publications
Permanent URI for this collection
Search Results
1 - 8 of 8
-
ItemTrajectory-based proofs for sampled-data extremum seeking control(IEEE, 2013)Extremum seeking of nonlinear systems based on a sampled-data control law is revisited. It is established that under some generic assumptions, semi-global practical asymptotically stable convergence to an extremum can be achieved. To this end, trajectory-based arguments are employed, by contrast with Lyapunov-function-type approaches in the existing literature. The proof is simpler and more straightforward; it is based on assumptions that are in general easier to verify. The proposed extremum seeking framework may encompass more general optimisation algorithms, such as those which do not admit a state-update realisation and/or Lyapunov functions. Multi-unit extremum seeking is also investigated within the context of accelerating the speed of convergence.
-
ItemOn sampled-data extremum seeking control via stochastic approximation methods(IEEE, 2013-01-01)This note establishes a link between stochastic approximation and extremum seeking of dynamical nonlinear systems. In particular, it is shown that by applying classes of stochastic approximation methods to dynamical systems via periodic sampled-data control, convergence analysis can be performed using standard tools in stochastic approximation. A tuning parameter within this framework is the period of the synchronised sampler and hold device, which is also the waiting time during which the system dynamics settle to within a controllable neighbourhood of the steady-state input-output behaviour. Semiglobal convergence with probability one is demonstrated for three basic classes of stochastic approximation methods: finite-difference, random directions, and simultaneous perturbation. The tradeoff between the speed of convergence and accuracy is also discussed within the context of asymptotic normality of the outputs of these optimisation algorithms.
-
ItemExtremum Seeking Methods for Online Automotive Calibration(Springer, 2014-01-01)The automotive calibration process is becoming increasingly difficult as the degrees of freedom in modern engines rises with the number of actuators. This is coupled with the desire to utilise alternative fuels to gasoline and diesel for the promise of lower CO2 levels in transportation. However, the range of fuel blends also leads to variability in the combustion properties, requiring additional sensing and calibration effort for the engine control unit (ECU). Shifting some of the calibration effort online whereby the engine controller adjusts its operation to account for the current operating conditions may be an effective alternative if the performance of the controller can be guaranteed within some performance characteristics. This tutorial chapter summarises recent developments in extremum seeking control, and investigates the potential of these methods to address some of the complexity in developing fuel-flexible controllers for automotive powertrains.
-
ItemMulti-agent gradient climbing via extremum seeking control(IFAC - International Federation of Automatic Control, 2014)A unified framework based on discrete-time gradient-based extremum seeking control is proposed to localise an extremum of an unknown scalar field distribution using a group of equipped with sensors. The controller utilises estimates of gradients of the field from local dithering sensor measurements collected by the mobile agents. It is assumed that distributed coordination which ensures uniform asymptotic stability with respect to a prescribed formation of the agents is employed. The framework is useful in that a broad range of nonlinear programming algorithms can be combined with a wide class of cooperative control laws to perform extreme source seeking. Semi-global practical asymptotically stable convergence to local extrema is established in the presence of bounded field sampling noise.
-
ItemMulti-agent source seeking via discrete-time extremum seeking control(PERGAMON-ELSEVIER SCIENCE LTD, 2014-09)Recent developments in extremum seeking theory have established a general framework for the methodology, although the specific implementations, particularly in the context of multi-agent systems, have not been demonstrated. In this work, a group of sensor-enabled vehicles is used in the context of the extremum seeking problem using both local and global optimisation algorithms to locate the extremum of an unknown scalar field distribution. For the former, the extremum seeker exploits estimates of gradients of the field from local dithering sensor measurements collected by the mobile agents. It is assumed that a distributed coordination which ensures uniform asymptotic stability with respect to a prescribed formation of the agents is employed. An inherent advantage of the frameworks is that a broad range of nonlinear programming algorithms can be combined with a wide class of cooperative control laws to perform extreme source seeking. Semi-global practical asymptotically stable convergence to local extrema is established in the presence of field sampling noise. Subsequently, global extremum seeking with multiple agents is investigated and shown to give rise to robust practical convergence whose speed can be improved via computational parallelism. Nonconvex field distributions with local extrema can be accommodated within this global framework.
-
ItemUnified frameworks for sampled-data extremum seeking control: Global optimisation and multi-unit systems(PERGAMON-ELSEVIER SCIENCE LTD, 2013-09)Two frameworks are proposed for extremum seeking of general nonlinear plants based on a sampled-data control law, within which a broad class of nonlinear programming methods is accommodated. It is established that under some generic assumptions, semi-global practical convergence to a global extremum can be achieved. In the case where the extremum seeking algorithm satisfies a stronger asymptotic stability property, the converging sequence is also shown to be stable using a trajectory-based proof, as opposed to a Lyapunov-function- type approach. The former is more straightforward and insightful. This allows for more general optimisation algorithms than considered in existing literature, such as those which do not admit a state-update realisation and/or Lyapunov functions. Lying at the heart of the analysis throughout is robustness of the optimisation algorithms to additive perturbations of the objective function. Multi-unit extremum seeking is also investigated with the objective of accelerating the speed of convergence.
-
ItemMultidimensional global extremum seeking via the DIRECT optimisation algorithm(PERGAMON-ELSEVIER SCIENCE LTD, 2013-07-01)DIRECT is a sample-based global optimisation method for Lipschitz continuous functions defined over compact multidimensional domains. This paper adapts the DIRECT method with a modified termination criterion for global extremum seeking control of multivariable dynamical plants. Finite-time semi-global practical convergence is established based on a periodic sampled-data control law, whose sampling period is a parameter which determines the region and accuracy of convergence. A crucial part of the development is dedicated to a robustness analysis of the DIRECT method against bounded additive perturbations on the objective function. Extremum seeking involving multiple units is also considered within the same context as a means to increase the speed of convergence. Numerical examples of global extremum seeking based on DIRECT are presented at the end.
-
ItemA non-gradient approach to global extremum seeking: An adaptation of the Shubert algorithm(PERGAMON-ELSEVIER SCIENCE LTD, 2013-03-01)The main purpose of this paper is to adapt the so-called Shubert algorithm for extremum seeking control of general dynamic plants. This algorithm is a good representative of the "sampling optimization methods" that achieve global extremum seeking on compact sets in the presence of local extrema. The algorithm applies to Lipschitz mappings; the model of the system is assumed unknown but the knowledge of its Lipschitz constant is assumed. The controller depends on a design parameter, the "waiting time", and tuning guidelines that relate the design parameter and the region of convergence and accuracy of the algorithm are presented. The analysis shows that semi-global practical convergence (in the initial states) to the global extremum can be achieved in presence of local extrema if compact sets of inputs are considered. Numerical simulations for global optimization in the presence of local extrema are provided to demonstrate the proposed approach.