Modelling and control of Hammerstein system using B-spline approximation and the inverse of De Boor algorithm

In this article a simple and effective controller design is introduced for the Hammerstein systems that are identified based on observational input/output data. The nonlinear static function in the Hammerstein system is modelled using a B-spline neural network. The controller is composed by computing the inverse of the B-spline approximated nonlinear static function, and a linear pole assignment controller. The contribution of this article is the inverse of De Boor algorithm that computes the inverse efficiently. Mathematical analysis is provided to prove the convergence of the proposed algorithm. Numerical examples are utilised to demonstrate the efficacy of the proposed approach.

Inverting recurrent neural networks for internal model control of nonlinear systems

In this paper, we show how a set of recently derived theoretical results for recurrent neural networks can be applied to the production of an internal model control system for a nonlinear plant. The results include determination of the relative order of a recurrent neural network and invertibility of such a network. A closed loop controller is produced without the need to retrain the neural network plant model. Stability of the closed-loop controller is also demonstrated.

Internal model control of nonlinear systems through the inversion of recurrent neural networks

Recurrent neural networks can be used for both the identification and control of nonlinear systems. This paper takes a previously derived set of theoretical results about recurrent neural networks and applies them to the task of providing internal model control for a nonlinear plant. Using the theoretical results, we show how an inverse controller can be produced from a neural network model of the plant, without the need to train an additional network to perform the inverse control.

Dynamic input/output linearization using recurrent neural networks

A dynamic recurrent neural network (DRNN) is used to input/output linearize a control affine system in the globally linearizing control (GLC) structure. The network is trained as a part of a closed loop that involves a PI controller, the goal is to use the network, as a dynamic feedback, to cancel the nonlinear terms of the plant. The stability of the configuration is guarantee if the network and the plant are asymptotically stable and the linearizing input is bounded.

A comparative study of multilayered and single layered neural network based predictive controllers

The authors compare the performance of two types of controllers one based on the multilayered network and the other based on the single layered CMAC network (cerebellar model articulator controller). The neurons (information processing units) in the multi-layered network use Gaussian activation functions. The control scheme which is considered is a predictive control algorithm, along the lines used by Willis et al. (1991), Kambhampati and Warwick (1991). The process selected as a test bed is a continuous stirred tank reactor. The reaction taking place is an irreversible exothermic reaction in a constant volume reactor cooled by a single coolant stream. This reactor is a simplified version of the first tank in the two tank system given by Henson and Seborg (1989).

Mismatch effects on discrete-time poles

The presence of mismatch between controller and system is considered. A novel discrete-time approach is used to investigate the migration of closed-loop poles when this mismatch occurs. Two forms of state estimator are employed giving rise to several interesting features regarding stability and performance.

Decentralized PID self-tuning control of industrial robots

A three degrees of freedom industrial robot is controlled by applying PID self-tuning (PID/ST) controllers. This control is considered as a corrective term to a nominal value, centrally computed from an inaccurate and/ or simplified dynamic model. An identification scheme on an assumed linear plant describing the deviation from the desired trajectory is employed in order to tune the controller coefficients and thus accomplish a behaviour prescribed through a desired pole placement. A salient feature of our approach is the decentralized nature of the controllers producing the corrective term for each joint. This opens the way to practical implementation, as recent computing requirement calculations for similar set-ups have shown in the literature. Numerical results are presented.

Reduced complexity set-point following in self-tuning control

An external input signal is incorporated into a self-tuning controller which, although it is based on a CARMA system model, employs a state-space framework for control law calculations. Steady-state set point following can then be accomplished even when only a recursive least squares parameter estimation scheme is used, despite the fact that the disturbance affecting the system may well be coloured.

Implementation of microprocessor based self-tuning control on a heating system

This paper describes the implementation, using a microprocessor, of a self-tuning control algorithm on a heating system. The algorithm is based on recursive least squares parameter estimation with a state-space, pole placement design criterion and shows how the controller behaves when applied to an actual system.

Discrete demand side control performance under dynamic building simulation: a heat pump application

This study presents the findings of applying a Discrete Demand Side Control (DDSC) approach to the space heating of two case study buildings. High and low tolerance scenarios are implemented on the space heating controller to assess the impact of DDSC upon buildings with different thermal capacitances, light-weight and heavy-weight construction. Space heating is provided by an electric heat pump powered from a wind turbine, with a back-up electrical network connection in the event of insufficient wind being available when a demand occurs. Findings highlight that thermal comfort is maintained within an acceptable range while the DDSC controller maintains the demand/supply balance. Whilst it is noted that energy demand increases slightly, as this is mostly supplied from the wind turbine, this is of little significance and hence a reduction in operating costs and carbon emissions is still attained.

The system identification and control of Hammerstein system using non-uniform rational B-spline neural network and particle swarm optimization

In this paper a new system identification algorithm is introduced for Hammerstein systems based on observational input/output data. The nonlinear static function in the Hammerstein system is modelled using a non-uniform rational B-spline (NURB) neural network. The proposed system identification algorithm for this NURB network based Hammerstein system consists of two successive stages. First the shaping parameters in NURB network are estimated using a particle swarm optimization (PSO) procedure. Then the remaining parameters are estimated by the method of the singular value decomposition (SVD). Numerical examples including a model based controller are utilized to demonstrate the efficacy of the proposed approach. The controller consists of computing the inverse of the nonlinear static function approximated by NURB network, followed by a linear pole assignment controller.

Adaptive recursive control of a soil testing load frame

Recursive Learning Control (RLC) has the potential to significantly reduce the tracking error in many repetitive trajectory applications. This paper presents an application of RLC to a soil testing load frame where non-adaptive techniques struggle with the highly nonlinear nature of soil. The main purpose of the controller is to apply a sinusoidal force reference trajectory on a soil sample with a high degree of accuracy and repeatability. The controller uses a feedforward control structure, recursive least squares adaptation algorithm and RLC to compensate for periodic errors. Tracking error is reduced and stability is maintained across various soil sample responses.

Predictive computed-torque control of a PUMA 560 manipulator robot

This paper describes the integration of constrained predictive control and computed-torque control, and its application on a six degree-of-freedom PUMA 560 manipulator arm. The real-time implementation was based on SIMULINK, with the predictive controller and the computed-torque control law implemented in the C programming language. The constrained predictive controller solved a quadratic programming problem at every sampling interval, which was as short as 10 ms, using a prediction horizon of 150 steps and an 18th order state space model.

A search for invariant relative satellite motion

This report presents the canonical Hamiltonian formulation of relative satellite motion. The unperturbed Hamiltonian model is shown to be equivalent to the well known Hill-Clohessy-Wilshire (HCW) linear formulation. The in°uence of perturbations of the nonlinear Gravitational potential and the oblateness of the Earth; J2 perturbations are also modelled within the Hamiltonian formulation. The modelling incorporates eccentricity of the reference orbit. The corresponding Hamiltonian vector ¯elds are computed and implemented in Simulink. A numerical method is presented aimed at locating periodic or quasi-periodic relative satellite motion. The numerical method outlined in this paper is applied to the Hamiltonian system. Although the orbits considered here are weakly unstable at best, in the case of eccentricity only, the method ¯nds exact periodic orbits. When other perturbations such as nonlinear gravitational terms are added, drift is signicantly reduced and in the case of the J2 perturbation with and without the nonlinear gravitational potential term, bounded quasi-periodic solutions are found. Advantages of using Newton's method to search for periodic or quasi-periodic relative satellite motion include simplicity of implementation, repeatability of solutions due to its non-random nature, and fast convergence. Given that the use of bounded or drifting trajectories as control references carries practical di±culties over long-term missions, Principal Component Analysis (PCA) is applied to the quasi-periodic or slowly drifting trajectories to help provide a closed reference trajectory for the implementation of closed loop control. In order to evaluate the e®ect of the quality of the model used to generate the periodic reference trajectory, a study involving closed loop control of a simulated master/follower formation was performed. 2 The results of the closed loop control study indicate that the quality of the model employed for generating the reference trajectory used for control purposes has an important in°uence on the resulting amount of fuel required to track the reference trajectory. The model used to generate LQR controller gains also has an e®ect on the e±ciency of the controller.

Assisted remote viewing of a teleoperation work cell

Intelligent viewing systems are required if efficient and productive teleoperation is to be applied to dynamic manufacturing environments. These systems must automatically provide remote views to an operator which assist in the completion of the task. This assistance increases the productivity of the teleoperation task if the robot controller is responsive to the unpredictable dynamic evolution of the workcell. Behavioral controllers can be utilized to give reactive 'intelligence.' The inherent complex structure of current systems, however, places considerable time overheads on any redesign of the emergent behavior. In industry, where the remote environment and task frequently change, this continual redesign process becomes inefficient. We introduce a novel behavioral controller, based on an 'ego-behavior' architecture, to command an active camera (a camera mounted on a robot) within a remote workcell. Using this ego-behavioral architecture the responses from individual behaviors are rapidly combined to produce an 'intelligent' responsive viewing system. The architecture is single-layered, each behavior being autonomous with no explicit knowledge of the number, description or activity of other behaviors present (if any). This lack of imposed structure decreases the development time as it allows each behavior to be designed and tested independently before insertion into the architecture. The fusion mechanism for the behaviors provides the ability for each behavior to compete and/or co-operate with other behaviors for full or partial control of the viewing active camera. Each behavior continually reassesses this degree of competition or co-operation by measuring its own success in controlling the active camera against pre-defined constraints. The ego-behavioral architecture is demonstrated through simulation and experimentation.