Showing papers in "International Journal of Modelling, Identification and Control in 2008"
TL;DR: Non-linear state estimation and some related topics like parametric estimation, fault diagnosis and perturbation attenuation are tackled here via a new methodology in numerical differentiation within the framework of differential algebra.
Abstract: Non-linear state estimation and some related topics like parametric estimation, fault diagnosis and perturbation attenuation are tackled here via a new methodology in numerical differentiation. The corresponding basic system theoretic definitions and properties are presented within the framework of differential algebra, which permits to handle system variables and their derivatives of any order. Several academic examples and their computer simulations, with online estimations, illustrate our viewpoint.
363 citations
TL;DR: A very efficient Integrated Forward Orthogonal Search (IFOS) algorithm, which is assisted by the squared correlation and mutual information, and which incorporates a Generalised Cross-Validation (GCV) criterion and hypothesis tests, is introduced to overcome limitations in model structure selection.
Abstract: Model structure selection plays a key role in non-linear system identification. The first step in non-linear system identification is to determine which model terms should be included in the model. Once significant model terms have been determined, a model selection criterion can then be applied to select a suitable model subset. The well known Orthogonal Least Squares (OLS) type algorithms are one of the most efficient and commonly used techniques for model structure selection. However, it has been observed that the OLS type algorithms may occasionally select incorrect model terms or yield a redundant model subset in the presence of particular noise structures or input signals. A very efficient Integrated Forward Orthogonal Search (IFOS) algorithm, which is assisted by the squared correlation and mutual information, and which incorporates a Generalised Cross-Validation (GCV) criterion and hypothesis tests, is introduced to overcome these limitations in model structure selection.
104 citations
TL;DR: This paper analyses performance degradation induced on this system of systems by the presence and spreading of failures in order to emphasise the most critical links existing among different phenomena using Fuzzy Numbers (FNs) to represent involved quantities.
Abstract: Welfare in developed countries strongly relies on many heterogeneous infrastructures generically named as critical infrastructures. These infrastructures, designed as autonomous systems, are actually more and more mutually dependent. This introduces new and extremely dangerous vulnerabilities in the overall system because an accidental or a malicious fault (e.g. terroristic attack) could exploit these 'connections'to unpredictably spread, amplifying its negative consequences and affecting unforeseeable and haphazard sets of users. In this paper, we analyse performance degradation induced on this system of systems by the presence and spreading of failures in order to emphasise the most critical links existing among different phenomena. Due to uncertainties that characterise these systems, we use Fuzzy Numbers (FNs) to represent involved quantities. This allows a modelling approach that can be set up using also qualitative information that are easier to obtain from experts and stakeholders. Moreover, this choice brings to a better characterisation of the level of confidence of our results. Preliminary results on a simple case study illustrate the effectiveness of the proposed approach.
83 citations
TL;DR: The decoupled multiple model is an interesting alternative to the popular Takagi-Sugeno multiple model because different dimensions of submodels can be considered as mentioned in this paper, but to our knowledge, the state estimation problem of non-linear systems represented by this structure is not thoroughly investigated.
Abstract: The multiple model approach is an elegant and a powerful tool for modelling real-world complex processes. In this modelling framework, a judicious combination of a set of submodels makes it possible to describe the behaviour of a non-linear system. Two different structures of multiple models can be distinguished according to whether the submodels share a common state vector (Takagi-Sugeno multiple model) or not (decoupled multiple model). This latter structure is an interesting alternative to the popular Takagi-Sugeno multiple model because different dimensions of submodels can be considered. The decoupled multiple model is nowadays increasingly used to perform the identification and the control of non-linear systems. However, to our knowledge, the state estimation problem of non-linear systems represented by this structure is not thoroughly investigated. The present paper deals with this worthwhile problem.
50 citations
TL;DR: The design of a pneumatically powered transtibial prosthetic device is presented and a first prototype has been built and provides a preliminary test bed for control algorithm development and testing with able-bodied subjects in laboratory conditions.
Abstract: Due to its high power-to-weight ratio, a pleated pneumatic artificial muscle (PPAM) offers an interesting alternative actuation source for robotic devices. Its inherent compliant behaviour excites another broad field of interest: assistive clinical devices such as powered exoskeletons and prosthetics. In this paper, the design of a pneumatically powered transtibial prosthetic device is presented. A first prototype has been built and provides a preliminary test bed for control algorithm development and testing with able-bodied subjects in laboratory conditions. The characteristics and working principle of a PPAM are described. The design specifications and the mechanical model of the prosthesis are discussed. The mechanical design and the control structure are outlined. Furthermore, some initial walking trials with an able-bodied subject wearing the prosthesis prototype are presented and discussed.
49 citations
TL;DR: To reduce the computational load required for the evaluation of the simulation error, a two-stage identification algorithm, that exploits the effect of the choice of the sampling time on structure selection is proposed.
Abstract: In non-linear model identification the problem of model structure selection is critical for the success of the identification process. This paper discusses this problem with reference to the class of polynomial NARX models. First it is shown that classical identification approaches based on (one-step-ahead) Prediction Error Minimisation (PEM) may lead to an incorrect or redundant model structure selection, especially in non-ideal identification conditions where the identification data are not adequately exciting or over-sampled. Then a more effective approach is introduced, based on the minimisation of the simulation (or model prediction) error. Finally, to reduce the computational load required for the evaluation of the simulation error, a two-stage identification algorithm, that exploits the effect of the choice of the sampling time on structure selection is proposed. A coarse identification of the model structure is initially performed using over-sampled input-output data, and then the structure is refined considering a decimated version of the data. Some simulation and experimental examples are also discussed.
44 citations
TL;DR: An efficient Common Model Structure Selection algorithm, called the Extended Forward Orthogonal Regression (EFOR) algorithm, is proposed to select a common parameter-dependent model structure that best fits the adjustable parameter properties for the underlying system.
Abstract: This study considers the identification problem for a class of non-linear parameter-varying systems associated with the following scenario: the system behaviour depends on some specifically prescribed parameter properties, which are adjustable. To understand the effect of the varying parameters, several different experiments, corresponding to different parameter properties, are carried out and different data sets are collected. The objective is to find, from the available data sets, a common parameter-dependent model structure that best fits the adjustable parameter properties for the underlying system. An efficient Common Model Structure Selection (CMSS) algorithm, called the Extended Forward Orthogonal Regression (EFOR) algorithm, is proposed to select such a common model structure. Two examples are presented to illustrate the application and the effectiveness of the new identification approach.
27 citations
TL;DR: Based on the techniques of high gain observer and adaptive estimation, an adaptive observer is proposed in this paper for sensor fault estimation in a class of uniformly observable non-linear systems.
Abstract: Based on the techniques of high gain observer and adaptive estimation, an adaptive observer is proposed in this paper for sensor fault estimation in a class of uniformly observable non-linear systems. It is first assumed that a high gain observer exists for the fault-free system. With a parametric model of sensor fault, a high gain adaptive observer is then designed for sensor fault estimation. In order to establish the global convergence of the adaptive observer, in addition to the usual conditions for high gain observer convergence, a persistent excitation condition is also required, like in most recursive parameter estimation problems.
26 citations
TL;DR: An Economic Order Quantity/Economic Production Quantity model with finite replenishment/production rate, taking into account the effect of the suppliers' trade credit and the retailers' promotional effort made by the advertising and the sales-team is derived.
Abstract: This paper derives an Economic Order Quantity (EOQ)/Economic Production Quantity (EPQ) model with finite replenishment/production rate, taking into account the effect of the suppliers' trade credit and the retailers' promotional effort made by the advertising and the sales-team. The rate of demand in market is assumed to be a function of the selling price and the budget for the advertising and the sales-team. The objective of this paper is to determine the optimal selling price, replenishment period and the above budget so that the average net profit is maximised. The associated constrained maximisation problem is solved by Interior-Penalty-Function Method (Fiacco and Mc Cormick). The model is illustrated with a result, using numerical example and it provides some insights into the effect of the retailers' promotional effort.
25 citations
TL;DR: This study concludes that a control parameter β (the magnitude of the effect of the fish population size on the fishing effort function E), changes not only the rate at which the population goes to equilibrium, but also the equilibrium values.
Abstract: In traditional harvesting models a fishing effort E is defined by the fishing intensity and does not address the inverse effect of fish abundance on the fishing effort. In this paper, based on a canonical differential equation model, we developed a new fishing effort model which relies on the density effect of fish population. This study concludes that a control parameter β (the magnitude of the effect of the fish population size on the fishing effort function E), changes not only the rate at which the population goes to equilibrium, but also the equilibrium values. To examine systematically the consequences of different harvesting strategies, we used numerical simulations and qualitative analysis of five standard fishery strategies.
24 citations
TL;DR: A non-linear adaptive observer-based fault diagnosis algorithm is proposed to diagnose the fault in the dynamic part of SDC systems, where it has been shown that a good output PDF tracking can still be realised when fault occurs in the system.
Abstract: Stochastic Distribution Control (SDC) systems are a group of systems where the outputs considered are the measured Probability Density Functions (PDFs) of the system output whilst subjected to a normal crisp input. The purpose of the control algorithm design of such systems is to choose a control input such that the PDF of the system output can follow a prespecified PDF as close as possible. Using the rational square-root B-spline model to represent the dynamics between the output PDF and the input, in this paper a non-linear adaptive observer-based fault diagnosis algorithm is proposed to diagnose the fault in the dynamic part of such systems. Using the estimation to the unknown fault, a fault tolerant control via a controller reconfiguration is proposed, where it has been shown that a good output PDF tracking can still be realised when fault occurs in the system. A simulated example is given to illustrate the use of the proposed algorithm.
TL;DR: Simulation results indicate that considerable improvements in the vehicle handling can be achieved when the vehicle is governed by the proposed fuzzy observer-based controller.
Abstract: This paper deals with the robust control for Four Wheels Steering (4WS) vehicle dynamics when the road adhesion conditions change and the sideslip angle is unavailable for measurement. The non-linear model of the vehicle is first represented by an uncertain Takagi-Sugeno model. Next, the robust output stabilisation of the vehicle is considered. The controller and the observer are designed in terms of Bilinear Matrix Inequalities (BMI) problem. A method is then proposed to get the LMI formulation resolved sequentially. The obtained simulation results indicate that considerable improvements in the vehicle handling can be achieved when the vehicle is governed by the proposed fuzzy observer-based controller.
TL;DR: A comparative study is presented to illustrate the benefits of the proposed algorithms so as to maximise the Region of Asymptotic Stability (RAS).
Abstract: This paper deals with two evolutionary algorithms for calculating the asymptotic stability region for discrete non-linear polynomial systems. These algorithms rest on two new approaches of inverting discrete direct polynomial state equation. Based on some topological considerations, the systematic computational algorithms are then applied on the second-order polynomial model of a synchronous generator power system. Hence, a comparative study is presented to illustrate the benefits of the proposed algorithms so as to maximise the Region of Asymptotic Stability (RAS).
TL;DR: An approach useful to analyse the performance of the product development model adopted by an organisation and the nature of the information and knowledge flows exchanged to coordinate tasks was the primary focus of this study.
Abstract: This paper presents an approach useful to analyse the performance of the product development model adopted by an organisation. A dynamic information management perspective of the development process was assumed while the nature of the information and knowledge flows exchanged to coordinate tasks was the primary focus of this study. In particular, the effect of information and knowledge flows ambiguity was taken into account. Coordination between tasks was modelled by using the Design Structure Matrix (DSM) tool. The dynamic behaviour of the Product Development Process (PDP) and the evolution of information and knowledge quality at the interface between tasks was foresighted through simulation. System dynamics was used as a modelling framework. This approach was implemented to analyse a specific subsystem - 'the climatic system' - of a new car model recently developed by a large Italian car manufacturer. Data were collected by administrating a questionnaire to technical managers and engineers involved in the development process of the subsystem. Continuous interaction and feedback from them was useful to build the process model simulated in our study and to test scales to measure model variables developed for the purpose in the first part of the study and discuss the simulation outcome and possible organisational solutions to correct bad performance.
TL;DR: The problem of exact model matching (EMM) for general left invertible neutral multi-delay systems, via proportional realizable state and output feedback, is extensively solved and the results are rather useful to adaptive control of distributed industrial processes described by neutral time delay models.
Abstract: The problem of Exact Model Matching (EMM) for general left invertible neutral multidelay systems, via proportional realisable state and output feedback, is extensively solved. The necessary and sufficient conditions for the problem to have a realisable solution are established. The general analytical expression of the controllers, solving the problem, is derived. The present results are rather useful to adaptive control of distributed industrial processes described by neutral time delay models.
TL;DR: This paper presents the design of a longitudinal control system for a platoon of vehicles relying on a non-linear vehicle model and the chosen control methodology is second-order sliding mode control.
Abstract: This paper presents the design of a longitudinal control system for a platoon of vehicles The synthesis of an automatic vehicle following system is an important aspect of an automated highway system design The main objectives of the control of a platoon of vehicles are the increment of the exploitation of the road capacity and the improvement of passengers' safety and comfort The proposed control system is designed relying on a non-linear vehicle model The chosen control methodology is second-order sliding mode control This choice is motivated by the well-known robustness features of the sliding mode control approach, which are particularly appropriate dealing with the automotive context Moreover, the proposed approach produces a considerable reduction of the chattering phenomenon, which can determine undesired mechanical wear in the actuators The individual vehicle stability and platoon stability are guaranteed by the proposed control system This latter is tested in simulation considering a platoon of three vehicles in a stop-and-go traffic situation
TL;DR: The proposed adaptive SMC is composed of the equivalent and the switching controls, and it is constructed on the basis of the information from the proposed estimator to take the estimates for the state and uncertainty.
Abstract: This paper is concerned with an adaptive Sliding Mode Control (SMC) for non-linear discrete-time systems with mismatched uncertainty. The proposed adaptive SMC is composed of the equivalent and the switching controls, and it is constructed on the basis of the information from the proposed estimator to take the estimates for the state and uncertainty. It is assumed that some of the state variables can be measured and the uncertainty in the discrete-time systems is parameterised linearly in the unknown constant terms. The estimates for the state and the uncertainty are taken by applying the proposed weighted least squares method. It is verified based on the Lyapunov scheme according to which the estimation errors converge to zero as time increases, and the trajectory of the system response is stable as the sliding surface is suitably selected. The effectiveness of the proposed adaptive SMC is illustrated by the simulation experiment in a simple numerical example.
TL;DR: A design methodology of an adaptive power system stabiliser (PSS) to achieve such task by targeting certain damping ratio and settling time by incorporating an adaptive neuro-fuzzy inference system (ANFIS).
Abstract: Well-damped transient response of power systems signifies a vital control task. The paper presents a design methodology of an adaptive power system stabiliser (PSS) to achieve such task by targeting certain damping ratio and settling time. A standard simple-structure controller is used with the plant, which includes the synchronous generator and exciter, and the overall transfer function is derived in terms of the PSS parameters. A multi-objective optimisation function is formulated in order to force the damping ratio and settling time of the system to desired values. Particle swarm optimisation (PSO) is applied to independently obtain the PSS parameters which minimise such objective function at selected load points covering a wide range of operation. The data obtained from PSO represent the training data of an adaptive neuro-fuzzy inference system (ANFIS), which could give the PSS parameters at any load within a wide region of operating conditions. Testing of the proposed PSS shows that the desired performance indices could be fulfilled from light load to over load under both lagging and leading power factor conditions.
TL;DR: An online approach for mapping with a mobile robot in dynamic and unknown environments is presented and some simulation results indicate that the approach is feasible.
Abstract: In this paper, we address the problem of mapping dynamic and unknown environments The static and moving objects are modelled as the components in a Gaussian mixture model (GMM) By recursive learning of GMM, the components corresponding to the static objects will have larger weights while the components corresponding to the moving objects will have smaller weights At each time step, a number of components with the largest weights are adaptively selected as the background map and the new observations which do not match with the background map are classified as the foreground map In addition, based on a Bayesian factorisation of simultaneous localisation and mapping (SLAM) problem, we present an online algorithm for SLAM with GMM learning Our contributions are employing GMM learning approach to model the dynamic environment with detection of moving objects and jointing the GMM learning with SLAM in unknown environment Consequently, an online approach for mapping with a mobile robot in dynamic and unknown environments is presented Some simulation results indicate that our approach is feasible
TL;DR: A hybrid approach that integrates the unified modelling language (UML) with the FB model; to semi automatically generate the design diagram in the form of a network of interconnected FB instances; and a heuristic-based approach to allocate FB instances to the execution environment, so as to satisfy constraints imposed by this kind of applications.
Abstract: The IEC61499 standard proposes the function block (FB) model for the next generation of distributed control applications. This model defines the FB type to be the basic construct in this paradigm. A control application is considered as a network of interconnected instances of FB types. These applications can be executed on one device such as a multitasking programmable logic controller (PLC), but they are usually executed on a network of interconnected devices. In this paper 1 , a methodology for the development and deployment of IEC61499-based control applications on a network of interconnected devices is proposed. The methodology is based on a hybrid approach that integrates the unified modelling language (UML) with the FB model; to semi automatically generate the design diagram in the form of a network of interconnected FB instances. A heuristic-based approach is described to allocate these instances to the execution environment, so as to satisfy constraints imposed by this kind of applications. A formal description of FB design models is proposed and a formal approach is described to assign FB instances to feasible OS tasks of devices while considering temporal constraints. A running example from industry is used to demonstrate the feasibility of the proposed approach.
TL;DR: The design and development of Artificial Neural Network (ANN)-based model for the fault detection in centrifugal pumping system is presented and the performance of the trained network is found to be satisfactory for the real-time fault diagnosis.
Abstract: Fault detection and diagnosis of technical plants is of great importance for the safe operation and long life. An early detection of faults may help to avoid product deterioration, performance degradation, damage to the machinery and damage to human operators. This paper presents the design and development of Artificial Neural Network (ANN)-based model for the fault detection in centrifugal pumping system. The network is developed to detect a total of 20 faults. The training and testing data required to develop the neural network model were generated at different operating conditions by running the pumping system and by creating various faults in real time in a laboratory experimental model. A principal component analysis-based feature extraction method is proposed to reduce the dimension of the input features. The performance of the trained network is found to be satisfactory for the real-time fault diagnosis.
TL;DR: This paper addresses the issue of Cartesian space trajectory tracking control of robot arms with link flexibility by proposing a control method using vision feedback based on dynamics of the robot and Lyapunov stability theory.
Abstract: This paper addresses the issue of Cartesian space trajectory tracking control of robot arms with link flexibility. A control method using vision feedback is proposed based on dynamics of the robot and Lyapunov stability theory. A CCD camera and video tracker are used as a vision system for the measurement of end-effector position and link flexural behaviour in the control process. Using this vision system the end-effector position is measured directly, whereas the link deflections are measured indirectly based on kinematics and inverse kinematics of the flexible robot. End-effector trajectory tracking control experiments are carried out using a two-link flexible robot system as the test bed. The results demonstrate effectiveness and usefulness of the proposed sensing and control methods.
TL;DR: Investigation of automotive engine air path faults from transient data is investigated using Neural Networks using Radial Basis Function NNs to detect and diagnose the faults, and also to indicate fault size, by recognising the different fault patterns occurring in the transient data.
Abstract: Classification of automotive engine air path faults from transient data is investigated using Neural Networks (NNs) A generic Spark Ignition (SI) Mean Value Engine Model (MVEM) is used for experimentation Several faults are considered, including sensor faults, Exhaust Gas Recycle (EGR) valve and leakage in intake manifold Consideration of different fault intensities for all the sensor and component faults is a unique feature of this research The identification of a fault and its intensity has been considered both equally important Radial Basis Function (RBF) NNs are trained to detect and diagnose the faults, and also to indicate fault size, by recognising the different fault patterns occurring in the transient data Three dynamic cases of fault occurrence are considered with increasing generality of engine operation: (1) engine accelerates or retards from mean speed, (2) engine runs at different steady speeds and (3) engine accelerates or retards from any initial speed The approach successfully classifies the faults in each case
TL;DR: A multiagent model of cell components involved in the metabolic pathways of Carbohydrate Oxidation (CO) is constructed, and the multiagent paradigm proved natural and useful not only in building simulations, but also for devising an engineering methodology.
Abstract: A cell consists of a large number of components interacting in a dynamic environment. The complexity of interactions among cell components makes design of cell simulation a challenging task. Multiagent systems can be considered a suitable framework for modelling and engineering complex systems organised as autonomous interactive components. The multiagent paradigm proved natural and useful not only in building simulations, but also for devising an engineering methodology. To evaluate the proposed approach, we constructed a multiagent model of cell components involved in the metabolic pathways of Carbohydrate Oxidation (CO).
TL;DR: This paper will focus on the local?global relationship in a Self-Organising MAS (SOMAS), and discuss the phenomena of self-Organised Criticality (SOC) and phase transition in SOMAS.
Abstract: In a Multi-Agent System (MAS), there usually exist system-level goals that agents are collectively trying to attain. Generally speaking, the local behaviours and performances of individual agents directly determine which global characteristics of an MAS, for example, the global performance and/or patterns, will emerge and how the system-level goals will be attained. In most cases, such a local?global relationship is neither obvious nor straightforward. In this paper, we will address several issues on modelling emergent complex behaviours in MAS. In particular, we will focus on the local?global relationship in a Self-Organising MAS (SOMAS), and discuss the phenomena of Self-Organised Criticality (SOC) and phase transition in SOMAS.
TL;DR: A novel prototype of reverse engineering equipment is characterised by combining a robotics laser surface measurement system with fuzzy neural network model reconstruction, which validates the expectation that the ANFIS model can match the real surface satisfactorily.
Abstract: This paper reports a design of a novel prototype of reverse engineering equipment, which is characterised by combining a robotics laser surface measurement system with fuzzy neural network model reconstruction. The major novelty of the intelligent equipment includes, firstly scanning complex freeform surface in 3D space using a six degree of freedom robot with laser scanning head, so that 3D point cloud data with specified accuracy can be quickly collected. Secondly, implementing digital model reconstruction using an adaptive network based fuzzy inference system (ANFIS). To demonstrate the efficient applicability of the developed prototype, a group of point cloud data on a surface are collected by the robot measure system, which the first section of the samples is used to train the ANFIS network; therefore a 3D digital model is reconstructed to describe the measured data in a concise equation. Comparison of the output from the reconstructed 3D digital model with the second section of the measured data validates the expectation that the ANFIS model can match the real surface satisfactorily.
TL;DR: The identification of the Generalised Maxwell Slip (GMS) friction model is presented and simulation results are presented to prove the efficiency of the observer in estimating the frictional force even though the model has been approximated.
Abstract: The identification of the Generalised Maxwell Slip (GMS) friction model is presented in this paper. An approximation is proposed to make the friction model linear over its unknown parameters. A bounded disturbance is used to model the approximation error. A robust observer is then applied to estimate the unknown parameters in spite of the approximation error. The approximated friction model is filtered to render it appropriate for use by the observer. Simulation results are presented to prove the efficiency of the observer in estimating the frictional force even though the model has been approximated.
TL;DR: This work presents a 3D agent-based simulation model of the emergent construction behaviour of Macrotermes termites, focusing on construction of the queen's royal chamber and protective trail galleries using available biological evidence.
Abstract: Social insects provide a unique example of emergence and self-organisation. Their ability to exploit environmental complexity to offset individual simplicity presents us with a novel and immensely powerful engineering methodology. In this work we present our 3D agent-based simulation model of the emergent construction behaviour of Macrotermes termites, focusing on construction of the queen's royal chamber and protective trail galleries using available biological evidence. Previous models have been somewhat lacking in modelling individual behavioural responses and in including factors intrinsic to the termite building process. We show how simple response-threshold functions can be combined with pheromone templates and stigmergic mechanisms to mirror experimental work and provide an explanation for poorly understood termite behavioural patterns. We also demonstrate the low-level disorder inherent in such a system and explain how such effects can add a layer of stochasticity to facilitate the self-organised building process.
TL;DR: An EA for Multi-Layer Perceptron (MLP) learning, called Perceptrons Learning using Genetic algorithm (PLG), that gets results comparably better than BackPropagation (BP) is introduced.
Abstract: It is shown through a considerably large literature review that combinations of Artificial Neural Networks (ANNs) and Evolutionary Algorithms (EAs) can lead to significantly better intelligent systems than relying on ANNs or EAs alone. Evolution can be introduced into ANNs at many different levels. This paper focuses on the evolution of connection weights, which provides a global approach to connection weight training especially when gradient information of the error function is difficult or costly obtained. Due to the simplicity and generality of the evolution and the fact that gradient-based training algorithm often have to be run multiple times in order to avoid being trapped in a poor local optimum, the evolutionary approach is quite competitive. This paper takes a step in that direction by introducing an EA for Multi-Layer Perceptron (MLP) learning, called Perceptron Learning using Genetic algorithm (PLG), that gets results comparably better than BackPropagation (BP).
TL;DR: Simulation results show the superiority of the proposed system relative to the classical passive suspension, and signify robustness of the active controller design.
Abstract: The paper presents a design technique for a fixed-structure PD robust controller of car active suspension systems. The design takes into consideration the uncertainty of system parameters, particularly tyre stiffness and body mass. Robustness is achieved by tuning the controller over a set of operating conditions covering the whole range of system parameters, e.g., body mass and tyre stiffness. Particle swarm optimisation (PSO) is used to attain different performance objectives of the system. Settling time of body displacement is minimised, system damping is maximised, and actuator saturation is avoided via control effort reduction. The design of controller parameters is cast in a multi-objective non-linear optimisation problem, and described to ensure the best possible performance. Simulation results show the superiority of the proposed system relative to the classical passive suspension, and signify robustness of the active controller design.