Phil Barber

Robust adaptive finite-time parameter estimation and control of nonlinear systems

This paper exploits an alternative adaptive parameter estimation and control approach for nonlinear systems. An auxiliary filter is developed to derive a representation of the parameter estimation error, which is combined with an adaptive law to guarantee the exponential convergence of the control error as well as the estimation error. The proposed method is further improved via a sliding mode technique to achieve the finite-time (FT) error convergence. The traditional persistent excitation (PE) is simplified as an a priori verifiable sufficiently rich (SR) requirements on the demand signal. The robustness of the control schemes with bounded disturbances is also investigated. The developed methods are finally tested via simulations.

Adaptive Observer-Based Parameter Estimation With Application to Road Gradient and Vehicle Mass Estimation

A novel observer-based parameter estimation scheme with sliding mode term has been developed to estimate the road gradient and the vehicle weight using only the vehicles velocity and the driving torque. The estimation algorithm exploits all known terms in the system dynamics and a low-pass filtered representation of the dynamics to derive an explicit expression of the parameter estimation error without measuring the acceleration. The proposed parameter estimation scheme which features a sliding-mode term to ensure the fast and robust convergence of the estimation in the presence of persistent excitation is augmented to an adaptive observer and analyzed using Lyapunov Theory. The analytical results show that the algorithm is stable and ensures finite-time error convergence to a bounded error even in the presence of disturbances. In the absence of disturbances, convergence to the true values in finite time is guaranteed. A simple practical method for validating persistent excitation is provided using the new theoretical approach to estimation. This is validated by the practical implementation of the algorithm on a small-scaled vehicle, emulating a car system. The slope gradient as well as the vehicles mass/weight are estimated online. The algorithm shows a significant improvement over previous results.

Robust Scheduling of Sampled-Data Networked Control Systems

In this brief, we consider the robust (in terms of model uncertainty) off-line scheduling problem for networked control systems (NCSs) with time-triggered communication. We model the NCS as a hybrid switched system and we propose an efficient method to calculate the continuous-time

An adaptive observer-based parameter estimation algorithm with application to road gradient and vehicle's mass estimation

{\cal L}_{2}

Controllability, Observability in Networked Control

-gain of the resulting closed-loop system for a given communication sequence. Finding a periodic scheduling policy that minimizes this norm is the (combinatorial) optimization problem that is solved by fast stochastic algorithms. A case study of a networked vehicle control system in a hardware-in-the-loop implementation is provided as an example of application.

Electric vehicles with individually controlled on-board motors: Revisiting the ABS design

A novel observer-based parameter estimation algorithm with sliding mode term has been developed to estimate the road gradient and vehicle weight using only the vehicles velocity and the driving torque from the engine. The estimation algorithm exploits all known terms in the system dynamics and a low pass filtered representation to derive an explicit expression of the parameter estimation error without measuring the acceleration. The proposed algorithm which features a sliding-mode term to ensure the fast and robust convergence of the estimation in the presence of persistent excitation is augmented to an adaptive observer and analyzed using Lyapunov Theory. The analytical results show that the algorithm is stable and ensures finite-time error convergence to a bounded error even in the presence of disturbances. A simple practical method for validating persistent excitation is provided using the new theoretical approach to estimation. This is validated by the practical implementation of the algorithm on a small-scaled vehicle, emulating a car system. The slope gradient as well as the vehicles mass/weight are estimated online. The algorithm shows a significant improvement over a previous result.

The new paradigm of an anti-lock braking system for a full electric vehicle: experimental investigation and benchmarking

Abstract We reconsider and advance the analysis of structural properties (controllability and observability) of a class of linear Networked Control Systems (NCSs). We model the NCS as a periodic system with limited communication where the non updated signals can either be held constant (the zero-order-hold case) or reset to zero. Periodicity is dealt using the lifting technique. We prove that a communication sequence that avoids particularly defined pathological sampling rates and updates each actuator signal only once is sufficient to preserve controllability (and observability for the dual problem of sensor scheduling). These sequences can be shorter than previously established and we set a tight lower bound to them.

Computation of an optimal communication schedule in a nonlinear networked control system using sum-of-squares

The paper introduces the anti-lock braking system (ABS) designed for the all-wheel drive full electric vehicle equipped with four on-board motors. The main features of the ABS under consideration are (i) continuous wheel slip control with feedforward and feedback controller parts, and (ii) versatile actuation architecture realizing pure electric braking, braking with electro-hydraulic decoupled brake systems, and blended braking with operation both of friction brakes and electric motors in regenerative mode. Results of ABS validation demonstrate essential benefits of the developed control strategy for brake performance, precise wheel slip tracking, and drive comfort at braking. Adaptive properties of the proposed ABS are discussed for test cases including the transient road friction.

Information fusion for vehicular systems parameter estimation using an extended regressor in a finite time estimation algorithm

The study presented in this paper discusses developments in the area of anti-lock braking control for full electric vehicles. The main contributions of the paper are the development and experimental validation of the combined electric and hydraulic brake system with application of a continuous anti-lock braking system, which is expected to be more effective than the existing industrial solutions. It covers the topic of high-performance braking and driving comfort under a direct slip control function. The research is related to the full electric sport utility vehicle equipped with four individual on-board motors and a decoupled electrohydraulic brake system. The brake controller architecture includes functions of the continuous anti-lock braking system strategy, a brake blending algorithm aimed at minimization of the friction brake torque and operational limitations of the electric brakes. The developed brake controller was subjected to different validation procedures but, within the framework of this paper, emergency braking tests on a wet surface with a low coefficient of friction are considered. The results obtained demonstrate significant improvements in the braking performance, the driving comfort and the control performance for continuous anti-lock braking control of the electric vehicle compared with those of diverse vehicle configurations and, in particular, with those of a sport utility vehicle of the same type equipped with an internal-combustion engine and a conventional hydraulic brake system.

Optimal and Robust Scheduling for Networked Control Systems

Abstract In this paper, we consider the design of a control optimal communication schedule in nonlinear networked control with nonlinear affine plant systems. We focus here on periodic communication schedules in control systems with time-triggered control communication and assume the existence of predesigned controllers which need to be optimally integrated via the control communication system. For this, we formulate a quadratic cost function for the nonlinear affine plant. The system and the cost are first discretized using an exact mixed discretization approach which allows to lift the problem from the continuous-discrete domain into a purely discrete-time representation. Periodicity due to the fixed communication schedule is dealt with the discrete lifting technique. To allow for the computation of the cost, nonlinearities are approximated in terms of a Taylor series which allows the use of the sum-of-squares framework. Once a cost function of the networked control system for a given fixed communication schedule is defined, it is then possible to compare and choose those sequences that minimize this cost for optimal controller integration.