Natsuki Takagi

Adaptive steering controller to improve handling stability for driver-combined-vehicles system

If the dynamics of combined vehicles such as tractor-semitrailer varies greatly, it may be very difficult for inexperienced drivers to achieve good handling stability. Moreover, once combined vehicles become unstable, it is very difficult for all drivers to stabilize vehicles. However, if the behavior of actual combined vehicles tracks a designed desired combined vehicle, the good handling property can be maintained even when the dynamics of actual combined vehicles varies large. In this paper, to achieve good handling property even for large variation of vehicle dynamics, a design method for a desired combined vehicle is shown, and then, an adaptive steering controller is developed so that the actual vehicle tracks the desired vehicle. The developed adaptive steering controller has strong robustness for the uncertainties of vehicle parameters. Moreover, a driver model is introduced to show the characteristics of driver. Carrying out numerical simulations, it is shown that the developed adaptive steering controller is very useful for the driver-combined-vehicles system.

Adaptive control scheme achieving smooth control input in the presence of input saturation

In this paper, the main attention is focused on transient property of control input signal, we propose a novel model reference adaptive control scheme for time-continuous single-input single-output linear systems with input saturation. To improve a control performance, a novel estimator using an obserer for the output tracking error signal is proposed. Using the estimator, it is shown theoretically that the tracking error between the controlled object output and the reference model output can converge to zero when the initial value of the tracking error satisfies a certain condition. Moreover, it is also shown in the case of zero initial tracking errors that the control input signal becomes close to the exact model matching input signal rapidly by setting only one design parameter, and then the input saturation does not occur.

Design of adaptive control systems for unstable plants in the presence of input saturation

It is important to consider input saturation in adaptive control systems to prevent the deterioration of control performance caused by excessive adaptation. This paper deals with a model reference adaptive control problem for a first order time-invariant continuous-time unstable plant in the presence of input saturation. We propose an algorithm considering error feedback and easing reference signal conditions. Moreover the adaptation law is switched according to the input conditions to inhibit excessive adaptation caused by the input saturation. Further, we give a necessary condition for locally asymptotic stability. By this algorithm, design flexibility of the system becomes higher than conventional algorithms. These results are confirmed by simulation studies.

A modified adaptive control scheme in the presence of input saturation

We have proposed a model reference adaptive control scheme (conventional control scheme) for continuous time single-input single-output linear systems with an input saturation. In the closed-loop system using the conventional control scheme, it has been shown theoretically that the tracking error between the controlled object output and the reference model output can converge to zero. Moreover, it has been also shown theoretically that tracking performance can be improved by setting only one design parameter. However, there is the indispensable condition that initial states in the tracking error system must satisfy. To meet the condition, many design parameters have to be adjusted. In the paper, in order to save the time spent to meet the condition for initial states, the main attention is focused on reduction of the design parameters to be adjusted. To achieve the reduction of the design parameters, we modify the conventional control scheme and propose a new adaptive control scheme. As a result of analysing stability of the closed-loop system using the new control scheme, a new condition for initial states is derived. It should be emphasised that we can apply the new control scheme in a larger region of initial states compared with the conventional control scheme.

Adaptive control of systems with input saturation: A scheme using output derivatives of order up to relative degree

In this paper, the main attention is focused on transient property of control input signal, we propose a novel adaptive controller for time-continuous single-input singleoutput linear systems with an input saturation in which i-th derivatives of the output signal (i =3D 1,···, relative degree) are available. To improve the control performance, a novel estimator using an observer for the tracking error signals is proposed. Using the estimator, it is shown theoretically that the tracking error between the controlled object output and the reference model output can converge to zero when the initial value of the tracking error satisfies a condition.

Dynamic state estimation using particle filter and adaptive vector quantizer

Particle filter (PF) is a method for discrete approximation of dynamic and non-Gaussian probability distribution by using numerous particles, and its procedure can execute at high speed and is suitable for on-line applications. However, in conventional methods, a weighted average value or a maximum weighted value of particles is used as a filter output, and information on most particles is disregarded. On the other hand, an adaptive vector quantization (AVQ) algorithm called competitive reinitialization learning (CRL) that can achieve high-speed adaptation without depending on initial conditions has been proposed. Then, in this research, a method for extracting information on shape of probability density distributions by combining PF with CRL is proposed. Moreover, a rapid adaptation performance and the robustness of the proposed method are shown by the simulations.

Design method of ideal vehicle models in adaptive steering driver-combined-vehicle systems

The behaviour of combined vehicles is easily affected by the uncertainties of vehicle parameters. It may be very difficult for the drivers to achieve good handling stability. If the behaviour of actual combined vehicles tracks the designed ideal vehicle models, the good handling property can be maintained for the large uncertain variation of vehicle parameters. Then, the authors have proposed the adaptive steering controllers to achieve good handling property. However, in the conventional schemes, just only one ideal vehicle model is designed for various drivers. For each driver, if an adequate ideal vehicle model is selected, the better handling stability can be achieved. In this paper, a new design method to design ideal vehicle models of adaptive steering driver-combined-vehicle systems for various drivers is proposed. Carrying out numerical simulations, it is shown that good handling performance can be achieved when the actual vehicles track the designed ideal vehicle models.

Robust oscillation control of wheeled mobile robots

In cases where a wheeled mobile robot runs fast on a rough surface, the sensors mounted on the robot’s body may be destroyed due to the body acceleration and oscillation. In this article, we propose a new scheme to reduce the body acceleration at any specified location for mobile robots with the actuators set on the wheel axes. To achieve this, a combined ideal robot model is designed. In the combined ideal robot model, the location where the acceleration performance is at its best can easily be moved by setting only two design parameters. Next, a robust model tracking controller is developed so that the behavior of an actual mobile robot can track the combined ideal robot model. The developed controller has the following useful properties. (1) The body acceleration at any specified location can easily be improved. (2) The developed controller has good robustness for uncertainties in robot mass, pitch and roll moment of inertia of the robot’s body, and the position of the center of gravity.

An adaptive controller to realize smooth input for systems with relative degree less than 3

In this paper, the main attention is focused on transient of control input signal, and then, we propose a novel model reference adaptive controller for SISO systems with relative degree less than three. When we control real devices by using the elementally conventional adaptive controller, we have an experience that high frequency oscillations with large amplitude occur in the control input signals. This is one of serious problems in actual applications. To generate a control input signal without oscillation, a novel estimator using an observer for the output tracking error signal is proposed. Using the estimator, it is shown theoretically that the control input signal converges with any convergent rate to an ideal input signal given by adding the exact model matching input signal and a smooth exponential damping signal. From the property it can be expected that. oscillations are hard to occur in the control input signals. Moreover, it is also proved theoretically that. the transient output error responses can be specified by setting design parameters.

A Design of Adaptive Control Systems with Input Saturation

This paper deals with model reference adaptive control problem for a linear single-input single-output time invariant continuous-time plant with input saturation. We propose an algorithm in which error feedback is effectively used and reference signal conditions are weakened. We have shown that globally asymptotic stability is guaranteed. Using this algorithm, we allow the adaptive system to have higher design flexibility than conventional ones. These results are confirmed by a simulation study