Platooning of vehicles is expected to be a solution for drivers shortage, traffic congestion, and thereby to reduce gas emission. In this technique, it is critically important to ensure a good response to the velocity change of the leader vehicle, as well as robustness against uncertainty and disturbance. To this end, we propose an adaptive two degree-of-freedom control method based on feedback error learning (FEL) in this paper. In this method, plant uncertainty is addressed by online tuning of feedforward controller by means of input and output signals of the plant. Robust FB control also addresses uncertainty/disturbance of the system. Numerical simulation and real-machine experiment have been carried out to verify effectiveness of the proposed method.

In this paper, we consider a control system design problem for multi-rate systems with long input periods. The proposed control system can be regarded as a periodically time-varying system. Then, the periodically time-varying system can be dealt with as a time-invariant system by using the cycling technique. This paper presents a method to design multi-rate state feedback gains for disturbance reduction. Then, analysis about the l₂-induced norm from the disturbance to the output with the proposed multi-rate state feedback gains is reduced to an LMI optimization problem. Then, the design problem of the time-varying gains is also described as an LMI optimization problem. The effectiveness of the multi-rate state feedback control is shown by numerical examples.

Human-assisted control using the Control Barrier Function (CBF) for nonlinear control systems such as two-wheeled and four-wheeled vehicles has been extensively studied in recent years. However, the effectiveness is only demonstrated in straight-line experiments, not steering ones. Further, it is not easy to design steering assist using the conventional method. In this paper, we design a control barrier function using the system resuscitation transformation proposed by Takai et al. and design steering assist control law with smooth assist characteristics for vehicle-type mobile robots. Then, a local absolute coordinate system is introduced to measure the positions of obstacles and robots. Finally, a steering assist control system is proposed based on the data obtained in the local absolute coordinate system. Experiments with a vehicle-type mobile robot confirmed the effectiveness of the proposed method.

We propose a design method of the tracking control Lyapunov function (TCLF) as a local trajectory tracking control method for a two-wheeled mobile robot. First, we derive an error system by taking an error between a state and a reference state. Second, we linearize the obtained error system by dynamic extension. Then, we design a control Lyapunov function (CLF) with a virtual state for the linearized system. Finally, we apply the minimum projection method to the obtained CLF to design a TCLF without a virtual state for the original nonlinear system. We show the effectiveness of our proposed method by actual experiments.

In this paper, we address real-time lighting optimization in a building based on so-called generalized primal-dual dynamics with a control barrier function. We first formulate a lighting optimization problem relying on time-varying human behavior as a linear programming by using an ideal mathematical model of the lighting system. The generalized primal-dual dynamics is then introduced as a solution to the linear programming. We then present a novel optimization algorithm which incorporates the control barrier function-based controller into the generalized primal-dual dynamics in order to satisfy the constraint of lower bound illuminance in transient response. In addition, we develop a lighting simulator that duplicates a floor in a real building, allowing to implement real-time lighting controller by combining the software UC-win/Road and MATLAB/Simulink. The effectiveness of the proposed algorithm is demonstrated through simulations on the developed simulator.

In this study, I expands the result on Virtual internal model tuning so as to address the input limitation explicitly. The proposed method can realize desired closed-loop response even when the initial control signal is saturated. I verified the validity of the proposed methods via experiment verifications.