This paper describes a partitioning method of region and index of choosing number for heterogeneous swarm of unmanned vehicles using multiplicatively weighted Voronoi tessellation and simple formulation. Although recent researches have verified the effectiveness of swarm of unmanned vehicles at the time of natural disaster etc. and many researches related to them have been made, most of them are homogeneous swarm. On the other hand, there are few researches related to heterogeneous swarm because of their complexity and difficulty of generalizing and formulating models. This paper proposes the method of generalizing heterogeneous swarm model and partitioning method of searching region which is one of the most possible missions for them. In this paper, Voronoi tessellation in computational geometry has been applied which is one of the representative partitioning methods in general.

The L_q/L_p Hankel norms of continuous-time LTI systems characterize the effect of the past L_p (p=1,2,∞) inputs on the future L_q (q=1,2,∞) outputs. Computing some of these norms of general LTI systems is known to be difficult, but recently it has been shown that we can avoid such difficulties in part when we deal with positive systems. On the other hand, studies on the l_q/l_p Hankel norms of discrete-time LTI systems are relatively scarce. In view of this fact, in this paper, we focus on the Hankel norms of discrete-time LTI systems. We first show that we face the same difficulties in computing the l_q/l_p Hankel norms of general discrete-time systems, and then we clarify that such difficulties can be avoided in part when dealing with positive systems. In particular, we show that some of the l_q/l_p Hankel norms of discrete-time positive systems can be described in closed form and hence easy to compute. Furthermore, we also characterize the l_q/l_p Hankel norms of positive systems in the form of semidefinite programming problems and linear programming problems. These allows us to compute the upper bounds of the l_q/l_p Hankel norms of uncertain positive systems by means of convex optimization.

In recent years, the labor force in construction industries is declining, and even inexperienced workers need to achieve high productivity. In particular, hydraulic excavators are widely used in construction sites, however, the operation is complicated and it takes time to learn the skills. Therefore, the productivity in the site is affected depending on the skills of operators, supportive measures are required. In a previous research, focusing on the velocity of the Center of Mass (CoM) of the attachment, a control to assist the digging operation was proposed, and it was confirmed that smooth digging work can be performed even by novice operators. However, the control performance for the nonlinear characteristics of excavation caused by reaction force from the soil is not sufficient. In this paper, the digging control using the database-driven control to respond the nonlinear characteristics for hydraulic excavators is proposed. The effectiveness of the proposed method is verified by implementation for a hydraulic excavator.

This paper considers decentralized demand optimization of the mixed continuous/discrete-valued output power equipment. Each power consumption equipment determines own set-point as a solution to the individual optimization problem. We investigate the real-time pricing strategy to align the individual decision making and the optimal solution that satisfies the power flow balance constraint. The proposed decentralized management methodology may not work and, due to the existence of discrete-valued output power equipment, the resulting time response may cause chattering phenomena. We investigate two modifications in the proposed decentralized management methodology which mitigate the chattering phenomena. The one refines the cost function for the continuous-valued output power equipment. The other switches the operating conditions of discrete-valued output power equipment. The numerical case studies show the effectiveness of proposed decentralized management methodology.

This paper constructs a model of wheelchair dynamics in a data-driven manner. In particular, we focus on the forward movement of the wheelchair, for which we adopt a Linear-Parameter-Varying Finite-Impulse-Response (LPV-FIR) model. To avoid the over fitting behavior, we employ the Bayesian estimation method. We also show the identified model is effective for the model predictive control (MPC).

In this paper, we present a decentralized controller for multi-agent systems that can autonomously set some of the control inputs by the agents to zero, depending on the relation between an output and the target. Specifically, a nonlinear oscillator is employed in our proposed controller of agents, inspired from the reflective interlimb coordination in quadruped locomotion. The control input is binary ON/OFF, and it switches depending on the oscillator state. Furthermore, our proposed controller does not require information of other agents, but it only uses the broadcasted control error. The effectiveness of our proposed controller is verified by the speed control experiment of a three-motor system.

This paper proposes a novel car sharing algorithm which changes OD (Origin-Destination) and departure time to the user in exchange for the charge discount, and aims to prevent the bias of the number of vehicles in one-way car-sharing stations. First, a one-way car sharing model based on discrete event systems is described. Next, we will explain how to calculate the optimal combination of the origin, destination, and the departure time in real time, and the method of deciding on incentives that meet strategy-proofness and individual rationality based on mechanism design. Finally, the increase in demand acceptance by the proposed method is confirmed by numerical simulation.

In this paper, we discuss a performance of a controller obtained by using VIMT. Here, we focus on a relationship between a controller embedded in the system and a class of controller that can achieve the specification defined by the user. We provide a method of performance improvement for VIMT by using pre-filter. Finally we give a numerical example to illustrate the validity of our results.

In this paper, we propose a novel Smart Parking System which determines optimal parking lot allocation based on matching theory. In this Smart Parking System, parking lot allocation is determined considering both driver's preference and parking lot manager's preference. Then, there is a possibility that a parking lot allocation is not possible due to no vacant parking space at current time and so on. For the driver who cannot be allocated a parking lot, Prior Reservation System (PRS) conducts the optimal parking lot reallocation considering waiting time of each parking lot until a vacant parking space is available. Also, the driver in PRS is satisfied with individual rationality. So, optimal parking lot allocation including rematching in PRS is determined based on matching theory. Finally, numerical simulation results show the effectiveness of proposed method.

In this study, assuming a case that two of the three reaction wheels mounted on the spacecraft failed, we examined a method of controlling the attitude with the remaining one reaction wheel. Since what can be done with one reaction wheel is limited, we propose an attitude control method for maximizing the amount of solar power generation using Monte Carlo economic model predictive control, and validated the effectiveness of the method by simulation. Extensive simulations were conducted to verify whether the proposed method is effective for various initial conditions, and it was confirmed that the power generation increased in most cases. Some examples of how the proposed control method behaved are shown and discussed.