Shinsaku Izumi

Randomized automated demand response for real-time pricing

Automated demand response (ADR) is an essential technology for demand management such as real-time pricing. A major issue for ADR is the design of the units which control electric supply to each electric devise according to the price, so as to benefit both the supplier and consumer sides. Although a kind of universal design principle must be useful for the issue, it has never been established so far. This paper thus attempts to derive a design principle of the ADR units for real-time pricing and proposes ADR units based on it. First, as a design principle, it is clarified that the heterogeneity of the consumer-side actions is essential to control the total electric power consumption. Based on it, we propose randomized ADR units to artificially produce the heterogeneity. In each proposed unit, a random number is generated and electric power is provided to the connected electric devise only if a price-dependent condition is satisfied for the resulting random number. The proposed units enable the consumers to automatically buy electricity with a low price and allow the supplier to control the total consumption. They also guarantee the scalability of the resulting real-time pricing system.

On a relation between graph signal processing and multi-agent consensus

This paper investigates the relation between graph signal processing and consensus of multi-agent systems. The graph signal processing is a technique to process graph signals, that is, signals whose values are on the vertices of graphs and whose structures are specified by the edges. By considering the combination of the states of agents and the graph describing the network structure between them as a graph signal, we show that the multi-agent consensus corresponds to low-pass filtering of the graph signal. This reveals a connection between the two distinct areas, i.e., the graph signal processing and the control of multi-agent systems. In addition, we provide a design method of consensus controllers based on the graph signal processing. In the proposed method, the controllers of agents are designed so that the spatial frequency of the states of the agents becomes a desired one. This enables us to construct controllers of multi-agent systems in the spatial frequency domain.

Robotic Mass Games by Distributed Hybrid Controllers

Abstract This paper addresses a problem of finding distributed controllers for the robotic mass games whose purpose is to let mobile robots organize themselves into a formation displaying a given grayscale image. As a solution to this problem, we develop three-mode hybrid controllers by combining ideas of the conventional coverage control and the halftone image processing. The roles of the three sub-controllers are to remove extra robots from the mass game field, to gather them unobtrusively, and to organize player robots into a formation displaying the image, respectively. The performance of the proposed controllers is demonstrated by numerical simulations with the standard images.

Real-Time Pricing by Data Fusion on Networks

This paper presents a framework of real-time pricing by data fusion of power consumption, i.e., estimating the total power consumption of consumers by sharing the information on local power consumption that each power source has. The problem addressed here is to find power estimators embedded in power sources (or embedded in access points provided in advance) and a power price controller such that the total power consumption tracks a given reference input. As a solution to this problem, we present power estimators with a consensus protocol and a proportional–integral type power price controller. We then show that our solution achieves the reference tracking of the total power consumption in a distributed manner.

Stability analysis of power systems with photovoltaic generators

This paper analyzes the stability of a power system in which a synchronous generator and a photovoltaic (PV) generator supply the power to an infinite bus. The infinite bus corresponds to a large system to which the two generators are connected. The problem considered here is to investigate the existence of the equilibrium points of the system and their stability. For the problem, we derive a sufficient condition on the magnitude of the PV current for the existence of the equilibrium points. In addition, we analyze the stability of the equilibrium points, and show that the equilibrium points found under the derived condition are stable. These results clarify the impact of the penetration of the PV generator on the existence of the stable equilibrium points of the system.

Stabilization by controller networks

This paper addresses a problem of finding a controller network to stabilize linear time-invariant plants. Here, the controller network means an architecture that networked controllers cooperatively determine the control input applied to a plant by exchanging the information with their neighbors. As a solution to this problem, we propose a controller network which acts as a state feedback controller by a dynamic consensus algorithm. By combining a coordinate transformation including a parameter and Gershgorin theorem, we derive a condition for the gain to stabilize the resulting feedback system. We also present a relation between the stabilizing gain and the network topology. These results will be useful for designing a controller network that is robust against changes in the network topology. Finally, the validity of these results is demonstrated by numerical simulations.

Coverage Control Inspired by Bacterial Chemotaxis

This paper considers distributed control of multi-agent systems, inspired by chemotaxis of bacteria. The chemotaxis is the biological phenomenon that organisms sense the concentration of a chemical in an environment and move toward or away from the highest concentration point. The problem considered here is a coverage problem, i.e., a problem of finding distributed controllers to steer agents so that they are placed uniformly on a given space. As a solution to this problem, we propose distributed controllers based on the mechanism of the chemotaxis. With the proposed controllers, each agent performs forward movement or random rotation depending on the difference between the current achievement degree of the coverage and the previous one. The performance of the proposed controllers is demonstrated by a numerical experiment. This result shows that multi-agent coordination in a distributed manner can be achieved by using the mechanism of the chemotaxis.

Real-time pricing based on networked estimators

This paper gives a framework of real-time pricing based on networked power demand estimators, that is, systems estimating total power demand by exchanging information with their neighbors. The problem considered here is to design the estimators and a power price controller such that the total power demand tracks a given reference input under a constraint on the range of the power price. As a solution to this problem, we propose consensus-based estimators and a proportional and integral controller, and derive conditions for the gains to achieve tracking the reference input. The result is demonstrated by a numerical simulation.

Nonlinear Control of Two-Link Planar Manipulator Using Unstable Property of Zerodynamics

In this paper, a control method for an underactuated two-link planar manipulator using a partial linearization method, which is a one of the nonlinear control method based on structure theory, is proposed. In proposed method, for controlling of 2nd link, sliding-mode type partial linearization method is used, and for 1st link, the unstable property of Zerodynamics(it is dynamics of 1st link) is used, respectively. By analyzing a relationship between behavior of Zerodynamics and desired value of 2nd link, we find the property of Zerodynamics is changing by desired value of 2nd link. Therefore, the control of both links is able to achieve by single control input by deciding the desired value of 2nd link based on behavior of Zerodynamics. Finally, numerical simulations are performed to show the effectiveness of the proposed method.