論文

Quantifying long-term risk in large-scale multi-agent systems is critical for ensuring safe operation. However, the high dimensionality of these systems and the rarity of risk events can make the required computations prohibitively expensive. To overcome this challenge, we introduce a graph-based representation and efficient risk quantification techniques tailored for stochastic multi-agent systems. A key technical innovation is a systematic approach to decompose the estimation problem of system-wide safety probabilities into smaller, lower-dimensional sub-systems with sub-safe sets. This decomposition leverages the graph Fourier basis of the agent interaction network, providing a natural and scalable representation. The safety probabilities for these sub-systems are derived as solutions to a set of low-dimensional partial differential equations (PDEs). The proposed decomposition enables existing risk quantification approaches but does so without an exponential increase in computational complexity with respect to the number of agents.

This paper is concerned with a new optimization problem named “phase change rate maximization” for single-input-single-output linear time-invariant systems. The problem relates to two control problems, namely robust instability analysis against stable perturbations and minimum-norm strong stabilization. We define an index of the instability margin called “robust instability radius (RIR)” as the smallest $\mathcal{H}_\infty$norm of a stable perturbation that stabilizes a given unstable system. This paper has two main contributions. It is first shown that the problem of finding the exact RIR via the small-gain condition can be transformed into the problem of maximizing the phase change rate at the peak frequency with a phase constraint. Then, we show that the maximum is attained by a constant or a first-order allpass function and derive conditions, under which the RIR can be exactly characterized, in terms of the phase change rate. Two practical applications are provided to illustrate the utility of our results.

Cells are small, closed spaces filled with various types of macromolecules. Although it is shown that the characteristics of biochemical reactions in vitro are quite different from those in living cells, the role of the co-existence of various macromolecules in cell-size space remains still elusive. Here, using a constructive approach, it is demonstrated that the co-existence of various macromolecules themselves has the ability to tune protein localization for spatiotemporal regulation and a biochemical reaction system in a cell-size space. Both experimental and theoretical analyses reveal that enhancement of interfacial effects by a large surface-area-to-volume ratio facilitates membrane localization of molecules in the cell-size space, and the interfacial effects are alleviated by competitive binding to lipid membranes among multiple proteins even if their membrane affinities are weak. These results indicate that competition for membrane binding among various macromolecules in the cell-size space plays a role in regulating the spatiotemporal molecular organization and biochemical reaction networks. These findings shed light on the importance of surrounding molecules for biochemical reactions using purified elements in small spaces.

In this study, we use graphical and analytical methods to study the collective behavior of a swarm of dynamic agents controlled by the generalized cyclic pursuit (GCP) scheme. The proposed graphical approach reveals the occurrence mechanisms of various formation patterns in terms of agents' dynamics and tuning parameters of the GCP scheme, including the rendezvous, diverging spirals, circular pattern, and spirograph-like patterns. Then, we develop an analytical tool to obtain the crucial information needed to formulate expressions of the circular and spirograph-like steady-state trajectories generated by the GCP-controlled dynamic multi-agent system. In addition, the steady-state formation trajectories of a swarm of dynamic agents are formulated by explicitly considering the agents' general continuous-time linear time-invariant model and the scenario in which more than two pairs of closed-loop poles exist on the imaginary axis of the complex plane. Numerical simulation cases of the GCP-controlled dynamic multi-agent system are presented to demonstrate the effectiveness of the proposed analysis methodology.

Molecular communication (MC) enables cooperation of spatially dispersed molecular robots through the feedback control mediated by diffusing signal molecules. However, conventional analysis frameworks for the MC channels mostly consider the dynamics of unidirectional communication, lacking the effect of feedback interactions. In this paper, we propose a general control-theoretic modeling framework for bidirectional MC systems capable of capturing the dynamics of feedback control via MC in a systematic manner. The proposed framework considers not only the dynamics of molecular diffusion but also the boundary dynamics at the molecular robots that captures the lag due to the molecular transmission/reception process affecting the performance of the entire feedback system. Thus, methods in control theory can be applied to systematically analyze various dynamical properties of the feedback system. We perform a frequency response analysis based on the proposed framework to show a general design guideline for MC channels to transfer signal with desired control bandwidth. Finally, these results are demonstrated by showing the step-by-step design procedure of a specific MC channel satisfying a given specification.

This paper is concerned with a robust instability analysis for the single-input-single-output unstable linear time-invariant (LTI) system under dynamic perturbations. The nominal system itself is possibly perturbed by the static gain of the uncertainty, which would be the case when a nonlinear uncertain system is linearized around an equilibrium point. We define the robust instability radius as the smallest H∞ norm of the stable linear perturbation that stabilizes the nominal system. There are two main theoretical results: one is on a partial characterization of unperturbed nominal systems for which the robust instability radius can be calculated exactly, and the other is a numerically tractable procedure for calculating the exact robust instability radius for nominal systems parametrized by a perturbation parameter. The results are applied to the repressilator in synthetic biology, where hyperbolic instability of a unique equilibrium guarantees the persistence of oscillation phenomena in the global sense, and the effectiveness of our linear robust instability analysis is confirmed by numerical simulations.

© 2017 IEEE. We consider chemical reaction systems that operate in a resource limited situation such as in a closed test tube. These systems can operate only for a finite interval of time before the system variables deviate from the normal operating region. Thus, in order to characterize the performance of the systems, it is important to analyze the duration of the normal operation, which is closely related to the notion of finite-time stability. In this letter, we propose an algebraic optimization approach to analyze finite-time stability of resource limited chemical reactions. Specifically, we present semidefinite programs that compute guaranteed lower and upper bounds of the duration of the normal operation, which we call survival time, for a given set of uncertain initial concentrations. The proposed semidefinite programs provide progressively tighter bounds of survival time by increasing the variables and constraints, allowing for the tuning of the balance between the computational time and the conservativeness of the bounds. We demonstrate the proposed method using the regenerator circuit of DNA strand displacement reactions.

In this paper, we describe our activities on student competitions conducted by connecting multiple universities online. By combining online lectures, tutorials, and various types of groupwork, undergraduate student participants from 11 universities worked together to propose their original ideas in a short period of time. The results were evaluated by faculty members from 16 universities and institutes based on objective criteria. Mixing the students and instructors of various backgrounds promoted open communication and formation of human network among them. Quantitative analysis of various indices elucidated the educational effect of the competition.

Programmable materials have artificially designed physical shapes responding to external stimuli, as well as high design capability and high flexibility. Here, we propose a microfiber-shaped programmable material with an axial pattern of stimuli-responsive (SR) and nonresponsive hydrogels. The SR pre-gel solution was mixed to sodium alginate pre-gel solution for instantaneous gelation with ionic crosslinking and solidified on a nonresponsive hydrogel microfiber with a valve-controlled microfluidic system. A design of microfiber-shaped programmable material (patterned position of SR regions) could be flexibly altered by changing a coded sequence program. We confirmed that the three-dimensional (3D) coil-like structures were self-folded at the patterned SR regions responding to the thermal stimulus and that the chirality of the self-folded 3D coil-like structures depends on the condition of the stimulus to the microfiber. Finally, interaction with objects using the programmable microfiber as a soft actuator was demonstrated. Our microfiber-shaped programmable materials expand possibilities of fiber-based materials in biomimetics and soft robotics fields.

© 2013 IEEE. The correlation between gas concentrations in human breath and diseases has been increasingly revealed in recent years, triggering the need for cheap and easy-to-use gas sensors that can detect diseases in their early stages. The gas sensors used in these clinical applications need to be portable and sensitive, and preferably provide on-demand measurements for prompt diagnosis. In this paper, we propose a portable and cost-effective sensor platform that can quickly identify gas concentrations in human breath. Specifically, we combined quartz crystal microbalance (QCM) sensors with a single board computer, Raspberry Pi, to enable real-time data processing and display of the sensor data. A web server was operated on the single board computer, and thus, on-demand visualization of the sensor data was possible using a web browser. A gas quantification protocol was also proposed based on Freundlich's adsorption isotherm. We demonstrated the real-time monitoring of methyl mercaptan (MM) gas over the internet using the developed small sensor device, which can easily be handled in one hand. The limit of detection (LOD) for MM gas was 107 ppb when the sensors were operated for 20 minutes after gas injection. The developed gas sensing platform is unmatched in terms of size and cost compared with analytical devices though less sensitive. These advantages would allow for wide distribution of the developed device to hospitals and individuals in large quantities, leading for behavioral changes for preventive care.

© 2019 IFAC-PapersOnLine. All rights reserved. In biomolecular communication networks, bacterial cells communicate with each other using a cell-to-cell communication mechanism mediated by diffusible signaling molecules. The dynamics of molecular concentrations in such systems are approximately modeled by reaction-diffusion equations. In this paper, we analyze the ability of cell-to-cell communication systems to attenuate impulsive disturbances with various spatial frequency profiles by computing the integrated squared concentration of molecules. In particular, we perform in-depth study of disturbance responses for an activator-repressor-diffuser biocircuit in the spatial frequency domain to characterize its spatial frequency gain.