OUR RESEARCH

Control Technology for Safe and Secure Society through Automation and Optimization

To achieve a sustainable society, innovation driven by science and technology across a wide range of fields is indispensable. In particular, system and control engineering, a multidisciplinary technology, plays a particularly significant role and is expected to contribute to solving various social challenges. Our laboratory has been conducting fundamental research and education in system and control engineering, emphasizing optimization and automation to realize a safe and secure society.

Specifically, we have been working on system and control engineering with the following three groups.
     Energy Management System (EMS) Group
     Vehicle Control (VC) Group
     Battery Management System (BMS) Group
We are also conducting research on fundamental system and control theory focusing on multi-agent control and optimization.

• What is System and Control Engineering ?
• Solve Energy Issues with the Power of Control!
  (EMS Group)
• Control Techniques Enabling Autonomous Vehicles to Operate in Uncertain Environments
  (VC Group)
• Safe and Efficient Operation of Batteries via Control Techiques
  (BMS Group)
• Control Theory of Multi-Agent Systems

---

What is System and Control Engineering?

• Just as a bicycle cannot run without a person riding it, any system requires not only hardware but also the control mechanism that drives it to function as intended. Control mechanisms are embedded in all systems, which is why it is often said that "control is a fundamental principle in science and engineering.
• Systems and Control Engineering is a field of technology that contributes to the development of a better society by
  • elucidating the mechanism of "control",
  • devising safe, accurate, and efficient control alogrithms, and
  • applying them to various systems around the world.
  
• The strong point of systems control engineering is that it is an interdisciplinary technology. In other words, the approach for problem solutions provided by system and control engineering is based on mathematical models and hence can be uniformly applied to problems in a wide variety of fields, including electrical and electronic engineering, mechanical engineering, robotics, aerospace engineering, information and communications, and process engineering, etc.
  
• Systems in our modern society have become increasingly large-scale and complex, and are now a combination of systems in various physical domains (e.g., robots, electric vehicles, sensor networks, smart grids, etc.). To deal with such large-scale complex systems, mathematical problem solution approaches such as systems control engineering play an important role as well as knowledge from individual fields.
  Against the above background, we are conducting research on the modeling and control of various large-scale complex systems aiming to contribute to the realization of a sustainable society.

---

Solve Energy Issues with the Power of Control!
 - EMS group -

• Distributed cooperative energy management of smart grids
  The large-scale introduction of renewable energy such as solar and wind power generation, as well as the liberalization of the power industry, are about to significantly change the nature of the power supply network. Current electricity consumers will be transformed into prosumers (producers and consumers) and will be able to make their own decisions about the amount of electricity they buy and sell. In addition, with the large-scale introduction of renewable energy, the power network will be greatly affected by fluctuations in renewable energy supply due to weather. In anticipation of such a near future, our laboratory is researching distributed cooperative algorithms that safely and optimally operate smart grid electricity from the perspective of distributed optimization and system control, making full use of information and communication networks and energy storage devices (batteries, EVs).
  
• Optimal operation of multi-energy systems
  In order to realize a decarbonized society, it is believed that ZEH (zero energy house), ZEB (zero energy building), and ZED (zero energy district), which are buildings or areas that are energy self-sufficient, are effective. In these frameworks, it is essential to integrate the management (procurement, conversion, storage, and consumption) of multiple forms of energy, including not only electricity but also natural gas and heat. In our laboratory, we are researching optimization methods for energy management that take into account the characteristics of each energy form for multi-energy systems that include electricity, gas, heat, etc., based on the concept of an energy hub.

Selected Publications of EMS Group (Click Here!) 

---

Control Techniques Enabling Autonomous Vehicles to Operate in Uncertain Environments
 - VC Group -

Autonomous vehicles such as mobile robots and drones are expected to play an important role in various fields including logistics, disaster rescue, home services, agriculture, etc. In these fields, autonomous vehicles are required to operate autonomously in uncertain environments, both indoors and outdoors. It will also be possible to accomplish a given task more precisely and more efficiently by having multiple vehicles work in cooperation,
We are conducting basic research on control and estimation algorithms that realize more efficient and reliable cooperative operation of multiple mobile robots or drones in uncertain environments.

• Simultaneous Localization and Mapping with Mobile Robots

  For a mobile robot to accomplish a given task in an uncertain environment, it is necessary for the robot to correctly recognize its own position and the surrounding environment (landmark positions). The technique for this is called SLAM (Simultaneous Localization And Mapping). In this research, we are working to improve the accuracy and efficiency of SLAM by using multiple robots. We are also working on the theory of distributed Kalman filters, which are the basis of multi-robot SLAM.

An efficient maze exploration by two robots based on SLAM information

This video is the simulation result of the following research.
W. Hashimoto and K. Takaba: "An efficient maze exploration method by multiple robots using SLAM information," 63rd Annual Conference of the Institute of Systems, Control and Information Engineers (SCI'19), 2019.

     Other videos of researches on SLAM (Click here!)

• Automatic Flight Control of UAVs

  UAV technologies have been rapidly developing in recent years. We are working on the development of control algorithms for enabling UAVs to autonomously execute given tasks in uncertain environments, including remote flight control with image feedback from built-in cameras, coordinated operations between ground vehicles and UAVs, cooperative transportation by multiple UAVs, and cooperative exloration with multiple UAVs.

Conceptual picture of coverage control with UAVs

Position control of a quadcopter with visual feedback

This is a experimental demonstration of the following paper:
J. Shirai, T. Yamaguchi and K. Takaba: "Remote visual servo control of drone taking account of time delays," Proc. of SICE Annual Conference 2017, pp. 1589-1594, 2017.

Cooperative control between an UAV and a ground vehicle (UAV's landing onto a moving ground vehicle)

By optimal control, the UAV (red) is landing onto the moving ground vehcile (black rectangle) with maintaining the desired approach angle (dashed line).

This video is the demonstration of the results in the following paper.
Y. Zhou and K. Takaba: "Optimal landing control of an unmanned aerial vehicle via partial feedback linearization," Proc. of ICAMechS2019, 2019.

       Other videos of control of UAVs (Click here!)

   ・Formation Control of Mobile Robots
        (Click Here!)

Selected Publications of VC Group (Click Here!) 

---

Safe and Efficient Battery Management via Control Techniques
- BMS Group -

Due to their high energy density and large capacity, lithium-ion batteries have been widely used in many electric devices (especially in electric vehicles). A battery management system (BMS) is essential for the safe and efficient operation of lithium-ion batteries, and the important functions of BMSs should include the estimation of the state of charge (SOC) and state of health (SOH) and the aging diagnosis.n particular, SOC should be estimated online with high accuracy in order to prevent firing due to over-charge and to detect anomalies.
Therefore, we are working to improve the accuracy of SOC and SOH estimations based on mathematical models of batteries, using techniques from the control theory such as Kalman filter and system identification.

Selected Publications of BMS Group (Click Here!) 

---

Mathematical Theory and Control of Multi-Agent Systems

A system in which multiple agents are dispersed across a network and autonomously and distributively control themselves through local interactions and information exchange with neighboring agents to perform cooperative tasks such as synchronization, consensus formation, and formation flight is called a multi-agent system.
Multi-agent systems have applications in various fields, including smart grids, sensor networks, vehicle control, distributed computing, physics, and systems biology. Consequently, the development of control theories for these systems is highly anticipated.

In this laboratory, we are conducting research on developing methodologies of multi-agent control systems design through understanding mechanisms of synchronization and consensus in multi-agent systems.

Our major research topics on multi-agent systems are
- Data-driven distribted cooperative control
- Distributed multi-agent optimization
- Robust synchronization of uncertain agents
- Distributed cooperative control via advanced information technologies such as cloud repositories, block chains, etc.
- Distributed Kalman filters

Synchronization of metronomes
The initially separate metronomes synchronize by interacting through the mobile stand which transmits the oscillations of the metronomes.
Notice that the stand is also slightly oscillating.

Simulation: Synchronization of nonlinear agents over a stochastically switching network
The four colored dots are the states of each nonlinear agent. When they overlap it means that synchronization has been achieved. We conducted a MATLAB simulation to observe how these agents synchronize under the condition that the communication links between agents (black lines) switch stochastically. When some communication links are connected, each agent controls itself based on local information from its neighboring agents.

This video is a demonstration of the following work.
K. Takaba: "Synchronization of passive agents over stochastically switching networks with imperfect prior information," 45th ISCIE Symposium on Stochastic Systems, Okinawa, November 1,2, 2013.

   Selected Publications on Multi-Agent Systems
    (Click Here!)

　

---