Optimisation of the energy system covering from energy resource mining to final utilization should be based on thermodynamics in the horizon over the entire life cycle of complex systems including the resource collection, transportation, conversion and accumulation processes.
On the other hand, for optimum operation and management of the system, optimum operation under prediction is performed.
It can be expected that feedback of information on actual operation to the phase on system design will contribute to better design.

On the other hand, if the reinforcement learning framework is incorporated in the control and management of the system, it is expected that the control/management system itself will “adaptively evolve”. However, for this purpose, it is necessary to incorporate a mechanism for periodic collection and analysis of sufficient amount of data for grasping the state. For this, it is desirable to gather sufficient information for state estimation as digital data to perform various analysis.

Currently, the digital communication standard represented by Fieldbus has been spreading in the field of process automation. Digitization of various forms of sensor information to connect to the host system via the gateway is in progress. In other words, this means that the sensor system has now become an information network system consisting of devices having both a sensing function, that is, a function of converting physical quantity of a real environment into a signal, ie, digital information, and a communicating function.

In order to effectively function a data driven system such as machine learning, access to a sufficient amount of trusted data is necessary. In addition, by adding various data which has not been used for control and management because it cannot be obtained so far, AI functions can be demonstrated. For this reason, the development of sensing technology for digitizing new real environments is also important in terms of system control and management.

Under these circumstances, we study the design theory for system management of the energy system. Specific research sub themes are as follows.

1. Optimum system design of large-scale distributed energy system in temporal and spatial expansion: How should we organize the energy systems distributed in a region, considering the life cycle for each technology? We will extract such a regional energy system proposal and packaging technology to be developed.

2. Derivation of algorithms to optimize large-scale hierarchical systems with minimum necessary components

3. Development of new sensing system for digitization of new environmental information using robotics technology

Digitization of environment by drone