CTI Engineering Co., Ltd. develops an AI-based linked operation model to provide support aimed at facilitating the suitable operations of multiple dams

To provide support to dam management sites dealing with issues that include responding to climate change and securing human resources, CTI Engineering Co., Ltd. has developed AI technology that predicts suitable dam operations. As a result, we have started testing technical services that can be utilized for various aspects of dam management, including the ability to predict linked operations for multiple dams in the same drainage basins--such as linked operations during floods to prevent or reduce downstream damage as well as linked operations during normal times to more efficiently use limited water resources--and then provide this information to dam managers.

To start providing full-scale service in the near future, we will work on improving our precision and propose the utilization of this new technology to dam management sites.

　

1. Dam operation issues related to responding to natural and social conditions

In cases where there are multiple dams in a river basin, to minimize the flood damage within the river basin as a whole, there is a need for operations that link multiple dams. In addition, even during normal times, there is a need to achieve optimal integrated management that considers the characteristics of each dam--including the water storage capacity and how easily water accumulates--to more efficiently use limited water resources given the increase in the frequency of droughts.

　

As the functions and operations necessary for dam management become more advanced as described above, it is becoming increasingly difficult to reduce costs and personnel at dam management sites in order to secure highly experienced operators.

　

2. Technical characteristics

To implement efficient, effective operations of multiple dams during floods and droughts, we have applied a reinforcement learning system with conditions that include the rainfall, water storage capacity, and downstream river flow rate in order to develop a system that utilizes AI to predict and facilitate optimal operations in complicated situations, including situations involving extreme weather conditions and the linked operation of multiple dams.　

　

　　

As shown in figures 1 and 2, utilizing this developed system enables linked operation in cases where multiple dams are set up in river basins, which is more effective than operating according to operating rules.

　

3. Provided services

Our recently developed system--which is intended to facilitate optimal dam checking--can be applied not only to individual dams but also to integrated management in cases where multiple dams have been set up in a river basin, and we provide it as a package system to achieve more efficient, effective dam operation during floods and droughts.

　

4. Future initiatives

To achieve implementation in the future, we are striving to further improve the prediction precision and realize the ability to respond to more complicated conditions as we work on achieving the utilization of our new model at dam management sites. In this way, we are endeavoring to reduce flood and drought damage.

　

　

　Figure 1. Example of linked operation prediction during a flood to prevent or reduce downstream damage

　

　

Figure 2. Example of linked operation prediction during normal times to efficiently use water resources