First of all
we need to construct an Artificial Intelligence system to perform two tasks:
1.Create any possible design of wind turbine based on certain constraints we applied to the system. These constraints are prepared from the conceptual design of the wind turbine, production capability, physical limitation and material properties. We called it computer generated modelling (CGM).
2.The system will combine some performance indicators (PI) and the information of the model being generated to predict the next candidate which will have high performance indicators.
Using this process loop
we are able to make use of the Artificial Intelligence to self-improve the wind turbine without any human involvement.
At this moment, the system is not able to understand the theory of aerodynamic so it needs help from external tools to provide the performance indicators. Here is where the simulation is applied.
We are using Xflow from Dassault Systèmes to understand the aerodynamic behavior of the wind turbine and create some performance indicators which will be input to our artificial intelligence system. One of the performance indicators is power coefficient (Cp). The power of the wind turbine can be calculated by using rotational speed and torque. In a range of rotational speed, the ratio between mechanical power to wind power will be varied. In general, the ratio will be lower in slow rotational and high rotational speed. Then, we can see a maximum ratio happening in one particular rotational speed in between two extreme cases. We call it the power coefficient.
At the end of the process loop inside Artificial Intelligence system, we will come up the final design, which will be sent to production team to create the overall supportive structure and incorporate into the product. Finally, Xflow will be used to study and minimize the performance impact of supporting structure.