A controller design and analysis using asymmetry triangular cloud models

The cloud model is one of the mathematical tools that realize the transformation of quantitative information from qualitative data. Therefore, cloud model theory is widely used in computer science, reliability estimation, nonlinear function approximation, controller design, etc. In general, cloud controller design is known to use Gaussian membership clouds. However, it is computationally expensive because membership cloud computing is a nonlinear operation, and it has a disadvantage that it is difficult to decompose the structure of the controller. This paper proposes a new asymmetric triangular cloud model consisting of linear operations instead of Gaussian functions and, on this basis, develops a controller design method to approximate the output of the controlled plant to the desired value. Furthermore, it is demonstrated that the proposed controller is capable of stability analysis even if the mathematical model of the plant is not given, and it is validated by simulation of electrode up and down control of ultra-high power electric arc furnace and stabilization control of inverted pendulum.