Granular Fuzzy Model with High Order Singular Values Decomposition and Hesitation Fuzzy Granularity

TheHigh order singular value decomposition granular fuzzy modelGranular fuzzy model (GFM) has been widely utilized in many different domains. In this article, a hesitation fuzzy granularity has been used in the proposed granular fuzzy model for quadrotor systems. Serving as a core component in the tensor product mode transformation (TPMT)Tensor product mode transformation (TPMT), the high-order singular value decomposition can generate different accuracy Takagi–Sugeno (TS) fuzzy systems with various reserved singular values. The antecedents and consequences are generated from the core tensor and weighting functions. For the TPMT-based control method, a linear matrix inequality (LMI) based convex optimization problem is utilized to find the feasible gain matrix. The information granularity in the controller design is depicted by the feasibility of LMI conditions related to the system performance. Additionally, the intelligent control strategy for quadrotors adopts a double-loop approach, where the position subsystem employs a fully actuated method and the attitude subsystem employs the hesitation fuzzy granularity-based TPMTTensor product mode transformation (TPMT) control methods. Moreover, a distance measureDistance measure function used to design the controller switching strategy is presented. Experimental results on the MFP450MFP450 quadrotor are provided to demonstrate the proposed intelligent control strategy in different flying stages (takeoff, tracking, landing) of the quadrotor, and the performance of the proposed control methods is analyzed.