Interval Type-2 Fuzzy Control Systems and Applications

Interval Type-2 (IT2) fuzzy control systems extend conventional Type-1 fuzzy logic by modelling uncertainty through upper and lower membership bounds, collectively termed the footprint of uncertainty. This richer representation enables enhanced robustness in the presence of noise, unmodelled dynamics and parameter variation. A typical IT2 controller comprises fuzzification, rule evaluation using interval‐valued firing strengths, type‐reduction to collapse the interval set and defuzzification to yield crisp control actions. Stability and performance are commonly assured through Lyapunov-based analysis, often recast as convex optimisation problems via linear matrix inequalities or sum‐of‐squares conditions. Recent methodological advances include membership‐function‐dependent approaches that directly exploit the shape of upper and lower membership functions to relax conservativeness, as well as hybrid frameworks that combine deep reinforcement learning with IT2 fuzzy inference to optimise type-reduction under stability constraints. Applications span nonlinear process control, robotics, marine systems, power electronics and biomedical devices, where uncertainty is pervasive and control performance must be maintained across varying operating regimes.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Researchers have advanced the theoretical foundations of IT2 polynomial-fuzzy-model-based control by integrating the properties of upper and lower membership functions into stability conditions. One approach partitions the operating domain into subdomains, applies sum-of-squares optimisation and derives less conservative Lyapunov conditions, while alternative methods adopt polynomial approximations of membership functions to minimise the number of stability constraints without sacrificing information. In the domain of stochastic nonlinear systems, actuator saturation and multiplicative noise have been addressed through interval Type-2 Takagi–Sugeno models. By converting Lyapunov stability criteria into linear matrix inequalities, designers can employ convex optimisation to ensure robust performance of a fuzzy controller under saturation limits, demonstrated in ship steering simulations. On the application front, depth control of remotely operated vehicles in complex underwater environments has been enhanced through an IT2 fuzzy–PID hybrid scheme. Comparative studies reveal that the IT2FPID controller achieves markedly reduced overshoot and faster settling times than both Type-1 fuzzy PID and conventional PID, underlining the practical benefits of interval uncertainty modelling in challenging real-world scenarios.

Interval Type-2 Fuzzy Control Systems and Applications publication trend

The graph below shows the total number of articles in interval type-2 fuzzy control systems and applications across all publications each year (not limited to Nature Index journals).

Technical terms

Interval Type-2 fuzzy set: A fuzzy set characterised by upper and lower membership functions, capturing uncertainty in membership grades.

Footprint of uncertainty: The region between upper and lower membership functions in an IT2 fuzzy set, representing the range of possible membership values.

Type-reduction: The process of converting an interval Type-2 fuzzy set into a Type-1 fuzzy set (or interval of crisp values) prior to defuzzification.

Linear Matrix Inequality (LMI): A convex constraint expressed as a matrix inequality, used to recast stability and performance conditions into tractable optimisation problems.

References

1. Stabilization of Interval Type-2 Polynomial-Fuzzy-Model-Based Control Systems. IEEE Transactions on Fuzzy Systems (2017).
2. Actuator Saturated Fuzzy Controller Design for Interval Type-2 Takagi-Sugeno Fuzzy Models with Multiplicative Noises. Processes (2021).
3. A Robust Control via a Fuzzy System with PID for the ROV. Sensors (2023).
4. Optimization for Interval Type-2 Polynomial Fuzzy Systems: A Deep Reinforcement Learning Approach. IEEE Transactions on Artificial Intelligence (2022).