A New Type-3 Fuzzy PID for Energy Management in Microgrids

Abstract

More recently, type-3 (T3) fuzzy logic systems (FLSs) with better learning ability and uncertainty modeling have been presented. On other hand, the proportional-integral-derivative (PID) is commonly employed in most industrial control systems, because of its simplicity and efficiency. The measurement errors, nonlinearities, and uncertainties degrade the performance of conventional PIDs. In this study, for the first time, a new T3-FLS-based PID scheme with deep learning approach is introduced. In addition to rules, the parameters of fuzzy sets are also tuned such that a fast regulation efficiency is obtained. Unlike the most conventional approaches, the suggested tuning approach is done in an online scheme. Also, a nonsingleton fuzzification is suggested to reduce the effect of sensor errors. The proposed scheme is examined on a case-study microgrid (MG), and its good frequency stabilization performance is demonstrated in various hard conditions such as variable load, unknown dynamics, and variation in renewable energy (RE) sources.

On a Class of General Type-n Normal Fuzzy Sets Synthesized From Subject Matter Expert Inputs

“…In the latter case, the th k SME would select a group of (typically adjacent) words spanning their ambiguity and prescribe a weight vector W  for aggregating their sentiment over these words. They would then use ( 11)- (12) with this vector to calculate their aggregate trapezoidal parameter sets k UMF and k LMF that would then be employed as their corresponding column entries of the matrices in (11) and (12) for aggregation over the SMEs.…”

“…In recent years, interest has arisen in the use of general type-2 and higher-order fuzzy systems because of their ability to model complex applications involving noisy or uncertain data [1]- [12]. System performance superior to that obtained with conventional fuzzy sets is possible because the underlying modelling better sculpts the state space [1].…”

“…Even within a class of MFs, system performance is highly dependent on shape, so a variety of methods have been proposed for parameterizing MFs. These range from arbitrary specifications and general methods such as histogram-fitting and gradient descent through to specialized methods designed to deal with particular situations such as non-linear behavior and non-Gaussian noise (e.g., [12]) or to optimize for a particular problem (e.g., prediction of the Mackey-Glass time series [13].…”

“…Improved performance of interval type-3 and general type-2 systems over type-1 fuzzy systems has been illustrated in diverse settings, including industrial applications, e.g., flowmeter fault reporting, product temperature prediction and battery management [12]- [16], and financial and health applications requiring time series predictions [17], [18]. Higher-order fuzzy sets have also been employed to represent group or consensus decisions (e.g., [19]- [21]).…”

Stability analysis with general fuzzy measure: An application to social security organizations

“…Moreover, a new type-3 fuzzy logic system with an online optimization scheme is proposed for stabilizing and synchronizing financial chaotic systems [ 41 ]. For more information about the new type-3 fuzzy logic system, see references [ 42 , 43 ].…”