TY - JOUR
T1 - Fuzzy-Observer-Based Predictive Stabilization of DC Microgrids With Power Buffers Through an Imperfect 5G Network
AU - Vafamand, Navid
AU - Asemani, Mohammad Hassan
AU - Dragicevic, Tomislav
AU - Blaabjerg, Frede
AU - Khooban, Mohammad Hassan
PY - 2020/9
Y1 - 2020/9
N2 - This article investigates the fuzzy model predictive control synthesis of a power buffer for dynamic stabilization of a dc microgrid (MG), which is controlled through a low-latency communication network, such as the one envisioned in fifth-generation (5G). The proposed approach employs a Takagi–Sugeno fuzzy model, a fuzzy observer, and a model predictive scheme to alleviate the effects of the 5G-network-induced delays and data loss of the sensor-to-controller and controller-to-actuator links on the dc MG plant response. By employing the so-called time-stamp technique and network delay compensator (NDC), the delays are computed, and the data loss effects are compensated, thus improving the effectiveness and robustness of the proposed controller. In addition, a time-delay-independent observer is proposed to estimate the states of the constant power loads (CPLs) and the power buffer, based on the measured information. Due to the usage of two NDCs, the presented approach is robust against the network delays and results in a small computational burden. To show the merits of the proposed approach, it is applied to a dc MG that feeds two CPLs. Results show the simplicity of designing the observer-based controller and better robustness against the network delays, compared with the state-of-the-art methods. Additionally, software-in-the-loop simulations are presented to prove the practical applicability of the proposed controller.
AB - This article investigates the fuzzy model predictive control synthesis of a power buffer for dynamic stabilization of a dc microgrid (MG), which is controlled through a low-latency communication network, such as the one envisioned in fifth-generation (5G). The proposed approach employs a Takagi–Sugeno fuzzy model, a fuzzy observer, and a model predictive scheme to alleviate the effects of the 5G-network-induced delays and data loss of the sensor-to-controller and controller-to-actuator links on the dc MG plant response. By employing the so-called time-stamp technique and network delay compensator (NDC), the delays are computed, and the data loss effects are compensated, thus improving the effectiveness and robustness of the proposed controller. In addition, a time-delay-independent observer is proposed to estimate the states of the constant power loads (CPLs) and the power buffer, based on the measured information. Due to the usage of two NDCs, the presented approach is robust against the network delays and results in a small computational burden. To show the merits of the proposed approach, it is applied to a dc MG that feeds two CPLs. Results show the simplicity of designing the observer-based controller and better robustness against the network delays, compared with the state-of-the-art methods. Additionally, software-in-the-loop simulations are presented to prove the practical applicability of the proposed controller.
KW - COMMUNICATION
KW - Computational modeling
KW - Constant power load (CPL)
KW - Delays
KW - MODEL
KW - NONLINEAR-SYSTEMS
KW - Observers
KW - Predictive control
KW - Predictive models
KW - Robustness
KW - Takagi-Sugeno (TS) fuzzy model
KW - Upper bound
KW - dc microgrid
KW - model predictive controller
KW - packet dropout
KW - random network delay
KW - software-in-the-loop (SiL)
UR - https://www.scopus.com/pages/publications/85090923074
U2 - 10.1109/JSYST.2019.2963788
DO - 10.1109/JSYST.2019.2963788
M3 - Journal article
SN - 1932-8184
VL - 14
SP - 4025
EP - 4035
JO - IEEE Systems Journal
JF - IEEE Systems Journal
IS - 3
M1 - 9027093
ER -