Fast nonlinear model predictive controller using parallel PSO based on divide and conquer approach

The high computationally expensive nature of nonlinear optimisation algorithms leads to limitations of its use in a real-time era. So, this is the main challenge against researchers to develop a fast algorithm that is used in real-time computations. This paper proposes a fast nonlinear model predict...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:International journal of control Ročník 96; číslo 9; s. 2230 - 2239
Hlavní autoři: Diwan, S. P., Deshpande, Shraddha S.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Abingdon Taylor & Francis 02.09.2023
Taylor & Francis Ltd
Témata:
Algorithms
asynchronous particle swarm optimisation
Control algorithms
Control theory
divide and conquer approach
fast dynamic systems
Hybrid systems
Nonlinear control
Nonlinear model predictive control
Particle swarm optimization
Predictive control
Real time
rotary inverted pendulum
ISSN:0020-7179, 1366-5820
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The high computationally expensive nature of nonlinear optimisation algorithms leads to limitations of its use in a real-time era. So, this is the main challenge against researchers to develop a fast algorithm that is used in real-time computations. This paper proposes a fast nonlinear model predictive control algorithm that utilises a parallel particle swarm optimisation with synchronous and asynchronous methods inclusion, that handles nonlinear optimisation problems with constraints. The additional divide and conquer approach of the proposed algorithm improves the speed of computation and disturbance rejection capability which proves its efficacy in real-time applications. A highly nonlinear fast dynamic real-time inverted pendulum system with hybrid embedded hardware platform (ARM + FPGA) is used to validate the performance of this algorithm under constraints. The solution presented in the paper is computationally feasible for smaller sampling times i.e. in miliseconds and it gives promising results with synchronous and asynchonous parallel PSO compared to the state-of-art PSO algorithm.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0020-7179
1366-5820
DOI:10.1080/00207179.2022.2087739