Intelligent Tuning of PID Controller to Balance the Shape Memory Wire Actuated Ball and Beam System

Abstract

Íàöèîíàëüíûé òåõíîëîãè÷åñêèé èíñòèòóò, Òèðó÷èðàïïàëëè, 620015, Èíäèÿ Ñèñòåìà «øàðèê áàëàíñèð» ïðåäñòàâëÿåò ñîáîé ñèëüíî íåëèíåéíóþ ìîäåëü, èñïîëüçóåìóþ â ñîâðåìåííîé êëàññè÷åñêîé òåîðèè àâòîìàòè÷åñêîãî ðåãóëèðîâàíèÿ.  ðàáîòå ðàññìîòðåíû âîïðîñû óïðàâëåíèÿ ïîëîaeåíèåì øàðèêà íà áàëàíñèðå ñ èñïîëüçîâàíèåì ïðîâîëî÷íîãî ìåõàíèçìà èç ñïëàâà ñ ïàìÿòüþ ôîðìû, ñêîíñòðóèðîâàííîãî íà ïðèíöèïå ïðîòèâîäåéñòâèÿ. Ïðèâîä èç ñïëàâà ñ ïàìÿòüþ ôîðìû îáåñïå÷èâàåò íåîáõîäèìîå óñèëèå äëÿ íàêëîíà áàëàíñèðà âëåâî è âïðàâî. Äëÿ ýôôåêòèâíîãî ðåãóëèðîâàíèÿ ëþáîé ñèñòåìû íåîáõîäèìî óñòðîéñòâî ñîîòâåòñòâóþùåé êîíñòðóêöèè, ðàçðàáîòêà êîòîðîãî ÿâëÿåòñÿ íåïðîñòîé çàäà÷åé ââèäó îñîáûõ òðåáîâàíèé ê íàñòðîéêå, ðåàëèçàöèÿ êîòîðûõ íåâîçìîaeíà â òðàäèöèîííûõ ðåãóëÿòîðàõ. Öåëüþ äàííîãî èññëåäîâàíèÿ ÿâëÿåòñÿ îïòèìèçàöèÿ êîíñòðóêöèè øèðîêî èñïîëüçóåìîãî ÏÈÄ-ðåãóëÿòîðà ñ èñïîëüçîâàíèåì èíòåëëåêòóàëüíûõ ìåòîäîâ íàñòðîéêè äëÿ îáåñïå÷åíèÿ íàèëó÷øåé ïðîèçâîäèòåëüíîñòè ñèñòåìû. Íàñòðîéêà ÏÈÄ-ðåãóëÿòîðà îñóùåñòâëÿåòñÿ íà îñíîâå ãåíåòè÷åñêîãî àëãîðèòìà è îïòèìèçàöèè ìåòîäîì ðîÿ ÷àñòèö. Äàííûå ÷èñëåííûå àëãîðèòìû øèðîêî èñïîëüçóþòñÿ äëÿ ðåøåíèÿ ñëîaeíûõ èíaeåíåðíûõ çàäà÷, ñâÿçàííûõ ñ ïîâûøåíèåì ïðîèçâîäèòåëüíîñòè. Ïðîâåäåíî ñðàâíåíèå ðåçóëüòàòîâ, ïîëó-÷åííûõ ñ èñïîëüçîâàíèåì ãåíåòè÷åñêîãî àëãîðèòìà è îïòèìèçàöèè ìåòîäîì ðîÿ ÷àñòèö, ñ äàííûìè äëÿ ÏÈÄ-ðåãóëÿòîðà òðàäèöèîííîé êîíñòðóêöèè.Êëþ÷åâûå ñëîâà: ñïëàâ ñ ïàìÿòüþ ôîðìû, ñèñòåìà «øàðèê áàëàíñèð», ÏÈÄ-ðåãóëÿòîð, îïòèìèçàöèÿ, ãåíåòè÷åñêèé àëãîðèòì, îïòèìèçàöèÿ ìåòîäîì ðîÿ ÷àñòèö

Angular control of differential shape-memory alloy spring actuator for underactuated dynamic system

“…A genetic algorithm (GA) is a technique used to find optimal solutions based on a natural selection process that mimics biological evolution. The algorithm modifies a population of solutions continuously and repeatedly and approaches an optimal solution using genetic operators called selection, crossover, and mutation (Ahn and Nguyen, 2008 and Banu Sundareswari et al, 2020). Particle swarm optimization (PSO) is inspired by particles moving in search space and is used to optimize a problem.…”

“…Particle swarm optimization (PSO) is inspired by particles moving in search space and is used to optimize a problem. It begins with a particle known as a swarm, initialized at the random value of its position and velocity, and each time they are updated with a new value of position and velocity based on the best fitness value (Aghababa 2016 and Banu Sundareswari et al, 2020). In both optimization methods, the objective function or the fitness function based on the error criterion is predefined to evaluate the performance of each particle which are the gain values of the PID controller.…”

“…Hence, controlling and stabilizing the ball on the beam at the required position by controlling the beam angle with the actuator output is a complicated task. Various control strategies for the control of SMA-based revolute joint/rotary actuators reported by literature include linear classic proportional–integral–derivative (PID) (Da Silva, 2007), PWM control (Ma and Song, 2003), and nonlinear controllers including fuzzy logic (Kumagai et al, 2006), neural network (Asua et al, 2008), genetic algorithm and particle swarm optimization (Banu Sundareswari et al 2020), feedback linearization, and sliding mode control (SMC) (Ashrafiuon and Jala, 2009). Most linear control systems are not guaranteed or possess the limited opportunity for stability and robustness for SMA actuated systems because of the nonlinearity and parameter uncertainties; therefore, nonlinear controllers are preferred over linear controllers.…”

“…The primary goals of this article are twofold; to claim improvements in a meritorious actuator configuration and eligible control actions regarding the study revealed in Banu Sundareswari et al (2019):The design and investigation of the driving performance of a diagonally connected antagonistic SMA spring actuator for the most significant underactuated unconventional ball on the beam control system.The design and implementation of the well-proven nonlinear sliding mode controller for the angular control of the SMA actuator, as well as optimized (GA, PSO) PID controllers to balance the ball on the beam in their integer and fractional forms.…”

Observer Design for a Variable Moment of Inertia System

“…Krishna [20] used gain-scheduled PID. Banu Sundareswari [21] used intelligent PID tuning. Recently, fuzzy control was also applied to control the ball-and-beam system in the works of Chang [22], Asadi [23], and Minh [24].…”