Robustness of Smooth Sliding Control to Unmodeled Dynamics: Overcoming First-Order SMC and Super-Twisting Algorithm

Author(s):  
Tiago Roux Oliveira ◽  
Liu Hsu
2017 ◽  
Vol 65 (2) ◽  
pp. 233-245
Author(s):  
Y. Wang ◽  
M. Sun ◽  
S. Du ◽  
Z. Chen

Abstract Target manoeuvre is one of the key factors affecting guidance accuracy. To intercept highly maneuverable targets, a second-order sliding-mode guidance law, which is based on the super-twisting algorithm, is designed without depending on any information about the target motion. In the designed guidance system, the target estimator plays an essential role. Besides the existing higher-order sliding-mode observer (HOSMO), a first-order linear observer (FOLO) is also proposed to estimate the target manoeuvre, and this is the major contribution of this paper. The closed-loop guidance system can be guaranteed to be uniformly ultimately bounded (UUB) in the presence of the FOLO. The comparative simulations are carried out to investigate the overall performance resulting from these two categories of observers. The results show that the guidance law with the proposed linear observer can achieve better comprehensive criteria for the amplitude of normalised acceleration and elevator deflection requirements. The reasons for the different levels of performance of these two observer-based methods are thoroughly investigated.


Author(s):  
Lasse Schmidt ◽  
Torben O. Andersen ◽  
Henrik C. Pedersen

This paper discusses the utilization of the so-called twisting algorithm when applied in output feedback position control schemes for electro-hydraulic cylinder drives. The twisting controller was the first second order sliding controller ever introduced, and can structure-wise be considered a straight forward extension of the simplest first order sliding controller, that is, a relay controller. Such a controller may be implemented without the knowledge of system time constants etc., as opposed to the surface based first order sliding controllers which has been presented in numerous contributions in literature. This paper considers the twisting algorithm when applied directly for output feedback control, and with the design based on a reduced order model representation of an arbitrary valve driven hydraulic cylinder drive. The consequence of implementing such a controller with the well-known saturation-, or boundary layer method is discussed, and the control operation inside- and out-side the boundary layer region is considered. Furthermore, the global stability of such a controller is discussed, with emphasis on possible local instability modes. Results demonstrate that the proposed output feedback controller may be successfully applied to hydraulic valve driven cylinder drives, with performance being on the level with a conventional surface based first order sliding mode controller.


2017 ◽  
Vol 40 (5) ◽  
pp. 1457-1470 ◽  
Author(s):  
Hasan Omur Ozer ◽  
Yuksel Hacioglu ◽  
Nurkan Yagiz

In this study, a new high order sliding mode controller (HOSMC), based on super twisting algorithm (STA), is proposed for vehicle active suspensions. It is well known that first order sliding mode controller (SMC) is insensitive to parameter variations and external disturbances. On the other hand, it suffers from chattering present in control signal that may harm the mechanical components of the system. Therefore, HOSMC is preferred in this study that attenuates chattering effectively while preserving its robustness. Proposed HOSMC uses an estimation for the equivalent part of the control signal and uses the STA for the discontinuous part of the control law. Additionally, the controller gains are obtained by offline multi-objective genetic algorithm search. Extensive simulations and experimental results are presented to reveal the performance of the proposed controller. First order SMC is also designed and used for comparison. The results indicate the superior performance of the proposed HOSMC.


2000 ◽  
Author(s):  
Supavut Chantranuwathana ◽  
Huei Peng

Abstract This paper presents adaptive robust controllers for force tracking application in a quarter-car active suspension system. In previous publications (Chantranuwathana and Peng 1999a, 1999b), an active suspension architecture was presented. The overall active suspension system was decomposed into two loops. At the main-loop, the desired force signal is calculated while the sub-loop force tracking controller tries to keep the actual force close to this desired force. An Adaptive Robust Control (Yao and Tomizuka 1997) design technique was used to achieve good force tracking performance in a robust manner under plant uncertainties. It was found that force-tracking of up to 5Hz can be reliably achieved. It is, however, found to be unreliable in experiments, especially when high frequency disturbances are present. In this paper, we will show that unmodeled dynamics and especially, the delay (first order lag) in implementing the control signal is a main cause of the problem. With this insight, three controller modifications are proposed to reduce the effect of the unmodeled dynamics, 1) include the actuator dynamics in the ARC design, 2) cancellation of the actuator dynamics and 3) online-adaptation of an ARC parameter. A number of simulation results will be presented to show the effect of these remedies. The last two modifications were found to be promising for actual implementations.


2019 ◽  
Vol 42 ◽  
Author(s):  
Daniel J. Povinelli ◽  
Gabrielle C. Glorioso ◽  
Shannon L. Kuznar ◽  
Mateja Pavlic

Abstract Hoerl and McCormack demonstrate that although animals possess a sophisticated temporal updating system, there is no evidence that they also possess a temporal reasoning system. This important case study is directly related to the broader claim that although animals are manifestly capable of first-order (perceptually-based) relational reasoning, they lack the capacity for higher-order, role-based relational reasoning. We argue this distinction applies to all domains of cognition.


1984 ◽  
Vol 75 ◽  
pp. 461-469 ◽  
Author(s):  
Robert W. Hart

ABSTRACTThis paper models maximum entropy configurations of idealized gravitational ring systems. Such configurations are of interest because systems generally evolve toward an ultimate state of maximum randomness. For simplicity, attention is confined to ultimate states for which interparticle interactions are no longer of first order importance. The planets, in their orbits about the sun, are one example of such a ring system. The extent to which the present approximation yields insight into ring systems such as Saturn's is explored briefly.


Author(s):  
Richard J. Spontak ◽  
Steven D. Smith ◽  
Arman Ashraf

Block copolymers are composed of sequences of dissimilar chemical moieties covalently bonded together. If the block lengths of each component are sufficiently long and the blocks are thermodynamically incompatible, these materials are capable of undergoing microphase separation, a weak first-order phase transition which results in the formation of an ordered microstructural network. Most efforts designed to elucidate the phase and configurational behavior in these copolymers have focused on the simple AB and ABA designs. Few studies have thus far targeted the perfectly-alternating multiblock (AB)n architecture. In this work, two series of neat (AB)n copolymers have been synthesized from styrene and isoprene monomers at a composition of 50 wt% polystyrene (PS). In Set I, the total molecular weight is held constant while the number of AB block pairs (n) is increased from one to four (which results in shorter blocks). Set II consists of materials in which the block lengths are held constant and n is varied again from one to four (which results in longer chains). Transmission electron microscopy (TEM) has been employed here to investigate the morphologies and phase behavior of these materials and their blends.


Sign in / Sign up

Export Citation Format

Share Document