Shape Memory Alloy Wire Actuators Research Articles

A comparative analysis between deep neural network-based 1D-CNN and LSTM models to harness the self-sensing property of the Shape Memory Alloy wire actuator for position estimation

Abstract The article presents a performance based comparative analysis of popular deep neu-
ral network(DNN) models such as 1D-CNN and LSTM for Shape Memory alloy(SMA)-
based wire actuator position estimation. These DNN models utilize the self-sensing
property(SSP) for position prediction of the SMA actuator. The phase dependent elec-
trical resistivity of SMA wire act as SSP, where the electrical resistivity is in the form
of resistance acts as inputs to the proposed models for precise estimation of current
position of the SMA actuator. For effective position control of the actuator, an ac-
curate sensor feedback is required, utilizing SSP results in the elimination of external
position sensor. This will improve the overall system in terms of more compactness and
reduced interface complexity. Coming to DNN models, 1-D CNN has been meagerly
explored in the current literature landscape for self-sensing prediction of SMA actua-
tors, these 1-D CNN models are becoming quite popular for time series prediction for
various applications and are emerging as an alternative to LSTM models. In this paper,
a novel implementation of a 1D-CNN model for SMA actuator position estimation has
been done. A comparative analysis between 1D-CNN and LSTM has been done for
prediction capability and inference speed based on performance measures such as Mean
Square Error(MSE), Mean Absolute Error(MAE), sMAPE(symmetric Mean Absolute
Percentage Error), data distribution and average inference speed. This comparison
shows that 1D-CNN has matching performance with the LSTM model with respect to
the prediction capability, however 1D-CNN offers faster inference speed. This analy-
sis can be useful for choosing suitable DNN model for deployment in low computing
platform such as micro-controller for SMA actuator based real time applications where
time latency is a critical parameter.

A comparative analysis between deep neural network-based 1D-CNN and LSTM models to harness the self-sensing property of the Shape Memory Alloy wire actuator for position estimation

Abstract The article presents a performance based comparative analysis of popular deep neu-
ral network(DNN) models such as 1D-CNN and LSTM for Shape Memory alloy(SMA)-
based wire actuator position estimation. These DNN models utilize the self-sensing
property(SSP) for position prediction of the SMA actuator. The phase dependent elec-
trical resistivity of SMA wire act as SSP, where the electrical resistivity is in the form
of resistance acts as inputs to the proposed models for precise estimation of current
position of the SMA actuator. For effective position control of the actuator, an ac-
curate sensor feedback is required, utilizing SSP results in the elimination of external
position sensor. This will improve the overall system in terms of more compactness and
reduced interface complexity. Coming to DNN models, 1-D CNN has been meagerly
explored in the current literature landscape for self-sensing prediction of SMA actua-
tors, these 1-D CNN models are becoming quite popular for time series prediction for
various applications and are emerging as an alternative to LSTM models. In this paper,
a novel implementation of a 1D-CNN model for SMA actuator position estimation has
been done. A comparative analysis between 1D-CNN and LSTM has been done for
prediction capability and inference speed based on performance measures such as Mean
Square Error(MSE), Mean Absolute Error(MAE), sMAPE(symmetric Mean Absolute
Percentage Error), data distribution and average inference speed. This comparison
shows that 1D-CNN has matching performance with the LSTM model with respect to
the prediction capability, however 1D-CNN offers faster inference speed. This analy-
sis can be useful for choosing suitable DNN model for deployment in low computing
platform such as micro-controller for SMA actuator based real time applications where
time latency is a critical parameter.

Trajectory tracking control of parallel manipulator actuated with shape memory wire

This study describes the design of a parallel spatial manipulator with four degrees of freedom actuated with shape memory alloy (SMA) wire to validate the use of SMA in complicated mechatronics systems. The manipulator has a closed kinematic structure, which includes a fixed base and a moving square platform (end effector). The four arms of the manipulator are SMA wires fastened between the fixed base and the end effector. SMA wire-based actuators replace bulky conventional revolute actuators. This work spotlights the development of an actuator model, dimensional analysis, and design of cascade control strategies of various PID and sliding mode controller in their integral and fractional order configurations. Experimental evaluation of the actuator is performed through trajectory tracking to quantify the different controller configurations. The experimental results indicate that the parallel manipulator associated with SMA wire actuators is the best alternative to conventional motion stages for highly precise micro-positioning and tracking applications in the fields of 3D printing, intricate surgical operations, the medical and pharmaceutical industries, and flight and gaming simulators.

The driving characteristics of bidirectional SMA wire actuators - Theoretical modeling and experimental testing

Bidirectional shape memory alloy (SMA) wire actuators play a critical role in driving applications thanks to their characteristics in memory shaping and good overall performance. SMA wire actuators were often studied at a single load type, and there is no comprehensive comparison investigation of different SMA wire actuators in the literature. In this paper, the driving characteristics of four available types of SMA wire actuators, namely, dead-load, biased spring, differential type, and biased superelastic SMA wire actuators are modelled mathematically and then tested in laboratory settings. The characteristics of SMA wire actuators are quantitatively measured and then compared under a range of application scenarios. The output displacement and thermal-electric models of NiTi SMA wires are derived based on the Brinson constitutive model with stress effects in consideration. A customized SMA wire actuator test device is developed to evaluate the theocratical models and benchmark testing four types of SMA wires and their driving characteristics, including structure, design, displacement, response and recovery time, energy efficiency and hysteresis properties. The SMA wire phase transformation characteristics of the four actuators are tested under different activation currents, and then quantitatively compared. In conclusion, this paper adds theoretical and experimental significance to the literature on SMA wire actuators. The reference value on the phase transformation characteristics of SMA wires can be used to guide SMA application in industrial designs and enable theoretical studies in complementary research areas.

Systematic Methodology for an Optimized Design of Shape Memory Alloy‐Driven Continuum Robots

Continuum robots stand out due to their high dexterity, which allows them to effectively navigate in confined spaces and dynamic environments. At present, motor‐controlled tendons represent the most prominent actuation method used in continuum robots which require large drive units, increasing the overall system size and weight. A potential alternative technology to overcome those limitations is represented by shape memory alloy (SMA) wire actuators, which are characterized by extremely high energy density and flexibility, leading to a reduction of the size, weight, and design complexity of continuum robots. The complex thermomechanical behavior of SMA wires, however, makes the design of SMA‐based applications a challenging task, and systematic approaches to design SMA‐driven continuum robots are poorly understood. To overcome this issue, this article presents a novel systematic methodology for designing SMA‐driven continuum robots capable of motion in a three‐dimensional environment. First, the kinematic relationship between SMA wires and continuum robot deformation as well as the required actuator force in quasi‐static conditions, is mathematically described based on the assumption of a constant curvature deformation. Subsequently, the model is validated by in‐plane experiments for different design parameters. Based on the results, a fully integrated, antagonistic SMA continuum robot is built and validated.

Effect of polyurethane coating on mechanical response of shape memory alloy wires

In micro-electro-mechanical systems, the risk of short circuit arises when shape memory alloy (SMA) wire actuators mistakenly come into contact with other metal parts. Coating the SMAs with insulating layer can solve this problem. However, unlike traditional conductive wires that experience minimal length changes with current, SMA wires undergo significant length variations during electrical execution. This can affect the bonding between the coating and the SMA wires, as well as the mechanical response of the actuators. Therefore, this study investigated the effect of an insulating coating on the mechanical response of SMA wires. The results show that a layer of polyurethane coating provides long-term insulation to the SMA wires. While this coating reduces the recoverable strain of SMA wires under passive heating, its effect on actuation strain under electrical heating is relatively minor. Additionally, the coating enables the SMA wire to work smoothly by relieving overshooting and vibration phenomena during execution, and helps the wire work in a wider current range than an uncoated wire. Moreover, the coating enhances the response speed of SMA wires during power-on but reduces the response speed during power-off, this influence weakens as driving currents increase. These distinct mechanical responses can be primarily attributed to the coating’s influence on heat transfer behavior of the SMA wires.

Actuation response of typical biased shape memory alloy wire under variable electric heating rates: Experimental investigation and modeling

Electric heating is commonly used to induce the inverse martensitic transformation in shape memory alloy (SMA) wire actuators. Different actuation responses can be achieved when adjusting the heating currents. In this study, typical biased NiTi SMA wire actuators were tested under the heating currents of 1.5 A, 3 A, 4 A, and 5 A. It was found that SMA wire actuators under three typical biases (dead weight, linear spring, dead weight & linear spring) behaved differently as the heating current increased. Moreover, under dead weight bias and mixed bias, high heating currents tended to cause dynamic responses during actuation. The maximum inertia load induced by dynamic responses was 2.65 times as much as the static load. In contrast, the actuation response under linear spring bias was relatively steady. A model based on the SMA constitutive equation and the dynamic equation was also established, which was capable of simulating the dynamic actuation responses of SMA wire actuators under high electric heating rates.

Owl-Neck-Spine-Inspired, Additively Manufactured, Joint Assemblies with Shape Memory Alloy Wire Actuators.

Nature provides a considerable number of good examples for simple and very efficient joint assemblies. One example is the enormously flexible cervical spine of American barn owls, which consists of 14 cervical vertebrae. Each pair of vertebrae produces a comparatively small individual movement in order to provide a large overall movement of the entire cervical spine. The biomimetic replication of such joints is difficult due to the delicate and geometric unrestricted joint shapes as well as the muscles that have to be mimicked. Using X-ray as well as micro-computed tomography images and with the utilisation of additive manufacturing, it was possible to produce the owl neck vertebrae in scaled-up form, to analyse them and then to transfer them into technically usable joint assemblies. The muscle substitution of these joints was realised by smart materials actuators in the form of shape memory alloy wire actuators. This actuator technology is outstanding for its muscle-like movement and for its high-energy density. The disadvantage of this wire actuator technology is the low rate of contraction, which means that a large length of wire has to be installed to generate adequate movement. For this reason, the actuator wires were integrated into additively manufactured carrier components to mimic biological joints. This resulted in joint designs that compensate for the disadvantages of the small contraction of the actuators by intelligently installing large wire lengths on comparatively small installation spaces, while also providing a sufficient force output. With the help of a test rig, the developed technical joint variants are examined and evaluated. This demonstrated the technical applicability of this biomimetic joints.

Development and Implementation of a Guideline for the Combination of Additively Manufactured Joint Assemblies with Wire Actuators made of Shape Memory Alloys

Smart Materials actuators in the form of wires made of shape memory alloys in combination with additively manufactured carrier components are used in a wide variety of prototype developments of innovative joint assemblies. This combination is relevant because of the same manufacturing costs of the additively manufactured components, which are independent of the quantity of parts, the free geometric design possibilities as well as the huge energy density of the aforementioned actuator technology. In particular, the focus is on the possibility of appropriately fitting large wire lengths on a compact part volume while taking into account acceptable force losses. Since there is no design guideline for such joint developments, each is individual, which results in unnecessarily long development times and a higher risk of errors. Based on selected in-house and third-party examples, integration possibilities of shape memory alloy wire actuators in additively manufactured carrier components are analysed and transferred into a universally applicable design guideline. These recommendations are brought into the framework of existing design guidelines of the VDI (Verein Deutscher Ingenieure – Association of German Engineers), namely VDI 2206 and VDI 2221 with extensions for additive manufacturing, for a better usability and integrability into existing processes. Finally, this results in a simplified access to the topic of the combination of additive manufacturing and shape memory alloys and a more efficient realisation of such joint developments.

Particle filter based self sensing Shape Memory Alloy wire actuator under external cooling

Shape Memory Alloy wires are widely used as actuators due to their ability to generate a wide range of force and displacement. In practice, these wires are activated following resistive heating by passing an electrical current through them, whereas deactivation is enacted through environmental cooling or forced cooling. It is observed that for fine wires, the temperature is highly sensitive to ambient conditions, i.e., temperature, the velocity of wind, humidity etc., thus severely affecting the performance of these wire actuators. In this study, a particle filter based state estimation technique has been developed to estimate the outcome of the Shape Memory Alloy wire actuated system, by using the change in electrical resistance data of the wire during actuation. Under stationary ambience, the developed particle filter based estimator performs as good as the EKF based estimators (Mohan and Banerjee (2021)). However, both estimators fail while an unwanted external cooling is introduced through a fan. To improve on it, both filters are modified, incorporating convective heat transfer coefficient as an unknown to be estimated in addition to the system’s state. The performance of both estimators are evaluated for various loading conditions in the presence of forced cooling of different extents and intensities. Under external cooling conditions, the proposed Particle Filter based estimator outperforms the same based on EKF in almost all loading situations.

A new model-free control method for SMA wire actuators based on DRL

Shape memory alloy (SMA) has been widely used in different applications due to its unique shape memory property. However, when used as an actuator, it exhibits a hysteresis behavior in its relation between temperature and strain, which is highly nonlinear and difficult to control. Although studies have been conducted on establishing various constitutive models of SMA, it is still difficult to achieve the precise control of the SMA wire with the existing models. In this work, a new promising approach regarding the SMA control task as a reinforcement learning (RL) problem is proposed to address this issue, which does not require accurate mathematical models. Both RL and an improved method named deep reinforcement learning (DRL) are used to solve the problem of precise control of a 1D SMA wire actuator, respectively. The simulation results indicate that with the DRL method, the agent can precisely control the output deformation of the SMA wire after only ten episodes of training. Compared with the DRL method, the RL agent can also achieve the same training target but with hundreds of training.

Experimental analysis of fiber-reinforced laminated composite plates with embedded SMA wire actuators

Shape memory alloy (SMA) wires are effective smart actuators since they provide relatively large actuation force and strain in a very limited space. Embedding SMA wires in flexible laminated composite plates or shells allows for creating active structures for morphing applications. However, the design of such structures is challenging due to the nonlinear thermomechanical coupling of SMA materials, the adhesion of the SMA actuators to the resin, the effect of temperature on the adhesion strength, the heat transfer between the SMA wires and the surrounding laminate, and the convection boundary conditions. This paper presents an experimental analysis of smart fiber-reinforced composite laminated plates with embedded SMA wires, focusing on the relationship between the applied electric current, measured temperature of the SMA wires, and the resulting tip deflection. A temperature-based control method was developed to achieve the required wire temperature and tip deflection as quickly as possible. Results showed that the fiber-orientation angle of the composite plies and the convection boundary conditions have significant effects on the structure’s actuation. With cyclic actuation, damage starts around the heated SMA wires and can grow to cause delamination at areas between neighboring SMA wires and yielding of the matrix material, resulting in permanent deformation.

Deep Neural Network-Based Physics-Inspired Model of Self-Sensing Displacement Estimation for Antagonistic Shape Memory Alloy Actuator

The paper introduces a self-sensing feature that utilizes differential resistance measurement of an antagonistic Shape Memory Alloy (SMA) actuator to estimate linear displacement. The external position sensor used for control feedback becomes extraneous while utilizing an electrical resistance as feedback. The self-sensing capability provides additional advantages such as the overall reduction in size, weight, and interface complexity of an actuator. SMA actuator wires were used in the antagonistic configuration for the bi-directional actuation of the targeted application. A Deep Neural Network (DNN) having Long Short Term Memory (LSTM) layers were used to estimate the dynamic relation of resistance and displacement. A novel DNN model was developed for estimation purposes inspired by the physics-based description of the SMA actuation phenomenon. Accordingly, DNN was developed using the first and last layers as LSTM and a middle layer of feedforward neural network. Estimation results are presented with performance evaluation in terms of the R-squared index and mean absolute error (MAE) of the proposed model; also, the model has been compared with the generic LSTM 2-layered model.

Coupled thermomechanical modelling of shape memory alloy structures undergoing large deformation

Shape Memory Alloys (SMAs) exhibit strain recovery capabilities even after undergoing large deformation, out of underlying stress and temperature dependent diffusionless martensitic transformation. Modelling its behaviour accurately requires coupled thermomechanical analysis, considering absorption and emission of latent heat during phase transformation, thermoelastic effects, thermomechanical loading and boundary conditions. Moreover, the variation of material properties, e.g., modulus of elasticity and thermal expansion coefficient, during phase transformation, significantly affects the response of the system. In addition, appropriate stress and strain measures are required to model the large deformation of SMA-based components under practical thermomechanical loads. In this study, a nonlinear finite element formulation based on Total Lagrangian (TL) approach has been developed to address geometric as well as material non-linearity arising out of the SMA behaviour. Furthermore, considering all the above-mentioned effects, both the stress and thermal equilibrium equations are solved simultaneously in an incremental–iterative finite element (FE) framework, simulating the coupled SMA response. The SMA constitutive model, using Green–Lagrange strain and Second Piola–Kirchhoff stress measures, proposed by Qidwai and Lagoudas (2000), has been implemented. Using the developed FE model, the thermomechanical responses of SMA wire actuator, beam and biomedical staple are simulated to demonstrate the need for the coupled analysis while considering large deformation of the SMA structures.

A self-sensing and robust resistance phase transition detection method for the displacement estimation of shape memory alloy wires

Shape memory alloy (SMA) wires are widely used as actuators because of their simplicity, high energy ratio, and Joule heat activation. The resistance behavior caused by the phase transition can predict the stress or strain of the SMA wire, which is called self-sensing. At present, there are researches on detecting the start and end points of phase transition based on resistance. However, the anti-interference ability of these researches is poor, and it is difficult to identify the thermal equilibrium points. Therefore, this paper proposes a novel resistance-based phase transition detection method with good robustness and displacement estimation. Firstly, the excitation signal is optimized for the thermal cycle pretreatment of the SMA wire. Secondly, the interferences of the resistance signal are determined, which are noise, small current signals, and step signals. Finally, using filtering and outlier processing, a phase transition region detection algorithm based on the difference of resistance extreme values is proposed. Moreover, a resistance-displacement model is established through linear regression. The experimental results show that the proposed method effectively resists the interference of signals and has good robustness. The mean relative error between the estimated and measured displacements is 4.43%. The resistance-displacement model effectively estimates the displacement of the SMA wire actuator and distinguishes the thermal equilibrium points and the end points of the phase transition. The proposed method can be used potentially for real-time phase transition detection, overheating protection, power consumption reduction, and self-sensing drive of sensorless actuators.

Integrated Temperature and Position Sensors in a Shape-Memory Driven Soft Actuator for Closed-Loop Control

Soft actuators are a promising option for the advancing fields of human-machine interaction and dexterous robots in complex environments. Shape memory alloy wire actuators can be integrated into fiber rubber composites for highly deformable structures. For autonomous, closed-loop control of such systems, additional integrated sensors are necessary. In this work, a soft actuator is presented that incorporates fiber-based actuators and sensors to monitor both deformation and temperature. The soft actuator showed considerable deformation around two solid body joints, which was then compared to the sensor signals, and their correlation was analyzed. Both, the actuator as well as the sensor materials were processed by braiding and tailored fiber placement before molding with silicone rubber. Finally, the novel fiber-rubber composite material was used to implement closed-loop control of the actuator with a maximum error of 0.5°.

GWO-Based Optimal Tuning of Controllers for Shape Memory Alloy Wire Actuators

The purpose of this paper is to design the optimal controllers for nonlinear processes with Shape Memory Alloy (SMA) wire actuators viewed as controlled processes. Optimal process models are first derived. A comparative analysis is done between the evolved Takagi-Sugeno-Kang (TSK) fuzzy models of SMA wire actuators, two linear dynamic system models with parameters optimally tuned using the recent metaheuristic Grey Wolf Optimiser (GWO) algorithm and one linear dynamic system model obtained in a previous paper using the System Identification Toolbox. Measured input-output data is involved in this comparison. The parameters of two Proportional-Integral-Derivative controllers, a type-1 fuzzy controller and an interval type-2 fuzzy controller are next optimally tuned using GWO, and a comparative analysis of these linear and nonlinear controllers is done. The simulation results illustrate that the type-2 fuzzy controller ensure good performance for these processes.

Differential resistance based self-sensing recurrent neural network model for position estimation and control of antagonistic Shape Memory Alloy actuator

The paper presents the development and experimental investigation of recurrent neural network (RNN) based self-sensing position estimation (SSPE) model for shape memory alloy actuator (SMA). RNN is used as an estimator in the position feedback control loop to replace the external additional position sensor. The model was inspired by the physics-based analogy of the Mass-Spring-Damper (MSD) system for the antagonistic SMA wire actuator. Actuator displacement presents the hysteresis and non-linear dynamic relationship with observed differential resistance (sensing signal) during phase transformation. The resistance variation causes due to dissimilar resistivity between primary phases of SMA material. The RNN based estimation model was considered because it consists of memory elements for storing the processed information and feedback connections for dynamic modeling of the system. RNN was trained with input sensing signal and target displacement datasets. The estimation accuracy of the model was real-time evaluated during trajectory tracking of reference signals in the feedback control loop. A quantitative performance analysis is assessed in terms of correlation (ℛ), mean absolute error (MEA), and root mean square error (RSME) of the learned model along with the developed actuator system. The tracking results confirm the close agreement between estimated and measured displacement at a reasonable accuracy.

Characterization and Analysis of a Flexural Shape Memory Alloy Actuator

Shape memory alloys (SMAs) are popular as actuators for use in soft robots due to their high work density and compatibility with miniaturized on-board batteries and power electronics. However, because SMA actuators are activated through electrical Joule heating, they exhibit poor energy efficiency and low actuator frequencies that arise from long cool-down times. Moreover, in the case of SMA wires that are subject to flexural loading, their load capacity and mechanical work output decrease exponentially with decreasing cross-sectional area. In this study, we perform analytic and numerical analyses to examine the thermal and structural design space around a particular class of flexural SMA wire actuators with the intention of increasing actuator operating frequency and actuation forces. Measurements obtained through experimental testing are consistent with theoretical studies of actuator force output and provide additional insight into the efficiency of electrical-to-mechanical energy conversion. Together, the theoretical and experimental studies provide insights that have the potential to inform SMA wire design and usage in soft robotic applications.

A driving strategy of shape memory alloy wires with electric resistance modeled by logistic function for power consumption reduction

To propose a novel driving strategy for power consumption reduction, this paper studies the electric resistance behaviors of shape memory alloy (SMA) wires, which are heated by the short-time large current. With a concern for the relationship between electric resistance and shape memory effect, an improved phase transformation equation with the Logistic function is proposed. Moreover, based upon electric resistance and its rate, a new detection method for the driving strategy of SMA wires is proposed. The proposed method is compared with the traditional deformation detection method. The results show that the method effectively resists the interference of system noise and captures the start and end points of phase transformation. To assess the deformation response of SMA wires by the proposed method, constant load tests are conducted, which are applied with the periodic excitation signals of triangular wave, sine wave, and square wave, respectively. The experimental results show the square wave is more suitable for driving the SMA wire actuator. The heating and cooling cycle experiment verifies the effectiveness of the proposed method. Compared with the method using pulse width modulation signals, the proposed method reduces power consumption obviously, and does not need external sensors or material parameters. Based on the proposed method, the power consumption can be reduced by the optimized driving strategy, which improves the lightweight and reliability of SMA actuators.