Window functions and sigmoidal behaviour of memristive systems

Georgiou, Panayiotis S.; Yaliraki, Sophia N.; Drakakis, Emmanuel M.; Barahona, Mauricio

doi:10.1002/cta.2186

Cited by 24 publications

(18 citation statements)

References 42 publications

(156 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4. Four-step Procedure to recover the constitutive relation of an Ideal Memristor from any one of its infinitely many Ideal Generic Memristor Siblings [12].…”

Section: Recovering Ideal Memristor From Its Sibling ?mentioning

confidence: 99%

“…22, namely, the entire x-axis. It follows that the above memristor is endowed with a continuum of stable (but not asymptotically stable) equilibrium points 12 . We will henceforth call a memristor whose POP coincides with the x-axis as a ContinuumMemory Memristor 13 .…”

Section: Example 93: Pop Of a Continuum-memory Memristormentioning

confidence: 99%

“…In practice, G 0 is a scaling constant chosen to fit the intrinsic memductance scale of the memristor. 12 An equilibrium point x(Q) of a differential equation ( ) x f x   is said to be asymptotically stable if a small ball placed initially at x(Q) will always return to x(Q) when it is displaced by an arbitrarily small perturbation of arbitrarily short duration by following the direction of motion indicated by the arrowhead along the dynamic route where the perturbed state ˆ( ) x x Q x    is located. If the perturbed state x did not return to x(Q), but remains motionless after the perturbation Δx became zero, then the equilibrium point is said to be stable.…”

Section: Example 93: Pop Of a Continuum-memory Memristormentioning

confidence: 99%

“…Consider the PTC thermistor defined in (12). To measure the DC V-I Plot of this memristor, we must substitute a set of constant voltages V = V 1 , V 2 ,…, V m for v in (12a) and (12b), and measure the corresponding constant current i = I = I 1 , I 2 ,…, I m after all transient had decayed to zero.…”

Section: Example 101 DC V-i Plot Of Ptc Thermistormentioning

confidence: 99%

“…The caveat here is how long should the experimentalist wait before he can proceed to measure the next point? To answer this fundamental question, let us suppose the PTC memristor behaves exactly according to (12). At zero transient, the measured state variable x = X Q must be a con-…”

Section: Example 101 DC V-i Plot Of Ptc Thermistormentioning

confidence: 99%

See 4 more Smart Citations

Everything You Wish to Know About Memristors But Are Afraid to Ask

2015

View full text Add to dashboard Cite

show abstract

“…4. Four-step Procedure to recover the constitutive relation of an Ideal Memristor from any one of its infinitely many Ideal Generic Memristor Siblings [12].…”

Section: Recovering Ideal Memristor From Its Sibling ?mentioning

confidence: 99%

Section: Example 93: Pop Of a Continuum-memory Memristormentioning

confidence: 99%

Section: Example 93: Pop Of a Continuum-memory Memristormentioning

confidence: 99%

Section: Example 101 DC V-i Plot Of Ptc Thermistormentioning

confidence: 99%

Section: Example 101 DC V-i Plot Of Ptc Thermistormentioning

confidence: 99%

See 3 more Smart Citations

Everything You Wish to Know About Memristors But Are Afraid to Ask

2015

View full text Add to dashboard Cite

show abstract

Parameters self‐tuning PID controller circuit with memristors

Liang

Huang

2017

Circuit Theory & Apps

View full text Add to dashboard Cite

SUMMARYTo achieve the satisfactory control performance of a dynamic system using a proportion-integrationdifferentiation (PID) controller, it is necessary that the three parameters of the PID controller are set to values that suit the system to be controlled. Most current applications employ fixed setting parameters that are not adequate for nonlinear time varying systems with uncertainty and disturbances. In this paper, we use neuron weighted tuning functions to develop self-tuning methods that can automatically tune the parameters of PID controllers. The weighting processes performed using voltage-controlled memristors and current-controlled memristor are discussed. It is proven that the output tracking error of a closed-loop system controlled by the proposed controllers and the weight update rules governed by the memristors descends toward its minimum. A simulation and experimental verification of the performances of these controllers in linear and nonlinear systems are included. This proposed PID controllers are shown to be more robust than conventional PID controllers.

show abstract

The influence of the activation function in a capsule network for brain tumor type classification

Adu

Cai

et al. 2021

Int J Imaging Syst Tech

View full text Add to dashboard Cite

Capsule network's hierarchical framework (CapsNets) consists of an initial standard convolution layer that uses an activation function at its core. The rectified linear unit (ReLU) activation function is widely used in CapsNet and brain tumor classification tasks among several existing activation functions. However, ReLU has some shortcomings where the zero derivatives of the function cause failure of neuron activation. Furthermore, the performance accuracy obtained by the ReLU with CapsNet on brain tumor classification is unsatisfactory. We proposed a new activation function called parametric scaled hyperbolic tangent (PSTanh), which enhances the conventional hyperbolic tangent by avoiding vanishing gradient, provides a small gradient with the introduction of λ and β parameters, and enables faster optimization. Eight standard activation functions (i.e., tanh, Memrister-Like Activation Function (ReLU), Leaky-ReLU, PReLU, ELU, SELU, Swish, ReLU-Memrister-Like Activation Function (RMAF), and the proposed activation) are analyzed and compared in brain tumor classification tasks. Furthermore, extensive experiments are conducted using MNIST, fashion-MNIST, CIFAR-10, CIFAR-100, and ImageNet datasets trained on CapsNets models and deep CNN models (i.e., AlexNet, SqueezeNet, ResNet50, and DenseNet121). The brain tumor's experimental results based on CapsNet and CNN model show that the proposed PSTanh activation achieves better performance than other functions.

show abstract

Window functions and sigmoidal behaviour of memristive systems

Cited by 24 publications

References 42 publications

Everything You Wish to Know About Memristors But Are Afraid to Ask

Everything You Wish to Know About Memristors But Are Afraid to Ask

Parameters self‐tuning PID controller circuit with memristors

The influence of the activation function in a capsule network for brain tumor type classification

Contact Info

Product

Resources

About