In this study, we design and test a complex network using our chaotic circuits with memristors that switch between periodic and chaotic oscillations. As a result, we observe the simultaneous presence of both periodic and chaotic oscillations, alongside instances where switching between these oscillation types occurs within the proposed complex network. In addition, we confirm that the synchronization state and the dynamics of the memristor change with coupling strength. Furthermore, this study investigates the ratio of periodic and chaotic oscillations by changing the memristor parameters to clarify the effect of memristors in the complex network.

Peristaltic pumps are widely used in many industrial applications, especially in medical devices. Their reliability depends on proper maintenance, which includes the total replacement of tubes regularly due to the aging of the materials. The proper use of predictive maintenance techniques could potentially improve the efficiency of maintenance interventions and prevent failures by having a way to determine when the tube has passed its replacement time. We recorded a dataset using six different sensors (three accelerometers, one gyroscope, one magnetometer, and one microphone) using several cassettes (three new units and three units with expired life span). The recording was done at the highest possible frequency (100–6667 Hz, different for each sensor) and then downsampled several times to obtain frequencies as low as 12 Hz. This dataset is now publicly available. We trained 939 different models, which were the result of combining all different sensors as inputs but the microphone, and four basic architectures of recurrent neural network: One or two layers of either gated recurrent unit or long short-term memory with different number of nodes per layer (from 2 to 64). Among all trained models, we selected the ten best performing networks in terms of both accuracy and complexity. All of them reached an F1 score of 0.99 or 1 with holdout cross-validation. Those models were deployed on four different edge AI devices. For all combinations of model and edge AI devices we obtained metrics of memory size (from 0.3% to 160.6% RAM, and from 0.9% to 21.3% flash), inference time (from 0.39 to 1463.91 ms), and average consumption (from 0.15 to 5.30 mA). Nine out of ten models were proven viable for deployment. We concluded that the four models based on magnetometer data were significantly better in terms of consumption and inference time. To the best of our knowledge, the use of magnetometer data is a very uncommon approach to failure detection in predictive maintenance applications, and this is probably the first time it has been used for peristaltic pump aging detection, so our results are very promising for future applications. Also, since most trained models use little resources, we have proved that our approach is perfectly compatible with running other communication and control algorithms on the same device, which is ideal for easy integration and scalability in industrial systems. Some limitations for real deployment include facing environmental factors (noise) and long-term monitoring, so we also proposed a protocol that should reduce the impact of those factors by taking measurements in a controlled way.

In this study, we applied sparse coupling to the coupling system of van der Pol oscillators and investigated the realization of associative memory. The phase difference of oscillator waveforms and the difference of output patterns by disconnecting the coupling between oscillators was clarified. As a result, it was confirmed that the realization of associative memory is possible using sparse coupling, however the maximum rate of coupling disconnection for which associative memory is possible depends on the combination of oscillators in the coupling system.

In this study, we propose the coupled oscillatory networks with the memristor couplings as a ring structure. We focus on the dynamics of the memristor couplings, and investigate synchronization phenomena. As a result, this study observes three different synchronization state types: multi-phase synchronization state, in-phase and anti-phase synchronization state and amplitude death. In addition, we analyze the power consumption of the memristor couplings to make clear the cause of phase sift.

We investigate neural activity of the neural network including delayed couplings by using attractor reconstruction images. The paper presents the effects on neural activity when the value and distribution of delayed coupling is changed. In particular, we focus on the complexity of attractor images and evaluate the relationship between delayed coupling and complexity.

This paper presents an investigation into networked reservoir computing using interconnected oscillators. Specifically, the Bonhoeffer-van der Pol (BVP) oscillator is used as the basis for the reservoir, and its performance is evaluated in speech recognition tasks. The results reveal that the BVP reservoir exhibits enhanced fidelity to input waveforms and improved accuracy in speech recognition compared to the van der Pol (VDP) reservoir. Furthermore, the impact of reservoir configuration on accuracy is examined by varying the number of oscillators, the connection probability, and the coupling strength. The results indicate that accuracy improves as the parameters are increased. These results emphasize the importance of carefully selecting and setting the dynamics of the physical system and parameters of the reservoir to achieve optimal performance.

If clustering can be performed in a continuous system, there is potential usefulness in processing speed for larger scale clustering. We have proposed clustering method of coupled chaotic circuit networks with learning. In this study, we propose a method to extract the center of a cluster by stabilizing synchronization through learning in networks where synchronization is not stable due to chaos.

In this study, we focus on an effect of frustration to triangular oscillatory network with stochastically coupling. We propose a coupled nonliear circuit network with stochastically coupling. Frustration as environmental factor is occurred by network topology which is composed from polygonal structure. We investigate synchronization of the proposed network using different frustration levels by changing the coupling strength. By using computer simulations, the effect of frustration to triangular oscillatory networks with stochastically coupling is shown.

In this study, we investigate synchronization phenomena in ladder-coupled chaotic circuits including ring structures. We focus on the effect of the position of the ring structures in the network to the synchronization of the whole network. In our proposed network model, chaotic circuits are coupled by resistors. We set the parameters of the circuits to generate periodic solutions or chaotic solutions. Computer simulations confirm that the various synchronization phenomena occur in different situations.

In this study, we investigate amplitude change and amplitude death in coupled van der Pol oscillators with different oscillation frequency. The network topology is inspired by the concept of a real brain network. We observe the amplitude death of all oscillators when the coupling strength reaches to the certain value. We also confirm that the characteristics of amplitude death depends on the position of oscillator with different oscillation frequency.

In this study, we focus on network structure with a hub of Scale-free network. We consider the synchronization of coupled chaotic circuit network whose hub nodes are split two nodes and we investigate two network models. In one of them, we replace a hub node with two distributed nodes connected by a positive resistor. In another, we replace a hub node with two distributed nodes connected by a negative resistor. From the simulation results, we can confirm that connection between hubs plays an important role in network structure.

In this study, we propose a method of generating complex networks by exploiting synchronization between coupled oscillatory circuits. To each node of a 2D fully connected network a van der Pol oscillator is assigned. We then study the topological evolution of the network in dependence on environmental conditions. These conditions are modeled by considering the distance between the oscillators and some small frequency errors that are added. By carrying out computer simulations, we confirm that different types of complex networks are obtained depending on different environmental conditions.

In this study, synchronization observed in dynamical oscillatory networks with non-uniform coupling distributions is investigated. The coupling states (on/off) of all connections are stochastically determined at every certain time with the coupling probability. We focus on the heavy tail type of coupling distribution for the network. By using computer simulations, we confirm that the dynamical network with the heavy tail type of coupling distribution can hardly achieve global synchronization.

A brain has neuron and glial cells. In the brain, these cells correlate each other and make a higher brain function. In this study, we propose a Multi-Layer Perceptron (MLP) with pulse glial chain having oscillatory excitation threshold. We connect artificial glia units with the neurons in a hidden-layer. When the connecting neuron output is larger than an excitation threshold of the connected glia, the glia excites and generates the pulse. This pulse transmits to the neighboring glias and the connecting neuron. The pulse increases a threshold of the connecting neuron, thus the glia gives energy for solving tasks. In this model, the excitation threshold is oscillating within a defined value. Even if the connecting neuron output does not change, the pulse generation occurs by the oscillation of the excitation threshold. The oscillation of the excitation threshold gives more energy to the network and improves a learning performance of the MLP. By computer simulation, we confirm that the oscillation of the excitation threshold improves a learning performance.

In this study, we investigate synchronization of parametrically excited van der Pol oscillators with small mismatch. In the case of two subcircuits, we confirm that the two subcircuits are synchronized at in-phase state when the adding mismatch is small. By increasing the small mismatch, we observe unsynchronous phenomena. Furthermore, we apply this circuit model to ten coupled oscillators as random network model with hub.

In this study, we propose a Multi-Layer Perceptron (MLP) with pulse glial chain based on individual inactivity period which is inspired from biological characteristics of a glia. In this method, we one-by-one connect a glia with neurons in the hidden-layer. The connected glia is excited by the connecting neuron output. Then, the glia generates the pulse. This pulse is input to the connecting neuron threshold. Moreover, this pulse is propagated into the glia network. Thus, the glia has a position density each other. In this network, a period of inactivity of the glia is dynamically changed according to pulse generation time. In the previous method, we fix the period of inactivity, thus the pulse generation pattern is often fixed. It is similar to the local minimum. By varied the period of inactivity, the pulse generation pattern obtains the diversity. We consider that this diversity of the pulse generation pattern is efficiency to the MLP performance. By the simulation, we confirm that the proposed MLP improves the MLP performance than the conventional MLP.

Epilepsy of the neuropsychiatric disorder is provoked from an imbalance in the long-term potentiation (LTP) versus long-term depression (LTD) of the synapses in the hippocampus. The LTP and LTD are replicated by using the Spike Timing Dependent Plasticity (STDP). Additionally, the spiking activity of the synapses in the hippocampus can be approximated by using the Rulkov maps. In our previous study, we considered some easy simulation models which are constructed by using Rulkov maps with STDP. Moreover, these simulation models consist of unidirectionally coupled neurons. In this paper, we consider some easy simulation models with bidirectionally coupled neurons. We explore the effect of unidirectional and bidirectional connection on spiking activity, as basic simulation for constructing the approximate simulation model of epilepsy. From these results, the unidirectional models show high accuracy in-phase/anti-phase synchronization, and it shows divergent relatively early. The bidirectional models show the stable waveform (i.e., non-divergent) for a long term compared to unidirectional models.

Chaos synchronization can be observed in various natural systems. In this study, we investigate synchronization phenomena on a ring of coupled chaotic circuits. We can observe N-phase chaos synchronization and simple synchronization phenomena depending on initial values in the system. This paper suggests ring topology and alternately shifted bias are essential factors of N-phase synchronization.

We have recently proposed a novel neural network structure called an “Affordable Neural Network” (AfNN), in which affordable neurons of the hidden layer are considered as the elements responsible for the robustness property as is observed in human brain function. In this study, we investigate the mechanism of learning process of the AfNN to make clear the reason of that the AfNNs can perform well for learning and generalization abilities and operate as usually against damaging neurons.

A glia is a nervous cell existing in a brain. This cell transmits signals by various ion concentrations. By the ion concentration, the glia correlates the neuron by ions. Thus, the glia composes the network different to the neural network. We proposed a four-layer Multi-Layer Perceptron (MLP) with glia connections of hidden-layer neurons which is inspired from characteristics of the biological glia. The proposed MLP has two hidden-layer. In this network, we connect the glia to the neurons in two hidden-layer. The glia receives the output of connecting neuron and the outputs are summed. When the summed value is over the glia excitation threshold, this glia is excited. The excited glia generates a pulse. This pulse input to the thresholds of connecting glias. We consider that the glia gives the energy to the MLP and the glia give the position relationship of neurons in the different layers. By the simulation, we confirm that the proposed MLP obtains a high solving ability. Moreover, we investigate the characteristics of the proposed MLP.

In this study, we research a new layer arrangement of three layer cellular neural network (CNN). In this paper, we investigate the output characteristics by using our proposed method to image processing of gray scale image and binary image and show its effectiveness with simulation results.

A glia is a nervous cell which is existing in a brain. This cell changes a Ca2+concentration. This ion affects a neuron membrane potential and it is propagated to the neighboring glia. Moreover, the Ca2; directly affects the human memory by increasing of a D-serine. From these functions, we propose a Multi-Layer Perceptron (MLP) including glial pulse and switching between a learning and non-learning. In this method, the neurons in the hidden-layer received the pulse from connected glias. The pulse is generated depending on the neuron outputs and it is propagated to the neighboring glias and neurons. Moreover, the neurons are separated to some groups. Each group periodically switches a learning term and a non-learning term. Each group starts the learning term having a small lag each other. We consider that a performance of the MLP improves by two different methods influencing each other. By two simulations, we confirm that the MLP obtains the high solving ability by using our methods.

Genetic Algorithm (GA) is modeling behavior of evolution in organic and known as one of method to solve Traveling Salesman Problem (TSP). However, GA obtains solution by overlaying generations because of being based on evolution in organic. Thus, it takes a long time to find approximate solution. While, Virus Theory of Evolution (VTE) can evolve by virus infection. VTE characteristic has sharing of information among same generation. If new algorithm is using both these characteristics of GA and VTE, convergence speed would be faster than GA. Thus, this study proposes Genetic Algorithm with Virus Infection (GAVI). GAVI algorithm is Virus Theory of Evolution (VTE) to be based on Genetic Algorithm (GA). We apply GAVI to TSP and confirm that GAVI obtains more effective result than GA.

In this study, synchronization phenomena observed in coupled polygonal oscillatory networks with frustration is investigated. We focus on the amplitude of each oscillator when the coupling strength is changed. By using computer simulations and circuit experiments, we confirm that the amplitude of the shared and the other oscillators obtains different value by increasing the value of the coupling strength. Furthermore, theoretical analysis is applied to solve the amplitude and the phase differences between the adjacent oscillators.

A glia is a nervous cell existing in a brain. The brain is composed of the relationship with glias and neurons. By an ion concentration, the glia transmits signal to neurons and neighboring glias. In this study, we propose the MLP with positive and negative pulse glial chain which is inspired from features of the biological glia.We add the MLP to the positive and negative pulse glial chain. In the positive and negative pulse glial chain, the glias are connected to the neurons one by one. The glia generates pulse when the glia is excited by the connected neuron's output. If the connected neuron has large amount of output, the glia generates positive pulse. Moreover, if the connected neuron has small amount of output, the glia generates the negative pulse. The positive and negative pulse are propagated to the connected neuron and neighboring glias. We consider that the positive and negative pulse glial chain give the relationships of position of neurons in a same layer. By solving a Two-Spirals Problem (TSP), we confirm that the proposed MLP has better a learning performance and a generalization capability than the conventional MLP.

Solving combinatorial optimization problems is one of the important applications of neural networks. Many researchers have proposed noise induced hopfield neural networks in which noises are induced state values of neurons. However, the noise inducing method to state values of neurons cause problems. In this study, we propose hopfield neural networks with periodic brake. In the proposed system, external noises are not induced to state values of neurons. Thus, the proposed system can avoid the problem caused in the noise induced system. We investigate the solving ability of the proposed system for quadratic assignment problems and designing of parameters.

A glia is nervous cell which exists in a brain. The glia can transmit signal to other glias and neurons by change of ions' densities. We have an interest in this feature of the glia. We consider that we can apply this feature to an artificial neural network. In this study, we propose a Multi-Layer Perceptron (MLP) with propagation of glial pulse to two directions. The proposed MLP has the glias in a hidden layer. The glias are connected with neurons and are excited by the outputs of neurons. The exciting glias generate pulses and the pulses affect neurons' thresholds and neighboring glias. We consider that the MLP obtains the relationships of position of neurons in the hidden layer and this information give good influence to the MLP leaning. We confirm that the proposed MLP has better learning performance than the conventional MLP. Moreover, we confirm that the performance of the proposed MLP is changed by some conditions of propagation of the glial pulse.

In this study, we analyze a synchronization phenomenon observed from a circuit network consisting of several number of oscillator chains which are one-dimensional arrays of weakly coupled van der Pol oscillators. Computer simulations and circuit experiments show interesting unexpected synchronization phenomenon and theoretical analysis explains the synchronization state.

In this study, we investigate synchronization phenomena in coupled chaotic circuits which are connected with distance information. We confirm that the chaotic circuits arranged in the near distance are synchronized at in-phase state, and the coupled circuits with the far distance could not be synchronized. Namely, clustering phenomena of coupled chaotic circuits is observed on two-dimensional place.

Honeybee Colony Optimization (HCO) is an optimization algorithm based on a particular intelligent behavior of honeybee swarms. In this study, we propose a new HCO containing a characteristic of guidepost pheromone that has the effect to attract other bees. Namely, many bees can move to the optimal place. We investigate the performance of the proposed HCO by using four benchmark problems. It discovered that the effect of guidepost pheromone works well for the high dimension problems. We consider that the proposed HCO with pheromone can leave from local minima more easily than the standard HCO.

We have proposed the glial network which was inspired from the feature of the brain. The glial network is composed by glias connecting each other. All glias generate oscillations and these oscillations propagate in the glial network. We confirmed that the glial network improved the learning performance of the Multi-Layer Perceptron (MLP). In this article, we investigate the MLP with the impulse glial network. The glias generate only impulse output, however they make the complex output by correlating with each other. We research the proposed networks' parameter dependency. Moreover, we show that the proposed network possess better learning performance and better generalization capability than the conventional MLPs.

In this study, we propose Cellular Neural Networks with switching two types of templates. In the CNN, space varying system is known that it can perform complex processing. Generally, the space varying CNN is not easy to design. However, we can set existing template on each cell of CNN by the proposed method. In binarization, complex portions of input image are not processed well by using the conventional CNN. On the other hand, the complex portion can be processed well by the proposed method. In the edge detection, the indistinct portion is not detected by the conventional CNN with “Edge detection” template of 3×3 matrix. It is difficult for CNN to recognize that it is the edge or not. Additionally, the detected edge is too bold and some noises are left with “Edge detection” template of 5×5 matrix. By switching these templates in case, we can detect edge in indistinct position. In pattern formation, generally, simple pattern is formed by using one template. On the other hand, some complex patterns are formed by the proposed method. From some simulation results, we confirm that the proposed method is effective for various image processing.

In this study, synchronization phenomena observed in coupled polygonal oscillatory networks with frustration is investigated. We focus on the power consumption of coupling resistors of the whole system. By using computer simulations and theoretical analysis, we confirm that the phase differences of the coupled oscillators are solved by finding the minimum value of the power consumption function.

In this study, we propose a space-varying cellular neural network (CNN) designed by Hopfield neural network (Hopfield NN). CNN is classified into two types of system like space-invariant system and space-varying system. Space-invariant means that all cells have identical template. On the other hand, space-varying means that all cells do not have identical template according to the state values of the cell and neighbor cells and so on. The proposed CNN is the space-varying system and it is designed by using an associative memory ability of Hopfield NN. In general, the design of space-varying systems is not easy. However, we can set one of prepared existing templates on each cell of CNN according to the retrieved pattern by Hopfield NN to which some typical local image structures are embedded. Namely, we need only some existing templates and their associated patterns. Some simulation results show the basic properties of the proposed CNN and its effectiveness.

We have recently proposed a novel neural network structure called an “Affordable Neural Network” (AfNN), in which affordable neurons of the hidden layer are considered as the elements responsible for the robustness property as is observed in human brain function. We have confirmed that the AfNN gains good performance both of the generalization ability and the learning ability. Furthermore, the AfNN has durability, because the AfNN still performs well even if some of neurons in the hidden layer are damaged after learning process. In this study, we study the characteristics of weights of the AfNN during the learning process to make clear the reason of that the AfNNs can perform well for learning and generalization abilities and operate as usually against damaging neurons.

Some methods using artificial neural network were proposed for solving to the Two-Spiral Problem (TSP). TSP is a problem which classifies two spirals drawn on the plane, and it is famous as the high nonlinear problem. In this paper, we propose a chaos glial network which connected to Multi-Layer Perceptron (MLP). The chaos glial network is inspired by astrocyte which is glial cell in the brain. By computer simulations for solving TSP, we confirmed that the proposed chaos glial network connected to MLP gains better performance than the conventional MLP.

We explore the behavior of two van der Pol oscillators coupled by a stochastically time-varying resistor. We observe switching between in-phase and anti-phase synchronization and analyze the statistics of the switching, with and without additional noise. We find stochastic resonant behavior of synchronization in both regimes. Correlating the two noise sources quickly destroys the stochastic resonance phenomenon in the anti-phase synchronization regime. In the in-phase regime, noise correlation first contributes to the synchronization before removing the resonance by taking over the synchronization by its own means.

One of the applications of neural network is solving combinatorial optimization problems. In our past study, the solving ability of the Hopfield Neural Network with noise for quadratic assignment problem is investigated. However, even if we injected the noise to the network, the optimal solution cannot occasionally be found. In this study, we propose the method adding scale-rule noise to the Hopfield Neural Network to achieve better performance. By computer simulations solving quadratic assignment problem, we evaluate the performance of the method.

Synchronization phenomena in coupled oscillatory systems are very important model to describe various higher-dimensional nonlinear phenomena in the field of natural science. In this study, we investigate synchronization phenomena in van der Pol oscillators coupled by time-varying resistors with timing mismatch as a ring. By carrying out computer simulations, we confirm the various interesting phenomena (wave propagation, clustering, complex phase pattern) which cannot be observed in simple oscillatory systems coupled by resistors.

Many people propose the method adding chaos noise to Hopfield neural network for solving combinatorial optimization problems. In our past study, we solved quadratic assignment problem by Hopfield neural network with various chaotic noises. However, the solution of the network is sometimes trapped in a certain area and can not find a good solution, especially for relatively larger problems. In this study, we propose a method changing the amplitude of the chaos noise. We investigate the performance of Hopfield neural network with the hopping chaos for quadratic assignment problem.

In this study, two chaotic circuits coupled by a time-varying resistor are investigated. We assume that the time-varying resistor is realized by switching positive and negative resistors alternately. By carrying out circuit experiments and computer simulations, we confirm that interesting switching phase states between in-phase and anti-phase is observed.

In recent years, many people have been trying to develop some applications to information processing by exploiting oscillatory phenomena in neural networks. Bifurcation and stability of equilibrium points in a simple neural oscillator consisting of two neurons have been analyzed in detail. On the other hand, oscillatory phenomena in the simple neural oscillator can be modeled by electrical circuit such as van der Pol oscillators. In this study, we propose a network model of simple oscillators coupled by time-varying resistor, which can explain some interesting complex phenomena observed in a large scale network of neurons coupled by both excitability and inhibitory synapses. By carrying out computer simulations and circuit experiments, we confirm the generation of various interesting phenomena which cannot be observed in simple coupled oscillatory networks.

In this study, we address the durability of the brain, which is able to operate in various imperfect situations. In our previous research, we have proposed a new network structure called the "Affordable Neural Network", where affordable neurons in the hidden layer of the feedforward neural network reflect important aspects of the real brain mechanism. We consider that the concept of the affordable neural network embodies the important feature of durability. In our contribution, we investigate the durability of the affordable neural network when some of the neurons in the hidden layer are damaged after the learning process.

Solving combinatorial optimization problems is one of the important applications of the neural network. Many researchers have reported that exploiting chaos achieves good solving ability. However, the reason of the good effect of chaos has not been clarified yet. In this study, we investigate a performance of chaotic switching noise injected to the Hopfield neural network for quadratic assignment problems. By computer simulation we confirm that the chaotic switching noise is effective for solving quadratic assignment problems as well as intermittent chaos near three-periodic window

Several people point out that the Hopfield neural network (abbr.NN ) with chaos injection gains the good performance for solving traveling salesman problems, which is one of combinatorial optimization problems. In this study, we investigate the performance of the intermittency chaos injected to the Hopfield NN working as an associative memory. The rate and the speed of the convergence to an embedded pattern are evaluated by computer simulations.. Furthermore, in order to confirm the reason of the good ability of intermittency chaos, we carry out the same simulation using the time series produced by the Markov chain model. Simulated results show that the Markov chain model is good enough to gain similar performance of the intermittency chaos.