Neural Simulation Research Articles

Box-Behnken Design for MAFM Precision Surface Finishing of Al-6063/SiC/B<sub>4</sub>C Composites: A Comparative Study with Nature-Inspired Algorithms

Precision polishing is difficult for advanced materials like silicon carbide and boron carbide. Magnetic abrasive flow machining (MAFM) has become an effective method for cleaning, deburring, and polishing metal and high-tech engineering parts. By finishing hybrid Al/SiC/B4C-metal matrix composites (MMCs), this research uses MAFM for experimental readings. The present work is innovative due to the aluminum workpiece fixture, hybrid composites, and response surface methodology (RSM) modeling. The neural simulation of the MAFM process and nature-inspired error reduction make it unique. Using six input and two output parameters, a generic framework is created. Box-Behnken design (BBD) of response surface methodology plans and executes 54 runs of experimentation. The hybrid artificial neural network (ANN) technique is used to compare the MAFM process systematically. ANN is used to model parameter input-output relations. To anticipate the created surface accurately, regression models must be precise. These hybrid particle swarm optimization (PSO)-genetic algorithm (GA)-simulated annealing (SA) algorithms optimize the MAFM process. Additionally, trained ANN models outperform the BBD model in prediction. For optimal error reduction, the neural network uses Bayesian regularization with 112 iterations. The ANN model regression graph shows a correlation between inputs and outputs. A scanning electron microscope (SEM) with 300-magnification examines the workpiece surface. According to SEM, MAFM provides fine surface textures, thus reducing abnormalities.

Inferring galaxy cluster masses from cosmic microwave background lensing with neural simulation based inference

Gravitational lensing by massive galaxy clusters distorts the observed cosmic microwave background (CMB) on arcminute scales, and these distortions carry information about cluster masses. Standard approaches to extracting cluster mass constraints from the CMB cluster lensing signal are either sub-optimal, ignore important physical or observational effects, are computationally intractable, or require additional work to turn the lensing measurements into constraints on cluster masses. We apply simulation based inference (SBI) using neural likelihood models to the problem. We show that in circumstances where the exact likelihood can be computed, the SBI constraints on cluster masses are in agreement with the exact likelihood, demonstrating that the SBI constraints are close to optimal. In scenarios where the exact likelihood cannot be feasibly computed, SBI still recovers unbiased estimates of individual cluster masses and combined constraints from multiple clusters. SBI will be a powerful tool for constraining the masses of galaxy clusters detected by future cosmic surveys. Code to run the analyses presented here will be made publicly available.

Neural Simulation of Actions for Serpentine Robots.

The neural or mental simulation of actions is a powerful tool for allowing cognitive agents to develop Prospection Capabilities that are crucial for learning and memorizing key aspects of challenging skills. In previous studies, we developed an approach based on the animation of the redundant human body schema, based on the Passive Motion Paradigm (PMP). In this paper, we show that this approach can be easily extended to hyper-redundant serpentine robots as well as to hybrid configurations where the serpentine robot is functionally integrated with a traditional skeletal infrastructure. A simulation model is analyzed in detail, showing that it incorporates spatio-temporal features discovered in the biomechanical studies of biological hydrostats, such as the elephant trunk or octopus tentacles. It is proposed that such a generative internal model could be the basis for a cognitive architecture appropriate for serpentine robots, independent of the underlying design and control technologies. Although robotic hydrostats have received a lot of attention in recent decades, the great majority of research activities have been focused on the actuation/sensorial/material technologies that can support the design of hyper-redundant soft/serpentine robots, as well as the related control methodologies. The cognitive level of analysis has been limited to motion planning, without addressing synergy formation and mental time travel. This is what this paper is focused on.

A sensory–neuromorphic interface capable of environmental perception, sensory coding, and biological stimuli

AbstractThe sensory–neuromorphic interface is key to the application of neuromorphic electronics. Artificial spiking neurons and artificial sensory nerves have been created, and a few studies showed a complete neuromorphic system through cointegration with synaptic electronics. However, artificial synaptic devices and systems often do not work in real environments, which limits their ability to provide realistic neural simulations and interface with biological nerves. We report a sensory–neuromorphic interface that uses a fiber synapse to emulate a biological afferent nerve. For the first time, a sensing–neuromorphic interface is connected to a living organism for peripheral nerve stimulation, allowing the organism to establish a connection with its surrounding environment. The interface converts perceived environmental information into analog electrical signals and then into frequency‐dependent pulse signals, which simplify the information interface between the sensor and the pulse‐data processing center. The frequency of the interface shows a sublinear dependence on strain amplitude at different stimulus intensities, and can deliver increased frequency spikes at potentially damaging stimulus intensities, similar to the response of biological afferent nerves. To verify the application of this interface, a system that monitors strain and provides an overstrain alarm was constructed based on this afferent neural circuit. The system has a response time of <2 ms, which is compatible with the response time in biological systems. The interface can be potentially extended to process signals from almost any type of sensors for other afferent senses, and these results demonstrate the potential for neuromorphic interfaces to be applied to bionic sensory interfaces.

Boosting Bidirectional Photoresponse with Wavelength Selectivity through Ambipolar Transport Modulation

AbstractNegative photoconductive phototransistors, referring to transistors that exhibit a decrease in photocurrent under illumination, have the potential to revolutionize optoelectronic applications involving light, such as optoelectronic logic circuits and visual neural simulation. Currently, achieving negative photoconductivity (NPC) requires complex material design or interface structure construction. However, achieving precise control over NPC behaviors poses a significant challenge. Herein, a simple yet effective strategy is demonstrated for realizing controllable NPC responses in organic phototransistors through ambipolar transport modulation. Due to the controversy between the preferred exciton dissociation/charge separation direction and the gate electric field driven charge drift direction, the main semiconductor channel (n‐ or p‐channel) exhibits NPC behavior under illumination. The validity of this mechanism has been confirmed through intensive studies by varying the component and combination of the p‐n heterostructure. Moreover, devices utilizing ambipolar transport exhibit a wavelength‐selectivity NPC response due to the absorption characteristics of the combined semiconductor materials. Most encouragingly, by incorporating both negative and positve photoconductivity along with wavelength‐selective responses, high‐contrast image sensing, information encryption and decryption, as well as optoelectronic logic circuit design is successfully achieved. This work promotes the design and development of bidirectional optoelectronic devices and offers a new route for developing attractive multifunctional optoelectronic devices.

Scaling neural simulations in STACS

As modern neuroscience tools acquire more details about the brain, the need to move towards biological-scale neural simulations continues to grow. However, effective simulations at scale remain a challenge. Beyond just the tooling required to enable parallel execution, there is also the unique structure of the synaptic interconnectivity, which is globally sparse but has relatively high connection density and non-local interactions per neuron. There are also various practicalities to consider in high performance computing applications, such as the need for serializing neural networks to support potentially long-running simulations that require checkpoint-restart. Although acceleration on neuromorphic hardware is also a possibility, development in this space can be difficult as hardware support tends to vary between platforms and software support for larger scale models also tends to be limited.In this paper, we focus our attention on Simulation Tool for Asynchronous Cortical Streams (STACS), a spiking neural network simulator that leverages the Charm++ parallel programming framework, with the goal of supporting biological-scale simulations as well as interoperability between platforms. Central to these goals is the implementation of scalable data structures suitable for efficiently distributing a network across parallel partitions. Here, we discuss a straightforward extension of a parallel data format with a history of use in graph partitioners, which also serves as a portable intermediate representation for different neuromorphic backends.We perform scaling studies on the Summit supercomputer, examining the capabilities of STACS in terms of network build and storage, partitioning, and execution. We highlight how a suitably partitioned, spatially dependent synaptic structure introduces a communication workload well-suited to the multicast communication supported by Charm++. We evaluate the strong and weak scaling behavior for networks on the order of millions of neurons and billions of synapses, and show that STACS achieves competitive levels of parallel efficiency.

Neural Physical Simulation with Multi-Resolution Hash Grid Encoding

We explore the generalization of the implicit representation in the physical simulation task. Traditional time-dependent partial differential equations (PDEs) solvers for physical simulation often adopt the grid or mesh for spatial discretization, which is memory-consuming for high resolution and lack of adaptivity. Many implicit representations like local extreme machine or Siren are proposed but they are still too compact to suffer from limited accuracy in handling local details and a long time of convergence. We contribute a neural simulation framework based on multi-resolution hash grid representation to introduce hierarchical consideration of global and local information, simultaneously. Furthermore, we propose two key strategies: 1) a numerical gradient method for computing high-order derivatives with boundary conditions; 2) a range analysis sample method for fast neural geometry boundary sampling with dynamic topologies. Our method shows much higher accuracy and strong flexibility for various simulation problems: e.g., large elastic deformations, complex fluid dynamics, and multi-scale phenomena which remain challenging for existing neural physical solvers.

Neuromorphic Signal Classification Using Organic Electrochemical Transistor Array and Spiking Neural Simulations

Neuromorphic Signal Classification Using Organic Electrochemical Transistor Array and Spiking Neural Simulations

Closed-Loop Deep Brain Stimulation With Reinforcement Learning and Neural Simulation.

Deep Brain Stimulation (DBS) is effective for movement disorders, particularly Parkinson's disease (PD). However, a closed-loop DBS system using reinforcement learning (RL) for automatic parameter tuning, offering enhanced energy efficiency and the effect of thalamus restoration, is yet to be developed for clinical and commercial applications. In this research, we instantiate a basal ganglia-thalamic (BGT) model and design it as an interactive environment suitable for RL models. Four finely tuned RL agents based on different frameworks, namely Soft Actor-Critic (SAC), Twin Delayed Deep Deterministic Policy Gradient (TD3), Proximal Policy Optimization (PPO), and Advantage Actor-Critic (A2C), are established for further comparison. Within the implemented RL architectures, the optimized TD3 demonstrates a significant 67% reduction in average power dissipation when compared to the open-loop system while preserving the normal response of the simulated BGT circuitry. As a result, our method mitigates thalamic error responses under pathological conditions and prevents overstimulation. In summary, this study introduces a novel approach to implementing an adaptive parameter-tuning closed-loop DBS system. Leveraging the advantages of TD3, our proposed approach holds significant promise for advancing the integration of RL applications into DBS systems, ultimately optimizing therapeutic effects in future clinical trials.

Effects of Numerical Method Selection on Fully-Pipelined FPGA Accelerators for Neural Simulations

Effects of Numerical Method Selection on Fully-Pipelined FPGA Accelerators for Neural Simulations

Neural Simulation of Digital Twin of Top Management Motivation Mechanism in Regional Government Agencies

Neural Simulation of Digital Twin of Top Management Motivation Mechanism in Regional Government Agencies

Structural and Parametric Synthesis of Neural Network Controllers for Control Objects with Limiters

The article presents a methodology for the synthesis of digital control systems for nonlinear objects with limiters under conditions of incomplete information. Closed-loop tracking systems with negative feedback are considered. Artificial neural networks are proposed to build a controller, which is included in series with the control object. This approach is effective when known classical methods do not allow to synthesize control. This is the case, for example, if the mathematical model is essentially nonlinear and is not fully defined. The developed methods allow us to expand the class of technical systems, for which the direct (without using various kinds of simplifications) synthesis of control laws that are close to optimal is possible. In addition, neural network controllers possess the properties of robustness, adaptivity, and are initially digital, i.e. those qualities, which are very much in demand in practice. In article main attention is given to such problems, as a choice of neural network structure for neural simulator and neural controller, construction of training sample, ensuring convergence of the process of weights correction. For training neural networks the method of back propagation of error is used as a basic one. The effectiveness of the proposed technique is demonstrated by the example of the synthesis of a neuroregulator for a specific technical object and its comparison with classical control systems. It should be noted that today neural network technologies are widespread enough in various spheres of activity. The successes demonstrated in sound processing, image processing, automatic translation, in navigation systems, in big data processing are impressive. However, their application in automatic control systems is not so widespread. The authors of this article believe that the potential of artificial neural networks can be used in this direction. It should be understood that the use of neural networks is effective only under certain conditions and properties of the control object.

Adaptive octree meshes for simulation of extracellular electrophysiology

Objective. The interaction between neural tissues and artificial electrodes is crucial for understanding and advancing neuroscientific research and therapeutic applications. However, accurately modeling this space around the neurons rapidly increases the computational complexity of neural simulations. Approach. This study demonstrates a dynamically adaptive simulation method that greatly accelerates computation by adjusting spatial resolution of the simulation as needed. Use of an octree structure for the mesh, in combination with the admittance method for discretizing conductivity, provides both accurate approximation and ease of modification on-the-fly. Main results. In tests of both local field potential estimation and multi-electrode stimulation, dynamically adapted meshes achieve accuracy comparable to high-resolution static meshes in an order of magnitude less time. Significance. The proposed simulation pipeline improves model scalability, allowing greater detail with fewer computational resources. The implementation is available as an open-source Python module, providing flexibility and ease of reuse for the broader research community.

SNS-Toolbox: An Open Source Tool for Designing Synthetic Nervous Systems and Interfacing Them with Cyber-Physical Systems.

One developing approach for robotic control is the use of networks of dynamic neurons connected with conductance-based synapses, also known as Synthetic Nervous Systems (SNS). These networks are often developed using cyclic topologies and heterogeneous mixtures of spiking and non-spiking neurons, which is a difficult proposition for existing neural simulation software. Most solutions apply to either one of two extremes, the detailed multi-compartment neural models in small networks, and the large-scale networks of greatly simplified neural models. In this work, we present our open-source Python package SNS-Toolbox, which is capable of simulating hundreds to thousands of spiking and non-spiking neurons in real-time or faster on consumer-grade computer hardware. We describe the neural and synaptic models supported by SNS-Toolbox, and provide performance on multiple software and hardware backends, including GPUs and embedded computing platforms. We also showcase two examples using the software, one for controlling a simulated limb with muscles in the physics simulator Mujoco, and another for a mobile robot using ROS. We hope that the availability of this software will reduce the barrier to entry when designing SNS networks, and will increase the prevalence of SNS networks in the field of robotic control.

Optimized monophasic pulses with equivalent electric field for rapid-rate transcranial magnetic stimulation

Objective. Transcranial magnetic stimulation (TMS) with monophasic pulses achieves greater changes in neuronal excitability but requires higher energy and generates more coil heating than TMS with biphasic pulses, and this limits the use of monophasic pulses in rapid-rate protocols. We sought to design a stimulation waveform that retains the characteristics of monophasic TMS but significantly reduces coil heating, thereby enabling higher pulse rates and increased neuromodulation effectiveness. Approach. A two-step optimization method was developed that uses the temporal relationship between the electric field (E-field) and coil current waveforms. The model-free optimization step reduced the ohmic losses of the coil current and constrained the error of the E-field waveform compared to a template monophasic pulse, with pulse duration as a second constraint. The second, amplitude adjustment step scaled the candidate waveforms based on simulated neural activation to account for differences in stimulation thresholds. The optimized waveforms were implemented to validate the changes in coil heating. Main results. Depending on the pulse duration and E-field matching constraints, the optimized waveforms produced 12%–75% less heating than the original monophasic pulse. The reduction in coil heating was robust across a range of neural models. The changes in the measured ohmic losses of the optimized pulses compared to the original pulse agreed with numeric predictions. Significance. The first step of the optimization approach was independent of any potentially inaccurate or incorrect model and exhibited robust performance by avoiding the highly nonlinear behavior of neural responses, whereas neural simulations were only run once for amplitude scaling in the second step. This significantly reduced computational cost compared to iterative methods using large populations of candidate solutions and more importantly reduced the sensitivity to the choice of neural model. The reduced coil heating and power losses of the optimized pulses can enable rapid-rate monophasic TMS protocols.

Neural simulation pipeline: Enabling container-based simulations on-premise and in public clouds.

In this study, we explore the simulation setup in computational neuroscience. We use GENESIS, a general purpose simulation engine for sub-cellular components and biochemical reactions, realistic neuron models, large neural networks, and system-level models. GENESIS supports developing and running computer simulations but leaves a gap for setting up today's larger and more complex models. The field of realistic models of brain networks has overgrown the simplicity of earliest models. The challenges include managing the complexity of software dependencies and various models, setting up model parameter values, storing the input parameters alongside the results, and providing execution statistics. Moreover, in the high performance computing (HPC) context, public cloud resources are becoming an alternative to the expensive on-premises clusters. We present Neural Simulation Pipeline (NSP), which facilitates the large-scale computer simulations and their deployment to multiple computing infrastructures using the infrastructure as the code (IaC) containerization approach. The authors demonstrate the effectiveness of NSP in a pattern recognition task programmed with GENESIS, through a custom-built visual system, called RetNet(8 × 5,1) that uses biologically plausible Hodgkin-Huxley spiking neurons. We evaluate the pipeline by performing 54 simulations executed on-premise, at the Hasso Plattner Institute's (HPI) Future Service-Oriented Computing (SOC) Lab, and through the Amazon Web Services (AWS), the biggest public cloud service provider in the world. We report on the non-containerized and containerized execution with Docker, as well as present the cost per simulation in AWS. The results show that our neural simulation pipeline can reduce entry barriers to neural simulations, making them more practical and cost-effective.

Efficient parameter calibration and real-time simulation of large-scale spiking neural networks with GeNN and NEST.

Spiking neural networks (SNNs) represent the state-of-the-art approach to the biologically realistic modeling of nervous system function. The systematic calibration for multiple free model parameters is necessary to achieve robust network function and demands high computing power and large memory resources. Special requirements arise from closed-loop model simulation in virtual environments and from real-time simulation in robotic application. Here, we compare two complementary approaches to efficient large-scale and real-time SNN simulation. The widely used NEural Simulation Tool (NEST) parallelizes simulation across multiple CPU cores. The GPU-enhanced Neural Network (GeNN) simulator uses the highly parallel GPU-based architecture to gain simulation speed. We quantify fixed and variable simulation costs on single machines with different hardware configurations. As a benchmark model, we use a spiking cortical attractor network with a topology of densely connected excitatory and inhibitory neuron clusters with homogeneous or distributed synaptic time constants and in comparison to the random balanced network. We show that simulation time scales linearly with the simulated biological model time and, for large networks, approximately linearly with the model size as dominated by the number of synaptic connections. Additional fixed costs with GeNN are almost independent of model size, while fixed costs with NEST increase linearly with model size. We demonstrate how GeNN can be used for simulating networks with up to 3.5 · 106 neurons (> 3 · 1012synapses) on a high-end GPU, and up to 250, 000 neurons (25 · 109 synapses) on a low-cost GPU. Real-time simulation was achieved for networks with 100, 000 neurons. Network calibration and parameter grid search can be efficiently achieved using batch processing. We discuss the advantages and disadvantages of both approaches for different use cases.

A comprehensive neural simulation of slow-wave sleep and highly responsive wakefulness dynamics.

Hallmarks of neural dynamics during healthy human brain states span spatial scales from neuromodulators acting on microscopic ion channels to macroscopic changes in communication between brain regions. Developing a scale-integrated understanding of neural dynamics has therefore remained challenging. Here, we perform the integration across scales using mean-field modeling of Adaptive Exponential (AdEx) neurons, explicitly incorporating intrinsic properties of excitatory and inhibitory neurons. The model was run using The Virtual Brain (TVB) simulator, and is open-access in EBRAINS. We report that when AdEx mean-field neural populations are connected via structural tracts defined by the human connectome, macroscopic dynamics resembling human brain activity emerge. Importantly, the model can qualitatively and quantitatively account for properties of empirically observed spontaneous and stimulus-evoked dynamics in space, time, phase, and frequency domains. Large-scale properties of cortical dynamics are shown to emerge from both microscopic-scale adaptation that control transitions between wake-like to sleep-like activity, and the organization of the human structural connectome; together, they shape the spatial extent of synchrony and phase coherence across brain regions consistent with the propagation of sleep-like spontaneous traveling waves at intermediate scales. Remarkably, the model also reproduces brain-wide, enhanced responsiveness and capacity to encode information particularly during wake-like states, as quantified using the perturbational complexity index. The model was run using The Virtual Brain (TVB) simulator, and is open-access in EBRAINS. This approach not only provides a scale-integrated understanding of brain states and their underlying mechanisms, but also open access tools to investigate brain responsiveness, toward producing a more unified, formal understanding of experimental data from conscious and unconscious states, as well as their associated pathologies.

The Effects of Environmental Scene and Body Posture on Embodied Strategies in Creative Thinking

ABSTRACT Accounts of embodied cognition suggest that environmental scene and motor information can be used in a type of neural simulation when generating creative uses for manipulable objects. A scarce amount of studies suggests that the state of the body and environment play a role in people’s ability to devise creative uses for objects. In this theoretically motivated study, we manipulated the environmental context and adopted a previously used body posture manipulation, while participants performed the alternative uses task (AUT) under general instructions or instructions to describe actions toward or away from the body. We show that the environment and body posture interact to shape performance on the AUT. Specifically, body posture facilitates fluent responding under high fixedness environmental contexts. In addition, describing actions away from the body results in more original AUT responses. Further, we replicate previous findings that higher-creative individuals use different “embodied strategies” than less-creative individuals when generating creative uses for objects. We propose that higher-creative individuals differentially access and navigate an abstract “action space” necessary for simulating creative uses of objects.

Оценка размера теневой экономики промышленного региона (на примере Донецкой Народной Республики)

Introduction. Combating tax evasion, the shadow economy and unofficial (illegal) employment are important policy goals in every region. Achieving those goals requires various resources; therefore, it is important to determine the size of the shadow economy to use the resources efficiently. Purpose. The paper aims at developing and testing a methodology for neural simulation estimates to find the size of the shadow economy in an industrial region taken as an example, i.e. the Donetsk People’s Republic. Materials and Methods. The study refers to the data from the World Bank Group’s Ease of Doing Business Index and data about the size of the shadow economy calculated with L. Medina and F. Schneider’s methodology. The neural network learns from the generated data array of 17,160 values for the economies of the world for 2006–2015 with the Statistica software package. The result was then used to estimate the value of the shadow economy in the industrial region. Results. The hypothesis that the Ease of Doing Business Index and the size of the shadow economy are correlated has been proposed and tested. The indicators of the Ease of Doing Business Index and the size of the shadow economy by country were sampled. The neural network was taught to assess the size of the shadow economy with the Statistica software package. The values of the Ease of Doing Business Index have been determined for the Donetsk People’s Republic. The taught neural network calculates the size of the shadow economy in the Donetsk People’s Republic and establishes that its size ranges between 30 and 40 %. The debatable issues of applying the suggested algorithm for calculating the size of the shadow economy in an industrial region are outlined. Conclusion. It is often not the irresponsibility of the entrepreneurs themselves that leads to the shadowing of the economy, but the difficulties with the business legalization and exorbitant taxes. The measures of minimizing the level of the shadow economy in the Donetsk People’s Republic are described. The suggested information will be useful for those who determine the economic policy of the region.