130 results about "Membrane potential" patented technology

The invention discloses a liquid state machine dynamics optimization system based on a self-feedback mask, and the system comprises a liquid state machine reservoir which is provided with a plurality of spike neurons; the liquid state machine reservoir updates the membrane potential state of the spiking neuron by receiving the time sequence input signal; the mask generation module is connected with a membrane potential state of a spiking neuron in a reservoir of the liquid state machine and is used for generating a self-feedback mask according to the membrane potential state; and the feedback adjusting unit is respectively connected with the mask generation module and the liquid state machine reservoir and is used for adjusting the circulating connection and the self-feedback calculation of the liquid state machine reservoir according to the self-feedback mask. The dynamic stability of the liquid state machine and the sufficiency of time sequence information utilization can be optimized, and the performance and robustness of the liquid state machine can be improved.

The invention discloses a pulse neural network lightweight method based on nonlinear pruning, and relates to the technical field of artificial intelligence and neural networks. The method comprises the following steps: S1, constructing an NDI-LIF spiking neural network model, and introducing a bilinear product term into an input current in a dynamic updating process of a neuron membrane potential; s2, constructing a nonlinear synaptic pruning mechanism, representing a connection weight as a re-parameterization function composed of a re-parameterization weight and a conversion gain coefficient, and pruning the connection weight based on the re-parameterization function; and S3, training the NSPDI-SNN model to obtain a lightweight pulse neural network. By introducing a nonlinear dendritic integration mechanism and a state-adjustable synaptic pruning mechanism, balance between network sparsity and high performance is realized while the expression ability of the model is enhanced, and a lightweight spiking neural network model with high spatial-temporal expression ability, highly sparse structure and reasonable biological mechanism is constructed. And the generalization ability and the efficient training performance under various tasks are ensured.

The invention relates to a battery life and safety collaborative prediction method and system based on an improved pulse neural network algorithm, and belongs to the technical field of electric tool battery management. The method is based on a four-level SNN architecture, an input layer receives voltage, current, temperature, stress and change rate parameters of a battery, and neural pulse conversion of the parameters is realized through rate coding and time coding; the synaptic layer simulates a physical hysteresis effect between parameters by using a double-index model, and the hidden layer passes through Adaptive Iamp; the F neurons integrate historical memory and nonlinear coupling, and the output layer generates health status, residual life and risk index. The synaptic weight is optimized through a supervised STDP learning rule, the contribution degree of each parameter is quantified, a membrane potential linear accumulation mechanism of the life dimension and an abnormal pulse triggering mechanism of the safety dimension are established, and a joint early warning decision is realized. The system is deployed in a battery management system, the battery state can be monitored in real time, and accurate decision support is provided for battery full life cycle management.

The invention discloses a myocardial transmembrane potential segmented time sequence reconstruction method based on a physical information neural network, and the method improves the model generalization ability: a physical information data generation step, especially a diversified generation strategy, can create large-scale training data covering wide physiological and pathological states, and improves the model generalization ability. According to the method, a deep learning model can learn robustness characterization of various complex electrocardio phenomena, and the generalization ability of the model and the applicability of the model in a real scene are greatly improved. According to the composite loss function, especially a physical consistency loss item, the physical law describing propagation of an electric signal from the heart to the body surface serves as a soft constraint to be embedded into the training process, a solution output by a forcing network must be capable of'explaining 'observed body surface potential BSP data, and the BSP data can be used as a soft constraint. The method greatly reduces the understanding space, and effectively inhibits the generation of artifacts and wrong solutions which do not accord with physical laws.

The embodiment of the invention relates to the technical field of artificial intelligence, and discloses an image analysis method and device based on a pulse neural network, computer equipment and a computer readable storage medium, and the method comprises the steps: obtaining a to-be-analyzed image; inputting the to-be-analyzed image into a pulse neural network based on suppression neurons; the neural network comprises an SNN module based on convolution, an SNN module based on Transform, a pulse self-feedback suppression module and a reasoning module. The suppression pulse self-feedback module comprises at least one suppression neuron and is used for dynamically issuing suppression pulses to adjust the membrane potential state of surrounding excitation neurons; neurons of the pulse neural network carry out signal transmission through integer pulses. The neurons comprise an SNN module based on convolution, and excitation neurons and inhibition neurons in the SNN module based on Transform; and outputting an analysis result. By means of the mode, the model prediction precision is improved, and computing resources are remarkably reduced.

The invention relates to the technical field of bionic control, and discloses a robot control method based on a pulse neural network, and the method comprises the steps: carrying out the spatial-temporal feature coding of the environment perception information and body state information of a robot, and obtaining a pulse feature sequence; performing pulse dependence fusion on the pulse characteristic sequence and the feedback pulse signal to obtain a bionic information processing process, and performing membrane potential integration on the bionic information processing process to obtain an integrated output quantity; performing pulse distribution processing on the integrated output quantity to obtain a driving pulse signal; performing path fitting on the driving pulse signal to obtain a continuous motion track; based on the continuous motion track and the bionic information processing process, performing joint space mapping on the continuous motion track to obtain an execution control instruction; the application executes the control instruction, synchronously collects a pulse response signal and feeds back the pulse response signal in real time so as to obtain an optimization control instruction of the robot; the pulse neural network-based robot control efficiency can be improved.

The invention relates to a control method of optical acupuncture, which comprises the following steps: a control system continuously applies a plurality of optical stimulation pulses lower than a nerve excitation threshold value, and target acupuncture point nerves enter a subcritical energy storage state through membrane potential superposition after synapse; laser output intensity is increased, and a directional threshold crossing pulse is emitted to trigger nerve discharge; judging the principal axis direction of local tissue fibers, and adjusting the laser incident angle, so that a light beam penetrates along the principal axis direction to trigger compliant tissue resonance; two beams of frequency differential laser are used for cross irradiation in a non-treatment acupoint area needing suppression to form fixed interference fringes, and a minimum power drop point is calculated to cover and suppress a target acupoint, so that the acupoint is always in an interference offset zone in the treatment process to form a selective function suppression window; a control system synchronously collects heart rate and respiration signals of a user, extracts respective periods to calculate a minimum common rhythm, sets a subsequent light stimulation rhythm as a physiological rhythm differential resynchronization frequency, and enables alternating resonance activation of different neural rhythms through phase staggering.

The invention provides a spiking neural network distributed system based on adaptive time sequence granularity, and relates to the technical field of computers, and the system comprises a spiking neural network splitting and cross-node communication module which is used for adaptively splitting a spiking neural network and supporting distributed communication on a GPU; the spiking neural network reconstruction module based on state-parameter decoupling is used for optimizing neuron internal state management during distributed deployment of the spiking neural network, and optimization comprises state parameter decoupling through a container and membrane potential cross-time sequence transmission through a serialization controller; and the timing sequence granularity self-adaptive spiking neural network communication module is used for dynamically adjusting the timing sequence granularity of the spiking neural network in a distributed environment and optimizing communication between GPU nodes according to the dynamically determined timing sequence granularity. By adopting the scheme, the time sequence granularity can be adaptively adjusted, the calculation efficiency and the communication overhead can be balanced, and the overall performance of the system can be improved.

The invention discloses an MI-EEG classification and recognition method based on a pulse neural network. The method comprises the steps that electroencephalogram signals are collected and preprocessed; converting the two-dimensional time-frequency image into a pulse sequence in a Poisson coding mode; generating a frequency histogram of excitation neuron pulse distribution by inputting the pulse sequence into a pulse neural network; and classifying the output pulses by adopting a voting method. According to the invention, the pulse neural network is used to identify and classify the image, so that the precision meets the requirement, and the expenditure of power consumption and the like is reduced; meanwhile, an STDP learning mechanism and a lateral inhibition mechanism are introduced, the connection strength is adjusted through the STDP learning mechanism, and excitation neurons of unissued pulses are inhibited through the lateral inhibition mechanism; the STDP learning mechanism and the side suppression mechanism jointly influence the neuron group, so that the neurons of the corresponding instructions can emit pulses more easily, the membrane potential of the inactive neurons is reduced, the cost of emitting the pulses is increased, and the pulses are less likely to be emitted.

A neural network device according to an embodiment includes a plurality of synapse circuits and a plurality of neuron circuits. A first neuron circuit out of the neuron circuits includes an input circuit, a charge holding circuit, a comparison circuit, a firing circuit, a charge control circuit, and a control signal output circuit. When a determination signal changes from a second value to a first value, the firing circuit outputs a spike signal. The control signal output circuit outputs a control signal indicating a comparison voltage that is based on an excess component of a membrane potential exceeding a threshold potential. In response to acquiring the spike signal from the first neuron circuit, a first synapse circuit out of the synapse circuits that acquires the spike signal from the first neuron circuit outputs a synaptic current of a current amount corresponding to the control signal and a synaptic weight.

This invention discloses a method for constructing a spiking neural network based on recurrent firing neurons, comprising: S1: constructing a recurrent firing spiking neuron model; S2: calculating positive and negative pulses based on the gradient function corresponding to each time step in the recurrent firing spiking neuron model to obtain cross-time-step positive and negative pulses; S3: backpropagating the cross-time-step positive and negative pulses back through the recurrent firing spiking neuron model to obtain a spiking neural network; S4: training the spiking neural network using a positive and negative membrane potential balance loss function to obtain a recurrent firing spiking neural network. This invention significantly improves the learning ability and information representation ability of the spiking neural network, while enhancing its robustness.

This invention discloses a liquid state machine dynamics optimization system based on a self-feedback mask, comprising a liquid state machine reservoir with multiple spike neurons; the liquid state machine reservoir updates the membrane potential state of the spike neurons by receiving time-series input signals; a mask generation module, connected to the membrane potential state of the spike neurons in the liquid state machine reservoir, is used to generate a self-feedback mask based on the membrane potential state; a feedback adjustment unit, connected to both the mask generation module and the liquid state machine reservoir, is used to adjust the cyclic connections and self-feedback calculations of the liquid state machine reservoir based on the self-feedback mask. This invention can optimize the stability of liquid state machine dynamics and the sufficiency of time-series information utilization, thereby improving the performance and robustness of the liquid state machine.

An artificial neuron includes a first capacitive node of application of a membrane potential of the neuron. A first transistor is configured to discharge the first capacitive node. A second capacitive node is driven according to the membrane potential and delivers a potential for controlling the first transistor. A second transistor is configured to discharge the second capacitive node. The second transistor is controlled according to a potential present at the second capacitive node.

The application provides a kind of method for enhancing robustness of pulse neural network based on neural activation and connection optimization, applied to artificial intelligence and neural network technical field, the method comprises: constructing pulse neural network model, the neuron of pulse neural network model adopts burst enhancement type pulse neuron, the input of pulse neural network model is image data, and the output of pulse neural network model is the image processing result corresponding to computer vision task;When the membrane potential of burst enhancement type pulse neuron exceeds the firing threshold, the part exceeding the membrane potential is converted into the pulse output within the burst window by quantization linear mapping function, and the number of pulse output is an integer between 0 and the preset maximum burst pulse number;When training pulse neural network model, add activation perception regularization term to total loss function;Through the application, the accuracy and robustness can be simultaneously improved while maintaining the high energy efficiency advantage of pulse neural network.

A neural network device according to an embodiment includes a plurality of synapse circuits and a plurality of neuron circuits. A first neuron circuit includes a first terminal to which a synaptic current is supplied. The first neuron circuit includes a secondary battery element, a spike generation circuit, and a reset control circuit. The secondary battery element accumulates a charge according to the synaptic current supplied to the first terminal. The spike generation circuit generates a spike signal when the membrane potential generated from the secondary battery element is larger than a threshold potential being a predetermined potential. The reset control circuit releases the charge accumulated in the secondary battery element during a refractory period being a predetermined time after generation of the spike signal.

The present application relates to the field of artificial neuromorphics, and provides an artificial neuron circuit based on a volatile threshold switching device. A membrane potential generation circuit is used for generating different membrane potentials under different input currents and different external reset voltages; a slow variable generation circuit is used for generating a slow variable; a membrane potential reset circuit comprises a volatile threshold switching device and an external reset voltage excitation source; the external reset voltage excitation source is used for providing different external reset voltages, so that the membrane potential and the slow variable are changed, and a spiking behavior of biological neurons is simulated; and the volatile threshold switching device is used for realizing a membrane potential reset function. Compared with traditional CMOS-circuit-based neurons, the neuron circuit provided by the present application requires a greatly reduced number of transistors and reduced hardware and power consumption overhead, has the characteristics of simple structure, high flexibility and rich functions, and is beneficial to large-scale integration in a hardware pulse neural network.

The invention relates to the technical field of luminescent materials, in particular to a nano material and a preparation method and application thereof. The invention provides a nanometer material, the nanometer material is NaYbF4: Er (at) NaYF4 (at) Cy7.5 (at) DPPC, the nanometer material comprises a core-shell structure with the component of NaYbF4: Er (at) NaYF4, and the core-shell structure is modified with DPPC and compounded with Cy7.5. When the nano material provided by the invention is used for an optical nano indicator or sensor for detecting the membrane potential under an NIR-IIb window, the sensor generates emitted light (1500-1700nm) under the NIR-IIb window under the excitation of 808nm near-infrared incident light. By utilizing the nano sensor, the membrane potential change of cells marked by the sensor can be visually detected by means of optical imaging.

The invention relates to the field of artificial intelligence and the field of integrated circuits, in particular to a chip for in-memory calculation of a spiking neural network, which comprises a synaptic weight storage array for storing synaptic weights; the threshold storage array is used for storing a neuron threshold voltage; the membrane potential storage array is used for storing a neuron membrane potential, a bit line of the membrane potential storage array is connected with a bit line of the synaptic weight storage array, and logic calculation results of the synaptic weight and the neuron membrane potential are directly generated on the connected bit lines; the membrane potential accumulation unit obtains a membrane potential accumulation result through processing of an addition logic circuit; the comparison unit is used for opening read sub-lines of the membrane potential storage array and the threshold voltage storage array and multiplexing an addition logic circuit to realize subtraction calculation of a membrane potential accumulation result and a threshold voltage; and the transmitting unit is used for judging whether the neurons output pulses or not and resetting the membrane potential. The method can achieve the push-training integration of in-memory calculation, remarkably improves the SNN operation energy efficiency, and reduces the power consumption.

The invention discloses a low-voltage nanosecond pulse platform waveform fitting method, and belongs to the technical field of high-frequency pulse control, and the method comprises the steps: obtaining a set voltage value, calculating the optimal pulse width corresponding to a target cell through a transmembrane voltage formula according to the set voltage value, and calculating the optimal pulse width according to a dose and transmembrane potential accumulation model. Calculating the number of equivalent nanosecond pulse sub-pulses capable of replacing the optimal pulse width, calculating the number of pulse strings according to the total dose set by the optimal pulse width and the number of the equivalent nanosecond pulse sub-pulses, fitting nanosecond pulses, and replacing the optimal pulse width, the calculated nanosecond fitting value, the calculated pulse string number and preset parameters are stored in a storage chip of the pulse platform, the pulse platform calls the storage parameters for energy emission, the low-voltage nanosecond platform can be used for achieving the combined effect of a microsecond platform and a high-voltage nanosecond platform, released bubbles are smaller than those of the microsecond platform, the occurrence rate of embolism is reduced, and the service life of the microsecond platform is prolonged. The released voltage is lower than that of traditional nanoseconds, and the problems of blood fusion and smooth muscle and nerve damage are reduced.

The invention discloses an FPGA-based nerve cell membrane potential detection method and a related device, and the device comprises an input excitation module which is used for inputting an analog current signal to an FPGA core processing module; a model IP core is deployed in the FPGA core processing module, a pipeline design is adopted in the model IP core, the model IP core is obtained by mapping a Hodgkin-Here differential equation corresponding to the Purkinje cell electrophysiology model through high-level synthesis, the Purkinje cell membrane potential value at the current moment is calculated in real time according to an analog current signal, and the Purkinje cell membrane potential value is calculated in real time according to the calculated Purkinje cell membrane potential value. A digital voltage signal is obtained; the digital-to-analog conversion module is used for receiving the digital voltage signal output by the FPGA core processing module and converting the digital voltage signal into an analog voltage signal; and the display module is used for displaying the membrane potential waveform in real time according to the analog voltage signal. On the basis, the real-time high-speed calculation of the membrane potential change can be realized, and the occupation of hardware resources is effectively reduced.

Disclosed herein are system, method, and computer program embodiments for an improved spiking neural network (SNN) configured to learn and perform unsupervised, semi-supervised, and supervised extraction of features from an input dataset. An embodiment operates by receiving a modification request to modify a base neural network, having N layers and a plurality of spiking neurons, trained using a primary training dataset. The base neural network is modified to include supplementary spiking neurons in the Nth or N+1th layer of the base neural network. The embodiment includes receiving a secondary training dataset and determining membrane potential values of one or more supplementary spiking neurons in the Nth or Nth+1 layer which learn features based on secondary training data set to select a supplementary / winning spiking neuron. The embodiment performs a learning function for the modified neural network based on the winning spiking neuron.

This invention provides a method, system, electronic device, and storage medium for analyzing the energy consumption of neural networks based on ion channel regulation. It constructs a neural network dynamics model containing excitatory and inhibitory neurons. The neuronal membrane potential dynamics equation is defined by a set of differential equations for various voltage-gated ion currents (fast sodium ion current, continuous sodium ion current, slow potassium ion current, etc.) and leakage current. By adjusting the unit area conductivity parameter and / or inactivation time constant parameter of the voltage-gated ion channels in the model, the external intervention effect is simulated. After obtaining membrane potential sequence data through simulation, the total energy consumption, synaptic energy consumption ratio, average discharge rate, and average synchronization rate between neurons of the model are calculated. Finally, a quantitative correlation model between parameter changes and the above indicators is established, clarifying the coupling effect of the target ion channel dynamics characteristics on the neural network discharge behavior pattern and energy metabolism.

The application provides a sparse pulse neural network accelerator based on a ping-pong architecture, which transmits compressed weight values to a compressed weight calculation module, uses a sparse pulse detection module to extract valid pulse indexes from pulse input signals, avoids each pulse signal from participating in operation, reduces the amount of calculation, and accumulates non-zero values in the compressed weight values to the membrane potential of neurons according to the valid pulse indexes, so as to finally determine whether to fire a pulse or not. Compared with the technical solution in which all synapses are activated and participate in operation in a conventional synaptic cross array, only the synapse weights corresponding to the valid pulse indexes are activated in the application, and other synapses do not participate in operation, so that the amount of calculation is reduced, the running power consumption of the whole chip is reduced, and the running speed, energy efficiency and area efficiency of the pulse neural network are improved.

This invention provides a multimodal human-machine interaction chip based on an adaptive threshold spiking neural network, comprising: a sensor module for real-time acquisition of input data from three modalities: visual signals, pressure signals, and surface electromyography (sEMG) signals; an acquisition module for analog-to-digital conversion and preprocessing of pressure signals and sEMG signals to generate feature tensors; and a recognition module comprising three adaptive threshold spiking neural networks (CNNs) that process the feature tensors of the three modalities respectively, perform cross-modal feature fusion, and output the robot's motion intention. The adaptive threshold spiking neural network dynamically adjusts the membrane potential threshold of the spiking neurons, reducing power consumption while maintaining computational accuracy. This invention, by combining the input data from the three modalities with the adaptive threshold spiking neural network algorithm for feature fusion, can comprehensively capture the operator's motion intention and the environmental interaction state, significantly improving the accuracy and adaptability of robot motion generation and enhancing the naturalness and reliability of human-machine collaboration.

The invention relates to a piezoelectric nano-material with ultrasonic responsiveness, which is a PVP (Polyvinyl Pyrrolidone) modified bismuth oxyiodate nano-sheet, the particle size is 100-200 nm, and a surface electric field can be generated under the ultrasonic action. The piezoelectric nano material can hyperpolarize the potential of a T cell membrane and inhibit an ion channel Kv1.3 regulated and controlled by the membrane potential, so that Th17 cell differentiation is reduced, and joint inflammation is relieved; the piezoelectric nano material is further prepared into a joint injection preparation, and the joint injection preparation can be combined with ultrasonic irradiation to be applied to treatment of rheumatoid arthritis. According to the invention, the piezoelectric nano material is successfully applied to immune microenvironment regulation and control of rheumatoid arthritis for the first time, and the product and technology blank in the field of'precisely regulating and controlling immune cell functions by using physical signals to treat autoimmune diseases' is filled.

Minimize the loss of information transmitted. [Solution] The neural network device according to the embodiment comprises a plurality of synaptic circuits and a plurality of neuron circuits. The first neuron circuit among the plurality of neuron circuits is supplied with synaptic current to its first terminal from each of the first synaptic circuits among the plurality of synaptic circuits. The first neuron circuit has a charge storage circuit, a spike output circuit and a cutoff circuit. The charge storage circuit stores charge according to the synaptic current and generates a membrane potential according to the stored charge. The spike output circuit outputs a spike signal when the membrane potential is greater than a preset threshold potential. The cutoff circuit stops the supply of synaptic current from the first terminal to the charge storage circuit during a cutoff period, which is a predetermined time after the spike signal is output.

The embodiment of the invention provides an artificial neural network model conversion method, a storage medium and a program product, and relates to the technical field of artificial intelligence, and the method comprises the steps: after obtaining an artificial neural network model obtained through pre-training, converting each nonlinear operator in the artificial neural network model into a corresponding pulse module, each pulse module comprises a difference expectation compensation module, the difference expectation compensation module is used for calculating an output increment according to the accumulated membrane potential and inserting a difference pulse neuron into each pulse module, the difference pulse neuron updates a coding activation value when issuing a pulse, otherwise, the coding activation value is kept unchanged, and the difference expectation compensation module is used for outputting the difference pulse neuron. According to the method, the bias term of the linear operator located on the previous layer of each nonlinear operator is removed, the initial membrane potential of the differential pulse neuron inserted into the pulse module corresponding to the nonlinear operator is set as the bias term, and the coding activation value is updated only when the pulse is emitted, so that the loss caused by updating the coding activation value no matter whether the pulse is emitted or not is avoided, and the accuracy of the coding activation value is improved. And the energy consumption is obviously reduced.

The application discloses a kind of video physiological signal dynamic extraction method and device, collect video and carry out face detection and posture correction to each frame video;The corrected face image is divided into micro-grid, and the brightness of each micro-grid is response-delay coding, and the pulse sequence of presynaptic pulse neuron is generated;Through membrane potential model and mutual inhibition between pulse neurons, competition is carried out and combined with connection weight, and the pulse emission time and pulse frequency of postsynaptic pulse neuron are obtained;According to the time difference between presynaptic and postsynaptic pulse neuron, dynamically adjust connection weight;After sorting and screening the pulse frequency of all postsynaptic neurons, it is mapped to the pulse sequence formed by corresponding micro-grid and weighted fusion, and the final pulse wave signal is generated;Finally, the physiological parameters of practitioner are obtained by calculating pulse wave signal, and the connection weight is adjusted by signal quality evaluation, and the strategy optimization of pulse wave signal extraction is realized continuously.

The specification discloses a training method and device of a reserve pool calculation model based on a pulse signal, comprising: inputting historical environment images as training samples into each neuron of a hidden layer of a reserve pool calculation model based on a pulse signal to be trained. According to the annotation of the training sample, a target vector is determined. For each neuron, according to the pulse signals fired by other neurons at the last moment and the initial connection weight, the current connection weight of other neurons to the neuron is determined, and other input potentials of other neurons input into the neuron are determined. According to the time scale of the decay of the membrane potential of the neuron, a decay potential is determined. According to the decay potential, the other input potentials and the training sample, a state vector composed of the membrane potentials of each neuron is determined. According to the state vector and the target vector, the readout weight of the readout layer is calculated, so that the trained model has good robustness, high output result accuracy and can achieve the expected effect.