27 results about "Cerebellum" patented technology

The invention relates to the technical field of robot control, and discloses a robot cerebellum reinforcement learning method and system based on adaptive environment change, and the method comprises the steps: collecting multi-modal sensing data in the operation process of a robot; based on the multi-modal sensing data, determining an environment sudden change increment and an environment characteristic variable quantity, and obtaining a main strategy network according to the environment sudden change increment and the environment characteristic variable quantity; outputting a basic action vector of the robot based on the main strategy network, and performing value evaluation on the state of the robot to obtain a value evaluation quantity; performing updating calculation on the basic action vector based on the value evaluation quantity to obtain a compensation action vector; and performing vector superposition fusion on the basic action vector and the compensation action vector to obtain a final control vector, and taking the final control vector as a cerebellum learning action instruction of the robot, thereby realizing dynamic adaptation of a reinforcement learning strategy and cerebellum rhythm control. And the motion stability and the learning convergence efficiency of the robot in the unstructured environment are optimized.

The invention discloses a robot motion control method and system based on cerebellum reinforcement learning, and the method comprises the steps: obtaining a current environment state vector, and inputting the current environment state vector to a main strategy channel and a cerebellum compensation channel in parallel; the main strategy channel outputs a basic action based on a long-term task target, and the cerebellum compensation channel outputs a compensation action responding to real-time dynamic through an efficient query mechanism; synthesizing the basic action vector and the compensation action vector into a synthesized action vector driving robot; feeding back latest data after the robot drives the motion action vector, and determining a sensory prediction error based on the latest data; and updating the original parameters of the cerebellum compensation channel based on the sensory prediction error, and optimizing the original strategy parameters of the main strategy channel based on the latest data. According to the invention, by constructing a parallel double-channel architecture, functional decoupling of advanced decision and rapid adaptation is realized, and unification of rapid adaptation and continuous optimization is realized through a double-loop hierarchical learning system.

The invention discloses a real-time and deterministic communication monitoring system and method for large and small brains of a robot, and relates to the technical field of communication monitoring of the large and small brains, and the method comprises the steps: collecting a communication time sequence and a walking phase state sequence of an EtherCAT master station for brain planning and cerebellum servo driving in the walking control process of the robot; constructing the phase tissue density of the initial monitoring time window, and performing multi-scale time unwrapping and phase resampling on the phase tissue density to generate a plurality of candidate tissue densities; based on phase-time entropy preserving mapping, the organization density is converted into an executable monitoring scheme which is respectively applied to master station periodic communication monitoring, communication feature extraction and time sequence return evaluation, and finally the optimal organization density is screened out, so that key phase high-precision real-time communication monitoring is realized. The problem that existing communication monitoring is insufficient in sensitivity to key phases or excessively sensitive to transient jitter is solved.

The application discloses a kind of robot motion control method and system based on cerebellum reinforcement learning, it includes: obtaining current environment state vector, and parallel input to main strategy channel and cerebellum compensation channel;Main strategy channel outputs the basic action based on long-term task target, and cerebellum compensation channel then outputs compensation action by efficient query mechanism to cope with real-time dynamics;The basic action vector and compensation action vector are synthesized into synthesized action vector to drive robot;After robot drive movement action vector, feedback latest data, and determine sensory prediction error based on latest data;Based on sensory prediction error, the original parameters of cerebellum compensation channel are updated, and the original strategy parameters of main strategy channel are optimized based on latest data.The application realizes the functional decoupling of high-level decision and rapid adaptation by constructing parallel double-channel architecture, and the hierarchical learning system of double loop realizes the unity of rapid adaptation and continuous optimization.

The present invention addresses the problem of providing an agent for preventing or treating amyotrophic lateral sclerosis (ALS), Parkinson's disease (PD), Huntington's disease (HD), spinocerebellar ataxia (SCA), aging-related degenerative or neurological disease, brain aging, or diseases associated with brain aging, as well as a more stable antibody that exhibits an effect of preventing or treating these diseases, Alzheimer's disease (AD), or frontotemporal lobar degeneration (FTLD). A human monoclonal antibody that specifically binds to human HMGB1, wherein the human monoclonal antibody (anti-human HMGB1 antibody) comprises a heavy chain CDR1, heavy chain CDR2, and heavy chain CDR3 each consisting of a specific amino acid sequence and a light chain CDR1, light chain CDR2, and light chain CDR3 each consisting of a specific amino acid sequence, is used as an agent for preventing or treating ALS, PD, HD, SCA, aging-related degenerative or neurological disease, brain aging, or diseases associated with brain aging. An antibody in which the light chain complementarity determining region (CDR) 3 of the anti-human HMGB1 antibody has been modified is used.

The utility model relates to the technical field of fixing devices, in particular to an integrated fixing device for manufacturing rodent pathological sections, which comprises a fixing device shell and a fixing device cover plate which are rotatably connected. The fixing device comprises a fixing device shell, a plurality of brain cutting clamping grooves are formed in the periphery of the fixing device shell, brain clamping grooves are formed in the bottom of the fixing device shell, the brain clamping grooves comprise a cerebellum fixing groove, a brain fixing groove and an olfactory bulb fixing groove, and the cerebellum fixing groove, the brain fixing groove and the olfactory bulb fixing groove are integrated. The closed fixing groove can be adopted, formalin volatilization is reduced, the brain tissue fixing function and the material taking function are integrated, and the operation process is simplified.

The invention discloses application of neurogranular protein in diagnosis and treatment of spinal cerebellar ataxia type 3, and belongs to the technical field of biology. Research finds that SCA3 patients and healthy control can be well distinguished by detecting the concentration of the nerve granule protein in plasma, and the diagnosis efficiency of the nerve granule protein in the SCA3 is proved. Furthermore, by overexpressing the neurogranular protein in an SCA3 mouse model, it is proved that the neurogranular protein can be used as a therapeutic target to treat SCA3. According to the invention, a new biomarker and a treatment target are provided for diagnosis and treatment of SCA3, and a technical support and a theoretical basis are provided for precise diagnosis and treatment of SCA3 and in-depth research of pathogenesis of SCA3.

Provided herein are targeting peptides and vectors containing a sequence that encodes the targeting peptides that deliver agents to specific substructures in the brain. Specifically, the targeting peptide is a component of a modified, sequence-specified adeno-associated virus (AAV) capsid protein further wherein the brain substructure may be the globus pallidus, putamen, internal capsule, caudate, claustrum, substantial nigra, motor cortex, insula,.temporal cortex, thalamus, hippocampus, subiculum, and deep cerebellar nuclei.

This invention overcomes the shortcomings of uncontrollable hardware interrupts in existing general control architectures and provides a hardware interrupt fault-tolerant emergency control method and system based on a brain-like independent architecture. It simulates the brain and cerebellum division of labor logic in the human brain, realizes complete isolation between the emergency control link and hardware interrupts, and completes millisecond-level emergency takeover when the main processor is abnormal, eliminating the risk of equipment loss of control.

The VEM-robot emotional right-brain model construction method decomposes the robot's brain into an emotional right brain, an intellectual left brain, and a motor cerebellum, forming the embodied or humanoid robot operating system VEM-ROS. This enables emotional communication between the robot and humans or other robots. Human output is used as the perceptual spectrum, and robot output as the deductive spectrum. Both the perceptual and deductive spectra are segmented, synthesized, and aligned using the emotional rhythm of the multimodal VEM-Token. LLM-Token decomposition of the large language model calculates lexicalized logical perception, and VEM-Token decomposition calculates the micro-expressions of multimodal emotional components. VEM-ROS also includes priority and interruption mechanisms, VEM memory mechanisms, dialogue relationships, craniofacial and limb sensors and actuators and interfaces, robot cloning, and operating system encapsulation. The right-brain model supports emotional micro-expression communication between the robot and humans, similar to the Turing test, and is expected to distinguish whether the emotional test subject is a robot or a human.

An aspect of the present description relates to compounds useful for improving pre‑mRNA splicing in a cell. In particular, another aspect of the present description relates to substituted heteroaryl compounds, forms, and pharmaceutical compositions thereof and methods of use for treating or ameliorating spinocerebellar ataxia type 3 (SCA3), also known as Machado–Joseph disease (MJD).

Provided is a method of treating spinocerebellar ataxia in a patient. The method includes administering to the patient in need of such treatment troriluzole at a dosage amount of about 200 mg per day and for a treatment period of at least one year, wherein the treatment is effective to provide a reduction in the patient's total f-SARA score of at least 50% compared to an untreated patient.

Provided are compounds, methods, and pharmaceutical compositions for reducing the amount or activity of ATXN3 RNA in a cell or animal, and in certain embodiments reducing the amount of ATXN3 protein in a cell or animal. Such compounds, methods, and pharmaceutical compositions are useful to ameliorate at least one symptom or hallmark of a neurodegenerative disease. Such symptoms and hallmarks include motor dysfunction, aggregation formation, and neuron death. Such neurodegenerative diseases include spinocerebellar ataxia type 3 (SCA3).

The application relates to the technical field of robot control, and discloses a robot cerebellum reinforcement learning method and system based on adaptive environmental changes, multi-modal sensing data in the running process of a robot is collected; based on the multi-modal sensing data, an environmental mutation increment and an environmental feature change amount are determined, and a main strategy network is obtained according to the environmental mutation increment and the environmental feature change amount; a basic action vector of the robot is output based on the main strategy network, the state of the robot is valued, and a value evaluation amount is obtained; the basic action vector is updated and calculated based on the value evaluation amount, and a compensation action vector is obtained; the basic action vector and the compensation action vector are subjected to vector superposition and fusion, a final control vector is obtained, and the final control vector is taken as a cerebellum learning action instruction of the robot, dynamic adaptation of a reinforcement learning strategy and cerebellum rhythm control is realized, and the motion stability and learning convergence efficiency of the robot in an unstructured environment are optimized.

This invention discloses a passive running device for a mouse with its head fixed, comprising a base plate, a support assembly, a head fixation component, and a running mechanism. The support assembly is disposed on the base plate. The head fixation component is used to fix the head of the mouse and is detachably connected to the support assembly. The running mechanism is disposed on the base plate and includes a driver and a running belt assembly. The head fixation component is located above the running belt assembly to restrain the mouse on the running belt assembly. The driver is driven by the running belt assembly to drive the running belt assembly, enabling the mouse on the running belt assembly to passively perform running movements. This invention, by fixing the head of the mouse and utilizing the movement of the running belt assembly to passively induce running movements, can assist in observing the two-photon neuronal electrical signals of the mouse during running using a two-photon microscopy system, and observe the cerebellar neuronal firing patterns during movement, thus contributing to neuroscience research.

The application provides a multi-functional operant conditioning reflex neural network circuit with blocking and competition effect, wherein a stimulating signal is connected with an input end of a hippocampus module, an exciting signal and an output signal are connected with input ends of a cerebellum module, the exciting signal, the hippocampus module and the cerebellum module are connected with an input end of a voltage control module, an output end of the voltage control module is connected with an experience module and a feedback module respectively, an output end of the experience module is connected with a decision module and the feedback module respectively, an output end of the feedback module is connected with the voltage control module and the experience module respectively, and an output end of the decision module is connected with the feedback module and obtains the output signal. The application realizes learning of operant conditioning reflex under multiple input signals, can simulate decision judgment of animals in a complex natural environment, and widens the possibility of the existing operant conditioning reflex circuit.

A robot arm brain-like control method and control apparatus based on a spiking neural network. The control method comprises: on the basis of a spiking neural network, constructing a cerebellar module, so as to use the cerebellar module to obtain an initial torque instruction for a target robot arm; on the basis of the initial torque instruction, constructing a thalamic module, and using the thalamic module to adjust the initial torque instruction, so as to obtain a final torque instruction; using the final torque instruction to construct a brainstem module, so as to obtain a feedforward torque; and on the basis of spike coding and the feedforward torque, constructing a spinal cord module, and combining the cerebellar module, the thalamic module and the brainstem module to construct a robot arm control model, so as to output a torque control instruction. The control apparatus is used for implementing such a control method. The present invention can solve problems such as reduced control accuracy, impaired universality and inability to adapt to control tasks under complex load conditions, which are caused by significant differences in working requirements when a robot arm executes different tasks due to the impact of the size, mass, etc., of an operated object in complex working scenarios.

This application provides a reinforcement learning system, method, and medium based on cerebellar circuit structure, relating to the field of artificial intelligence technology. The system includes: a state input module for acquiring state information in the reinforcement learning environment; a feature extraction module for extracting features from the state information to obtain state feature representation information; a cerebellar structure processing module for performing high-dimensional expansion, sparse connectivity, sparse activation, and dendritic modulation processing on the state feature representation information by simulating the information processing mechanism of the biological cerebellum to obtain a fused feature representation vector; a reinforcement learning decision module for inputting the fused feature representation vector into a reinforcement learning decision network to obtain an action space and corresponding action value vectors; and a policy output module for selecting the action with the maximum action value based on the action value vector output by the reinforcement learning decision network to generate a target action policy. This system can improve the efficiency of action decision-making tasks.

The application provides a cerebellar Purkinje neuron rapid identification system and method based on multi-modal deep learning, relates to the cross technical field of biomedical engineering and computer vision, and fuses specific fluorescence features and multi-dimensional morphological features, constructs a special multi-modal deep learning model, solves the interference problem of other neurons in a mixed culture system, and greatly reduces the false positive rate; a small sample learning architecture solves the industry pain point of a small amount of biological sample labeled data, improves the model generalization capability, and adapts to multi-scene sample identification. The identification system fuses Purkinje neuron specific fluorescence labeling features and cell morphological features, constructs a multi-modal deep learning recognition model based on small sample learning, is matched with a full-automatic microscopic imaging and analysis system, realizes rapid and high-accuracy identification of Purkinje neurons under a mixed culture system, does not need a complex cell purification step, greatly shortens the experimental period, and reduces the technical threshold.

Provided are compounds, methods, and pharmaceutical compositions for reducing the amount or activity of ATXN1 RNA in a cell or subject and, in certain cases, reducing the amount of ATXN1 in a cell or subject. Such compounds, methods, and pharmaceutical compositions can be used to improve at least one symptom or marker of a neurodegenerative disease. Such symptoms and markers include gait and limb ataxia, cognitive impairment, difficulty speaking and swallowing, atrophy of the cerebellum and brainstem in magnetic resonance imaging (MRI), neurochemical abnormalities in the cerebellum and brainstem detected via magnetic resonance spectroscopy (MRS), and death within 10-15 years after symptom onset. Such neurodegenerative diseases include spinocerebellar ataxia type 1.

Provided are compounds, methods, and pharmaceutical compositions for reducing the amount or activity of ATXN3 RNA in a cell or animal, and in certain embodiments reducing the amount of ATXN3 protein in a cell or animal. Such compounds, methods, and pharmaceutical compositions are useful to ameliorate at least one symptom or hallmark of a neurodegenerative disease. Such symptoms and hallmarks include motor dysfunction, aggregation formation, and neuron death. Such neurodegenerative diseases include spinocerebellar ataxia type 3(SCA3).

The application discloses a kind of " artificial muscle " soft driver bionic control method based on pulse neural network, the method is with rough adjustment based on feedforward controller of rough inverse model, with fine adjustment based on cerebellum controller of pulse neural network, real-time compensation modeling uncertainty and environmental interaction uncertainty is realized through online learning method.The application uses representative " artificial muscle " soft driver-pneumatic artificial muscle, realizes a kind of pulse neural network that imitates mammal cerebellum neural working mechanism, realizes the efficient, robust, adaptive control for highly nonlinear, time-varying " artificial muscle " soft driver.

The invention relates to a multi-agent robot cooperative grabbing system and method based on a brain-cerebellum layered architecture. The system comprises an industrial-grade body brain coordination module, a cross-body execution module and a shared memory bus module. And the industrial-grade body brain coordination module constructs a memory-driven reasoning mechanism based on a visual language multi-modal large model, extracts a context from space-time memory through an RAG retrieval enhancement generation mechanism, generates a structured reasoning chain and a subtask DAG, and realizes dynamic task allocation and exception recovery. And the cross-implement execution module establishes a standardized robot file, and realizes unified scheduling and safe execution of heterogeneous robots through an adaptation layer. And the shared memory bus module constructs a real-time shared space-time memory system based on Redis, constructs a three-level layered scene graph by adopting multi-view RGB-D input and SAM and CLIP technologies, and realizes deterministic information exchange and conflict resolution among multiple robots. The method has the advantages of fast response, strong compatibility, high cooperation precision and the like.

The application belongs to the technical field of motor speed regulation control, and designs a permanent magnet synchronous motor speed regulation control method based on an improved cerebellar model neural network controller, a storage medium and equipment, which comprises the following steps: firstly, a mathematical model of a permanent magnet synchronous motor is established, and then an improved cerebellar model neural network controller is designed and applied to a speed loop in a vector control structure of the permanent magnet synchronous motor; the controller improves the weight adjustment mode and learning rate of the original network, adopts a gradient descent method combined with an inverse relationship of learning times to update the weight of the cerebellar model neural network, replaces a calculation mode of division by a network generalization parameter C for average distribution, and improves the learning efficiency of the network; meanwhile, the value of the network learning rate is dynamically adjusted according to the motor operating state, and the rapidity and stability of the system are improved. Results show that the improved cerebellar model neural network controller effectively improves the response speed of the system and greatly improves the dynamic stability of the permanent magnet synchronous motor.

A method and apparatus for preventing brain death by using rapid and safe cooling of the brain is disclosed. The fourth ventricle is accessed through the patient's neck by using a specially designed needle / cannula that enters the cistema magna with an entry through the patient's neck, allowing the cooled artificial cerebral spinal fluid (aCSF) to flow around the brain. The aCSF exits from an opening in the skull where temperature / pressure sensors are placed. Data is sent to a computer controlled motorized system that pumps the cooled aCSF to the needle / cannula placed in the cistema magna. The pumping of the aCSF is controlled to maintain a predetermined temperature and / or pressure of the outlet aCSF.

This application relates to the field of robotic arm control technology, and particularly to a brain-like control method and device for robotic arms based on spiking neural networks. The method includes: constructing a cerebellum module based on a spiking neural network to obtain initial torque commands for the target robotic arm; constructing a thalamus module based on the initial torque commands and adjusting the initial torque commands using the thalamus module to obtain final torque commands; constructing a brainstem module using the final torque commands to obtain feedforward torque; constructing a spinal cord module based on pulse coding and feedforward torque, and combining the cerebellum module, thalamus module, and brainstem module to construct a robotic arm control model to output torque control commands. This solves the problems in related technologies where, in complex working scenarios, the size and mass of the manipulated object significantly affect the working requirements of the robotic arm when performing different tasks, leading to reduced control accuracy and impaired versatility, making it unable to adapt to control tasks under complex load conditions.

The application can detect the whole brain blood flow change of the patient with spinocerebellar degeneration and optimize the cerebellum visualization. The application carries out CBF reconstruction on ASL data, carries out nonlinear co-registration on the whole brain CBF image to the PET template, and spatially standardizes to the MNI space, divides the CBF value of each voxel by the whole brain average value to obtain the normalized CBF and smooth; construct a general linear model in SPM8, and carry out whole brain statistical analysis at the voxel level; extract the t value image of the significant cluster and its mask, and use the mask to extract the cluster average CBF value for regional analysis. The significant cluster t value image of the cerebellum is overlaid to the coronal SUIT template, the t value image is mapped to the SUIT plane graph in the SUIT toolbox, and the outline of the t value plane graph is overlaid to the cerebellar anatomical and functional orientation plane graph respectively, so that the anatomical and functional detail visualization of the cerebellar region is realized.