2463 results about "Non linearite" patented technology

Devices and techniques are generally described for nonlinear tensor compression for neural networks. In various examples, a first tensor associated with a first layer of a neural network may be determined. One or more neural processing units of accelerator hardware may generate a first compressed tensor by applying a nonlinear compression function to the first tensor. The first compressed tensor may be stored in a first memory of the one or more computer-readable media. A first operation associated with a second layer of the neural network may be determined, where the first operation uses output of the first layer. The first operation may be performed based on the first compressed tensor.

This invention discloses a stepwise data assimilation method for a set subspace in a nonlinear inverse problem, relating to the field of data processing technology. The invention constructs an initial prior physical set based on the physical state variables to be inverted and optimized, and historical observation data. It extracts the static subspace basis anomaly matrix, initializes the latent variable set and particle weights, calculates fractional-step incremental reweighting, evaluates the current likelihood mismatch penalty using predicted data, updates and normalizes the particle weights in the logarithmic domain, obtains the effective sample number, performs system resampling operations in conjunction with a preset resampling tolerance coefficient, eliminates low-weight particles and replicates high-weight particles, synchronously updates the latent variable set and predicted data, executes a dynamic MCMC mutation loop to generate proposed latent state vectors and affinely maps them to a high-dimensional physical space, and updates the latent variable set by evaluating the annealing target energy within the latent variable subspace until all fractional steps are traversed. Finally, it outputs the latent variable set and maps it back to the physical space.

The application discloses a kind of high-pressure fuel filter prediction model construction method and flow prediction method, including two prediction models, method specifically includes: using computational fluid dynamics method carries out fluid dynamics simulation calculation to high-pressure fuel filter, constructs working condition sample data set;The size of filter hole, quantity, filter working pressure, fuel viscosity and filter pressure drop are used as input variable, and filter flow is used as output variable, two new high-pressure filter general nonlinear flow prediction models are established using artificial neural network and regression prediction method respectively, the flow characteristics of different models of filter in working interval can be accurately predicted using the neural network prediction method proposed in the application at any working condition point, and the error is low;The flow of different models of filter under different conditions can be quickly and accurately obtained, the accuracy meets the engineering needs, while greatly shortening the experimental time and calculation time, saving design cost.

The application discloses a correction method and device of a seismic disaster personnel casualty assessment model and a storage medium, and relates to the technical field of disaster emergency management. By constructing an emergency preparedness index and an emergency response index as a nonlinear attenuation correction factor, the traditional assessment model's defects of relying on expert experience for manual correction and lacking in automation and intelligence are overcome. The coupling correction formula takes the nonlinear attenuation correction factor as the core, and through the multiplication coupling of independent double indexes, the synergistic attenuation effect of emergency preparedness and emergency response is accurately reflected. Through a closed-loop optimization mechanism, highly reliable data support is provided for emergency command. The correction process not only outputs personnel casualty prediction values, but also quantitatively diagnoses regional resilience by generating an emergency preparedness index and an emergency response index, so that emergency plans can be optimized, rescue force deployment can be adjusted or popular science drills can be strengthened, thereby directly serving the active prevention and control and rapid recovery goals of a resilient city.

The application discloses a tunnel lining thermal-mechanical coupling damage identification and positioning method and system based on a distributed optical fiber, the application is characterized in that a distributed optical fiber network is arranged along the circumferential direction and the longitudinal direction at the inner surface of the lining and the key positions between layers, and the full-distributed temperature and strain data are acquired in combination with a demodulator; the core of the application is a built-in decoupling algorithm of a deep residual network (ResNet) based on a sparrow search algorithm (SSA) optimization; the algorithm introduces residual connection, a LeakyReLU activation function and a SSA global optimization strategy to construct a high-precision nonlinear mapping model in view of the cross-sensitivity problem; the application can accurately remove the non-damage deformation caused by temperature, and restore the real mechanical strain with high fidelity, so that the early identification of the stress concentration area and the micro crack is realized; meanwhile, the thermal expansion and cold contraction and the structural actual damage are effectively distinguished, the defects of the limited monitoring range and the easy misjudgment of the traditional monitoring method are overcome, and the accurate damage positioning and reliable monitoring of the full scale of the tunnel are realized.

This invention discloses a real-time attitude estimation system for non-cooperative industrial targets based on strong tracking filtering, relating to the fields of industrial machine vision and automation control technology. This system decouples the complete six-degree-of-freedom attitude estimation problem into a linear translational sub-state space and a nonlinear rotational sub-state space. This decomposes the nonlinear calculations that would normally be performed in six dimensions into a three-dimensional linear problem and a three-dimensional nonlinear problem. The translational sub-state only performs a simplified Kalman gain iteration with closed-form solutions, and matrix operations are simplified to low-order linear recursion. Although the rotational sub-state uses strong tracking filtering, it only needs to perform nonlinear updates in the reduced low-dimensional space. The Jacobian matrix sparsification pre-computation and incremental update unit pre-stores the constant part of the observation equation's derivative calculation offline and only incrementally updates the sparsely changing part online. This allows the system to be fully deployed on ARM architecture edge devices and stably meet millisecond-level processing requirements.

This invention relates to the field of thermal power generation cost accounting technology, and proposes a dynamic modeling method and system for the cost per kilowatt-hour based on the forward and reverse balance method. The method includes: real-time acquisition of data on instantaneous flow rate of the coal feeder, coal pile volume, density, coal analysis upon entering the furnace, unit operation, and fuel price; calculation of dynamic coal quality data using a multimodal fusion algorithm; calculation of forward fuel cost based on flow rate, coal quality, and price data; calculation of reverse balance theoretical cost by inferring the theoretical standard coal quantity based on unit operation data; comparison and verification of forward and reverse costs to determine the final fuel cost; and calculation of dynamic cost per kilowatt-hour by combining variable cost and power generation data. This invention achieves a nonlinear dynamic mapping from multi-dimensional real-time data to accurate cost per kilowatt-hour, enabling high-precision, real-time, and automated dynamic modeling of the cost per kilowatt-hour for thermal power generation enterprises.

This invention discloses a method for multi-source data fusion and power prediction of wind turbine units, belonging to the field of power prediction technology. It addresses the technical problems of poor multi-source data fusion and power prediction analysis in existing solutions. By filtering and retaining key features through mutual information, redundant information can be effectively reduced. Dynamic weighting via an attention mechanism makes the fused features more adaptable to real-time operating conditions. Differential weight adjustments for wind speed, terrain, and equipment status ensure high relevance of the fused features even in complex scenarios. Spatial feature extraction captures local spatial correlations, temporal features capture temporal dependencies, and random forests handle nonlinear mappings, solving the problem of insufficient generalization ability of single models. The attention mechanism automatically assigns weights to different time steps and features, avoiding complex mathematical processes such as matrix operations and noise covariance estimation.

This invention discloses a multi-objective optimization method for high-pile piers and abutments based on an improved Grey Wolf algorithm. The method includes the following steps: establishing a design variable with pile inclination and torsion angle as the design variables, and simultaneously minimizing the standard deviation of pile driving force, maximum pull-out force, and maximum pile bending moment as the optimization objective; designing an integer index discrete encoding strategy to map discrete inclination and torsion angle parameters in the project to integer indices; using an improved multi-objective Grey Wolf optimization algorithm for solving the problem, efficiently searching for Pareto optimal solutions by introducing a nonlinear convergence factor, external archive maintenance, and a leader selection mechanism based on crowding distance; and finally outputting a series of pile placement schemes that achieve a balance between pile driving force uniformity, pull-out force, and bending moment control. This invention achieves collaborative optimization of multiple engineering objectives, with strong algorithm adaptability and significant optimization effects, providing scientific and efficient design decision support for high-pile piers and abutments and other pile foundation projects.

The application discloses a kind of encryption control information nonlinear mapping methods based on chip self-heating, it is related to chip circuit's encryption information nonlinear mapping technical field, comprising: chip is divided into several physical area blocks according to two-dimensional matrix form, each physical area block is correspondingly provided with a hotspot unit and a network node, each physical area block corresponds different data representation;Each network node is numbered in disorder;Each array of heat intensity control information obtained according to encryption control information is transmitted to each corresponding network node;For a certain network node, the heat intensity control information received is transmitted to the corresponding hotspot unit, and the hotspot unit generates heat;According to the thermal imaging of chip and the data representation corresponding to each physical area block of chip, the nonlinear mapping result of encryption control information is obtained.The application improves the effectiveness of information mapping, better protects the transmission process of control information, and improves the security of information mapping.

Hardware unit (10) for determining a non-linear ground surface, the hardware unit (10) comprising: at least one data input interface (1) via which the hardware unit (10) receives measurement points (M) from a 3D point-supplying sensor (20); a computing device (2) which maps the measurement points (M) into a two-dimensional grid (G) and determines a height difference for each grid cell (C) based on the measurement points (M) assigned to that grid cell (C), and classifies one of these grid cells (C) as invalid if the respective height difference exceeds a predetermined threshold; if the respective height difference of one of these grid cells (C) does not exceed the threshold and the respective height difference of at least two grid cells (C) directly adjacent to that grid cell (C) also does not exceed the threshold, classifies that grid cell (C) as a flat surface;wherein the computing device (2) performs a connected component analysis stepwise over the grid cells (C) classified as planar surfaces, separating planar surfaces belonging to obstacles from planar surfaces belonging to a ground surface and obtaining the nonlinear ground surface; a data output interface (3) that outputs the nonlinear ground surface.

The present invention relates to a method for measuring the observability of a cat qubit. The quantum system includes a command circuit and a nonlinear superconducting quantum circuit including a nonlinear element and a resonant portion. Each of the nonlinear superconducting quantum circuits has a resonant frequency f a and f b It has first and second modes. The method is a) by a command circuit, frequency f b The method includes (1) delivering microwave radiation to drive a second mode, thereby causing a nonlinear element to operate a jump operator [Equation 1] that stabilizes a two-dimensional manifold hosting a cat qubit (900), (2) mapping a measurement of the observable under measurement to a Pauli operator Z in a cat qubit manifold of a size value whose square coefficient is 2 or greater (910), and (3) measuring the value of the Pauli operator Z (920).

This invention provides a simulation method based on the functional attributes of forest and grassland communities, comprising: acquiring community survey data and individual plant functional trait data of the target forest and grassland vegetation; calculating the functional attributes of the plant community; acquiring the actual values ​​of key hydrological parameters corresponding to the target forest and grassland vegetation; training the model using a random forest algorithm to construct a nonlinear mapping model; integrating the nonlinear mapping model into a parameterized process and outputting the simulation results. The hydrological simulation method based on the functional attributes of forest and grassland communities provided by this invention establishes a nonlinear mapping relationship between these ecological attributes and key hydrological parameters using a random forest algorithm, thereby dynamically driving the simulation of regional hydrological processes. This achieves a refined characterization of the mechanism by which vegetation community structure and function affect hydrological processes, significantly improving the simulation accuracy of the model in scenarios such as vegetation restoration and configuration optimization.

This invention discloses a control method, device, and system for suppressing current ripple in a coreless brushed DC motor, belonging to the field of DC motor control. The method includes: using an H-bridge Buck converter as the driving topology to provide smooth voltage; designing a state feedback controller based on its state-space model to adjust the dynamic current response; simultaneously designing nonlinear disturbance observers to compensate for matched disturbances in the inductor branch and unmatched disturbances and their derivatives in the capacitor branch; innovatively designing a multi-channel quasi-resonant proportional-integral resonant observer to observe and compensate for the unmatched periodic disturbances caused by motor component commutation, with higher precision; and finally synthesizing a composite control law to drive the motor. The device and system correspondingly implement the above method. Compared with higher-order disturbance observers, the multi-channel quasi-resonant proportional-integral resonant observer proposed in this invention significantly improves the compensation capability for unmatched periodic disturbances, effectively suppressing current ripple in the coreless motor and improving torque stability.

A recursive RISE control method and system for maglev train levitation systems is disclosed. The control method includes: constructing the dynamic equations and voltage balance equations of the electromagnets based on a single-point levitation system, and then constructing a dynamic model of the single-point levitation system accordingly; designing a virtual control law, and then designing a recursive RISE controller based on the backstepping method by combining the dynamic model of the single-point levitation system and the virtual control law; performing stability analysis on the RISE controller to obtain its stability conditions; and, based on satisfying the stability conditions, using the RISE controller to perform real-time control of the levitation gap of the maglev train levitation system. This invention has excellent anti-interference capabilities, reducing the impact of external interference or model uncertainty on levitation accuracy. Furthermore, by handling the nonlinear relationship between voltage and current through the backstepping method, this invention avoids the model mismatch problem caused by linearization.

The application relates to a code detection method and device, computer equipment and a storage medium. The method comprises the following steps: obtaining a target code to be detected; performing string division on the target code to obtain a target substring sequence; obtaining a substring vector corresponding to each target substring in the target substring sequence to form a vector sequence; performing linear feature extraction on the vector sequence to obtain first extracted features, and performing nonlinear feature extraction on the vector sequence to obtain second extracted features; performing fusion processing on the first extracted features and the second extracted features to obtain fusion features; and performing code detection based on the fusion features to obtain a code detection result corresponding to the target code. A cloud server can use an artificial intelligence-based malicious code detection model to implement the code detection method of the application, thereby achieving the purpose of reducing network attacks. The method can improve the accuracy of malicious code detection.

The application provides a multi-intention interactive game decision method and system for automatic driving path planning, which fuses current environment information and current state information by acquiring the current state information and the current environment information, extracts state features and intention features of a moving target from current scene features, fuses the intention features of the state features of the moving target, constructs a game relationship graph, calculates attention weights between an automatic driving vehicle and the moving target, weights and aggregates a value vector by a prediction probability and the attention weights of different preset intention features executed by the automatic driving vehicle, performs feature splicing, nonlinear conversion and decoding on the aggregated features, evaluates each game candidate trajectory, and selects a game candidate trajectory with the maximum game value score as the automatic driving path; the application reduces uncertainty in the interactive process, avoids the risk of sudden braking or collision caused by misjudgment, and has stronger environmental adaptability and generalization ability.

The application discloses an intelligent management equipment adaptive control method based on RFID tag state feedback, which is applied to a logistics sorting line deployed in a metal environment interference, and comprises the following steps: controlling an RFID reader-writer to emit an excitation signal at a preset carrier frequency, and collecting multi-dimensional state parameters of a return signal, including a received signal strength indication (RSSI) value and a phase offset; constructing an environmental disturbance factor based on the phase offset, and fusing the normalized RSSI value, to generate a health degree comprehensive score of a tag through a nonlinear weighting algorithm; dynamically adjusting the transmission power of the RFID reader-writer and the tag query frequency according to the health degree comprehensive score; constructing an energy consumption-reliability cost function, solving the minimum value of the cost function under the constraint condition of meeting the minimum recognition confidence, and generating an optimal control instruction. The application solves the problems of poor control precision, high energy consumption and low reliability of the RFID intelligent management equipment in a metal environment, and improves the comprehensive operation efficiency of the logistics sorting line.

This invention discloses a method and system for magnet pose estimation based on magnetic dipole constraints. The method includes estimating the initial orientation and position of the magnet based on the triaxial magnetic field vectors measured by each magnetic sensor in a magnetic sensor array to determine the initial estimated orientation unit vector and initial estimated position of the magnet; iteratively solving the initial pose state parameters of the magnet using a nonlinear optimization algorithm based on a magnetic dipole model to obtain the final estimated orientation unit vector; obtaining the position residual and orientation residual using a pre-trained residual neural network model; and compensating the initial estimated position and the final estimated orientation unit vector based on the position residual and orientation residual to obtain the final magnet pose estimation result composed of the magnet orientation. This invention aims to overcome the influence of initial value deviation, sensor error, and environmental interference factors, improve robustness and generalization ability in complex environments, and meet the high-precision pose detection requirements of magnetic positioning.

This invention discloses a wide-temperature-range magnetic powder core loss calculation model, method, system, storage medium, and device, relating to the field of magnetic core loss technology. It embeds temperature as an independent variable into the magnetic powder core loss calculation model; considering the loss characteristics of magnetic powder cores over a wide temperature range, it establishes a piecewise function conductivity model based on the material properties of the magnetic powder core; it establishes a full-band dynamic calculation framework model covering hysteresis, eddy currents, and residual losses; and it calculates the wide-temperature-range magnetic powder core loss by substituting the collected data into the full-band dynamic calculation framework model. This invention utilizes a multi-stage nonlinear conductivity model based on the insulation failure threshold, overcoming the limitations of traditional linear or single-function models. By dividing conductivity into two dominant stages through the critical temperature Tc, this dual-stage switching mechanism realizes the abrupt change in conductivity response when the temperature crosses the critical value, accurately capturing the surge in eddy current paths caused by insulation failure, and providing a quantitative basis for high-temperature fault early warning.

The application provides a UAV cluster formation tracking method and system, computer equipment and medium, belongs to the UAV cooperative control technical field, and the method comprises the steps that firstly, a leader-follower second-order integral model is established, and position and speed tracking errors are calculated based on a preset time-varying formation; a fractional power nonlinear feedback term is introduced to construct a finite time control law, so that the errors converge to zero in a finite time; a dynamic event triggering mechanism is constructed at the same time, the control signal is updated and communication is triggered only when the current state and the measurement error at the last triggering time exceed the dynamic threshold, otherwise the old instruction is followed; in addition, theoretical analysis proves that there is a strict positive lower bound between adjacent triggering intervals, which fundamentally eliminates Zeno behavior. Under the premise of ensuring the finite time convergence performance, the application significantly reduces the communication frequency and on-board energy consumption of the UAV cluster, and is especially suitable for cluster cooperative operation scenarios with severely limited bandwidth and energy.

The application discloses an oil blending component oil proportioning optimization control method, system, device and medium based on soft measurement. The application predicts the quality attribute of the finished oil in real time through a soft measurement module, obtains an attribute prediction value, determines a target gain with the original proportioning data, and corrects the model gain of a multivariable controller online by using the target gain, generates adjusted target proportioning data by the multivariable controller, and issues a formula control instruction to a control device, so that the component oil proportioning is accurately adjusted. When the component oil formula or attribute changes, the application can automatically and timely adjust the controller gain, effectively overcome the nonlinearity of the blending process, improve the control precision and system robustness, thereby guaranteeing the stability of the finished oil quality and avoiding the waste of high-value component oil. By introducing the soft measurement technology, the method can replace the traditional online analyzer to a certain extent, reduce the dependence on expensive instruments, and improve the economy and maintenance convenience of the system.

This invention discloses a multivariate time series forecasting method based on prediction domain transformation and dual-path fusion. A time series forecasting model, PDT, is proposed. It utilizes data-adaptive prediction domain transformation to project the time series onto the energy-concentrated optimal latent space for prediction. A dual-path architecture is designed to simultaneously achieve efficient extrapolation of dominant linear trends and capture complex nonlinear dynamics. A masked channel dependency strategy and a lightweight linear encoder are used to adaptively fusion and filter inter-variable dependencies and deeply extract intra-channel features, respectively. The proposed time series forecasting model significantly solves the problem of insufficient nonlinear expressive power of traditional linear models and enhances noise resistance and computational efficiency in high-dimensional multivariate scenarios, comprehensively improving the prediction accuracy and robustness of the model in multivariate time series forecasting tasks.

A device, method, and hybrid computing architecture for a Symbolic Photonic Logic Gate (SPLG) system that performs data processing primarily in the optical domain. The system comprises a photonic substrate, including silicon-photonics, indium phosphide, or similar optically active materials, patterned with optical waveguides and interferometric structures configured to implement logic operations through controlled optical interference, phase modulation, or non-linear optical effects. The SPLG system is configured to execute Boolean and symbolic logic functions without requiring intermediate optical-to-electrical signal conversion. In operation, information is encoded into one or more optical parameters including phase, polarization, wavelength, or amplitude, and processed within a Symbolic Optical Core optimized for linear algebraic operations such as matrix multiplication and vector transformation. The architecture further comprises an electronic control layer configured to initialize, configure, and monitor the photonic logic elements while allowing the majority of computational operations to occur within the optical domain, thereby reducing electrical interconnect congestion and resistive heating relative to fully electronic processors.