74 results about "Function approximation" patented technology

Systems and methods for performing hardware approximation of functions are provided. In one example, a system comprises a controller, a plurality of multiplexors, configurable arithmetic circuits, and a mapping table that stores a set of function parameters. According to a mode of operations, the controller may configure the plurality of multiplexors to forward the set of function parameters or a subset of the function parameters to the arithmetic circuits to compute an approximation result. In a case where the subset of the function parameters is forwarded to the arithmetic circuits, the controller may configure the arithmetic circuits to perform post-processing, such as quantization, of the approximation result.

The invention discloses a self-adaptive cooperative control strategy based on a variable speed reaching law and a neural network, belongs to the field of intelligent control, and mainly aims at estimating and compensating uncertainty, external disturbance and actuator saturation existing in a queue in queue control and improving the stability, robustness and control performance of the whole queue. The method comprises the following steps: establishing a queue system model with uncertainty, external disturbance and actuator saturation; establishing a Gaussian error function approximation strategy; establishing a tracking error; establishing a variable speed reaching law; establishing a queue integral sliding mode error; establishing a queue coupling sliding mode error; establishing an adaptive parameter estimation mechanism for external disturbance and Gaussian function approximation errors in the queue; establishing a neural network estimation mechanism for uncertainty in the queue; and establishing a queue integral sliding mode control strategy based on a variable speed reaching law, a Gaussian error function approximation strategy, an adaptive parameter estimation mechanism and a neural network estimation mechanism. The method is used for queue cooperative intelligent control.

The invention discloses a mobile robot autonomous navigation planning method based on brain-like pulse deep reinforcement learning. Firstly, a brain-like pulse neural network (SNN) is constructed as a Q function approximator, and the energy efficiency of path planning is improved by using the time sequence processing capability and biological rationality of the SNN. Secondly, designing a double-priority experience playback mechanism, dividing experience samples into a common experience pool and an elite experience pool according to a reward threshold value, adjusting the sampling proportion of the two experience pools by adopting a dynamic self-adaptive sampling strategy, and intelligently balancing the sample utilization rate in the training process according to a loss function and a training stage; in addition, the reward function design is optimized through Manhattan distance reward and a repeated access punishment mechanism, and the robot is guided to efficiently explore the environment.

An example operation may include at least one of loading an Artificial Intelligence (AI) model from the storage, wherein the AI model is one of a diffusion-based model or a flow-based model, receiving tabular input data for execution by the AI model, wherein the tabular input data is scaled with a class-conditional scaler, creating a multi-output Gradient Boosted Tree (GBT), creating a Scalable AI (SAI) model from the AI model by using the multi-output GBT as a function approximator, and generating synthetic data by at least one of: executing the SAI model on the tabular input data or implementing a trained SAI model with the tabular input data, wherein the generating synthetic data reduces processing and memory resources.

The invention relates to a new energy power station configuration optimization method based on an agent model and multi-objective optimization. The method comprises the steps of obtaining historical power system configuration parameters and corresponding historical stability characteristic values, and performing preprocessing; inputting the preprocessed historical power system configuration parameters into a neural network agent model based on a multi-layer perceptron architecture, performing offline training in combination with corresponding historical stability characteristic values and physical prior constraints, and outputting predicted stability characteristic values; taking the prediction stability characteristic value as an optimization target, and constructing a target function of multi-target Bayesian optimization; and iteratively solving a Pareto optimal solution set based on the objective function, and determining an optimal new energy power station configuration scheme. According to the method, multi-objective optimization and Bayesian optimization are combined, a multi-objective acquisition function is adopted to approach the Pareto frontier between stability characteristic values, precision and multi-objective tradeoff can be considered, and a more flexible and scientific configuration scheme is provided for new energy station construction.

The invention discloses a friction torque prediction method based on deep learning, and belongs to the technical field of intelligent manufacturing and machining, and the method comprises the following steps: analyzing the kinematic characteristics of the overall structure of a numerical control machine tool machining center, and describing the dynamic response characteristics of a servo motor by combining the milling process of a numerical control machine tool and referring to a classical Stribeck friction model; the method comprises the following steps: respectively constructing system dynamic models related to X and Y feed axes, determining a dynamic relationship between excitation and response under the action of time-varying disturbance, solving the problems of nonlinear dynamic system modeling and parameter identification under a time-lag condition, and taking a servo system mechanism as a model architecture, introducing a deep neural network as a function approximation model, so as to realize the nonlinear dynamic system modeling and parameter identification. And the problem of quantitative characterization of friction parameters is effectively solved. Prediction conditions of different driving shaft friction torques are systematically evaluated from a time domain perspective, and results show that the method has better prediction precision and stability.

The invention belongs to the field of parameter self-adaptive control, and particularly discloses a laser welding parameter self-adaptive optimization method based on machine learning, and the method comprises the steps of hardware equipment configuration, weld bead width prediction and laser welding parameter optimization. According to the scheme, a deep neural network is established as a weld bead width prediction model, a Markov decision process is established, a discount factor is introduced, an SAC algorithm is adopted to maximize an entropy regularization reinforcement learning target, a neural Q network is utilized as a function approximator, Q function parameters are optimized by minimizing Bellman residual errors, and a weld bead width prediction result is obtained. And a loss function is defined through KL divergence to optimize welding strategy parameters, a weld joint square error minimization formula is rewritten into a space discount form, the weld bead length and the laser position serve as variables, an integral award function is defined in combination with the instantaneous welding speed, the influence of accumulated errors is reduced through a space discount mechanism, and space self-adaptive adjustment of the welding speed is achieved.

The embodiment of the invention provides a multi-modal large model-oriented security evaluation method, device and equipment and a storage medium. The method comprises the following steps that: a model owner performs quantization processing and nonlinear function approximation processing on a multi-modal large model to obtain an initialized multi-modal large model, and encrypts and sends the weight of the initialized multi-modal large model to different security agents; and the client performs hierarchical vectorization on the constructed mixed query data set to obtain an inference query data set, and encrypts and sends the inference query data set to different security agents. Different security agents carry out security collaborative reasoning on the weights of the reasoning query data set and the initialized multi-modal large model respectively to obtain reasoning results, and the reasoning results are sent to the client. And the client decrypts and reconstructs the reasoning results sent by the different security agents, obtains and verifies the reasoning prediction result, and judges whether the service provider is safe and available. In this way, threats of malicious service providers can be effectively avoided.

A secret share value [y]=[δx2+ax] is obtained through secure computation using a secret share value [x] of a real number x, and a secret share value [func(x)]=[y(ζy+b)+cx] of an elementary function approximation value z=func(x) of the real number x is obtained and output through secure computation using secret share values [x] and [y]. Here, x, y, and z are real numbers, a, b, c, δ, and ζ are real number coefficients, and a secret share value of · is [·].

The present disclosure relates to exponential and logarithmic function approximations for floating point numbers. The methods presented herein enable efficient approximation of exponential or logarithmic functions using values in half-precision floating point format. In at least one embodiment, an activation function of a neural network can perform tasks such as converting raw scores from a network into probabilities using an exponentiation of each output. A fixed point representation of each raw value can be generated that includes a bias factor and used to generate an intermediate representation having a determined number of bits. The intermediate representation can be incremented such that the mantissa corresponds to the fixed point representation. The decimal point can be shifted to obtain a value that is an approximation of the exponent of the raw value and can be used to determine a probability (or other such value) for a corresponding computational operation. At least some of these operations can be reversed to similarly obtain an approximation of a logarithmic function.

The invention discloses a security constraint unit commitment method based on tensor completion approximate dynamic programming, and the method comprises the following steps: firstly, building a security constraint unit commitment model based on a Markov decision process; secondly, decoupling the multi-period security constraint unit commitment model into a single-period sub; then, obtaining a value function and a decision function by adopting an approximate dynamic programming algorithm based on tensor completion; according to the approximate dynamic programming algorithm based on tensor completion, the approximate dynamic programming algorithm is improved from the angle of tensor completion, and all value functions and decision functions are approximately obtained by sampling a small number of state variables and decision variables. Decision is made from the angle of tensor complementation, in the face of a discrete state space, the value function of the whole state space can be approximated only by sampling a small number of state points, the calculation burden of an approximate dynamic programming algorithm in the aspect of value function approximation is effectively relieved, and the calculation efficiency of the algorithm can be greatly improved.

The present application provides a flow field prediction method, including: collecting sample data under different airfoils and different flow conditions; preprocessing the sample data; inputting the preprocessed data into a neural network for prediction to obtain a prediction result; iteratively refining the prediction result using a physical solver to obtain a flow field prediction result that meets the convergence constraint. The present application applies a neural network for prediction, and calls OpenFOAM as a physical solver for iterative refinement, coupling the neural network with the physical solver to obtain a new data-driven framework that can be used for flow field prediction, making full use of the function approximation ability of the neural network and the characteristics of the physical solver to ensure convergence constraints, accelerating the convergence process, and being able to adapt to different flow conditions and geometric shapes to achieve high-precision and high-efficiency predictions. The present application also provides a flow field prediction system, a computer-readable storage medium, and an electronic device with the above-mentioned beneficial effects.

The application provides a vibration isolation method, system and device of a control moment gyroscope system, which can be applied to the technical field of control. The method comprises the following steps: determining current kinematic parameters based on disturbance forces received by multiple components during operation; constructing a differential equation representing a dynamic relationship based on the current kinematic parameters and coupled dynamic characteristics; inputting a system state vector into a disturbance estimation network to output a disturbance estimation value; inputting the system state vector into the differential equation to obtain a disturbance theoretical value; inputting the disturbance estimation value and the disturbance theoretical value into a Hamilton function to obtain an initial control input quantity corresponding to a Hamilton function value satisfying a preset constraint condition; and updating the initial control input quantity in a gradient descent direction of a performance index function approximation value corresponding to the initial control input quantity until a difference between adjacent two performance index function approximation values is less than a preset threshold value, so as to obtain a target control input quantity.

The application discloses a kind of based on attention mechanism and reinforcement learning's ultra-short-term DNI prediction method, comprising: collecting solar thermal power station historical measured direct normal irradiance data and meteorological characteristic data;Periodic time characteristics and historical lag term are introduced to enhance time series dependence;Utilize bidirectional long short-term memory network to extract DNI time series features, construct attention mechanism based on dynamic change of cloud amount, strengthen the response capability of model to weather mutation;Adopt KAN to replace traditional full connection to high-dimensional nonlinear characteristics Adaptive function approximation is carried out;Introduce PPO reinforcement learning algorithm, dynamically optimize model feature representation and parameter, realize error self-feedback correction;Output DNI prediction result, for the dispatching and operation control of concentrating solar thermal power generation system.The method can effectively improve the DNI prediction accuracy and robustness under complex weather conditions, provide high-precision prediction support for photo-thermal power generation and renewable energy grid connection.

The invention belongs to the technical field of robot intelligent control, and particularly relates to a mechanical arm tracking control method based on transferable depth incremental reinforcement learning, which effectively solves the nonlinear optimal tracking control problem of a mechanical arm in a model-free mode by constructing a transferable incremental reinforcement learning framework. The method comprises the following steps: firstly, constructing a 1-degree-of-freedom depth increment model by utilizing one-step forward data offline learning, and providing universal dynamic representation for a mechanical arm which is difficult to accurately model; furthermore, an asynchronous depth value network is designed for the one-degree-of-freedom mechanical arm, stable and rapid convergence value function approximation is achieved through a separation base layer and an adaptive layer, and a cross-mechanical-arm migration mechanism is established, so that a pre-training model and a network base layer on the one-degree-of-freedom mechanical arm can be directly migrated to a high-degree-of-freedom mechanical arm subsystem; and the system difference is compensated only by updating the adaptive layer online, so that the repeated training overhead is remarkably reduced, and meanwhile, rapid, robust and adaptive tracking control on the mechanical arms with different degrees of freedom is realized.

The invention provides a power equipment simulation analysis system based on machine learning. The system comprises a data acquisition module, a data processing module, a power equipment simulation module and a fault analysis module. By fusing electro-thermal coupling modeling, orthogonal basis reduced order analysis and modularized ROM assembly, a rapid simulation model is constructed, and the modeling efficiency and response speed of power equipment under complex working conditions are remarkably improved; the simulation module introduces a GL-MLP-Trans model to realize intelligent prediction and stability regulation and control of a simulation scheme; the fault analysis module adopts a low-rank-Bi-LSTM model, combines a spectrum projection gain mechanism and a submodule function approximate projection strategy, and improves the fault identification precision while reducing the model complexity; the system is suitable for scenes of intelligent power grids, power operation and maintenance, industrial equipment state monitoring and the like, and has high efficiency, robustness and engineering adaptability.

The application relates to the technical field of computer vision and image processing, in particular to a light, fast and accurate self-supervised depth estimation method and system, which can solve the problem of recovering the depth of a scene from a two-dimensional image and needing a complex depth estimation network model to a certain extent. The method comprises the following steps: acquiring an initial image data set; constructing a weighted coefficient matrix calculation module and a sparse coding module; based on the weighted coefficient matrix calculation module and the sparse coding module, a small depth estimation network model is constructed to speed up the algorithm reasoning speed and reduce the parameter quantity of the model; a large depth estimation network model is constructed and trained by using luminosity loss as a supervision signal; the small depth estimation network model is trained by using luminosity loss and function approximation loss as supervision signals to optimize the depth estimation precision of the small depth estimation network model; finally, input data is input into the small depth estimation network model to obtain an estimation result.

The invention relates to an industrial process control method based on reinforcement learning of a second-order value gradient model. Aiming at the characteristics of nonlinearity, strong coupling, hysteresis and the like of industrial process control, the invention provides a control method based on reinforcement learning of a second-order value gradient model on the basis of a value perception model. Firstly, in the model training process of the method, second-order gradient information of a state value function is added, so that the method has more accurate function approximation capability and higher robustness, and the learning iteration efficiency is higher; and secondly, by adopting a new state sampling strategy, the model can be more efficiently utilized to carry out strategy learning. Finally, an experiment in a penicillin production fermentation industrial process scene shows that compared with a traditional maximum likelihood estimation model, an environment model prediction error based on a second-order value gradient model is remarkably reduced; the learning efficiency of the reinforcement learning method based on the second-order value gradient model is superior to that of an existing strategy optimization method based on the model, better control performance is achieved, and the oscillation phenomenon in the control process is reduced. The method has theoretical and practical significance for automatic control of the industrial process.

The invention provides an approximate dynamic programming control method based on support vector regression, relates to the technical field of approximate dynamic programming, and provides a method for solving the optimal control problem of a continuous nonlinear system. The method integrates a continuous time IRL Bellman equation and a support vector regression SVR technology; the use of IRL allows the algorithm to estimate a value function through an integral term under the condition that part of system dynamics is unknown, a powerful tool is provided for online learning, SVR is applied to ADP for the first time, a complex function approximation problem is converted into a convex optimization problem, and the existence of an optimal solution and the stability of the algorithm are ensured. By adopting the HESVR, the algorithm can more accurately approach the cost function and the control strategy, and meanwhile, the weight of the HESVR is adjusted by utilizing the least square method, so that the efficiency and the accuracy of the algorithm are improved.

A computer-implemented system and method is provided for determining a risk assessment. The method comprises receiving a plurality of vehicle behaviour data over a defined data collection period. This data is input into a supervised learning prediction model which is trained on historical vehicle behaviour data over a past time period, to generate a predicted value of a frequency of expected claim submissions for the policyholder of the vehicle in a future time period. Then a Shapley estimate is computed for each feature of the behaviour data applied to the model for determining a contribution of each said feature to the predicted value. A spline approximation is applied to the Shapley estimate for each said feature to estimate the contribution of each said feature. Then, a sum of the spline approximation for each said feature is calculated and a corresponding risk score determined based on the sum.

This application belongs to the field of delay differential equation modeling, specifically disclosing a delay differential equation modeling method based on Bayesian optimization and neural networks. The method includes: generating multiple trajectory data of the system; searching for the optimal delay term using a Bayesian optimization algorithm, where the delay term to be solved is the optimization variable, and the error based on neural network simulation is the objective function, obtaining the optimal delay term through iterative optimization using a surrogate model; constructing a state matrix and a delay matrix from the trajectory data based on the optimal delay term, using the concatenated matrix as input to a neural network to construct a neural network for approximating a nonlinear function; integrating a linear multi-step method into the loss function of the neural network, training the neural network using the trajectory data to obtain a nonlinear function approximation model; and separating the numerical discretization error and the neural network approximation error based on this model to construct a total error bound. This application can achieve high-precision, high-efficiency delay differential equation modeling with quantization error guarantee.

The invention discloses a nonlinear causal discovery method (KAN-NOTEARS) based on a Kolmogorov-Amold network, and belongs to the technical field of artificial intelligence and causal inference. The method comprises the following steps: acquiring an observation data set containing a plurality of variables; a causal model with the learnable weighted adjacency matrix and a set of KAN functions as parameters is constructed, the KAN functions are based on the Kolmogorov-Arnold representation theorem, and a causal mechanism between variables is modeled through a learnable one-dimensional nonlinear function; constructing a target function containing structural equation model fitting loss and a sparse regular term; converting a directed acyclic graph constraint into a continuous differentiable algebraic constraint; solving the constraint optimization problem by adopting an augmented Lagrangian method to obtain an optimal adjacent matrix and function parameters; and performing threshold processing on the adjacent matrix to obtain a final cause and effect graph structure. According to the method, the theoretical advantages of the KAN network in the aspect of function approximation are utilized, the problems of insufficient accuracy and low calculation efficiency when a traditional method is used for processing a complex nonlinear causal relationship are solved, and the causal discovery precision and efficiency are remarkably improved.

A computer-implemented system and method is provided for determining a risk assessment. The method comprises receiving a plurality of vehicle behaviour data over a defined data collection period. This data is input into a supervised learning prediction model which is trained on historical vehicle behaviour data over a past time period, to generate a predicted value of a frequency of expected claim submissions for the policyholder of the vehicle in a future time period. Then a Shapley estimate is computed for each feature of the behaviour data applied to the model for determining a contribution of each said feature to the predicted value. A spline approximation is applied to the Shapley estimate for each said feature to estimate the contribution of each said feature. Then, a sum of the spline approximation for each said feature is calculated and a corresponding risk score determined based on the sum.

The invention discloses a DNN-based adaptive optimization control method for a fractional order single-machine infinite bus system, and relates to the field of single-machine infinite bus system control. The method comprises the following steps: establishing a dynamical model of a single-machine infinite bus system, and converting the dynamical model of the system into a state model; designing a function approximate DNN architecture, approaching an unknown function in the system, and designing a weight updating law based on a first-order Taylor series to reduce the mathematical difficulty; in the backstepping process, a virtual controller and an actual controller are constructed by utilizing an optimization backstepping technology, and the overall control optimization of the system is realized; an event trigger function is designed, and consumption of system communication resources is reduced; and carrying out Lyapunov analysis to ensure that each signal of the system is bounded. The adaptive optimization backstepping controller based on the DNN architecture is designed for a fractional order single-machine infinite bus system, unknown nonlinear terms in the system can be compensated, and a single-machine power angle tracks a given reference signal.

Methods and devices of a function approximation unit configured to approximate a nonlinear function within a neural processing unit are described. According to one embodiment, the method includes storing an input value through an input register of the function approximation unit, transmitting the input value to a selected one of a plurality of preprocessing circuits of the unit according to a control signal, generating a preprocessing result corresponding to the input value by the selected one of the preprocessing circuits, transmitting the preprocessing result to a programmable function approximation circuit and a selected one of a plurality of post-processing circuits of the unit, generating an approximated function output based on the preprocessing result in the programmable function approximation circuit, and generating a final output value by post-processing the preprocessing result or the approximated function output in the selected one of the post-processing circuits.

This invention relates to the field of voice coil motor servo control technology, specifically to a fractional-order high-order sliding mode control method for voice coil motors based on a third-order superspiral sliding mode observer (STSMO) and an adaptive reaching law. The method includes the following steps: S1, establishing a mathematical model of the voice coil motor, deriving the dynamic equations of the voice coil motor including parameter uncertainties and external disturbances, defining the position tracking error, and deriving the error dynamic equation; S2, designing a fractional-order high-order integral sliding surface, combining it with an improved Oustaloup filter to achieve rational function approximation of the fractional-order calculus operator, and constructing the first-order derivative equation of the sliding surface; S3, designing an adaptive hyperbolic tangent reaching law that integrates dynamic feedback of the error amplitude, and deriving the initial expression of the sliding mode control law based on the first-order derivative equation of the sliding surface and the adaptive hyperbolic tangent reaching law. This invention achieves real-time accurate observation and feedforward compensation of lumped disturbances through a third-order superspiral sliding mode observer (STSMO).

The invention discloses an adaptive function approximation method for financial curve prediction. The method comprises a financial data feature analysis step, an adaptive network structure design step and a network training optimization step. The method comprises the following steps: performing feature analysis on input financial data, respectively calculating feature indexes such as oscillation degree, smoothness and Lipschitz constant of a corresponding function, and calculating a comprehensive complexity score based on the feature indexes; and adaptively designing a network structure according to the comprehensive complexity score, determining key parameters such as the number of network layers, the width of each layer, the dropout rate and the like, and carrying out key optimization on a high-error region by dynamically adjusting the parameters in a network training process. The network model output by the method realizes a function corresponding to high-precision approximation financial data, so that various financial curves are accurately predicted. According to the method, the calculation efficiency can be improved while the approximation precision is ensured, the limitation of designing a network structure depending on artificial experience in a traditional method is avoided, and the method is used in the fields of financial data analysis and the like.

The invention relates to the field of computer data processing, and discloses a CUDA (compute unified device architecture)-based fast exponential function approximate calculation method, which is used for improving the calculation efficiency while ensuring the precision. The CUDA-based fast exponential function approximate calculation method comprises the following steps: acquiring a to-be-calculated input floating-point number set; for each input floating-point number x, executing floating-point multiplication and addition fusion operation, simultaneously completing multiplication operation of the input floating-point number x and a preset coefficient a, type conversion from a floating point to an integer and addition operation of adding a preset offset b through a single floating-point multiplication and addition instruction, and generating a corresponding intermediate calculation result; performing numerical value conversion operation from a floating point to an integer on each intermediate calculation result to obtain a corresponding approximate exponential function value; and performing element-by-element difference calculation on each approximate exponential function value and the corresponding standard exponential function exp (x), and judging whether a relative error between each approximate exponential function value and the corresponding standard exponential function value is within a preset tolerance range or not.

This invention provides a machine learning-based power equipment simulation and analysis system, comprising a data acquisition module, a data processing module, a power equipment simulation module, and a fault analysis module. By integrating electro-thermal coupling modeling, orthogonal basis order reduction analysis, and modular ROM assembly, a rapid simulation model is constructed, significantly improving the modeling efficiency and response speed of power equipment under complex operating conditions. The simulation module introduces a GL-MLP-Trans model to achieve intelligent prediction and stability control of the simulation scheme. The fault analysis module employs a low-rank-Bi-LSTM model, combined with a spectral projection gain mechanism and a sub-modulus function approximation projection strategy, reducing model complexity while improving fault identification accuracy. This system is applicable to smart grids, power operation and maintenance, and industrial equipment condition monitoring scenarios, possessing high efficiency, robustness, and engineering adaptability.

This application proposes a method and device for obstacle avoidance control of an unmanned vehicle in a limited time domain, including: describing the safety of the control system based on forward invariance, and constructing an obstacle function of the control system based on the safety of the control system and the obstacle; integrating the obstacle function into the cost function of the control system, and converting the collision avoidance control problem of the control system into an optimal control problem based on the cost function; establishing a control system model with unknown parameters based on a neural network function approximation method, and using an integral parallel learning method to design a parameter adaptive law to identify the unknown parameters in the model; based on the control system model, using a quadratic optimization framework, and using an adaptive iterative learning method to solve the optimal control input that meets the requirements of the collision avoidance control task. This application adopting the above scheme does not rely on the precise parameter model of the system, takes into account both the optimization performance and safety of the control system, and can effectively solve the collision avoidance control problem in a limited time domain.