28 results about "Discretization error" patented technology

The invention relates to an automatic guided vehicle (AGV) adaptive path tracking method based on MPC. The method comprises the following steps: S1, establishing a kinematic discretization error model based on the kinematic characteristics of the two-wheel differential AGV, processing a continuous kinematic equation by adopting an Euler discretization method, expressing a dynamic change relationship between a transverse deviation distance and an angle deviation in a state-space equation form, and generating a kinematic discrete state-space model; the method has the advantages that the continuous equation is processed by establishing the kinematics discretization error model and adopting the Euler discretization method, the deviation dynamic relation is expressed through the state space, the calculation process is simplified, the precision is kept, sensor data are fused, the wheel type odometer, IMU and laser data are integrated through the extended Kalman filtering algorithm, and the precision is improved. The real-time position and angle deviation are calculated, the positioning accuracy is improved, a model prediction controller objective function is designed, a weight matrix and boundary constraint are introduced according to constraint conditions, and the effect of obstacle avoidance constraint is combined.

Provided are computer-implemented methods and quantum computing systems for preparing computational states representing quantum states of a physical system, including performing a computational evolution of the state and then determining physical properties of the system using the time-evolved computational state. Hamiltonian dynamics of observables are computed on a quantum computer to provide information about the physical system represented by the Hamiltonian. Low energy electronic structure states of a physical system are prepared using adiabatic evolution. The energy of an equilibrium quantum state of a physical system is determined using a time-evolution operator. The energy of an eigenstate of a physical system is indirectly determined by evaluating expectation values of a time-evolution operator averaged across multiple shots for randomly-generated quantum circuits. Example methods enable calculation of the energy of an evolved computational state with chemical accuracy, due to avoiding discretization errors, with a smaller circuit depth than known alternatives.

A system may be configured to train an artificial intelligence model (AI model). For example, the system, using a training framework with both a teacher layer and a student layer provides target outputs from the teacher layer, while the student layer learns to approximate these outputs. The system may obtain input data and, using the training framework, train the AI model using the input data. For example, the training framework defines a quantizer with clipping and rounding functions to convert full-precision data to quantized data. The framework calculates a first output distribution from the teacher layer and a second output distribution from the student layer. A training loss is computed between these distributions, and parameters of the quantized network are adjusted to minimize discretization error and prediction discrepancy. The trained quantized network is then used to generate new predictive outputs from the AI model.

The invention provides a hydrate-containing sediment depressurization exploitation numerical simulation method, medium, equipment and product, and relates to the technical field of natural gas hydrate exploitation numerical simulation, the method comprises the following steps: gridding a hydrate-containing sediment depressurization exploitation area based on a voronoi grid, constructing a CVT grid, and carrying out boundary treatment; establishing a control equation under the natural gas hydrate depressurization exploitation condition; establishing the relationship between the density of each phase and the temperature and pressure; and substituting the relationship between the density of each phase and the temperature and the pressure into a control equation, establishing a numerical model by using a finite volume theory according to the divided CVT grid, carrying out discretization, and solving the control equation. According to the method, discrete errors caused by traditional grid centroid deviation are effectively improved, and the calculation efficiency and the solving stability are remarkably improved.

The application discloses a physical neural network-based electrical property tomography method, which spatially obtains permittivity and conductivity of all points through four steps.The application does not need to simplify the core equation of MR-EPT to solve, and does not need to use a numerical method to solve by discretizing the derivative of the dielectric property in space, thereby avoiding the discretization error, so that the application has higher precision in obtaining the permittivity and the conductivity.The physical neural network-based electrical property tomography method uses a neural network to construct a mapping relationship between a transmit field and its spatial derivative and electrical properties, so that the precision of the obtained result is higher.

The present application relates to the technical field of double three-phase permanent magnet synchronous motor harmonic current control, and discloses a double three-phase permanent magnet synchronous motor harmonic current control method and device, the present application is through to the motor voltage equation implementation modeling domain conversion and discretization processing, constructs the harmonic current loop controlled object of adaptation digitization control, complex vector current regulator and low pass filter discrete domain model, combines current signal multi-dimension pretreatment and accurate regulation transformation, fully compensates the delay of digital control, effectively offsets the influence of harmonic coupling and speed correlation. Compared with the prior art, the present application avoids the adaptive deviation of continuous domain design and digital control, solves the system instability problem caused by low pass filter delay and discretization error, and ensures that the motor has high harmonic current control precision and fast dynamic response in the full speed range.

The application discloses a high-speed permanent magnet synchronous motor position identification method, samples motor a, b, c three-phase currents, carries out Clarke transformation on the obtained currents to obtain the currents of alpha axis and beta axis on a two-phase static coordinate system, estimates the output voltage of alpha axis and beta axis according to the SVPWM conduction time calculated in the last control period, obtains a position estimation reference quantity by directly processing the current voltage, and uses a phase-locked loop to calculate the angle and speed of the position estimation reference quantity and thereby calculates motor position information, the application uses the output characteristics of a voltage source inverter to obtain a calculation formula not influenced by discretization errors, and improves the position identification precision under a low load frequency ratio.

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 satellite navigation receiver tracking loop discretization error compensation method and system. The method comprises the following steps: acquiring a ranging code phase error measurement value, a carrier frequency error measurement value and a carrier phase error measurement value; for I and Q channels, ideal responses under a carrier frequency error measurement value and a carrier phase error measurement value are calculated, errors caused by sampling and frequency selective filtering are calculated, and I channel compensation amount and Q channel compensation amount are calculated in a combined mode; and accumulating the I-channel compensation quantity to the output of the I-channel integral zero clearing device, and accumulating the Q-channel compensation quantity to the output of the Q-channel integral zero clearing device. The invention aims to compensate the discretization error of the calculation result of the integral zero clearing device and improve the precision of the tracking loop of the satellite navigation receiver.

The invention relates to the technical field of ultrasonic tomography, and discloses a three-dimensional ultrasonic tomography sound velocity imaging method based on a multi-template fast marching method, which comprises the following steps: acquiring parameters of a three-dimensional ultrasonic tomography system and initial sound velocity distribution of biological tissues, and establishing a curved ray model based on an eikonal equation; discretizing the propagation time gradient in the eikonal equation by adopting second-order finite difference approximation, and covering 18 neighborhoods of the grid by adopting templates in four directions to carry out combined solution, so as to obtain ultrasonic propagation time distribution of the whole grid; and iteratively reconstructing a three-dimensional sound velocity distribution image through a maximum posterior probability estimation algorithm based on the ultrasonic propagation time distribution. According to the method, the propagation time gradient in the second-order finite difference approximation eikonal equation is adopted to reduce discretization errors, a plurality of direction templates are used, more accurate difference approximation is carried out on gradient operators, and the non-orthogonal propagation path of the ultrasonic waves in the three-dimensional space is completely captured.

The application discloses a smooth finite element-Newmark calculation method for underwater target transient acoustic scattering, and belongs to the technical field of underwater transient acoustic scattering numerical simulation. First, a calculation domain of an underwater transient acoustic scattering field is discretized into a standard quadrilateral element grid, and an acoustic pressure field interpolation format is constructed. The quadrilateral element is divided into multiple smooth domains, a smooth acoustic pressure gradient field is obtained through acoustic generalized gradient smoothing technology, and a smooth acoustic pressure gradient matrix is obtained through Gaussian integration. A system matrix equation is constructed in combination with a transient acoustic scattering control equation and a Galerkin weighted residual value method. A parameter-optimized Newmark time integration method is used to complete time domain discretization and recursive solving, and numerical results of the transient acoustic scattering field are obtained. The application softens the system stiffness matrix through gradient smoothing technology, reduces spatial discretization error, eliminates time domain false numerical damping, and has higher precision than the standard finite element method under the same grid, thereby providing an efficient and reliable scheme for underwater transient acoustic scattering engineering calculation.

This invention discloses a smoothed finite element adjustable damping implicit integral algorithm for underwater transient acoustic scattering, belonging to the field of underwater transient acoustic scattering numerical simulation technology. First, the computational domain of the underwater transient acoustic scattering field is discretized into a standard quadrilateral element mesh to construct an interpolation scheme for the sound pressure field. Then, each quadrilateral element is divided into multiple smooth domains, and the smoothed sound pressure gradient matrix is ​​obtained through acoustic generalized gradient smoothing technology. Combining differential control equations, the virtual displacement principle, and the Galerkin weighted residual method, the system matrix equation is constructed. An adjustable damping implicit time integral algorithm is used to complete the time-domain discretization, and the numerical results of the transient acoustic scattering field in the entire time domain are obtained through recursive solution. This invention can soften the "overly stiff" stiffness matrix to reduce spatial discretization errors, suppress high-frequency errors through adjustable numerical damping matching, and achieve coordinated control of spatiotemporal errors. Under the same mesh, its accuracy is superior to the standard finite element method, providing an efficient and reliable solution for underwater transient acoustic scattering engineering simulation.

The invention discloses a ball screw contact stress calculation method, and solves the problems of large assembly and discrete errors, disordered loading rules, low solving efficiency and the like in the existing entity steel ball simulation. According to the method, a rolling body is not needed, an assembly is built only through modeling of a lead screw and a nut, a spring is adopted to replace a steel ball, a finite element model is built in cooperation with Python batch processing, contact points are accurately controlled through functions, spring normal contact force and coordinates before and after deformation are exported in batches, and after actual contact angles are converted, the contact stress is calculated in combination with the Hertz contact theory. According to the method, initial assembly and discrete errors are avoided, the number of contact pairs is reduced, the solving speed and stability are improved, repeated model debugging is not needed, the research and development cost is greatly reduced, the product iteration period is shortened, and the method is suitable for ball screw performance analysis in the automobile field of EMB braking systems and the like.

The invention relates to the technical field of video security, solves the problem that the statistical security of a key stream is affected by periodic degradation of a pseudo-random sequence due to discretization errors in the prior art, and particularly relates to a monitoring video encryption transmission method, which comprises the following steps of: obtaining a monitoring video stream to be transmitted and performing framing processing to obtain an original video frame; the method comprises the following steps: performing block frequency domain transformation processing on an original video frame, generating a frequency domain coefficient matrix containing a plurality of transformation coefficient blocks, extracting an image feature value of the original video frame, mapping the image feature value into an initial state parameter, and initializing a preset high-dimensional hyper-chaotic system model by using the initial state parameter. According to the method, a one-time pad dynamic encryption mechanism is realized by establishing the nonlinear coupling relationship between the video content and the encryption key, so that the encryption process has extremely high sensitivity, the static periodicity of the key stream is fundamentally eliminated, plaintext attacks can be effectively resisted, the anti-attack capability is enhanced, and the security of the key stream is improved.

The invention provides a phase-shifted full-bridge model predictive current control method based on predictive error compensation, relates to the technical field of power electronics, and effectively compensates errors in a phase-shifted full-bridge converter system by constructing a predictive error recursive relation from a moment k to a moment k + 1 and utilizing a correction coefficient of a threshold judgment mechanism. On the premise that calculation complexity is not remarkably increased, current prediction deviation caused by factors such as inductance parameter change, sampling errors, discretization errors and inaccurate duty ratio loss compensation can be effectively restrained, and robustness and control precision of a system under parameter disturbance and load change are improved.

This invention relates to the field of precision optical inspection, and particularly to a method for selecting and evaluating the performance of an LC-SLM based on a multi-parameter coupled constraint model. The method includes determining the input variables of the inspection system based on the LC-SLM, which are divided into two main categories: parameters of the surface to be measured and basic requirements of the inspection system; determining the order of the blazed grating based on its diffraction efficiency requirements; determining the bit depth of the LC-SLM based on the accuracy requirements of the inspection system; determining the pixel parameters based on the wavefront modulation requirements of the LC-SLM; establishing a comprehensive error evaluation model for the inspection system; and calculating the total residual error, which is the square root of the sum of quantization error, pixel discretization error, LC-SLM surface shape error, nonlinearity error, and thermally induced phase drift error. This invention enables a systematic and quantitative design and evaluation of the hardware parameter selection, software configuration, and final compensation performance of the LC-SLM.

The application discloses a smooth finite element-Bathe calculation method for underwater target transient acoustic scattering, and belongs to the technical field of underwater transient acoustic scattering numerical simulation. Firstly, a sound scattering field calculation domain is discretized into a quadrilateral element grid, and an acoustic pressure field interpolation format is established; then, an element smooth domain is divided, a smooth acoustic pressure gradient matrix is obtained through acoustic generalized gradient smoothing technology, and a system matrix equation is constructed in combination with a transient acoustic scattering control equation and a Galerkin weighted residual value method; time domain discretization is completed through the Bathe time integration method, and iterative solution is carried out, so that full-time domain numerical results of the transient acoustic scattering field are obtained. Through the gradient smoothing technology, the spatial discretization error is reduced, error cooperative suppression is realized in combination with the frequency self-adaptive numerical damping of the Bathe integration method, the precision is higher and the stability is better than those of the standard finite element method under the same grid, and an efficient and reliable scheme is provided for underwater transient acoustic scattering engineering solution.

The application discloses a kind of based on array structure's magnetic field coil hybrid optimization design method, the multiple saddle coils are combined into coil array by the method, and on the basis of considering the coupling effect of high magnetic permeability material to coil magnetic field, the mathematical model of the magnetic field of single coil is solved, and the mathematical model of the magnetic field of saddle coil array is obtained according to superposition principle.The mathematical model is used to combine particle swarm optimization algorithm to optimize the best current integer ratio between the saddle coils in array, the current integer ratio of coil is converted into the linear combination of the number of turns and current of each coil, and particle swarm optimization algorithm is used again to optimize the minimum number of turns and optimal current required by each coil.Finally, the optimized magnetic field coil is generated by the coil path planning of depth-first search strategy.The application greatly improves the uniformity of coil, has no discretization error and simple structure, and is convenient for practical application.

The invention belongs to the field of delay differential equation modeling, and particularly discloses a delay differential equation modeling method based on Bayesian optimization and a neural network, and the method comprises the steps: generating multiple pieces of trajectory data of a system; an optimal delay term is searched by adopting a Bayesian optimization algorithm, the Bayesian optimization takes a delay term to be solved as an optimization variable and an error based on neural network simulation as a target function, and the optimal delay term is obtained through iterative optimization of an agent model; constructing a state matrix and a delay matrix from the trajectory data based on the optimal delay term, taking the spliced matrix as neural network input, and constructing a neural network for approaching a nonlinear function; fusing a linear multi-step method into a loss function of the neural network, and training the neural network by using the trajectory data to obtain a nonlinear function approximation model; based on the model, a numerical discretization error and a neural network approximation error are separated, and a total error bound is constructed. According to the method, high-precision and high-efficiency delay differential equation modeling with quantization error guarantee can be realized.

The invention discloses a sensor data discrete fusion modeling method, a fusion method, equipment and a storage medium, and the modeling method comprises the steps: constructing a sensor data continuous fusion model based on a complementary filtering algorithm; converting the sensor data continuous fusion model into a continuous state space equation; converting the system matrix of the continuous state space equation into a transfer function; converting the transfer function to a time domain to obtain a system matrix time domain expression, and further calculating an input matrix time domain expression; and obtaining a discrete state space equation according to the system matrix time domain expression, the input matrix time domain expression and the continuous state space equation, namely obtaining a sensor data discrete fusion model. According to the method, the dynamic performance and the precision of the original continuous system are kept to the maximum theoretically, and additional discretization errors caused by damage of module separation to the integrity of the original continuous system design in traditional complementary filtering discretization are avoided.

The invention belongs to the technical field of signal processing, and particularly relates to an underwater target detection method based on multi-source data fusion, a program, equipment and a storage medium. According to the method, confidence is quantified through a fuzzy set and a membership function, and compared with a hard decision, ground surface channel information is finer through soft decision fusion, so that a 0-1 discretization error of the hard decision is avoided, and the target judgment accuracy is improved; a membership function with adjustable parameters is introduced, and a mapping relation between a non-confidence interval and confidence can be dynamically adjusted according to the characteristics of the sensor to adapt to different noise environments and signal ranges; a uniform quantizer is adopted to replace an optimal quantizer, a complex optimization process is avoided, meanwhile, the calculation complexity is reduced through a log-likelihood ratio fusion rule, and the real-time processing requirement is met; and a confidence interval division mechanism is adopted, so that the sensor outputs a hard decision at high confidence, a soft decision is output in a middle region, and the noise robustness and the information utilization rate are balanced.

Provided are computer-implemented methods and quantum computing systems for preparing computational states representing quantum states of a physical system, including performing a computational evolution of the state and then determining physical properties of the system using the time-evolved computational state. A described computer-implemented method enables computing Hamiltonian dynamics of observables on a quantum computer to provide information about the physical system represented by the Hamiltonian. A described computer-implemented method prepares low energy electronic structure states of a physical system using adiabatic evolution. The method determines the energy of an equilibrium quantum state of a physical system using a time evolution operator that enables adiabatic evolution of a prepared electronic structure state on a quantum computing system. A described method involves indirectly determining the energy of an eigenstate of a physical system by evaluating the expectation values of a time-evolution operator averaged across multiple shots for randomly-generated quantum circuits. Example methods enable calculation of the energy of an evolved computational state with chemical accuracy, due to avoiding discretization errors, with a smaller circuit depth than known alternatives.

Provided are computer-implemented methods and quantum computing systems for preparing computational states representing quantum states of a physical system, including performing a computational evolution of the state and then determining physical properties of the system using the time-evolved computational state. A described computer-implemented method enables computing Hamiltonian dynamics of observables on a quantum computer to provide information about the physical system represented by the Hamiltonian. A described computer-implemented method prepares low energy electronic structure states of a physical system using adiabatic evolution. The method determines the energy of an equilibrium quantum state of a physical system using a time evolution operator that enables adiabatic evolution of a prepared electronic structure state on a quantum computing system. A described method involves indirectly determining the energy of an eigenstate of a physical system by evaluating the expectation values of a time-evolution operator averaged across multiple shots for randomly-generated quantum circuits. Example methods enable calculation of the energy of an evolved computational state with chemical accuracy, due to avoiding discretization errors, with a smaller circuit depth than known alternatives.

The invention discloses a predictive control method based on an SVG and H-bridge integrated model, belongs to the technical field of cascaded H-bridge SVG control, and is used for solving the problems of slow response and easy instability of traditional linear control under the load of an electric arc furnace. The method comprises the following steps: firstly, sampling an SVG output voltage, a power grid voltage, a compensation current and a reference current, converting a coordinate system, and establishing a discretization mathematical model of the cascaded H-bridge SVG under an alpha-beta coordinate system; considering the influence of H-bridge delay, and combining an SVG transfer function to construct an integrated current prediction model; equally dividing a prediction period to obtain a current prediction model considering a discretization error; and finally, pre-estimating a current prediction value by adopting a pre-estimation correction method, generating an alternating current side voltage reference value through error correction, and modulating the alternating current side voltage reference value into a PWM signal to act on a switching tube. According to the method, through integrated modeling and error compensation, the dynamic response speed and the control stability are improved, and reactive power is effectively compensated.

The application discloses a navigation control method and related device of a tracked vehicle, and relates to the field of automatic driving control. A discretization error tracking model is used to build a linear relationship between a discretized form of a pose tracking error vector and a tracked speed error vector, so that the control variable and the state variable are in a linear relationship. Solving the linear problem can obtain the left tracked speed error and the right tracked speed error. The expected value of the left tracked speed is calculated based on the left tracked speed error and the reference value of the left tracked speed, and the expected value of the right tracked speed is calculated based on the right tracked speed error and the reference value of the right tracked speed. The reference value of the bilateral tracked speed is calculated according to the left tracked speed error, the right tracked speed error and the steering curvature of the target reference point on the preset reference path, so that the expected value of the bilateral tracked speed can make the actual driving path of the tracked vehicle approach the reference path. In summary, the scheme improves the adaptability and control accuracy of the navigation control method for the tracked vehicle.

This invention discloses a long-term anti-drift lidar-inertial odometry method and system. Addressing the long-term drift problem caused by IMU bias model mismatch and low-order discretization errors in existing lidar-inertial odometry systems, this invention constructs a time-varying error model based on a first-order Gaussian-Markov process, modeling the gyroscope and accelerometer bias as time-varying stochastic processes and augmenting them into state vectors; simultaneously, it constructs a discretized continuous motion model based on a second-order prediction-correction method using Lie groups, constraining the system state to an SO(3) manifold; embedding these two improvements into an iterative error state Kalman filter framework, utilizing lidar point cloud features to construct observation residuals, and achieving tight coupling fusion through iterative optimization and backpropagation. This invention effectively suppresses pose drift during long-term operation, improves self-sustaining positioning capabilities in high-dynamic and degraded scenarios, and does not affect real-time performance, making it suitable for robot autonomous navigation and positioning.