38 results about "Velocity estimation" patented technology

A finite time dynamic mechanical arm control method based on a speed reconstruction strategy comprises the steps that firstly, a mechanical arm system physical model containing joint torque is constructed, joint tracking errors are defined, and a mechanical arm system state space model is obtained; 2, constructing a finite-time high-bandwidth gain speed estimator, defining a state estimation error, and obtaining an estimation error dynamic equation; 3, defining an auxiliary state to obtain an auxiliary state dynamic equation; 4, constructing a finite time non-recursive adaptive controller; and 5, obtaining a final input torque of the controlled system, and outputting the final input torque as a joint torque. According to the method, through the processes of error modeling, state estimation, auxiliary conversion, control law design and closed-loop feedback, the dynamic tracking problem of the mechanical arm is converted into the finite time convergence problem of the error state, meanwhile, dynamic gain and robust compensation are introduced, the adaptability of a system to parameter uncertainty and external disturbance is improved, and the system reliability is improved. And a high-precision and high-robustness control target is realized.

The invention provides an IMU (Inertial Measurement Unit) speed estimation method and system based on multi-scale time sequence feature coding and Mama. The method belongs to the field of IMU speed estimation, and comprises the following steps: S1, acquiring original time sequence data which is generated by an IMU and contains three-axis acceleration and three-axis angular velocity, and intercepting a section of historical time sequence data with a fixed length by taking a current moment as an end point to construct a historical window; s2, inputting the original time sequence data in the historical window into a multi-scale one-dimensional convolution encoder, processing the input data by the encoder by using at least two convolution kernels with different sizes in parallel so as to extract time sequence characteristics under different receptive fields, and fusing output characteristics of each branch to obtain a potential characteristic sequence; and S3, inputting the potential feature sequence into a Mama model to carry out time sequence dependency relationship modeling, and extracting an output state of the Mama model at the end of the sequence as a solved three-dimensional speed output. According to the method, the IMU speed estimation precision can be improved.

The invention discloses a multi-target optimization tracking control method for a heterogeneous unmanned cluster system based on MADDPG, and the method comprises the following steps: building a dynamic model of the heterogeneous unmanned cluster system, and carrying out the isomorphism of the heterogeneous unmanned cluster system through speed estimation; establishing a communication topology model of the heterogeneous unmanned cluster system, designing an adjacent matrix weight, and quantifying the information interaction strength between the heterogeneous unmanned cluster systems; describing a plurality of optimization objectives of the heterogeneous unmanned cluster system and executor saturation constraint conditions of the optimization objectives; modeling a multi-target optimization tracking control problem of the heterogeneous unmanned cluster system into a reinforcement learning problem, and designing MADDPG-based tracking control algorithm elements; and a learnable weight parameter is established, a target weight coefficient is updated online by using gradient back propagation, and dynamic optimization of multi-target weight is realized. Through an adaptive target weight adjustment mechanism and a saturation constraint and heterogeneous unified modeling framework, the tracking control problem under the influence of multi-target conflict, actuator saturation constraint and heterogeneous characteristics is solved, and the cooperative tracking control effect and robustness of a heterogeneous unmanned cluster system in a dynamic environment are improved.

A method for adapting a causal FIR filter is provided. The causal FIR filter is configured to estimate velocities by filtering position measurement data in accordance with a frequency response of an electro-mechanical system involved in a task. The method uses a processor coupled with stored instructions to implement the method. The stored instructions, when executed by the processor, carry out steps of the method. The method includes collecting a sequence of position measurement data and a corresponding sequence of ground truth velocities of a component of the electro-mechanical system performing the task. The method further includes adapting the causal FIR filter for estimating velocities of the component performing the task by updating one or more coefficients of the causal FIR filter to reduce a difference between the sequence of ground truth velocities and estimated velocities produced by filtering the sequence of position measurement data by the causal FIR filter.

The invention relates to a traffic flow speed estimation method based on a Gaussian mixture model and hierarchical Bayesian, and belongs to the technical field of intelligent traffic systems. The method comprises four steps of data preparation and preprocessing, traffic state clustering division, hierarchical Bayesian random parameter model construction, and model verification and performance evaluation. Through deep fusion of the Gaussian mixture model and the hierarchical Bayesian model, data-driven traffic state identification and random parameter estimation are realized, traffic flow heterogeneity can be effectively described, model parameter explosion is avoided, and model prediction precision and generalization ability are improved.

The invention belongs to the technical field of automatic driving and environmental perception, and particularly relates to a 4D millimeter-wave radar-camera omnidirectional speed estimation method based on a CAN bus, and the method specifically comprises the steps: S1, data collection and coordinate system unification; s2, vehicle motion information provided by a vehicle CAN bus is utilized to carry out pre-compensation on motion of the vehicle, and the absolute radial speed is obtained; s3, constructing a speed constraint condition by using optical flow and depth information, simplifying an inter-frame transformation matrix into a unit matrix in combination with motion compensation, and greatly simplifying an inter-frame motion transformation relationship; and S4, establishing three constraint conditions by combining the optical flow, the depth and the radial speed of the radar, constructing a linear equation set about the omnidirectional speed of the target, and directly solving the linear equation set to obtain a closed form solution of the three-dimensional omnidirectional speed of the target. According to the method, the accurate target depth and radial speed provided by the 4D millimeter wave radar and the optical flow information provided by the camera are fused, the CAN bus is innovatively introduced to carry out pre-compensation on the motion of the vehicle, a linear constraint model of multi-source information fusion is constructed, and complete estimation of the three-dimensional omnidirectional speed of the target is achieved. The method effectively overcomes the limitation of a single sensor in speed estimation, and improves the accuracy and robustness of speed estimation.

The present invention provides a terahertz near-field MIMO-OFDM multi-target position and velocity joint estimation method, including a terahertz near-field MIMO-OFDM model establishment step: establishing a near-field MIMO-OFDM signal model, representing the received signal in the near-field MIMO-OFDM signal model as a third-order tensor; a tensor decomposition step: decomposing the received signal tensor, and based on the uniqueness of the tensor decomposition, using the factor matrix after the tensor decomposition to estimate the estimated value of the intermediate parameter; a parameter preliminary estimation step: using the estimated value of the intermediate parameter and the least squares LS algorithm to obtain a preliminary solution to the iterative initial point to establish an optimization equation; a precise positioning and velocity estimation step: performing a Taylor expansion on the nonlinear term of the optimization equation at the initial point to perform a second-order Taylor approximation, converting the optimization equation into a convex optimization equation, and using the cvx toolbox to iteratively solve the convex optimization equation to obtain the estimated value of the target position and velocity. The present invention improves the accuracy of target parameter estimation by jointly processing the inter-antenna phase difference and the inter-subcarrier phase difference through tensor decomposition.

The present invention discloses a slope-compensated acceleration estimation method based on trajectory data, comprising the following steps: 1. obtaining trajectory data of a vehicle traveling on a slope and a horizontal section; 2. calibrating the vehicle's free flow speed using the trajectory data of the horizontal section; 3. analyzing the slope trajectory data to construct a slope acceleration model; 4. assuming that the expected acceleration of the horizontal section and the compensated acceleration of the slope conform to a normal distribution, according to the acceleration model, obtaining the probability density of the observed acceleration, the posterior probability of the compensated acceleration, and the joint probability distribution of the two; 5. Based on the joint probability distribution, constructing a compensated acceleration estimation model using an expectation-maximization algorithm and performing iterations to obtain the mean and variance of the expected acceleration of the horizontal section and the mean and variance of the compensated acceleration of the slope. The present invention can identify the additional compensated acceleration given to the vehicle by the driver on a slope, thereby better explaining the impact of driving behavior on road capacity.

The invention provides a UUV rapid iteration positioning and speed measuring method based on one-way communication, and relates to the technical field of acoustic positioning. The method comprises the specific implementation steps that unknown UUV nodes acquire known information through one-way communication with anchor nodes, a ranging equation is constructed, a linear-nonlinear iterative optimization algorithm is executed according to the known information, and coordinate data and speed data of the unknown UUV nodes are updated according to the result of the algorithm. According to the technical scheme, rapid positioning and speed measurement of the unknown UUV node can be completed only through one-way communication, the exposure risk of the unknown UUV node is reduced, and the accuracy of positioning and speed measurement is improved through the linear-nonlinear iterative optimization algorithm.

Methods and apparatus, including computer program products, implementing, and using techniques for estimating wind velocity. Measurements are obtained, which indicate two or more trajectories traversed by flying balls. Individual wind velocity estimates are determined for the two or more trajectories, wherein the determination comprises comparing a model acceleration of the ball with an observed acceleration of the ball, derived from the measurements. The aggregated wind velocity estimate is calculated as a weighted average of the determined two or more individual wind velocity estimates. The aggregated wind velocity estimate is used for generating ball trajectory information to be presented on an output device.

The application relates to a traffic flow speed estimation method based on a Gaussian mixture model and a hierarchical Bayesian method, and belongs to the technical field of intelligent transportation systems. The method comprises four steps of data preparation and preprocessing, traffic state clustering division, hierarchical Bayesian random parameter model construction, model verification and performance evaluation. Through the deep fusion of the Gaussian mixture model and the hierarchical Bayesian model, the application realizes data-driven traffic state recognition and random parameter estimation, can effectively depict traffic flow heterogeneity, avoids model parameter explosion, and improves the model prediction precision and generalization ability.

The present invention provides a multi-target position and velocity joint estimation method based on terahertz near-field MIMO-OFDM, comprising: a terahertz near-field MIMO-OFDM model establishment step: establishing a near-field MIMO-OFDM signal model, and representing a received signal in the near-field MIMO-OFDM signal model as a third-order tensor; a tensor decomposition step: decomposing the received signal sensor, and on the basis of the uniqueness of the tensor decomposition, performing estimation by using a factor matrix obtained after the tensor decomposition to obtain estimated values of intermediate parameters; a parameter preliminary estimation step: using the estimated values of the intermediate parameters and a least squares (LS) algorithm for preliminary calculation so as to obtain iteration starting points, and establishing an optimization equation; and a precise positioning and velocity estimation step: performing Taylor expansion and second-order Taylor approximation on nonlinear terms of the optimization equation at the starting points, converting the optimization equation into a convex optimization equation, and using a cvx toolbox to iteratively solve the convex optimization equation to obtain estimated values of the position and the velocity of a target. The present invention improves the accuracy of target parameter estimation by jointly processing an inter-antenna phase difference and an inter-subcarrier phase difference by means of tensor decomposition.

This disclosure provides systems, methods, and devices for wireless communications that support speed estimation, such as speed estimation with radio frequency (RF) sensing. In a first aspect, a velocity of an object is determined based on Doppler measurements from a plurality of transmit / receive points (TRPs). In a second aspect, a velocity of the object is determined based on angle of arrival (AoA) information received from the plurality of TRPs. The AoA information may include distance, AoA, time, or a combination thereof. In a third aspect, a velocity of the object is determined based on an orientation of the object during a sensing operation performed by the TRP. Other aspects and features are also claimed and described.

This invention discloses a method for estimating the remaining average velocity and controlling the flight time of a variable-speed missile, comprising the following steps: constructing a missile motion model and a differential geometry guidance model; predicting the remaining flight path of the missile under proportional guidance, and constructing a flight path error dynamic equation in the form of biased proportional guidance; designing a flight path control guidance law based on the optimal error dynamics method; predicting the missile velocity profile and estimating the remaining average velocity; converting the desired remaining flight time into the desired remaining flight path, and achieving flight time control through the flight path control guidance law. This invention relates to the field of aircraft dynamics, guidance, and control, and can effectively improve the accuracy of existing missile remaining average velocity estimation algorithms and reduce flight time control errors. It can provide an effective solution to the problem of time-varying multi-missile cooperative guidance.

The invention provides an inertial navigation error suppression method based on an LSTM-MMAE hybrid architecture, and the method comprises the steps: carrying out the data preprocessing of original IMU data, and obtaining a standardized time sequence; capturing a long-term time sequence dependency relationship of the standardized time sequence through a gating mechanism to obtain a speed increment predicted value; the noise covariance matrix of the speed increment predicted value and the observation noise covariance matrix of the theoretical trajectory speed are adjusted and fused through a plurality of parallel filtering models with different parameters, and optimal fusion estimation coping with different operation tasks is obtained; and carrying out speed prediction on the optimal fusion estimation to obtain optimal fusion speed estimation, and carrying out iteration on the optimal fusion speed estimation to obtain a target motion track. According to the method, through fusion of the LSTM neural network and multi-model adaptive estimation, IMU error accumulation is dynamically inhibited in a GPS failure scene, and the positioning precision and the system reliability are improved.

This invention discloses an AUV velocity estimation method based on a centrosymmetric multicell Kalman filter. First, a systematic analysis of the state-space model of the autonomous underwater vehicle (AUV) is performed. Then, combining the properties of centrosymmetric multicells and the concept of Kalman filtering, the reachable set of the AUV system state is calculated and optimized. The velocity range of the AUV is calculated using the optimal reachable set, enabling effective velocity monitoring. This method is applicable to AUVs with various configurations and different motion states. Compared with existing technologies, the proposed method can not only accurately estimate the velocity of the AUV but also provide velocity range estimates, thereby achieving effective velocity state monitoring and significantly improving the reliability and stability of the AUV.

The application provides a sparse point cloud optimization method, system, electronic device and storage medium. The method comprises: performing first position estimation and velocity estimation on a first point cloud cluster detected in a first period; obtaining motion estimation of the first point cloud cluster based on the velocity estimation and a time difference between the first period and a second period; obtaining second position estimation of the first point cloud cluster in the second period based on the motion estimation and the first position estimation; and superimposing the first point cloud cluster and a second point cloud cluster detected in the second period based on the second position estimation to form a dense optimized point cloud cluster. The first point cloud cluster in a historical period is predicted in velocity, so as to estimate the position of the first point cloud cluster in the current period, and then combined with the second point cloud cluster in the current period, so that the density of the point cloud cluster in the current period is increased, and the problem of sparse point cloud cluster obtained by the sensor is solved.

The invention discloses a robot control method and device based on linear speed, electronic equipment and a storage medium. The method comprises the following steps: performing attitude prediction and action simulation on first data through a first model to obtain second data; processing the second data through a second model to obtain third data; performing linear velocity estimation on the third data to obtain fourth data; and controlling the robot according to the fourth data. According to the method, the second data is processed through the second model, real noise data in the second data can be effectively removed, and the third data is obtained. And linear velocity estimation is carried out on the third data, so that stable linear velocity output can be obtained, and the control precision of the robot is improved.

The invention discloses an ultrasonic vector velocity imaging method, device and equipment, a medium and a product, and relates to the field of vector velocity measurement, and the method comprises the steps: determining a complex signal real part and a complex signal imaginary part according to ultrasonic radio frequency echo data; respectively carrying out weighted sliding normalization processing on the complex signal real part and the complex signal imaginary part to obtain a normalized complex signal real part and a normalized complex signal imaginary part; estimating by using a vector speed estimator according to the normalized complex signal real part and complex signal imaginary part to obtain a horizontal speed component and a vertical speed component of the vector speed; the loss function of the vector speed estimator is a speed estimation loss function; the speed estimation loss function is determined based on a mode length error, a direction angle error and a component level error of a speed vector; and determining a vector velocity image according to the horizontal velocity component and the vertical velocity component. According to the invention, the precision, real-time performance, stability and robustness of vector velocity imaging can be improved.

The application discloses a center-symmetrical polytope Kalman filtering AUV velocity estimation method based on a T-N-L structure. First, the state space model of the autonomous underwater vehicle is identified according to experimental data, then the reachable set of the AUV system state is predicted and corrected in combination with the properties of the center-symmetrical polytope and the Kalman filtering thought, the velocity interval of the autonomous underwater vehicle is calculated through the corrected reachable set of the system, and the effective monitoring of the velocity is realized. The method is suitable for autonomous underwater vehicles of various configurations and motion conditions, compared with the prior art, the application provides a system effective velocity estimation method, the method adopted by the application can generate more accurate velocity estimation value and more compact velocity estimation interval, not only the accuracy of the velocity state monitoring is improved, but also the behavior of the robot in the complex underwater environment is ensured to be more stable and predictable, the overall reliability and stability of the underwater robot system are significantly enhanced.

The application discloses a kind of robust frequency-modulated stepped frequency ISAR imaging radar velocity estimation methods, belong to radar technical field, first construct multiple frames frequency-modulated stepped frequency data model;Again, the velocity estimation and compensation of rough accuracy are realized using the same frequency point offset estimation of multi-frequency point synthesis;Then, the high-resolution imaging of the coarse resolution distance unit where the special point is located is obtained by the IFFT of intra-frame sub-pulse dimension, and the high-precision velocity estimation and compensation are realized using the high-resolution image of multiple special point synthesis;Finally, ISAR imaging is realized through the FFT between frames.The application realizes the estimation and compensation of velocity gradually using two-step method, first realizes the estimation and compensation of large motion amount by rough estimation, then realizes the residual motion estimation and compensation of high precision by synthesis high resolution;Offset estimation is realized using envelope cross correlation, motion parameter estimation is realized using offset fitting, and the amount of calculation is small;The robustness of algorithm is improved using the characteristics that multiple frequency points are not easily disturbed simultaneously, and the robustness of high-resolution velocity estimation is improved using the motion identity of multiple special points.

Embodiments of the present application provide a target pose estimation method and device and an automatic driving device. The method comprises: obtaining pose information corresponding to a plurality of target time instants of a target to be processed and a speed estimation result; estimating a pose of the target to be processed at a current time instant according to the pose information corresponding to each target time instant and the speed estimation result; and performing fusion processing on the pose corresponding to each target time instant to obtain a predicted pose of the target to be processed at the current time instant. The plurality of target time instants of the target to be processed are obtained, and a plurality of poses of the target to be processed at the current time instant are estimated according to the pose information and the speed estimation result, and then the plurality of poses are fused to obtain the predicted pose of the target to be processed at the current time instant. The influence of the error of a single pose information or a speed estimation result on the pose estimation of the target to be processed is reduced, and the stability and accuracy of the estimation result of the target pose are improved.

Embodiments of the present application provide a base speed estimation method and device for a dual-wheel-foot robot arm combined system, computer equipment, computer readable storage medium, computer program product, which belong to the field of robot control. The method comprises: obtaining a body perception history sequence corresponding to a target control period; grouping the body perception history sequence according to a physical subsystem to obtain a plurality of subsystem time sequence features; respectively encoding the plurality of subsystem time sequence features to obtain a mark sequence corresponding to the physical subsystem and a historical time step; performing interactive processing on the mark sequence based on a multi-head attention mechanism to obtain a context representation for representing cross-subsystem time sequence association; and outputting a base speed estimation value based on the context representation. The above technical solution can retain the physical subsystem grouping structure in the speed estimation process and utilize the cross-subsystem time sequence association to improve the adaptability of base speed estimation to system state changes.

The invention discloses an intelligent safe cooperative target hunting method and system for uncertain multiple water surface unmanned ships, and the method comprises the steps: building a nonlinear system model of an underactuated water surface unmanned ship with model uncertainty, defining the attitude state, speed state and control input variables of the underactuated water surface unmanned ship, and building a nominal system model which ignores the model uncertainty; estimating a target speed by using a finite time speed observer, and estimating model uncertain items of the unmanned surface ship system by using a finite time extended state observer; constructing a nominal controller based on a convex hull theory and a barrier function theory by utilizing state information obtained by estimation, and realizing preliminary stable surrounding and safety performance of the target; on the basis of nominal control, a flexible behavior-commentator reinforcement learning algorithm based on an entropy item is introduced, and a reinforcement learning controller used for compensating the uncertainty of the model is designed; and designing a quadratic programming problem including control of a Lyapunov function and control of a barrier function constraint, and combining a combined controller to implement safety key constraint on final control input so as to realize system safety control in a target encircle process. According to the method, speed estimation can be carried out on a target with an unknown speed, the uncertainty of a water surface unmanned ship model is eliminated, and cooperative safe hunting of the target is realized.

The invention discloses a target relative speed estimation method based on a laser radar point cloud, and belongs to the technical field of rail transit safety. In order to solve the problems of error accumulation, poor real-time performance and insufficient robustness to a sheltered scene in an existing target speed estimation method based on laser radar point cloud, the invention constructs an improved VoxelNeXt network. The improved VoxelNeXt network comprises a multi-scale sparse three-dimensional convolutional backbone network, a time sequence fusion module based on an attention mechanism and an all-voxel prediction head capable of integrally outputting a target detection frame and a relative speed, and training in an end-to-end mode to obtain a target relative speed estimation model; during application, real-time continuous multi-frame point clouds are input into the target relative speed estimation model, and the three-dimensional detection frame and the three-dimensional relative speed of the target can be directly output at the same time. According to the method, deep fusion of detection and speed estimation is realized, and the estimation precision, the processing efficiency and the system robustness are improved.

Systems, devices, processes, and computer-readable media for estimating impact severity during handling or delivery of assets are disclosed. For example, an example of a process includes capturing, by a device, a measured acceleration of an asset associated with the device; de-clipping, by the device, the measured acceleration of the asset to determine a reconstructed acceleration of the asset; determining, by the device, a velocity estimate of the asset based on the reconstructed acceleration; and determining, by the device, whether the asset has experienced a severe impact based on the velocity estimate.

The present application relates to the technical field of data processing, and especially relates to a parameter identification method for releasing residual velocity estimation deviation of less than 20% of TM. First, in the global range, the values of each group of to-be-identified parameters in the inertia sensor test quality release residual velocity estimation model are randomly initialized as the initial individual position in the population; and a plurality of dominant individuals are selected therefrom as the learning starting point; then, the learning ability is updated to make each individual generate a new solution around the dominant individual in different ranges, so as to take into account the global search ability and the local optimization ability. The method provided by the present application reduces the dependence on a single starting point, improves the stability and applicability of the algorithm, and ensures that the test quality release residual velocity estimation error is less than 20%.

The invention discloses a multi-base-station cooperative target position and speed estimation method based on an OTFS signal. The method comprises the following steps: establishing a multi-base-station cooperative sensing signal model based on the OTFS signal; converting a parameter estimation problem into a sparse vector reconstruction problem based on the measurement matrix; designing a subspace tracking algorithm oriented to a multipath environment, and solving the sparse vector; performing target position and speed coarse estimation by using the sparse vector corresponding time delay and Doppler frequency; and establishing a parameter search space, and fusing the sensing information of multiple base stations to obtain the accurate position and speed of the target. According to the method, the target position and speed estimation under the multipath interference is realized by utilizing the characteristics of the time delay-Doppler domain of the OTFS signal and the multi-base-station cooperative gain, and the method has the advantages of high estimation precision, low complexity, multipath interference resistance and the like, and has a wide prospect in practical applications such as a communication sensing integrated system and multi-base-station cooperative sensing.

The invention discloses a multi-mode IMU (Inertial Measurement Unit) data reconstruction and speed estimation method and a multi-mode IMU data reconstruction and speed estimation device based on adaptive feature generalization, which are used for constructing sharing features of acquired data in different wearing modes and advancing states through IMU data acquired by a dynamic sliding window and solving the problem of feature difference caused by motion mode difference. The method comprises the following steps: performing feature extraction on IMU sensor data in different modes through a convolutional neural network, reconstructing IMU sensor data at the same speed from a feature map through dimension raising of a decoder, and aligning the reconstructed IMU sensor data with IMU sensor data in a certain specific simple attitude, and data feature differences caused by different motion modes are eliminated. And finally, inputting the feature map into a linear network to realize speed prediction. According to the method, the influence of data feature differences on motion analysis and speed estimation in different motion modes is effectively solved, the speed estimation precision is remarkably improved, and meanwhile, the method has the adaptive capacity to the wearing mode and the advancing state.

Methods and devices for odometry are disclosed. A previous position estimate and a current position estimate for a device are obtained. A current velocity estimate for the device is also obtained. A first distance estimate is calculated based on the position estimates, and a second distance estimate is calculated based on the velocity estimate. These two distance estimates are combined to generate a combined or hybrid distance estimate for the distance travelled by the device.