484results about How to "Avoid Cumulative Errors" patented technology

The invention discloses an estimation method of state of charge (SOC) of a power battery. The method comprises the following steps: estimating the SOC estimated value SOC2 of the power battery by an electricity accumulative method; taking a Sigma-point Kalman filter (SPKF) as a basic estimation tool, taking a dual-RC circuit battery model as a time and measurement update engine of the SPKF, and estimating the SOC estimated value SOC1 of the power battery by a Kalman filter method; and obtaining the final SOC estimated value SOC based on the SOC estimated value SOC2 by the electricity accumulative method and the SOC estimated value SOC1 by the Kalman filter method by utilizing a weighted average method. Correspondingly, the invention further discloses an estimation system of the SOC of thepower battery. The invention has the advantages of high SOC estimation accuracy, stable operation and the like, and is convenient in real-time estimation, thus being applicable to pure electric vehicles and hybrid electric vehicles in need of the power battery.

The invention provides a sequence image splicing method of a low-altitude unmanned vehicle, comprising the following steps: acquiring sequence images, and correcting camera distortion; extracting characteristic points, and matching the images; generating an image splicing sequence; carrying out global optimization splicing, and splicing the images in accordance with splicing strategies; and carrying out global error compensation. A sequence image splicing system of a low-altitude unmanned vehicle comprises a characteristic point extraction module, an image matching module, a splicing sequencegeneration module and an image splicing module, wherein the characteristic point extraction module is used for extracting SURF characteristic points and HARRIS-AFFINE characteristic points of the sequence images; the image matching module is used for calculating a homography matrix between adjacent images in accordance with the characteristic points, and eliminating deteriorated images which are off-course completely; the splicing sequence generation module is used for determining the splicing sequence of the sequence images; and the image splicing module is used for carrying out splicing andglobal error compensation in accordance with the splicing sequence.

The invention discloses an lsd-slam-based real-scene three-dimensional reconstruction method. The method comprises the steps of S1, recording a to-be-modeled scene, and collecting video data that covers the to-be-modeled scene; S2, resolving a key frame gesture from a key frame to obtain the restored images of the three-dimensional information based on the lsd-slam modeling technology, and generating a three-dimensional space point cloud; S3, selecting images for being spliced into a panoramic image out of all the restored images of the three-dimensional information according to the space location relationship, and generating the panoramic image; S4, putting the panoramic image and the three-dimensional space point cloud into the same spatial reference system, and subjecting three-dimensional data to the monomer treatment process. In this way, the key frame can be obtained through acquiring and analyzing video data. After that, the three-dimensional space point cloud and the panoramic image are generated through analyzing the key frame. The three-dimensional reconstruction is achieved based on the three-dimensional space point cloud and the panoramic image. Therefore, the calibration of equipment is not required and the method can be implemented without any hardware equipment. Meanwhile, panoramic data and model data are collected, so that the problems during the existing three-dimensional reconstruction process are solved.

The invention relates to an aircraft positioning method based on a positioning and sensing device. The method comprises the following steps: acquiring aircraft attitude information by virtue of an inertial sensor unit, collecting relative altitude information by utilizing an ultrasonic sensor, and acquiring real-time ground image information by utilizing a camera; acquiring relative displacement information of the aircraft by utilizing an improved optical flow method, and acquiring global position and attitude information of the aircraft by utilizing grid information; and realizing an accurate and robust aircraft positioning method by virtue of multi-sensor information fusion and acquisition by utilizing a filtering algorithm. Long-term accurate positioning of the aircraft can be realized in case of absence of GPS signals and any other external aided navigation, the robustness is high, and an accumulative error is avoided.

Embodiments of the invention provide a character recognition model training method, device and system, a character recognition method, device and system, and a storage medium. The character recognition model training method comprises the steps of acquiring at least one sample image and labeled data corresponding to the at least one sample image, wherein the labeled data corresponding to each sample image is used for indicating the position of an actual character region of the sample image and actual characters in the actual character region; for each sample image in the at least one sample image, inputting the sample image into a region positioning network in a character recognition model to determine the position of a predicted character region of the sample image; inputting image information corresponding to the position of the predicted character region into a character recognition network in the character recognition model, so as to identify predicted characters in the predicted character region; and training the character recognition model by at least using the labeled data corresponding to the at least one sample image and the predicted characters. The accumulative error of character recognition can be effectively avoided.

The invention provides a method and system for laser-IMU external parameter calibration. The method comprises steps of acquiring IMU measurement data and laser radar measurement data; carrying out IMUpre-integration of the obtained IMU measurement data, carrying out calculation to obtain an IMU pose transformation estimated value of the IMU relative to the IMU initial pose at the next moment, andaccording to the estimated value and an actual measurement value of the IMU at the next moment, obtaining an associated residual error associated with the data; processing measurement data of the laser radar, utilizing IMU pre-integration to obtain projection coordinates of reprojecting a plurality of laser radar points to a world coordinate system, and calculating a reprojection error from eachlaser radar point to a calibration target map; adopting a nonlinear least square method to iteratively optimize laser radar-IMU external parameter calibration so that the external parameter calibration result can be obtained. The method is advantaged in problems that in laser-IMU external parameter calibration, mechanical external parameters are not easy to obtain, manual measurement errors are large, and measurement is troublesome are solved, defects of the laser radar and the IMU are mutually overcome to a certain extent, and pose solving precision and the speed of the SLAM method can be improved.

The invention discloses a visual information-based robot line-walking navigation method along a guide line, comprising the following steps of: (a) obtaining an environment image by a visual sensor arranged on a robot; (b) obtaining route information from the environment image, and transmitting the obtained route information to a controller; and (c) setting the linear velocity and the angular velocity of the robot by the controller according to the route information. According to the method, the robot can walk along an appointed route.

The present invention provides 3D mirror object surface outline measuring device and method. The measuring device consists of LCD panel display, industrial video camera, computer and guide track. The LCD panel display is fixed on the guide track and connected via electric cable to the computer, and the industrial video camera is connected via video line to the computer. The measuring method includes device regulation, data acquisition, data processing and other steps. The device of the present invention is simple in composition, and the measuring method is simple and has data processing with less complexity.

The invention discloses a mobile robot navigation map generation method based on ORB_SLAM2. The method comprises the following steps: step one, carrying out visual information collection; to be specific, enabling a mobile robot to move freely in a three-dimensional environment and collecting image information and depth information through a depth sensor; step two, on the basis of an ORB_SLAM2 algorithm, estimating pose information of the robot, carrying out local BA pose information optimization, screening key frames, selecting one key frame, outputting the pose information of the key frame, providing a sensor reference transformation matrix for creating a navigation map, and updating pose information of each key frame according to the global BA optimization under the condition of detecting a loop; and step three, carrying out map updating; to be specific, generating a three-dimensional map based on a skip list tree according to the pose information and carrying out updating continuously according to pose information of the new key frame. According to the invention, an efficient three-dimensional map is created by utilizing the depth image information; and the map uses the skip list tree structure as a data structure of the three-dimensional map, so that positioning, navigation and obstacle avoidance of the mobile robot can be realized in real time.

The invention discloses a sparse direct method-based monocular visual odometry (VO) method of a quadrotor unmanned-aerial-vehicle. The method is characterized by: carrying out depth estimation on a key frame, wherein feature points of the key frame are determined by a feature point method, an eigenmatrix between two adjacent frames is calculated, the eigenmatrix is decomposed and a rotation matrix and a translation matrix between the two adjacent frames are calculated to obtain an external parameter matrix, and then depths of the feature points are calculated according to a triangulation method; and after obtaining depth values of the feature points, obtaining the pose of the quadrotor unmanned-aerial-vehicle by solving through a sparse-matrix direct method, and carrying out motion estimation on all frames, wherein sparse feature points are extracted, the direct method is used to calculate a position of each feature point in the next frame, and grayscale information of each pixel point in pixel blocks which have a fixed size and are around the feature points is utilized to optimize grayscale differences between the two adjacent frames to obtain the motion pose of a camera. The method has the advantages that cumulative errors are avoided, higher accuracy is maintained for a long period, and the calculation amount can also be reduced.

The invention discloses a speech recognition method and a speech recognition device. A specific embodiment of the method comprises the following steps: segmenting speech information to be recognized into multiple frames of speech segments; performing acoustic model scoring and language model score-checking on the speech segments frame by frame through a preset decoding network; and determining a word sequence corresponding to at least one decoding path in the decoding network as the speech recognition result based on the score result, wherein first language model score-checking and second language model score-checking are carried out in sequence during language model score-checking on a frame of speech segment. By implementing the technical scheme, accurate and efficient speech recognition is realized.

The invention provides a shale gas reservoir brittleness earthquake prediction method and device. The method comprises the following steps: carrying out calculation by means of the generalized linear inversion thought, based on a Bayesian theory and by utilizing actual seismic data of a work area and a prior distribution function and a Zoeppritz equation of a plurality of elastic parameters, to obtain a disturbance quantity solution expression of the elastic parameters; carrying out calculation according to the solution expression and an initial model of the elastic parameters to obtain disturbance quantity value of the elastic parameters; carrying out calculation according to the disturbance quantity value and the initial model to obtain a current model of the elastic parameters; carrying out calculating based on the actual seismic data, the current model and the Zoeppritz equation of the plurality of elastic parameters to obtain seismic inversion residual error; and if the seismic inversion residual error is smaller than a set threshold value, serving the current model as the value of the elastic parameters, and carrying out calculation according to the value of the elastic parameters to obtain shale gas reservoir brittle index of the work area. The method and device can improve shale gas reservoir brittle index predication accuracy.

The invention discloses a disposable arc-rolling and forming die, relating to the technical field of motor production. The disposable arc-rolling and forming die comprises a lower template (1) and a lower template accessory device, an upper template (21) and an upper template accessory device, and also comprises a movable plate device, a slideway device, a pressing device and a positioning device. The disposable arc-rolling and forming die can ensure that a plate cylinder motor shell is once coiled and formed, thus accumulated errors are avoided, collision and deformation of the products in machining and conveying processes are avoided, uniformity of sizes of final products is greatly improved, mechanical automation degree is increased, working efficiency is improved, yield and quality ofthe products are increased, and input of equipment resources and manpower resources is also reduced.

The invention discloses a method for reconstructing a three-dimensional curve face through point cloud. The method comprises the steps of inputting point cloud data P possibly carrying noise and abnormal values and the number m of peaks of the curve face needed to be reconstructed; initializing a dictionary matrix V and a connection matrix B; updating the dictionary matrix V and the connection matrix B in an iteration mode till convergence; outputting a reconstructed triangular net to complete reconstruction of the three-dimensional curve face. By means of the method, the triangular net is directly reconstructed through the point cloud, the noise input into the point cloud can be removed well, non-uniform sampling is processed in a robust mode, and the characteristics in the point cloud can be recovered well. Compared with a conceal method, the triangular net is directly reconstructed through the point cloud so as to avoid accumulated errors caused by multi-step optimization in the conceal method. Compared with a combination method, the reconstruction errors serve as a target function, so that the reconstruction errors of the curve face reconstructed through the method are usually smaller than those reconstructed by the existing combination method.

The invention discloses a machining method for a tenon tooth of a shrouded turbine blade. The method includes the steps of A, selecting six positioning points for a blank part and guaranteeing the position dimensions of the positioning points in the process of producing the blank part; B, providing a clamp according to the position dimensions of the positioning points determined in the step A; C, clamping the blank part produced in the step A on the clamp provided in the step B and machining a tenon tooth blank part to form the tenon tooth; and D, carrying out clamping with the tenon tooth as the clamping and locating datum, and machining a lower flange plate and an upper flange plate to complete machining of the turbine blade. According to the machining method for the tenon tooth of the shrouded turbine blade, datum conversion errors can be reduced, accumulated errors caused by datum conversion can be avoided, and the qualification rate of delivery of each time can reach above 90% in actual production.

The invention provides a carrying platform elevating gear, a reaction cavity, and a plasma processing device. A bracket driving source drives a shaft lifting bracket to do lifting motion. A bellows shaft passes through the shaft lifting bracket and is fixedly connected with the shaft lifting bracket, so as to do lifting motion with the shaft lifting bracket in a linear bearing, thereby realizing lifting motion of the carrying platform which is fixed on the bellows shaft. In the carrying platform elevating gear, a guiding mechanism is used to limit axial rotation of the bellows shaft in a lifting process. The bellows shaft passes through the linear bearing, a guiding shaft installing assembly, and a shaft lifting bracket in sequence. The guiding shaft installing assembly tightly holds the bellows shaft. The upper end of the guiding shaft is fixed on the cavity wall of the reaction cavity. The lower end of the guiding shaft passes through the guiding shaft installing assembly, so that the guiding shaft installing assembly does lifting motion with the shaft lifting bracket along the guiding shaft. The carrying platform elevating gear can limit rotation degree of freedom of a connecting rod, so as to ensure a carrying device from deflecting.

The invention discloses a stepper motor out-of-step compensation method which is used to control an ultrasonic probe. The ultrasonic probe comprises a transducer and a stepper motor. The method comprises the following steps: acquiring the actual position information of the transducer; comparing the actual position information with target position information to get out-of-step data of the stepper motor; comparing the out-of-step data with a preset threshold to determine the threshold scope to which the out-of-step data belongs; and determining a compensation mode corresponding to the out-of-step data according to the threshold scope, and implementing a compensation operation corresponding to the compensation mode. According to the method, adaptive compensation is made by use of a compensation mode corresponding to the out-of-step condition of the stepper motor according to the out-of-step condition to prevent accumulative error of the relative position of the transducer, so that an image scanned by the ultrasonic probe is more accurate and does not drift. The invention further discloses a stepper motor out-of-step compensation device which has the same beneficial effect.

The invention provides a method and a device for calculating a passenger vehicle passenger flow congestion degree. The method comprises steps that images respectively acquired by at least two cameras in a passenger vehicle are acquired, the acquired images are spliced to form a panorama image in the passenger vehicle, body identification and counting are carried out on the basis of the panorama image, and the counting result is used for carrying out passenger vehicle passenger flow analysis. The method is advantaged in that the images acquired by the multiple cameras are spliced according to an image coordinate transformation matrix, the panorama image in the vehicle is determined, body detection of the panorama image in the vehicle is carried out to determined the quantity of passengers in the vehicle, the passenger vehicle passenger flow congestion degree is analyzed according to the passenger quantity detected from the panorama image but not according to statistics on the basis of accumulation difference of passengers getting on and off of vehicle gates, an accumulation error is avoided, and statistics precision is improved.

The invention discloses a Chinese electronic medical record named entity extraction method and system. The method comprises the following steps: mapping each word in an input statement into a vector through a character embedding layer; inputting the character sequence by adopting an LSTM model to obtain an implicit representation vector; taking each word as the last word of the semantic unit withthe word as the end, and carrying out semantic segmentation to obtain all grid codes corresponding to the word; carrying out linear combination on all grid codes of each character to obtain a final vector of each character, wherein a weight in the linear combination is given by a self-attention mechanism; taking a deep residual network as an output decoding layer to decode a named entity chain. According to the Chinese electronic medical record named entity extraction method and system, various medical named entities can be automatically extracted from the Chinese electronic medical record, the extraction efficiency is improved, and the labor consumption in a medical record structuralization process is avoided.

The invention provides a clamping device for processing a mushroom-shaped blade having a Z-shaped blade shroud, and can solve the problems that in a conventional processing course, different processing equipment is adopted to process a blade root and a blade shroud in operation procedures, so that the accumulated error is large, and the requirement for processing dimension cannot be met. A blade root positioning and clamping assembly and a blade shroud positioning and clamping assembly are respectively mounted on the lateral left side and the lateral right side of a bottom plate; the blade shroud positioning and clamping assembly comprises a blade shroud tightly-propping seat and a blade shroud positioning seat; the blade shroud tightly-propping seat is detachably mounted at the lateral right side end of the bottom plate; the blade shroud positioning seat is located at the lateral left side end of the blade shroud tightly-propping seat; a positioning groove is formed in a front side wall of the blade shroud positioning seat; supporting screws are separately arranged at a groove bottom surface and a front side groove wall of the positioning groove; floating supports are separately arranged on the rear side groove wall and the front side groove wall of the positioning groove, and are separately located on the right side of the corresponding supporting screws; a profile compacting block is located on the bottom plate and in front of the blade shroud positioning seat; and compacting screw facing the blade shroud positioning seat is arranged on the profile compacting block.

The invention discloses a wind power climbing event probability prediction method and system based on a Bayesian network, and the method comprises the steps: mining the dependency relationship betweena wind power climbing event and related meteorological influence factors such as wind speed, wind direction, temperature, air pressure, humidity, and the like, and building a Bayesian network topological structure with the highest fitting degree with sample data; quantitatively describing a conditional dependency relationship between the climbing event and each meteorological factor, estimating the value of each conditional probability in a conditional probability table at each node of the Bayesian network, and forming a Bayesian network model for predicting the wind power climbing event together with a Bayesian network topological structure; deducing the probability of occurrence of each state of the climbing event according to the numerical weather forecast information of the mastered prediction time; the value of the corresponding conditional probability at each node is adaptively adjusted, so that the inferred conditional probability result of each state of the climbing event is optimized, and the compromise between the reliability and the sensitivity of the prediction result is realized.

The invention provides a monitoring method for remaining power of a wireless sensor network node. The method comprises steps of building a lithium battery RC?equivalent model to obtain a state equation after discretization, estimating the current state variable value, updating an estimation variance and calculating Kalman gain according to the variance, updating state?variable battery remaining power?SOC and battery open circuit voltage, and comparing voltage in a register with the battery open circuit voltage obtained in the former step. A device using the method comprises a data acquisition module for sampling voltage, current and temperature of the battery, a Flash memory module for storing data, a voltage conversion module for providing power for the device and the sensor, a battery charging module, and an embedded microprocessor for receiving and processing relative battery data of the data acquisition module, and the data acquisition module is connected with the embedded microprocessor via an I2C bus.

The invention discloses automatic spot welding equipment with the advantages that the welding production efficiency is improved, and the work intensity is reduced. The automatic spot welding equipment comprises a welding machine body and an installing base, wherein a rotating device driven by a servo motor is vertically arranged on the installing base, an installing disc is detachably arranged at the upper part of the rotating device, at least four material support welding devices are uniformly distributed and arranged on the installing disc in the periphery direction, each material supporting welding device comprises a welding frame arranged on the installing disc, a lower electrode is arranged in the welding frame, a material jacking rod is movably arranged in the welding frame, the top of the material jacking rod is provided with a material jacking frame, the bottom of the material jacking rod is hinged to a crow bar, the middle part of the crow bar is hinged onto a bottom plate of the welding frame, the end part of the crow bar extends out of the welding frame, an inclined wedge block which is provided with the bottom surface downwards inclined in the rotating direction of a turntable and is used for downwards pressing a part, extending out of the welding frame, of the crow bar is arranged on the installing base, the top of the welding frame is provided with a corresponding lower electrode according to the position to be welded of the materials, and a detection device for detecting whether materials exist on the welding frame or not is arranged in a position, corresponding to the material support welding device in a former position of the welding station, on the installing base.

A length measurement precision calibrating method and device for a laser tracker belongs to the measurement and calibration field of optical instruments, and are aimed at solving the problems of the prior art that the calibrating precision of an ultra-long high-precision guide rail is low, the machining cost is high, the length measurement range is small and the occupation space is large. An interference length of a double-frequency laser interferometer is taken by the calibrating device as a length basis. Platforms spaced from each other at certain intervals within the measurement range and an adjustable pedestal above the platforms together form a high-precision virtual guide rail. Right-angle prism units are utilized to adjust the space of two measurement light beams, and five prism units are utilized for turning of light paths. In this way, length measurement precision calibration is conducted on a laser tracker within the ultra-long measurement range in a limited space. The calibrating method includes the steps: leveling a light beam of a laser interferometer; adjusting the virtual guide rail to realize flatness and linearity that are needed by calibration; adjusting an emitting light beam of the laser tracker to enable the emitting light beam to be parallel to the light beam of the laser interferometer; forming a measurement optical path; fixing a reflector, and adjusting the two measurement light beams and enabling the two light beams to parallel to each other so that an Abbe error is eliminated; conducting measurement; and performing data processing.

The invention discloses a constant-volume liquid feeding metering method and device for COD detection; the upper end of a metering tube is connected with a piston pump, and the lower end of the metering tube is connected with a multi-way solenoid valve; the tube wall of the metering tube is externally provided with a photometer fixing ring, the photometer fixing ring emptily sleeves the metering tube and can move up and down along the metering tube, the photometer fixing ring is fixedly connected with a liquid level photometer light source emitting module and a liquid level photometer light source accepting module which are symmetric relative to a metering tube center shaft, the photometer fixing ring is horizontally fixedly connected with one end of a horizontal rod, the other end of the horizontal rod is vertically fixedly connected with a laser distance measurement sensor located below the horizontal rod, and the middle section part of the horizontal rod is fixedly connected with a linear stepping motor located below the horizontal rod through a linear stepping motor shaft; a sample valve and a plurality of reagent values of the multi-way solenoid valve are connected with corresponding reagent bottles respectively; the solution volume is arbitrarily given in a metering pipe measuring range, addition of one part of water sample or a reagent can be completed through one-time extraction, and accumulated error of many times of constant-volume liquid taking can be avoided.