58 results about "Path integration" patented technology

The invention provides a robot navigation map construction method based on rat brain hippocampus spatial cells. According to an information transfer loop of spatial navigation related cells in a hippocampus structure of a mammal, a robot acquires a current autonomic movement clue and color depth image information through environment exploration. The autonomic movement clue progressively forms coding to the spatial environment through path integration and characteristic extraction of the spatial cells in the hippocampus structure. The place field of place cells is gradually formed in an exploration process and furthermore covers the whole spatial environment, thereby forming a cognitive map. Meanwhile, Kinect acquires color depth image information of a view scene right in front of the current position as an absolute reference for performing closed-loop detection on the path and correcting a path integration error. At a closed-loop point, a system resets spatial cell discharging activity and performs correction on the path integration error. Nodes on a final navigation map comprise place cell mass coding information, corresponding visual cues and place topological relationships.

The invention discloses a multi-modal action recognition method based on a depth neural network. The method comprehensively utilizes multi-modal information such as video images, optical flow diagramsand human skeleton. The specific steps are as follows: firstly, a series of preprocessing and compression are performed on the video; Obtaining an optical flow graph based on adjacent frames of the video; The human skeleton is obtained from the video frame by using the attitude estimation algorithm, and the path integral characteristics of the skeleton sequence are calculated. The obtained optical flow graph, skeleton path integration feature and original video image are input into a depth neural network with multi-branch structure to learn the abstract spatio-temporal representation of humanmotion and correctly judge its motion category. In addition, the pooling layer based on attention mechanism is connected to the video image branch, which enhances the abstract features closely related to the final motion classification results and reduces the irrelevant interference. The invention comprehensively utilizes multimodal information, and has the advantages of strong robustness and high recognition rate.

The invention discloses a contour-based two-dimensional fragment splicing method which comprises the steps of: a first step, obtaining a scanned fragment image and extracting a fragment contour curve; a second step, performing polygonal approximation and feature extraction on the contour, thereby obtaining two feature sequences, namely a rotation angle sequence and a side length sequence; a third step, searching for partial candidate matching segments between random two fragments based on the feature sequences, thereby searching for partial matching among the N fragments, comparing random two fragments of the N fragments, thereby obtaining a final partial matching set; and a fourth step, in the final partial matching set, searching a global path for partial matching based on a rotation angle which is required for splicing the fragments according to partial matching, and integrating all determined path to a path which contains all fragments, thereby realizing splicing. Compared with a best priority method, the contour-based two-dimensional fragment splicing method is advantageous in that the global path searching method based on partial matching splicing rotation angle measurement sufficiently utilizes local matching pairs; a correct matching path and a fragment reconstruction path are determined; and the matching efficiency is improved.

The invention discloses an indoor localization method based on a foot inertia sensor. According to the indoor localization method, a framework of a grid cell path integrating method integrated and simulated on the basis of extended Kalman filter estimation is used; the information of localization positions and directions is obtained by calculating data of the inertia sensor; a plurality of condition judging criteria are used by zero-velocity correction for judging whether the feet touch the ground or not; the current velocity is adjusted to zero and the Kalman filter estimation is triggered when the conditions are met; the calculated nonlinear error increase of the velocity and the position information are corrected; then the velocity information is used as input and a grid cell path integration method is used for further eliminating drift error of the position information.

The invention provides a hydraulic control loop for a garbage truck. The hydraulic control loop is composed of an oil source, an oil path integration block and an actuator. The oil source comprises a metering pump, an air cooler, a returned oil filter and other hydraulic auxiliary components. The oil path integration block comprises a high-pressure filter, an electromagnetic reversing valve, a balance valve and an overflow valve. The actuator comprises a feeding hydraulic cylinder set and a material pressing hydraulic cylinder set. The high-pressure filter, the electromagnetic reversing valve, the balance valve, the overflow valve, the feeding hydraulic cylinder set in the hydraulic oil source oil path integration block are sequentially connected through a pipeline to form a hopper feeding control loop, and are connected with the material pressing hydraulic cylinder set to form a material pressing control loop. According to the system, the number of the valves in the oil path integration block is reduced, the throttling loss is reduced, and the energy saving is achieved. A main oil path is internally provided with the high-pressure filter to guarantee cleanliness of oil liquid, and blockage of the valve is avoided. A returned oil path is provided with the air cooler, and heat dissipation of the system is improved. A hydraulic lock and a flow valve are replaced with the bi-directional balance valve, the self-locking function is achieved, and the stability of hopper moving of the garbage truck is guaranteed.

The invention discloses a departure moment optimization method based on a demand response type bus, which carries out optimization on a departure moment of the demand response type bus for the demandresponse type bus of which service objects have differences in spatial and temporal distribution. The optimization method aims at a maximum weight combination value of a loading rate and a passenger average satisfaction level of the demand response type bus, considers constraint conditions of a demand time interval, a departure moment time interval, a demand type, passenger waiting time, vehicle capacity and the like, and establishes a demand response type bus departure moment optimization model to carry out optimization. According to the invention, the departure moment, the demand response and a driving path are integrated in one uniform optimization model and synchronous optimization is carried out, so that service timeliness and the operation level of the demand response type bus are improved.

The invention discloses a continuous state routing algorithm of a satellite link network, which comprises the following steps: (1) registering the time of switching between a link connected state and a link unconnected state within the entire simulation time, and establishing a link state switching table; (2) searching and recording paths within the entire simulation time; (3) filtering the pathsand storing the optimal one; and (4) performing full-path integration by using the Dijkstra shortest path algorithm. The routing algorithm disclosed by the invention is different from the traditionalsatellite link network routing algorithm in that: the routing algorithm continuously records and uses the time-varying dynamic topology of the satellite link network within the entire simulation timewithout simplifying the topology into discrete static topology by dividing time intervals. The routing algorithm stores the most complete network dynamic information, thus the best routing design result can be provided; and moreover, the numerical form is simple, the paths are filtered in the path registration process, thus the workload of full-path integration is reduced.

The invention relates to system and method for realizing wavelength reusing and broadcasting functions of a wavelength division multiplexing passive optical network. An OLT (Optical Line Terminal) is adopted in the system and connected with an RN (Remote Node ) through two single-mode optical fibers, and the RN is connected with two groups of ONUs (Optical Network Units, wherein the RN mainly comprises two optical filters, three optical circulators, an optical coupler and two 2*N AWGs (Array Waveguide Gratings); the separation and the path integration of uplink and downlink signals can be realized at the position of the RN, and each ONU can simultaneously receive broadcasting signals and point-to-point data signals; and the invention realizes the transmitting and wavelength reusing functions of the broadcasting signals only by changing the structure of the system and enables the system to achieve an equilibrium between the cost and the performance.

The invention provides a laser light source device, which comprises a light source exciter, a light path integration unit, a first color wheel having a light-transmitting area and a fluorescence area, a second color wheel having two kinds of color light filter areas in alternative distribution, an optical path guidance unit for guiding an excitation light source, a relay lens group, a collimating lens group, a reflector and a light source reception unit. Consumption of the light source is reduced through the optical path guidance unit, and brightness is improved, thereby improving image quality; and the structure makes the color wheels easier to realize.

The invention discloses a high resolution imaging device used for a cold atom system. The device comprises a disordered optical lattice generation path and an imaging light path, and is characterized in that the disordered optical lattice generation path comprises a laser beam and is composed of a first lens, a second lens, a diffusion barrier, a reflection mirror and a dichroic mirror along the laser beam direction in sequence; the first lens and the second lens form a beam expanding system; the reflection mirror and the dichroic mirror are in 45 degrees with the beam direction; the imaging light path comprises an imaging lighting beam and is composed of a vacuum pool, a third lens, a dichroic mirror, a fourth lens and a CCD (Charge Coupled Device) along the imaging lighting direction in sequence. According to the device, a laser speckle generation light path and an imaging light path are combined together, the purpose of loading disordered optical lattices to cold atom groups is realized by using an aberration elimination lens, the high resolution imaging of the atom groups is also realized, and the high resolution imaging device has the advantages of simple structure, high light path integration degree, less occupied space, adjustment convenience and the like.

The invention provides a track integration method and a device thereof. The method comprises the steps of performing statistic calculation of the occupation of a maximal spreading factor layer, and constructing a set of occupied tracks; performing set partition of the maximal spreading factor layer to obtain a set to be selected; selecting the set to be selected with the minimum number of users as an adjustment source, and considering the users therein as users to be adjusted; and performing physical channel reconfiguration to the users to be adjusted to reallocate physical resources. The inventive method and the device can obtain physical channels meeting requirements via integration, thereby remarkably improving the utilization ratio of physical resources.

The invention relates to an improved minimum cost flow insar phase unwrapping method. According to the invention, a minimum cost flow phase unwrapping model is globally optimized. During the calculation of the phase gradient, the error superposition is easily caused when the path integration is carried out. When a target terrain obtained by a synthetic aperture radar is too steep, the elevation difference of the target between adjacent pixels exceeds an fuzzy range. In this way, a real phase value can be caused to jump, so that an unwrapping error transmission of the minimum cost flow algorithm is caused. A ground control point can assist in INSAR phase unwrapping, and the unwrapping success rate is improved. However, according to the existing phase unwrapping algorithm, the correction iscarried out by utilizing a control point after an unwrapping result is obtained, wherein the information of the control point is not really merged into the minimum cost flow unwrapping algorithm. According to the invention, a reasonable optimization objective function is constructed and the noise superposition caused by path integral is reduced due to the fact that the minimum cost flow unwrappingmodel is reduced. Meanwhile, the information of the control point is introduced into the model of the unwrapping algorithm, so that the unwrapping result is more accurate.

The invention provides an SOA (service oriented architecture)-based clothing enterprise collaboration flow path integration method which comprises the steps: 1) building an SOA integration development system platform which comprises at least one of message buses; 2) modeling and analyzing a collaboration flow path, and offering a target collaboration flow path by combining with the current situation of the existing ERP (enterprise resource planning) system of each enterprise; 3) under the SOA architecture, integrating the existing ERP system of each enterprise at a resource layer, and integrating the message buses at a service layer; 4) developing a Mediation procedure of an adapter; 5) connecting the ERP of each enterprise to the message buses WESB (websphere enterprise service bus) by the websphere process server deploy in the manner of Web Services; and 6) searching service routing information by the WESB according to the content of an SOAP (simple object access protocol) packet, dynamically routing a request to a corresponding service, and routing the calling operation of an ERP system resource service and analyzing a service data pair by the Mediation of the adapter to realize the clothing sale, the clothing production, and the assist of the collaboration of the production enterprise flow path. The method is used for developing a cross-industry, cross-region and cross-platform integrated collaboration management system in the clothing enterprises.

The invention discloses a time-optimal trajectory planning method and a time-optimal trajectory planning device aiming at a robot operation space. The method comprises the steps of calculating a constrained curve MVC according to a path of the operation space and all-order derivatives of various axial joint displacement of a robot on the time; carrying out path integration in the operation space according to the path of the operation space and the constrained curve MVC; obtaining an integrated path; and the like. The device comprises a memorizer and a processor. According to the time-optimal trajectory planning method and the time-optimal trajectory planning device aiming at the robot operation space provided by the invention, the limitation of the all-order derivatives of the various axial joint displacement of the robot on the time is converted into the limitation of all-order derivatives of arc length parameters, so that a multi-dimension constraint condition is converted into a single-dimension constraint condition, and the whole trajectory planning process is simple and high efficient; various axial joints are always at a saturation condition of constraint, so that the runningefficiency of the robot is improved; and after introducing higher-order derivatives of displacement such as accelerated speed and acceleration jerk of the various axial joints, the shaking of the tail end of the robot can be effectively reduced through an obtained trajectory planning result, and the stability of the robot is improved.

The invention relates to an airplane wake flow circulation inversion method and device based on path integration. The method comprises the following steps of: establishing a plane-coordinate system according to the position of a detection device; detecting the aircraft wake flow according to a detection result of the detection device; acquiring doppler velocity distribution in a plane coordinate system; depending on Doppler velocity distribution, determining a left vortex center coordinate and a right vortex center coordinate generated by aircraft wake flow in the plane coordinate system; according to the left vortex center coordinate and the right vortex center coordinate, eliminating Doppler velocity corresponding to background wind acquiring actual Doppler velocity distribution throughupdating, conducting integration on the actual Doppler velocity distribution in the detection direction of the detection device, and acquiring aircraft wake flow circulation. The aircraft wake flow circulation can be calculated by considering the calculation amount, the accuracy and the robustness.

The invention discloses a dual-path integration based speed model establishment method and system. The method comprises the steps of acquiring target data; determining a plurality of to-be-selected speed models required for processing the target data; obtaining a common imaging point gather corresponding to each to-be-selected speed model, and then obtaining a coherence spectrum and a shift result of the gather corresponding to each to-be-selected speed model; performing two-time weighted stacking on each shift result by utilizing the coherence spectrum corresponding to the shift result to obtain each dual-path integration imaging profile; calculating a coherence value of the common imaging point gather corresponding to each to-be-selected speed model and a coherence value of the path integration imaging profile separately; and based on the coherence values, determining a target speed model in the to-be-selected speed models, wherein the target speed model is used for establishing a speed model. According to the speed model establishment method and system, the high-precision multi-signal detection coherence spectrums are used for path weighting in path integration, so that the influences of high and low speed models on a stacking result are balanced, the quality, namely the accuracy, of speed modeling is improved, and the calculation efficiency is improved.

The invention discloses a brain-imitating navigation method based on an optical flow UWB combination. The brain-imitating navigation method based on the optical flow UWB combination comprises the following steps: dividing a movement region of an intelligent vehicle into a path integration domain and a position cell joint domain; in the path integration domain, measuring the speed of the intelligent vehicle by an optical flow sensor, and carrying out integration on speed to obtain position information; in the position cell joint domain, detecting the speed and position of the intelligent vehicle through UWB, subtracting the speed and position, which are detected by UWB, of the intelligent vehicle from the speed and position which are acquired by an optical flow method, and establishing a model among a speed error, a position error and running time; and when the intelligent vehicle moves to the path integration domain again, carrying out real-time correction on the speed and the positioninformation by using an error model. The brain-imitating navigation method based on the optical flow UWB combination can be used for application scenarios such as navigation and position of the intelligent vehicle which carries out tasks in a certain fixed domain, the autonomy of navigation and positioning can be improved effectively, the degree of intelligence of the intelligent vehicle is increased.

The invention discloses a method and device for realizing high dynamic navigation tracking. The method for realizing high dynamic navigation tracking comprises that: a receiver determines information of choosing a loop as information of choosing a loop I and respectively generates a jth-order I path local carrier signal ad a Q path local carrier signal, wherein the frequency of the Jth order I path local carrier signal is Jth regulation frequency; the receiver performs frequency mixing on a correlated signal and a generated I path local carrier signal and the Q path local carrier signal and performs integration on signals of the I path and the Q path after frequency mixing; the receiver performs frequency discrimination processing on a signal after I path integration and a signal after Q path integration to obtain a jth frequency error; the receiver determines that frequency errors of frequencies from a (j+j0)th frequency to a (j+j0+N)th frequency are smaller than a first locking threshold and determines frequency errors of frequencies from j0th frequency to the (j0+N)th frequency or are smaller than or equal to a first preset threshold, and changes information of choosing a loop into information of choosing a loop II.

The invention relates to a cognitive map construction method based on space and motion joint coding, and belongs to the technical field of robot navigation. A sniffing layer-hippocampus neural circuitin an internal positioning system of the brain of a mammal is considered; a continuous attractor neural network model-spatial memory network is formed by utilizing spatial navigation coding characteristics of position cells, head orientation cells, grid cells and velocity cells, space and motion characteristics of animals are coded at the same time, and path integration is carried out on angularvelocity and linear velocity. In the network, the asymmetric connection weights among the neurons generate system inherent activity peaks capable of spontaneously moving, once speed input exists, theattractor network can steadily integrate the linear speed and the angular speed, and stable robot head direction and position coding is formed. And local view cells are formed through single-camera visual input, so that the accumulative error of path integration can be corrected. According to the invention, a consistent half-meter topological map can be stably constructed by using a single camera.

The invention discloses a mobile meter reading system. The mobile meter reading system comprises meter reading equipment and a plurality of preset meter reading places, wherein each preset fixed meter reading place has fixed GPS (Global Positioning System) geographic information; the meter reading equipment is provided with a wireless communication module and a GPS; each preset meter reading place covers a plurality of remote meters and the covering ranges of all the preset meter reading places do not interfere with each other; each remote meter is provided with a wireless module; the data transmitting end of each wireless module is matched with the data receiving end of the wireless communication module. By virtue of the mobile meter reading system, all the individual remote meters are subjected to path integration to form the plurality of preset meter reading places, all the preset meter reading places have the fixed GPS geographic information, and the mobile meter reading system can perform meter reading after positioning one preset meter reading place with the GPS geography information by virtue of the GPS; the quantity of the remote meters covered by the preset meter reading places is large, the paths of meter reading are reduced and the meter reading intensity of workers is lowered.

The invention discloses a magneto-optic material gap waveguide magnetic surface fast wave direction-controllable photodiode which comprises a light input port, a light output port, two magneto-optic material layers, a dielectric layer and two bias magnetic fields. The photodiode and isolator is composed of the magneto-optic material layers and the dielectric layer, the light input port or the light output port is arranged at the left end of the photodiode and isolator, and the light output port or the light input port is arranged at the right end of the photodiode and isolator; the dielectric layer is arranged in a gap between the magneto-optic material layers, magnetic surface fast waves are arranged on the surfaces of the magneto-optic material layers and the dielectric layer, the bias magnetic fields which are opposite and controllable in direction are arranged at the magneto-optic material layers, and a magnetic surface fast wave photodiode is composed of magneto-optic material gap waveguide. The magneto-optic material gap waveguide magnetic surface fast wave direction-controllable photodiode is simple in structure, high in light transmission efficiency, small in size, convenient to integrate, suitable for large-scale light path integration and wide in application prospect.

The invention discloses a high-accuracy DEK multi-parameter water quality analyzer, which comprises a power pump and a measuring cell. A metering device is arranged under the power pump; a metering sensor is arranged on one side of the metering device; a flow divider valve island is arranged under the metering device; a drain valve is arranged under the flow divider valve island; a reference receiving lamp is arranged on one side of the measuring cell; a projection lamp is arranged on one side of the reference receiving lamp; a pellicle mirror is arranged at a position, under the reference receiving lamp, on one side of the measuring cell; an upper high-pressure valve is arranged above the measuring cell; the measuring cell is arranged on the other side of the metering device; a measuringreceiving lamp is arranged on the other side of the measuring cell; a light path integration ring is arranged in the measuring cell; a lower high-pressure valve is arranged under the measuring cell. The high-accuracy DEK multi-parameter water quality analyzer provided by the invention has the advantages of low cost, high stability and high resolution ratio.

The invention belongs to the field of devices for blood component separation, and relates to a double-layer-flow path integration device for a centrifugal bag type blood component separator. The integration device comprises an upper cover plate, a middle integration body and a lower cover plate from top to bottom, the middle integration body comprises an outer shell, the interior of the outer shell is partitioned into an upper layer and a lower layer by a middle partition plate, anticoagulant pipes, a brine pipe, a blood collection pipe, a blood return pipe, centrifugal pipes, plasma pipes, erythrocyte pipes, PRP pipes, blood platelet pipes, a whole blood pipe and mixing pipes are distributed on the upper layer and the lower layer, a plurality of connectors distributed in a two-layer mode are arranged on the periphery of the outer shell, and the connectors communicate with corresponding pipelines arranged in the upper layer and the lower layer. The integration device has the advantages that the structural design is reasonable, the pipeline structure is simple, in actual use, the number of parts which are required to be matched with the device for use is few, the safety is high, the damage rate is low, and the service life is long; and in the circulation process of blood, air is not prone to mixture, so that the damage probability of a blood donor is reduced.

The invention provides a whole-brain nerve tract projection path automatic tracking method and system, and the method comprises the following steps: data preprocessing; signal extraction; data space partitioning; signal density calculation; projection path calculation; path integration and visualization. The method and system can achieve the automatic tracking of a nerve tract projection path fora whole-brain data set, and a calculation result can highly accord with a signal, and the accuracy is high. A calculated vectorized nerve tract path can quantitatively describe the connection relationof a brain region.

The invention discloses a low-loss type magneto-optical gap magnetic surface quick wave any-angle unidirectional turning waveguide. The turning waveguide comprises a light input port (1), a light output port (2), two magneto-optical material layers (3, 4), a dielectric layer (5) and two bias magnetic fields in opposite directions. The magneto-optical material layers (3, 4) and the dielectric layer (5) for an optical waveguide with a three-layer structure. The three-layer structure is bent in any angle. The bias magnetic fields in opposite directions are arranged at the magneto-optical material layers (3, 4). A gap between the magneto-optical material layers (3, 4) is the dielectric layer (5). The port (1) of the unidirectional turning wave is the light input port and the port (2) is the light output port. The dielectric layer (5) is in the shape of a ring at the bending part of the waveguide. Magnetic surface quick waves are arranged on the surfaces of the magneto-optical material layers (3, 4) and of the dielectric layer (5). The turning waveguide provided by the invention is simple in structure, easy to integrate, low in loss, high in transmission efficiency, suitable for large-scale light path integration and widely applicable to all kinds of optical waveguide designs.