105 results about "Data fusion algorithms" patented technology

Each of multiple probability distribution outputting units computes a probability distribution of a data value detected by the corresponding sensor or algorithm in an image recognition processing or the like. The respective probability distributions of the multiple probability distribution outputting units are given as outputs to a synthetic determination processing unit. Data formats of the outputs to the synthetic determination processing unit can be thereby standardized. Hence, the synthetic determination processing unit is exempted from considering which type of sensor or algorithm each of the outputs is based upon. Even when a sensor or algorithm is added or changed, the same data-fusing algorithm in the synthetic determination processing unit can be uniformly used.

The invention discloses a control system for a four-rotor aircraft. The control system comprises a master controller, an inertial measurement unit, a geomagnetometer, a distance measuring sensor, a GPS module, a camera module, an image processing unit, a motor drive module and four motors, wherein the inertial measurement unit, the geomagnetometer, the distance measuring sensor and the GPS module are connected with the master controller, the camera module is connected with the master controller through the image processing unit, the master controller is connected with the four motors through the motor drive module, and the four motors drive four airscrews of the four-rotor aircraft respectively. The invention further comprises a control method of the control system. The control method involves an attitude control algorithm, a height determination control algorithm and a data fusion algorithm. Data fusion of the inertial measurement unit and a positioning system is adopted for a control structure and the control method, and stability and reliability of the system are improved.

A system and method for monitoring and controlling an aquatic environment thus regulating the aquatic environment and maximizing the stability of the aquatic ecosystem. This system uses sensor and data fusion algorithms to perform system anomaly analysis and predictive failure diagnostics based on the output of sensors associated with the aquatic environment control devices. Based on the system anomaly analysis and / or predictive failure diagnostics, the system may continuously adjust the system parameters to maintain an efficient and stable aquatic environment, alert local or remote monitors of failed or impending failure of control devices. In addition, the system may present the system anomaly analysis and predictive failure diagnostics information at a local or remote location to enable the monitor to address any problems presented in situ.

Provided is a human motion tracking system based on Zigbee / institute of electrical and electronic engineers (IEEE) 802.15.4. A wireless wearable sensing network is covered on nine independent limbs of a person, and a wireless wearable sensing node is adhered onto each independent limb for monitoring free motion of the limb in the three-dimensional space. In each wireless wearable sensing node, an inertia measuring unit which comprises a three-axis acceleration sensor, a three-axis gyroscope and a three-axis magnetic force sensor is used for calculating the position of the independent limb in the three-dimensional space through an on-line data infusion calculation method. In the wearable wireless sensing network, real-time location data can be transmitted to a substation through the low-consumption wireless communication technology Zigbee / IEEE 802.15.4. The substation forwards the obtained wireless wearable sensing network to a computer in real time. Human motion can be reconstructed in the computer according to a forward motion model. The reconstructed human motion can be transmitted to a remote user through the internet for tememedicine. The human motion tracking system totally eliminates ligature between sensor nodes, do not need a light source for assisting and facilitates system expansion and application development by means of the low-consumption wireless operation system.

The invention provides a ship power station fault diagnosis method based on data fusion. The method comprises the steps of respectively adopting different data fusion algorithms on different levels of a fault diagnosis, using a plurality of sensors for detecting fault messages in multiple aspects, conducting stage treatment on multi-source messages, accurately and timely judging the state of a system, and giving the correct judgment on whether the system fails or not and the correct judgment on the fault mode. According to the ship power station fault diagnosis method based on the data fusion, intelligent monitor of a ship power station unit is effectively achieved, the reliability and safety of the operation of the unit are improved, and the phenomena of false alarms, misinformation and information missing are reduced. The sensors respectively monitor parameters of the fault messages and integrate the monitored data to conduct a first-stage fusion to detect layer data, a second-stage fusion is carried out on the data passing through the first-stage fusion, a third-stage fusion is carried out on the data passing through the second-stage fusion, the data after the three-stage fusion are matched with data inside a fault diagnosis knowledge database, and the fault diagnosis result is output.

Disclosed is a quick chain data fusion algorithm based on virtual cell grids. A whole sensor zone is divided into virtual cell grids; data are sent to an appointed node for fusion by the nodes in the cell grids, the chain data fusion process is started in the vertical direction of the cell grids and the data are sent to the Sink node each time by only one node in each chain. The nodes play the role of data fusion in turn for balancing the energy consumption of the nodes. The invention has the advantages that the data transmission chain constructed in the CFDASC algorithm is short, therefore the time delay in data transmission is short. The algorithm has good robustness, that is, each chain transmits data to the Sink node; and has good energy saving property, that is, all the data in the wireless sensor network zone are finally fused into the data transmitted by a plurality of chains, the data account for transmission is largely reduced in the data transmission process and most of the energy is effectively saved.

The invention discloses a ZigeBee-based industrial environment dangerous gas leakage detection system. The invention can be mainly divided into three parts: a wireless sensor network composed of sensor nodes, a data processing center of an ARM platform, and a data storage and monitoring center based on a cloud server. In order to improve the installation flexibility and detection efficiency of the system, and reduce deployment costs and power consumption, ZigBee technology is used to build a wireless sensor network, and the solar charging function based on Texas Instruments' single battery management chip is added to the sensor nodes. The data processing center uses the ARM processor as the control core to realize data reception, leakage judgment, data upload to the cloud server, early warning SMS sending, and data display synchronously. Combined with the homogeneous sensor data fusion algorithm and fragment anomaly detection algorithm, the industrial Fast and accurate alarm of environmental hazardous gas leakage.

The invention provides a full-automatic parking method and system. The full-automatic parking method comprises the steps of acquiring space parking space data and obstacle data recognized by an ultrasonic radar and look-around parking space data and obstacle data recognized by a fisheye camera; performing fusion based on a data fusion algorithm to obtain parking space fusion data and obstacle space fusion data; according to a free track planning algorithm, carrying out planning in combination with the parking space fusion data and the obstacle space fusion data to obtain a parking track; and acquiring moving position information of a current vehicle, and making the vehicle run along the parking trajectory in real time based on a high-precision positioning algorithm and a high-precision track tracking algorithm. In the scheme, the parking track is determined by calculating the parking space fusion data and the obstacle space fusion data obtained by fusion by using the free track planning algorithm. And the vehicle is controlled to run along the parking track by using the high-precision positioning algorithm and the high-precision track tracking algorithm. Precision of parking spacerecognition and automatic parking control is improved so that the method and the system adapt to more parking scenes.

The invention relates to a multi-individual multi-parameter multi-sensing intelligent fireman physical ability early-warning and monitoring device, which mainly comprises heat-insulation fireproof clothes, a monitoring system network, a human body information collecting module, an upper computer display module, a phone APP, and an EnOcean energy collecting unit, wherein the monitoring system network comprises multiple wireless transmission nodes, a relay and a monitoring display terminal; the human body information collecting module is internally provided with a central processing unit, a human body temperature detection unit, a heart rate detection unit, a human body blood pressure detection unit, an ambient gas detection unit, and a falling detection unit. The device has the advantages that the environments and positions of the fire places where multiple firemen are as well as the human body key physiological parameters of the firemen can be monitored in real time, and early warning is sent, so that the harm to the firemen can be avoided furthest; meanwhile, a proper data fusion algorithm is adopted, so that the accuracy and reliability of collecting the human body data of the firemen are improved, and the condition that the best rescue time is lost because of delayed decision is avoided.

The invention discloses a wearable Tai Chi movement gait evaluating and training system based on a cloud platform. The system comprises a gait parameter collecting module, a gait parameter processingmodule, a gait movement function evaluating and training module and a cloud platform module. An attitude sensor, a plantar pressure sensor and a surface electromyography sensor are adopted to acquiregait data in Tai Chi movement of a human body in real time, and transverse and longitudinal comparison of user Tai Chi movement gait training is realized based on a cloud platform data fusion algorithm. Meanwhile, a big data deep mining algorithm is adopted to realize reverse analysis from electromyographic signals to gait specifications. Based on the human-computer interaction function of multiple users, clear digital visual guidance can be provided for Tai Chi teaching, and objective quantitative reference can also be provided for scoring of competitive Tai Chi competition judges. Through the packaging integration design of the devices, the system is portable, wearable, easy to use and reliable, and has good market value.

Provided are an earth satellite attitude data fusion system and method, applicable to an earth satellite space environment to estimate attitude data of a satellite. When the earth satellite attitude data fusion system of the present invention is used to perform the earth satellite attitude data fusion method, the first step is to perform a body rates / quaternion attitude data processing operation. Then, the next step is to perform an attitude / rates data fusion processing operation, wherein an attitude data fusion algorithm module receives a first IAE result data from a first EKF, and a second IAE result data from a second EKF, and performs an attitude / rates data fusion algorithm in a subsystem level to evaluate an attitude estimation IAE performance based on the first IAE result data, and the second IAE result data.

The invention discloses a disaster cellular alarm linkage system based on interval argumentation multi-sensor fusion of particle swarm optimization. The system comprises a fire monitoring alarm linkage system, a public fire alarm system and a property management platform, and is characterized in that the fire monitoring alarm linkage system comprises access control terminals of building doors of all units of a community and cellular user terminals, and emergency alarm switches are arranged on both a real estate management platform and the fire monitoring alarm linkage system; the public fire alarm system comprises a CO sensor, a smoke sensor, a temperature sensor and an alarm controller which are arranged in a resident of a community, and a fire judgment data fusion algorithm based on IPSO optimization interval demonstration multi-sensor fusion is arranged in the alarm controller. According to the present invention, the fire monitoring and alarming linkage system is matched with the public fire alarming system to work, the automatic fire alarming and the manual fire alarming are integrated, and the uncertainty of the sensor measurement data is eliminated based on the fire judgment data fusion algorithm based on IPSO optimization interval demonstration multi-sensor fusion.

The invention provides a tracking filtering method based on data fusion. The method comprises the following steps: firstly, calculating each Sagna sampling point and a corresponding weight value; propagating each Sigma sampling point through a nonlinear function; obtaining a mean value estimation and a covariance estimation of an observation vector, decomposing a nonlinear system into nonlinear filters with a plurality of subsystems, carrying out state estimation based on local filtering of an insensitive Kalman filtering method, and obtaining Kalman filter state estimation based on a data fusion algorithm after calculating the weight of each local filter. According to the obtained filtering estimation value, the estimation precision of the filtering algorithm can be improved, and the reliability of a filtering system is also improved.

The invention discloses a data fusion algorithm for accurate positioning based on GNSS, vision and IMU. A GNSS (Global Navigation Satellite System) positioning technology is easily limited by external observation conditions, the satellite signal receiving capability is sharply weakened in sheltered areas such as viaducts and urban canyons, and the requirements of positioning and navigation cannot be met. Pure vision position estimation cannot solve the scale problem, and a matching relation cannot be established in a weak texture scene and a rapid motion scene, so that tracking is easy to lose; the IMU can reflect dynamic changes in a short time (millisecond), but accumulative errors of the IMU can be continuously increased under long-time operation (second). According to the complementarity among different sensor data sources, the global positioning precision of the GNSS is improved by using local data sources (vision and IMU), local accumulative errors are eliminated by using the global data sources of the GNSS, and short-term high-precision global positioning is kept by depending on local vision and IMU data when satellite signals are shielded. And the performance and robustness of the positioning navigation system are integrally improved.

A satellite attitude data fusion system and method is disclosed, applicable to the earth satellite environment to estimate attitude data of the satellite. When the satellite attitude data fusion system of the present invention is used to perform the satellite attitude data fusion method, the first step is to perform a body rates quaternion attitude data processing operation. Then, the next step is to perform an attitude / rates data fusion processing operation, wherein an attitude data fusion algorithm module receives the first IAE result data from the first EKF, and the second JAE result data from the second EKF, and performs an attitude / rates data fusion algorithm in a subsystem level to evaluate an attitude estimation JAE performance.

The invention relates to a multi-mode intelligent inertial navigation sensing system and a data processing method thereof. The multi-mode intelligent inertial navigation sensing system comprises a sensor module, a processing module and an interface module, wherein the sensor module is connected with the processing module, and the processing module is connected with the interface module; the processing module comprises an engine kernel module and a selector, the engine in the engine kernel module comprises two or more engine modes, every engine mode corresponds to one data fusion algorithm; thesensor module is used for measuring acceleration, angular speed, axial geomagnetic component, external temperature, external air pressure and / or positioning information; the processing module is usedfor receiving measured data, and according to user or system configuration, performing data fusion on a part of or all the measured data from the sensor module according to the data fusion algorithmcorresponding to one of the engine modes; the selector selects and sends one data steam from data steams fused according to different engine modes to the interface module; the interface module is usedfor outputting processed data streams to an external module or a host; the processing module is embedded with a real-time operation system, in which a plurality of threads are arranged, and the plurality of threads are rounded at a preset time slice interval and used for controlling execution of different tasks.

The invention discloses a data fusion algorithm and system for multi-band radar target recognition. The system comprises: a data acquisition module used for acquiring electromagnetic echoes of an identified target body; a feature extraction module used for extracting echo data of the electromagnetic echoes; a data fusion module used for fusing the plurality of extracted feature data into feature data F by using a specific fusion algorithm; a category identification module used for classifying the identified target body according to the fused feature data F; and an image processing module usedfor converting the fused feature data F into a visual image by utilizing a technology and an image conversion principle. A radar system is used for emitting multiband electromagnetic waves, then linear discriminant analysis is carried out on multiband electromagnetic echoes of the identified target body to extract the features of the target, the features of the multiband electromagnetic echoes arefused, the fused features are classified through adoption of a classification algorithm, and therefore, the categories of the different identified target bodies can be obtained.

The invention discloses an Internet of Things multi-dimensional data fusion system and method. The method comprises the steps that a data supply direction data fusion sharing center provides Internetof Things data or publishes a data service, a data demand direction data fusion sharing center submits a data demand, and the data fusion sharing center integrates an Internet of Things spatio-temporal data description model and a data fusion model; the Internet of Things spatio-temporal data description model takes a Beidou grid code as a spatial domain identifier, and performs mapping processingon Internet of Things data of different data standards to complete data resource normalization; and the data fusion model uniformly describes detailed features of data required by a data demander, extracts feature vectors based on a geographic space information domain according to the space-time data demand description of the Internet of Things, and performs data fusion according to the data features by taking Beidou grid codes as space domain identifiers and adopting a data fusion algorithm. According to the invention, the fusion efficiency of the Internet of Things spatio-temporal data in the geographic space information domain is improved.

The invention discloses a water surface unmanned surface ship path tracking method based on intelligent predictive control, aiming at the influence of under-actuated performance and time-varying wind wave flow disturbance of an unmanned surface ship on the maneuverability of the unmanned surface ship. The method aims at solving the problems that during the power-on period, data deviation is large, and during high-speed navigation, data drifting occurs to the gyroscope. A data fusion algorithm is provided, a complementary filtering algorithm is improved by setting threshold increment constraint, and the navigation data precision is improved. The problems that an unmanned ship is prone to being interfered in a severe environment, a ship body shakes greatly, and steering is advanced or lagged when the unmanned ship enters a receiving circle are solved. The invention provides a self-adaptive sight line method, and the radius of a receiving circle is automatically adjusted, so that the unmanned ship has sufficient time to adjust the course, steers gently, and is prevented from deviating from a target path. The method aims at solving the problem that the solving precision is not high when a traditional objective function is used in a model prediction controller. The invention provides an improved artificial fish swarm algorithm for global optimization. And the anti-interference capability of the control system and the convergence speed of the algorithm are improved.

The invention relates to a fault diagnosis method for a horizontal attitude angle of an unmanned aerial vehicle, and the method comprises the steps: obtaining two groups of unmanned aerial vehicle horizontal attitude angle sensor information from a horizontal attitude angle hardware dual-redundancy sensor, carrying out the cross combination of the two groups of sensor information, and generating four-path gyroscope and accelerometer combination information; respectively resolving the four paths of combined information into four-redundancy horizontal attitude angle information through a data fusion algorithm; and carrying out information diagnosis on the four-redundancy horizontal attitude angle through motion over-limit judgment and redundancy voting, and carrying out fault diagnosis on the horizontal attitude angle sensor according to a four-redundancy horizontal attitude angle information diagnosis result. According to the method, a voting fault diagnosis mechanism which can be carried out only by existing redundancy hardware redundancy can be completed, wider sensor faults can be effectively diagnosed, the hardware structure of the system is simplified, the requirements for theinstallation space and the loading capacity are reduced, the system cost is reduced, and the reliability of the flight control system is improved.

The invention discloses a head posture recognition method based on data fusion. The method comprises the steps of firstly, collecting a precise head X-axis rotation angle and a precise head Y-axis rotation angle during the head rotation process; acquiring a Z-axis rotation angle during the head rotation process, wherein the Z-axis rotation angle comprises a gyroscope Z-axis rotation angle and an infrared feature point Z-axis rotation angle, and the gyroscope Z-axis rotation angle and the infrared feature point Z-axis rotation angle are used as the input quantities of a data fusion algorithm; obtaining the accurate head Z-axis rotation angle through data fusion, and finally outputting the head posture parameters composed of the accurate head X-axis rotation angle, the accurate head Y-axis rotation angle and the accurate head Z-axis rotation angle. According to the method, the measurement cost is low, the drift distance of the measurement result of a gyroscope can be corrected by adopting the infrared scanning measurement result through the fusion algorithm, and the purpose of high-precision measurement is achieved on the basis of keeping the original advantages of the two measurement methods.

The invention discloses a software platform applying a CART algorithm to realize electric energy meter detection data management. The invention relates to the technical field of electric energy metering detection. A cloud big data management platform is adopted, and information in mass data is fully mined to realize big data analysis, mining and processing, a cloud computing technology is adoptedto process hundreds of millions of data types within a few seconds, the classification capability and calculation precision of different data are improved through a CART algorithm and an improved CARTalgorithm, the construction of a distributed hierarchical model of big data can be realized through the CART algorithm, and the data identification capability is simplified. According to the invention, a data fusion algorithm is adopted to reflect various data of the electric energy meter with diversified information representation forms, huge information quantity and complicated information relationship timely, accurately and reliably, through a BP neural network algorithm model, the error of a direct leading layer of an output layer is estimated by using an output error, and the learning algorithm precision is improved.

The embodiment of the invention discloses a motion trail determination method, operation equipment, a device and a storage medium, and the method comprises the steps: obtaining the point cloud data of a current frame of a to-be-operated area and the global map of the current frame, the point cloud data of the current frame being the point cloud data collected by the operation equipment in the operation area, and the global map of the current frame being the point cloud data of the current frame being the point cloud data collected by the operation equipment in the operation area; the global map of the current frame is determined based on point cloud data of a previous frame of the current frame; determining first pose change information of the operation equipment based on the point cloud data of the current frame; determining second pose change information of the operation equipment based on the point cloud data of the current frame and a global map of the current frame; and based on a preset data fusion algorithm, performing data fusion on the first pose change information and the second pose change information to obtain pose information of the current frame of the operation equipment.

The invention discloses a GIS-based refined electric power weather forecast data display method and system. The method comprises the steps of 1, collecting electric power weather information; 2, cleaning the collected data and storing the data into a data storage center; 3, according to the selection of the user, searching whether a generated primitive meeting the selection of the user exists on the server, if so, directly calling the primitive and superposing the primitive on the GIS map to complete data display, and if not, entering the step 4; 4, extracting data in the corresponding data storage center according to the user selection area for data fusion; 5, generating corresponding primitives according to the fused data and outputting the primitives to a server; and 6, superposing the generated primitives on the GIS map. The system disclosed by the invention is based on the method. According to the method and the system, a data fusion algorithm and a method for repeatedly utilizing the primitives drawn before can greatly improve the display effect and the display efficiency of the refined numerical forecasting data on the GIS map, and the user experience is effectively improved.

The invention discloses a seismic attribute multi-scale fusion method and device, electronic equipment and a medium. The method comprises the following steps: performing spectral decomposition processing on three-dimensional superposition seismic data through generalized S transformation to obtain a multi-scale seismic data volume; calculating multi-scale attribute data according to the multi-scale seismic data volume; comparing the geological information with the multi-scale attribute data, and determining a source attribute of data fusion; and performing seismic attribute image fusion through a visual saliency image fusion algorithm to obtain a fused attribute image. A data fusion algorithm and a generalized S-transform spectrum decomposition method are combined to perform seismic attribute slice image fusion, detail information of source attribute slices can be reserved, and abnormal signals in the source attribute slices can be highlighted.

The invention provides a user-transformer relation abnormity identification method based on a spatial position, and the method comprises the steps: achieving the consistency matching of transformers and power customers of production, GIS and marketing through a data fusion algorithm, and constructing a static user-transformer relation; acquiring geographic coordinate data of the transformers and power supply range data of the transformers, and acquiring geographic coordinate data of the power customers at the same time; and calculating the distance between the transformers and the power customers to which the transformers belong through a spatial position coordinate distance calculation method, judging the calculated distance by taking the power supply range of the transformers as a reference, and identifying a suspected abnormal user-transformer relationship. According to the invention, the accuracy of power grid user-to-change relation identification is improved, the user-to-change relation accuracy is improved at the lowest cost, the quality of power grid basic data is effectively improved, the user-to-change relation is updated in time, and the accuracy and convenience of user-to-change relation verification are improved to a certain extent.

The invention discloses an MARG attitude calculation method with motion acceleration compensation, and belongs to the field of navigation. The method comprises the following steps of: firstly, carrying out attitude one-step prediction by utilizing a gyroscope, then, detecting an accelerometer and a magnetometer by utilizing a one-step prediction value, and finally, executing a corresponding data fusion algorithm according to different detection results. By using the method to estimate the attitude of the MARG sensor, the motion acceleration and the magnetic field interference of the carrier are effectively detected; directed at the dynamic characteristics of the motion carrier, sliding window data is introduced to fit a motion acceleration change rule and carry out corresponding model compensation; and an analytical solution of a gravity-magnetic (acceleration and magnetic induction intensity measured values) vector pair under the Wahba problem is derived so as to accelerate the attitude solving speed and remarkably improve the MARG attitude estimation efficiency.

The invention relates to the technical field of data processing, in particular to a database fusion association system, which comprises a plurality of sensors, a real-time database, a historical database, a fusion management module and a filtering tracking module; the fusion management module comprises a calibration module, a correlation module, an identification module and an estimation module; the real-time database is used for data entry and maintenance, real-time data updating, historical database communication and application program communication. The fusion model and the database are combined into a unified whole, the database is designed from top to bottom, an integrally optimized database strategy is achieved, great superiority is achieved, the performance and the reliability of afusion system can be guaranteed, the working efficiency of the database can be greatly improved, and starting from the requirement of data fusion for the database, a database model supporting a datafusion algorithm is established, and a special database of a data fusion system is developed on the basis of the model, so that the real-time performance is met, and the storage amount of data in a memory is reduced.

The invention provides a meadow environment monitoring system based on the Internet of Things, and belongs to the technical field of ecological environment monitoring. According to the meadow environment monitoring management information system based on the Internet of Things, the Internet of Things technology is applied to the field of grassland meadow environment monitoring, the NB-IoT wireless communication technology and the image processing technology are combined and applied to the system, and management of meadow environment monitoring information is scientifically achieved. According to different weights of various environmental factors in grassland information for herdsman grazing, characteristics of the factors are analyzed and compared, a multi-sensor data fusion algorithm is adopted, grassland environment data are analyzed and processed, an evaluation matrix is constructed, and grassland environment grades are scientifically evaluated.

The invention discloses an antenna monopulse performance evaluation method under a complete machine condition, and the method comprises the following steps: firstly, positioning the position of a target relative to a radar, then correcting the deviation between the zero depth of a difference channel of an antenna of the radar and the main beam of a sum channel, obtaining the normalization information of the difference channel and the sum channel of the antenna at different pointing angles, analyzing the normalized information to obtain an evaluation result of the monopulse performance of the antenna. According to the method, the antenna is controlled to move through a control mechanism of the radar according to the condition of the whole machine, normalized information at different angles is obtained by adopting a signal power estimation algorithm and a data fusion algorithm, and the monopulse performance of the antenna is rapidly evaluated.