Product
Patsnap Eureka
For R&D, Patsnap Eureka makes reading and utilizing patents & technical documents easy.
Patsnap Eureka AIR
Designed for self-driven R&D workflows. Generate viable solutions, solve complex R&D challenges, empower your innovation with AI.
Patsnap Eureka Materials
Designed for material experts only. Revolutionize your material R&D, from search, analyze, to developing new materials.
Feature
TechResearch
Generate reliable direction feasibility study reports for your R&D in just a few steps.
TechSeek
Discover and master advanced knowledge NOW. Basics, ideas, possibilities, all at once.
TechMind
As an expert in R&D Theories, TechMind can generates customized viable solutions instantly.
TechRisk
Analyze your overall solution with one click, know your potential R&D risks in advance.
TechMonitor
Get weekly tech updates, stay abreast of the latest tech innovations and key insights.
Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
70 results about "Behavior learning" patented technology
Efficacy Topic
Property
Owner
Technical Advancement
Application Domain
Technology Topic
Technology Field Word
Patent Country/Region
Patent Type
Patent Status
Application Year
Inventor
BEHAVIORAL LEARNING a process in which experience with the environment leads to a relatively permanent change in behavior or the potential for a change in behavior.
Driving behavior prediction method and apparatus
ActiveUS20080120025A1Improve reliabilityImprove forecast accuracyVehicle testingInstruments for road network navigationObject pointAlgorithm
A driving behavior prediction apparatus includes, for accurately predicting a driving behavior, a position calculation unit for a subject vehicle position calculation, a route setting unit for setting a navigation route, a distance calculation unit for calculating a distance to a nearest object point, a parameter storage unit for storing a template weighting factor that reflects a driving operation tendency of a driver, a driving behavior prediction unit for predicting a driver's behavior based on vehicle information and the template weighting factor, a driving behavior recognition unit for recognizing driver's behavior at the object point, and a driving behavior learning unit for updating the template weighting factor so as to study the driving operation tendency of the driver in a case that a prediction result by the driving behavior prediction unit agrees with a recognition result by the driving behavior recognition unit.
Owner:DENSO CORP
Automatic capture of diagnostic data based on transaction behavior learning
ActiveUS8938533B1Digital computer detailsMultiprogramming arrangementsDiagnostic dataApplication software
The present technology may determine an anomaly in a portion of a distributed business application. Data can automatically be captured and analyzed for the portion of the application associated with the anomaly. By automatically capturing data for just the portion associated with the anomaly, the present technology reduces the resource and time requirements associated with other code-based solutions for monitoring transactions. A distributed business application is monitored to detect an anomaly based the learned behavior of the business application. The behavior may be learned for each of one or more machines which implement a distributed business transaction. The present system may automatically collect diagnostic data for one or more business transactions and / or requests based on learned behavior for the business transaction or request. Data for a distributed business application transaction may be collected by sampling one or more threads assigned to handle portions of the distributed business transaction. Data regarding the distributed transaction may then be reported from agents monitoring portions of the distributed transaction to one or more controllers and assembled by one or more controllers into business transactions.
Owner:CISCO TECH INC
Virtual image and real scene combined stage interaction integrating system and realizing method thereof
ActiveCN103578135ARealize the stage effectInput/output for user-computer interactionCharacter and pattern recognitionLearning unitSimulation
The invention discloses a virtual image and real scene combined stage interaction integrating system and a realizing method thereof. The system comprises a scene behavior acquiring unit, a behavior learning unit, a performance intention analyzing unit and a scenery unit, wherein the scene behavior acquiring unit acquires training data or scene data for generating performance behaviors of performers; the behavior learning unit uses a support vector machine classifier to learn the training data to obtain preset behaviors of the performers, and uses the scene data to obtain scene behaviors of the performers; the performance intention analyzing unit obtains performance intentions of the performers according to corresponding relations between the present scene behaviors of the performers and corresponding preset behaviors; the scenery unit stores a plurality of scenery frames corresponding to the performance intentions of the performers, and chooses corresponding scenery frames to project on a stage according to the performance intentions. The system can realize the interaction between the performers and the stage scenes to achieve the stage effect of combining virtual images with real scenes.
Owner:恒德科技有限公司
User behavior based cross-cloud authentication service method
ActiveCN104202339AImprove securityMake up for one-sidednessTransmissionBehavioral analyticsInternet privacy
The invention provides a user behavior based cross-cloud authentication service method. The method includes the steps of collecting user behavior data, performing user trust management based on user behaviors, and providing a user behavior based identity authentication service. In order to achieve mutual trust between users and service providers in the cloud computing mode, a handling scheme is provided in terms of user behavior analysis, user behavior learning, user behavior authentication and the like, and the user behavior based cross-cloud authentication service method is established. Behaviors of users under different cloud service platforms are collected, overall trusts of the users are estimated, and accordingly, the problem that the users have different trusts in different public clouds is solved and reliable basis is provided for access control for users crossing multiple cloud service platforms. Meanwhile, in order to solve the problem that user identities are trustable but behaviors are not necessarily trustable, a solution for user behavior identity authentication according to the behaviors of users under different cloud service platforms is provided, and a trusted safeguard infrastructure is provided for a cloud application system.
Owner:GUANGXI UNIV
Driving behavior prediction method and apparatus
ActiveUS7805240B2Improve accuracyImprove reliabilityVehicle testingInstruments for road network navigationAlgorithmObject point
A driving behavior prediction apparatus includes, for accurately predicting a driving behavior, a position calculation unit for a subject vehicle position calculation, a route setting unit for setting a navigation route, a distance calculation unit for calculating a distance to a nearest object point, a parameter storage unit for storing a template weighting factor that reflects a driving operation tendency of a driver, a driving behavior prediction unit for predicting a driver's behavior based on vehicle information and the template weighting factor, a driving behavior recognition unit for recognizing driver's behavior at the object point, and a driving behavior learning unit for updating the template weighting factor so as to study the driving operation tendency of the driver in a case that a prediction result by the driving behavior prediction unit agrees with a recognition result by the driving behavior recognition unit.
Owner:DENSO CORP
Method for detecting exception target behavior in intelligent vision monitoring
The invention discloses a method for detecting abnormal target behavior in intelligent visual surveillance. The method includes two processes, namely, target behavior learning and target behavior detection. The target behavior obtains the abnormal threshold value Epsilon of every sub-track distribution mode and the abnormal threshold value Epsilon of every track distribution mode of target trajectory which describes occurred target behavior; the target detecting processes calculate the distance from the target trajectory which describes the target behavior to be measured to every sub-track distribution mode and every track distribution mode so as to compare the distance with the abnormal threshold value and judge whether the target behavior is abnormal. The method uses the local and integral behavior modes of self-organizing map network acquisition target respectively, which can detect not only the integral abnormal behavior of a target, but also the local abnormal behavior of the target in moving process.
Owner:ZHEJIANG UNIV
Multi-target tracking method and system based on behavior learning
ActiveCN105957105AKeep trackReduce tracking error rateImage enhancementImage analysisProbit modelMulti target tracking
The invention provides a multi-target tracking method and system based on behavior learning. The method comprises steps of: acquiring and detecting a target video sequence, and acquiring the size and the positional information of a tracked target candidate frame depending on a detection result; modeling a multi-target real-time tracking problem and creating a production probability model of the multi-target real-time tracking problem; for the global conditional probability items in the production probability model, performing offline training on a correctly-marked training set in order to perform global behavior prediction applicable to various scenarios, and for local conditional probability items in the production probability model, on-line training local behavior prediction for each target in real time by using the tracking data of the target prior to a current frame; obtaining the behavior prediction of the target in combination with the global behavior prediction and the local behavior prediction, and tracking the multiple targets depending on the predicted target behavior. The method and the system may keep a tracking rate while tracking the multiple targets and may significantly reduce a tracking error rate.
Owner:TSINGHUA UNIV +1
Intelligent robot movement comparing method and robot
InactiveCN106022305AImprove experienceImprove execution efficiencyProgramme-controlled manipulatorCharacter and pattern recognitionComputer visionBehavior learning
The invention discloses an intelligent robot movement comparing method and a robot. The method disclosed in the invention comprises the following steps: current movement behaviors of an interaction object are collected, standard movement behaviors corresponding to the current movement behaviors are obtained, deviation data is obtained after the current movement behaviors are compared with the standard movement behaviors, and multi-modal motion adjusting instructions are generated and output according to the deviation data. According to the method disclosed in the invention, the intelligent robot is enabled to finish movement comparing operation, a user can be instructed to learn and imitate the movement behaviors, execution efficiency of movement behavior learning and imitating can be greatly improved, an application scope of the intelligent robot can be expanded, and user experience of the intelligent robot can be improved.
Owner:BEIJING GUANGNIAN WUXIAN SCI & TECH
User behavior learning method based on PST in wireless network
InactiveCN103338467AImprove quality of experienceImprove network performanceMathematical modelsData switching networksQuality of serviceWireless mesh network
The invention provides a user behavior learning method based on a PST (Probabilistic Suffix Tree) in a wireless network. The method comprises the steps that a user behavior based on a business is divided into no business, a session business, an interaction business and a streaming media business according to different network QoS (Quality of Service) requirements of the business in the wireless network; a quaternary user behavior state sequence is generated; and an appropriate network resource can be selected for providing a high-quality service for a user according to a predicted business behavior by learning to construct the PST to train the user behavior sequence, and adopting the possible user behavior during a time period of a variable-length Markov model prediction. The method can improve the accuracy of user behavior prediction, is simple, and convenient to realize, and has a very good application prospect.
Owner:NANJING UNIV OF POSTS & TELECOMM
Driver longitudinal car-following behavior model construction method based on deep reinforcement learning
ActiveCN112201069ASolving decision problems on continuous action spacesRealize verificationRoad vehicles traffic controlNeural architecturesDriver/operatorNetwork architecture
The invention discloses a driver longitudinal car-following behavior model construction method based on deep reinforcement learning, and belongs to the field of automobile intelligent safety and automatic driving. The method comprises the steps of based on the actual road working condition of China, collecting vehicle state information and surrounding environment information, meeting the road characteristics of China, of a driver in the vehicle driving process, counting and analyzing the collected data, and giving behavior characteristics and influence factors of the driver in the car following driving process; determining reference information representing actions taken by the driver at a certain moment, and establishing a mathematical model for describing the iterative relationship of the driver car-following behavior state; designing a neural network structure of the driver longitudinal car-following behavior model based on the competitive Q network architecture; designing a driverlongitudinal car-following behavior learning process of a neural network based on the competitive Q network architecture; and designing a training method of the driver longitudinal vehicle following behavior model based on deep reinforcement learning. The car following behavior characteristics of the driver under different working conditions can be accurately described, and the reproduction capability of the car following behavior of the driver is achieved.
Owner:XIAMEN UNIV
Intelligent vehicle automatic driving control method and system
ActiveCN110347155AImprove riding performanceImprove comfortPosition/course control in two dimensionsVehiclesOnline learningSelf adaptive
The invention relates to an intelligent vehicle automatic driving control method and system and belongs to the field of intelligent driving technologies. Through the control method and system, the problem that online learning cannot be well completed self-adaptively in existing automatic driving is solved. The intelligent vehicle automatic driving control method comprises the steps that a global travel planning path of an intelligent vehicle is acquired, the global travel planning path is decomposed into different travel segments, and the different travel segments are dived into correspondingdriving subtasks according to a driving task; according to the current driving subtask, environment information corresponding to the driving subtask is collected, and the environment information is processed to obtain a state quantity corresponding to the driving subtask; the state quantity is input into a trained driver behavior learning model, and an action quantity is output in real time through processing by use of the driver behavior learning model; and a bottom control quantity of the intelligent vehicle is obtained according to the action quantity, and the intelligent vehicle is controlled to run based on the bottom control quantity. Through the control method and system, self-adaptive online learning of automatic driving of the intelligent vehicle is realized.
Owner:BEIJING INSTITUTE OF TECHNOLOGYGY
Robot obstacle avoidance behavior learning and target searching method based on deep belief network
ActiveCN107818333AHigh cost feasibilityGood automatic obstacle avoidance learning abilityProgramme-controlled manipulatorForecastingDeep belief networkAngular velocity
The invention discloses a robot obstacle avoidance behavior learning and target searching method based on a deep belief network. The robot obstacle avoidance behavior learning and target searching method based on a deep belief network includes the steps: controlling a robot to realize obstacle avoidance in the environment, acquiring the color, the deep image data, and the linear velocity and the angular velocity corresponding to a mobile matrix of the robot at the same time, and based on the data, constructing a network model of implementing automatic obstacle avoidance behavior learning of the robot; during the automatic target searching process of the robot, randomly searching the target in the environment through the automatic obstacle avoidance function, and once searching the target,directly approaching the target, wherein if the obstacle appears during the approaching process, the robot can avoid from the obstacle and perform path planning again, and if the target is lost duringthe approaching process, the robot randomly searches again; and continuously repeating the above process until the robot arrives at the target position. The robot obstacle avoidance behavior learningand target searching method based on a deep belief network only uses a single RGB-D camera to realize path planning and target searching with the automatic obstacle avoidance function, and has higherfeasibility and practicality in the cost aspect and the application aspect.
Owner:爱极智(苏州)机器人科技有限公司
Behavior status switching mode identification method of application program for Android-based smart phone
InactiveCN102647409ADetect security risksSubstation equipmentTransmissionData centerHide markov model
The invention discloses a behavior status switching mode identification method of an application program for an Android-based smart phone and belongs to the field of phone safety. The invention particularly relates to a behavior status switching mode identification method of an application program, in order to solve the problems that whether application is infected with a virus can not be detected out in the prior art and hidden potential safety hazards can not be effectively detected. The identification method comprises the following processes: a system monitoring module intercepts, filters and switches a status, records a generated composite status sequence and sends the composite status sequence into a data center module; a behavior learning module reads a sequence to be learned and an initial model, repeated learning is finished by the convergence criteria, and the result is stored in the data center module; and a detection strategy is set by a behavior detecting module, if application is a known type, an HMM (Hidden Markov Model) is selected for carrying out once complete evaluation; and if the application is an unknown type, whether an unsafe behavior exists is detected, all HMMs representing malicious behaviors are utilized for carrying out complete evaluation for multiple times, and then a result is output. The identification method is used for safety detection of the smart phone.
Owner:HARBIN INST OF TECH
Personalized environmental perception privacy protection method based on Android
ActiveCN106650485AMeet individual privacy needsUnderstanding behaviorDigital data protectionUser needsPersonalization
The invention provides a personalized environmental perception privacy protection method based on Android. According to the method, in order to meet personalized privacy demands which are changed due to the environment of a user, a system needs to percept the service environment and the user habit of mobile application, such as time, position, user interaction conditions, application types, concrete behaviors and the like, conduct deep semantic parsing and user behavior study on the environment in which mobile equipment is located, timely adjust a delegated strategy, and meet the requirements of the user. The invention further provides an automatic and controllable privacy management mode at the same time. The user can express his or her privacy requirements by setting a privacy strategy, delegated strategies are set aiming at specific contexts and privacy requests, and automatic and controllable privacy protection with fine grit is achieved.
Owner:PANSOFT
Method and apparatus for learning behavior in software robot
Disclosed is a method and apparatus for learning behavior in a software robot. The method includes detecting a kind of an object in cyberspace related to a kind of presently manifested action, and a kind and the variation of at least one state among percept states or emotional states preset so as to change in relation to the kind of the action; finding episodes respectively corresponding to each of one or more objects in the cyberspace, each of one or more emotional states and each of one or more percept states, respectively defined in the software robot, a kind of an object in cyberspace related to the detected kind of the action among multiple episodes for responding a combination of kinds of respective one or more actions and for storing variation related to each state, and a kind of at least one state among percept states or emotional states preset so as to change in relation to the kind of the action; using variation stored in response to the found episode and variation generated in response to the manifested action, and calculating a representative variation; and storing the representative variation as a variation of the found episode.
Owner:SAMSUNG ELECTRONICS CO LTD
Automatic driving algorithm training system and method
ActiveCN112180921AImprove accuracyCharacter and pattern recognitionPosition/course control in two dimensionsDriver/operatorMachine control
The invention discloses an automatic driving algorithm training system and method. The system comprises a vehicle computing device, a decision device and a server. In a driving test scene, the vehiclecomputing device generates an automatic driving instruction according to vehicle driving data and environment data and sends the automatic driving instruction to the decision device, the decision device can also receive a driver control instruction while receiving the automatic driving instruction, compares the automatic driving instruction with the driver control instruction, indicates that theautomatic driving is inaccurate if the automatic driving instruction is inconsistent with the driver control instruction, and controls the to-be-tested vehicle by adopting the driver control instruction. The server determines the machine control data and the driver control data and performs behavior comparison on the machine control data and the driver control data so as to train the automatic driving algorithm, and the effect of learning the driving behavior from the machine to the person is achieved through training, so that a plurality of persons do not need to be arranged on the vehicle; and the accuracy of automatic driving target recognition is improved through test training.
Owner:ANHUI JIANGHUAI AUTOMOBILE GRP CORP LTD
System and method for preventing attacks by self-learning in cloud environment
ActiveCN106570400AReduce difficultyReduce complexityPlatform integrity maintainanceMachine learningSecurity managementBehavior rule
The invention relates to the cloud security management technology, and aims to provide a system and method for preventing attacks by self-learning in a cloud environment. The system for preventing the attacks by self-learning in the cloud environment comprises a management end and a monitoring end; the management end and the monitoring end use a C / S architecture; the management end is deployed at a Server end, and the monitoring end is deployed in each host in the cloud environment. The system for preventing the attacks by self-learning in the cloud environment provided by the invention forms a behavior rule base after host behavior learning, which greatly reduces the difficulty and complexity of host attack prevention in the cloud environment, thereby improving the efficiency of the host attack prevention in the cloud environment; and besides, the system can flexibly adjust the behavior rule base, which greatly improves the accuracy of the host attack prevention in the cloud environment.
Owner:HANGZHOU ANHENG INFORMATION TECH CO LTD
Client virtual machine memory dynamic isolation and monitoring method and system
ActiveCN110058921ADefense stealingDefense against arbitrary overwrite attacksHardware monitoringSoftware simulation/interpretation/emulationVirtualizationPage table
The invention provides a client virtual machine memory dynamic isolation and monitoring method and system. A user request processing module, an extended page table exception interception and processing module, a virtual machine monitor interaction module and an extended page table communication module are respectively deployed in a client virtual machine; and an extended page table management module and an application behavior learning module are respectively deployed in a virtual machine monitor. The kernel address space isolation technology of the virtual machine is utilized to effectively defend stealing of kernel key data and random overwriting attacks of a memory. Even if a vulnerability of a certain module in the kernel is grasped, protected key data cannot be randomly tampered. Different isolation execution environments are provided for different modules, so that the kernel is protected from being stolen by unauthorized data and memory overwriting attacks. A virtualization mechanism provided by hardware is used to accelerate the switching function of the extended page table, so that the performance overhead is reduced.
Owner:SHANGHAI JIAO TONG UNIV
Intelligent charging distribution adjusting method, device and equipment for electric vehicle, and medium
ActiveCN113103905AOrderly and normal chargingMeet charging needsCharging stationsElectric vehicle charging technologyNetwork modelElectric vehicle
The embodiment of the invention discloses an intelligent charging distribution adjustment method for an electric vehicle, and the method comprises the steps: obtaining the feature information of an access vehicle through a charging pile, and inputting the feature information of the access vehicle into a vehicle charging behavior learning module; wherein the vehicle charging behavior learning module comprises a plurality of neural network models; determining a neural network model corresponding to the accessed vehicle by the vehicle charging behavior learning module so as to obtain predicted charging time and initial charging power of the accessed vehicle; monitoring the charging power of the charging pile system in real time to obtain an operation coefficient for ensuring the normal operation of the charging pile system; and when the operation coefficient is greater than a preset overload value or less than a preset low load value, determining a charging power value needing to be adjusted by the accessed vehicle according to a preset parallel system so as to realize charging distribution of the accessed vehicle.
Owner:山东大卫国际建筑设计有限公司
Robot behavior learning model based on utility differential network
InactiveCN102063640ASolve access difficultiesSolve the problem of difficult knowledge acquisitionBiological modelsOffline learningDecision networks
The invention relates to a robot behavior learning model based on a utility differential network, which comprises a utility fitting network unit, a differential signal calculating network unit, a confidence evaluating network unit, an action decision network unit, an action correcting network unit and an action executing unit. The model realizes the offline learning process and the online decision process. The utility fitting network unit calculates and obtains a utility fitting value of a state after action is executed; the differential signal calculating network unit is used for calculatinga differential signal; the confidence evaluating network unit outputs the confidence obtained by calculating to the action correcting network unit; the action decision network unit outputs an action selecting function; and the action correcting network unit corrects the action selecting function by utilizing confidence, calculates a probability value selected by each action and outputs the actionwith largest probability to the action executing unit for executing. The invention can more favorably ensure the completeness of a robot for obtaining environmental knowledge and more favorably ensure the timeliness and effectiveness of robot behavior decision.
Owner:BEIHANG UNIV
Cross-platform data matching method, device, computer device and storage medium
ActiveCN107688605AData matching implementationData processing applicationsDatabase queryingData matchingComputer terminal
The invention relates to a cross-platform data matching method, which particularly includes steps of receiving a data matching request sent by a terminal; acquiring group behavior data corresponding to a first user group from a first social networking service platform; learning the group behavior data to obtain a group characteristic distribution function; acquiring associated users and corresponding behavior data of an appointed root node user from a second social networking service platform; learning the behavior data of the root node user, and generating the group characteristic distribution function after matching with the root node user; carrying out the behavior learning on the behavior data of the associated user; calculating the maximum entropy of the group characteristic distribution function after matching with associated users, and confirming the associated user with the maximum entropy correspondingly as a matching user of the first user group; taking the confirmed matchinguser as the current root node user, and confirming the next matching user until the confirmed matching user meets the condition of setting number; completing the group matching. The method can realize the data matching of different social networking service platforms.
Owner:PING AN TECH (SHENZHEN) CO LTD
Multi-index anomaly detection method based on neural network
ActiveCN110674940AImprove forecast accuracyPreserve topological propertiesNeural learning methodsManufacturing computing systemsAnomaly detectionSpace mapping
The invention discloses a multi-index anomaly detection method based on a neural network. The multi-index anomaly detection method comprises the following specific steps: 1, defining a data format; 2,carrying out model training on the system by utilizing SOM, and defining the system as a learning process; 3, performing anomaly detection on the input data, and defining the anomaly detection as a mapping process; and 4, when the model is mapped to be abnormal, carrying out root cause positioning. According to the multi-index anomaly detection method, the induction behavior model can be used forpredicting the unknown performance abnormality and providing an abnormality reason prompt, and the model can obtain higher prediction precision in a benchmark test result; the high-dimensional inputspace is mapped into the low-dimensional map space by using the SOM; and meanwhile, the topological property of the original input space is reserved, so that expandability and effective system behavior learning can be realized.
Owner:上海擎创信息技术有限公司
Method and apparatus for learning behavior in software robot
Owner:SAMSUNG ELECTRONICS CO LTD
Cluster behavior learning method based on gene programming
Owner:NORTHWESTERN POLYTECHNICAL UNIV
Implementation method of dynamic intelligent self-adaptive honeynet
The invention discloses an implementation method of a dynamic intelligent self-adaptive honeynet, which comprises a dynamic intelligent self-adaptive honeynet architecture. The dynamic intelligent self-adaptive honeynet architecture comprises an intelligent honeynet architecture basic hardware facility layer, an intelligent honeynet architecture component recognition engine layer, an intelligent honeynet architecture decision distribution engine layer, an intelligent honeynet architecture flow cleaning engine layer, an intelligent honeynet architecture forwarding decision engine layer and an intelligent honeynet architecture behavior learning engine layer. Through mutual configuration and cooperation of multiple engines, the flexibility of the honeynet is improved, and the score of the attacker connection activity, the self-adaptive attack and the network environment are calculated according to the algorithm weight. Aiming at the problems of a traditional honeypot, resources are reasonably called and a honeynet most suitable for a client is deployed according to the actual situation of the client and the strength degree of an attack technique, and a balance point is achieved between the threat perception effect and the resource utilization efficiency.
Owner:江苏天翼安全技术有限公司
Industrial control safety auditing system and method based on artificial intelligence
ActiveCN112437041AImprove securityMeet industry compliance audit requirementsTransmissionTotal factory controlInformation transmissionAttack
Owner:北京珞安科技有限责任公司
UUV agent behavior learning and evolution model based on chaotic immune genetic mechanism
InactiveCN110619394AFast learningIncrease diversityChaos modelsNon-linear system modelsAntigenVaccination
The invention belongs to the technical field of underwater unmanned system modeling and simulation, and particularly relates to a UUV intelligent agent behavior learning and evolution model based on achaotic immune genetic mechanism. The method comprises the following steps: firstly, loading a to-be-solved problem and constraint conditions as antigen Ag, and generating an initialized antibody population according to a vaccine population, a memory population and a chaotic mechanism; secondly, controlling the convergence direction of the learning process by utilizing a vaccination mechanism according to an antibody fitness calculation result, and completing updating of an antibody memory bank; and finally, sequentially designing a selection operator based on roulette, a crossover operator based on adaptive adjustment and a mutation operator based on Gaussian and polynomial mixing to realize diversity of the antibody population, and performing premature suppression, thereby realizing updating and iteration of the antibody population. The model combines the advantages of the global search capability of a basic genetic algorithm and the local search capability of an immune and chaoticmechanism, and promotes the quick learning and evolution of behavior rules by continuously adjusting and optimizing the search space of a problem solution.
Owner:SHAANXI NORMAL UNIV
Robot obstacle avoidance behavior learning method based on deep learning
PendingCN111429515ASimplicity advantageCost feasibility advantageImage analysisCharacter and pattern recognitionPattern recognitionData set
The invention discloses a robot obstacle avoidance behavior learning method based on deep learning, and the method comprises the steps: controlling a robot to carry out the obstacle avoidance movementin an unknown environment, collecting RGB-D image data at a fixed frame rate, and carrying out the naming and storage according to a time sequence; constructing an RGB image and Depth image fusion neural network model, and inputting the collected RGB-D image data set into the RGB image and Depth image fusion neural network model; setting hyper-parameters of the RGB image and Depth image fusion neural network model, and training the RGB image and Depth image fusion neural network model through a neural network model training framework to obtain a trained fusion neural network model; inputtingthe RGB-D image data set collected in the S1 into the trained fusion neural network model, and outputting a fused feature image. The implementation of complete functions can be only suitable for a single RGB-D camera to serve as an input sensor, and in practical application, the method has certain advantages in cost feasibility and simplicity of robot structural design.
Owner:FOSHAN NANHAI GUANGDONG TECH UNIV CNC EQUIP COOP INNOVATION INST +1
Computer implemented method for tracking and checking measures and computer programs thereof
InactiveUS20140355412A1Telemetry/telecontrol selection arrangementsError preventionBehavioral learningTransmission network
A computer implemented method for tracking and checking measures and computer programs thereof. A master node receiving from a plurality of slaves nodes messages related with measures generated by the slaves nodes, the method including: capturing, a traffic driver unit, the messages sent by the slaves nodes and further sending them to a monitor unit; analyzing, the monitor unit, the received messages so as to detect, by a behavioral learning technique, anomalies in the messages; when an anomaly is detected, sending, the monitor unit, the detected anomaly to a regenerator unit for regenerating at least the detected anomaly by a prediction technique; and injecting, said traffic drive unit, measures regenerated by the regenerator unit to the transport network.
Owner:TELEFONICA DIGITAL ESPANA
Industry characteristics analyzer with artificial behavior learning capability
InactiveCN105512191AQuick analysisRealize screeningMetadata text retrievalSpecial data processing applicationsInformation processingData science
The invention discloses an industry characteristics analyzer with an artificial behavior learning capability, pertaining to the technical field of intelligent information processing technology and big data analyses. The industry characteristics analyzer comprises a dynamically-supplemented industry characteristics sample library. The analyzer is used for extracting industry rules out of two samples in a concentrated mode in the industry characteristics sample library according to the certain strategy in order to form an industry analysis rule library. When receiving analysis tasks, an analysis engine is used for analyzing inputted unknown characteristics tests according to the industry analysis rule library, adjusting analysis results, recognizing characteristics and achieving learning capability.
Owner:NANJING LES INFORMATION TECH
Who we serve
- R&D Engineer
- R&D Manager
- IP Professional
Why Patsnap Eureka
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com