149results about How to "Realize self-learning" patented technology

The invention provides a building method of a knowledge map based on a vertical field. The method comprises the following steps of (1) extracting the word realization of classes of an on-line encyclopedia and the hyponymy between classes; (2) merging the field knowledge information, defining the data attribute and the relationship attribute of the field, and further setting the statute on the definition domain and the value domain of the attributes; (3) studying an entity layer, i.e., extracting an entity and filling the attribute value of the entity; performing mass processing on structurized and semi-structurized data by D2R or data collecting tools; and for non-structurized text data, defining the classes and the attributes of the upper layer body and the relationship between the classes and the attributes, and recognizing examples according to the relationship between the classes and the attributes. The method has the advantages that by using the method, the built knowledge classification of the vertical field knowledge map is clear; the self study and the automatic expansion of the knowledge map are realized; and the key effects are achieved on the information retrieval and semantic analysis of the vertical field.

The invention provides a self-learning method and a self-learning device of a hybrid electric vehicle and a clutch of the hybrid electric vehicle, which can automatically finish self-learning of a key position of the clutch so as to simplify an adjusting process. The self-learning method comprises the following steps: judging whether a self-learning condition is met or not; if so, controlling the clutch to be separated and combined; recording positions where the clutch is completely separated, is completely combined and starts to slide and wear, and taking the positions as a complete separation point, a complete combination point and a sliding and wearing point. The self-learning device comprises a position sensor which is in signal connection with a controller, wherein the controller is provided with a judging module for judging whether the self-learning condition is met or not, and controlling the separation and combination of the clutch when the self-learning condition is met, so that the self-learning of the complete separation point, the complete combination point and the sliding and wearing point are realized; the position sensor is used for detecting the position of the clutch and transmitting position information to the controller; the controller is provided with a storage module for storing the position information.

The invention provides a data mining based intrusion detection system with self-learning and classified early warning functions. The system comprises a clustering analysis module, an anomaly detection engine, a rule base, a correlation analysis module, a rule generalization module, a rule management module, a log record and a classified early warning module. The data mining based intrusion detection system has the advantages that a data mining technique is applied to intrusion detection, and existing data mining algorithms and network attack characteristics are utilized fully, so that self-learning and classified early warning of the intrusion detection system are realized, detection accuracy and efficiency are improved effectively and substantial economic value and use value are achieved.

The invention discloses an ACC / AEB system based on machine learning. The ACC / AEB system comprises an environment perception module, a data fusion module, a machine learning and decision controlling module and an execution module. The technical scheme is adopted to obtain control parameters adapted to driving habits of a driver by integrating a convolutional neural network with an ACC / AEB control algorithm and training and learning continuously, also to realize self-learning and self-correction during operation of ACC / AEB, namely, to self-learn operating conditions not encountered and self-correct unsatisfactory operating conditions, to continuously correct a weight value between each neuron and each parameter to output optimal control parameters, to realize intelligent vertical control, and to achieve comfort, safety and robustness.

The invention is suitable for the IT technical field and provides a distributed computation frame parameter optimizing method, device and system. The method comprises the steps of obtaining currently-submitted parallel computation frame operation; retrieving historical distributed computation frame operation similar to the distributed computation frame operation in a distributed computation frame operation historical operation database, wherein the distributed computation frame operation historical operation database comprises execution information and configuration parameters of the distributed computation frame operation; retrieving historical distributed computation frame operation similar to the distributed computation frame operation during the similar distributed computation frame operation; performing optimizing configuration on the configuration parameters of the distributed computation frame operation according to configuration parameters of the similar historical distributed computation frame operation. By adopting the distributed computation frame parameter optimizing method, device and system, parameter configuration rationalization, automatization and self learning of Mapreduce operation can be achieved.

The invention discloses an intelligent operating system and method for a train examination robot of a railway vehicle, and belongs to the field of technical inspection of railway freight trains. The intelligent operating system for the train examination robot of the railway vehicle comprises a train examination robot system, a train examination operation, monitoring and management platform and a train examination wireless network platform; a wheel type robot and an air-control big gate are controlled to combine with an automatic control technology, a laser detection technology, an inactive radio frequency transmission technology, a position self-sensing technology, a machine visual technology and a digital image analysis technology to acquire holographic data of a to-be-detected vehicle; a mode identification technology is utilized to perform treatment and analysis as well as decision-making, so that brake test automation, part quality state diagnosis informatization, fault treatment monitoring and reexamination traceability after train maintenance in train examination operation of the railway vehicle are realized. The intelligent operating system realizes replacing manual examination by robot examination, can be suitable for different environments of train examination application places across the country, can comprehensively improve quality and efficiency of train examination operation, and has relatively great social benefits and economic benefits.

The invention relates to a non-intrusive load identification method with a self-learning ability and a smart electric meter, aiming to overcome the defect that existing non-intrusive load identification technology cannot realize self-learning. A metering core of the smart electric meter provides voltage and current sampling data; a load identification core intercepts transient variation and peripheral steady state data and calculates a load characteristic, and a load identification and load characteristic library is updated and modified; and a management module interacts with a user so as to determine uncertain identification results, and calculates electric appliance running data. The load identification method realizes self-learning; when a new electric appliance is added, the existing identification method is still usable; the characteristic library size is kept within a certain range; and load identification operation amount is within a certain range.

The invention discloses a system for measuring granularity of pulverized coal of a blast furnace coal powder injection middle-speed milling system. The system comprises a storage device, a sensor, a retrieve device, and a calculation correcting device, wherein the storage device is used for storing a case library, the case library comprises a plurality of cases, and each case represents a corresponding relation between the granularity of pulverized coal and an auxiliary variable influencing the granularity of the pulverized coal; the sensor is used for detecting a current production working condition in a blast furnace middle-speed milling production process; the retrieve device is used for retrieving in the case library according to a present production working condition and finding out the closest case of the auxiliary variable and the present production working condition; and the calculation correcting device is used for modifying the granularity of the pulverized coal according to an auxiliary variable of the retrieved closest case and a present production working condition difference value to obtain the granularity of the pulverized coal under the present working condition. The system disclosed by the invention can realize online detection on the granularity of the pulverized coal so as to carry out real-time regulation on control variable of the middle-speed milling production process according to the granularity of the pulverized coal to ensure that the granularity of the pulverized coal of pulverized coal production index meets process requirements and to create beneficial production conditions for following pulverized coal injection and blast furnace production.

The invention discloses a reinforcement learning reward self-learning method in a discrete manufacturing scene. The method comprises the following steps: 1, refining the process of the current production line, wherein g belongs to G = {g1, g2,..., gN}, and the intelligent agent reaches a preset target g and is recorded as an interaction sequence episode; according to the initial parameters, obtaining multiple sections of episodes corresponding to the g1 as a target, taking state actions in the episodes and a state difference value delta as a training data set to be input into a GPR module, andobtaining a system state transition model based on state difference; enabling the intelligent Agent to continue to interact with the environment to obtain a new state st, and enabling a Reward network to output r (st), enabling an Actor network to output a (st), enabling a Critic network to output V (st) and enabling a GPR module to output value function Vg as the updating direction of the whole;when the absolute value of Vg-V (st) is smaller than epsilon, considering that award function learning under the current procedure is completed, and carrying out parameter storage of the Reward network; continuously carrying out interaction, and generating the following sub-target g < n + 1 > as the episodes of the updating direction for updating the GPR; and when the set target G = {g1, g2,...,gN} is all realized in sequence, finishing the process learning of the production line.

The invention provides an information auditing method, an information auditing device, a computer-readable medium and electronic equipment. The information auditing method comprises the steps of acquiring an auditing strength factor set by a user; acquiring at least one piece of picture information from stored picture information, wherein the similarity coefficient of the picture information and a template picture is greater than or equal to the auditing strength factor; comparing information of a picture to be audited with the at least one piece of picture information; if the comparison result of the information of the picture to be audited and any picture information meets the auditing strength factor, determining that the picture to be audited passes the audit. After the technical scheme is adopted, the picture can be automatically audited, and the problems of slow audit speed and uncontrollable audit results which are caused by a manual audit way can be solved.

The invention provides a flow type high-precision and high-speed automatic counting device for particle matters based on machine vision. The flow type high-precision and high-speed automatic counting device adopts a flow line for automatically counting and detecting the particle matters, and the flow line comprises a camera unit, a microprocessor, a feed mechanism, a pouring mechanism, a rotating mechanism and a slight shock mechanism, wherein the microprocessor comprises an image reading module, an image processing module, an automatic counting module, a majority voting module, a result output module, a production line control module and a human-computer module. The automatic counting device for the particle matters, provided by the invention, is high in detection efficiency, excellent in distinguishing and counting precision, wide in adaptive scope of detection objects, and high in self-leaning capacity and improvement capacity.

The invention discloses an automatic software test system and method constructed based on an artificial intelligence algorithm, and the system can autonomously learn test specification requirements and previous test data and results, and summarize and refine a test specification requirement model. The system can autonomously learn programming language design content and generate a programming language model. The system can be combined with different software platforms and hardware platforms to carry out self-adaption on a programming language model to generate a target system model, and the target system model is used for guiding automatic software testing to be carried out on different software and hardware platforms. By adopting the test system, a system and a method for autonomously learning test requirements and programming specifications, autonomously adapting to different software and hardware platforms to automatically generate test cases, automatically carrying out software tests and automatically providing software test reports can be provided.

The invention provides an APT multi-dimensional detection and defense system and method. The system comprises two or more detection modules and a dimension expansion module, wherein the detection modules are allocated to all dimensions and used for detecting objects to be detected in the current dimensions, recording behaviors of the objects to be detected and detection results, screening out dimension detection expansion information according to preset screening rules and transmitting the dimension detection expansion information to the dimension expansion module; the dimension expansion module acquires dimension detection expansion information transmitted by the detection modules, performs projection association on the dimension detection expansion information according to preset dimension association rules to generate detection rules and screening rules available to other dimensions, and transmits detection rules and screening rules to the detection modules in corresponding dimensions. The invention further provides the corresponding detection and defense method. By means of the APT multi-dimensional detection and defense system and method, the detection and defense system can be updated automatically to obtain detectability of other dimensions.

The invention discloses an electric vehicle power battery SOC intelligent detection device which is characterized in that the intelligent detection device includes a battery parameter acquisition platform and a battery SOC estimation system, the battery parameter acquisition platform collects real-time parameters of voltage, current and temperature of an electric vehicle power battery group, the battery SOC estimation system estimates an SOC value through the collected parameters, and a battery SOC system is a nonlinear, time-delay and multivariable coupling complex real-time system with a high requirement. According to the device, a problem that a conventional detection device can not obtain an ideal effect of the intelligent detection of the electric vehicle power battery SOC.

The invention belongs to the technical field of wireless communication and discloses a terminal autonomous network selection system and a method of wireless heterogeneous network. The terminal detectsand obtains service node information by using ANDSF to form a service node list of the terminal. The terminal obtains the RMI of each node in the service node list by using pilot signal measurement and carrier sensing. The terminal triggers the feature learner, inputs the measured RMI of each node to the feature learner, and outputs the measured RMI as an evaluation value of the link quality of each node. The terminal uses 802.11 u standard and ANDSF to obtain the load information of each node in the service node list. The terminal triggers the policy learner, inputs the obtained link qualityevaluation value and load information of each node to the policy learner, and outputs the obtained link quality evaluation value and load information to the policy learner to select an access node for the terminal. The terminal records the selected node and updates and records the network selection benefit. The invention overcomes the shortcomings of additional signaling overhead and delay causedby adopting server assistance or network assistance.

The invention discloses an optimization method and device of a dialog system knowledge base. The optimization method comprises the following steps: obtaining target question and answer information: judging whether assessment parameters of question and answer information to be analyzed conform to an optimization standard or not, and taking the question and answer information to be analyzed as the target question and answer information when the assessment parameters of the question and answer information to be analyzed conform to the optimization standard; and generating an update answer set: on the basis of question information of the target question and answer information, obtaining a corresponding answer information set, independently calculating a relevance parameter of each piece of answer information in the answer information set, and generating the update answer set which aims at the question information in the target question and answer information according to the relevance parameters. The method can realize automatic update of the dialog system knowledge base, so that a dialog system can output answers which more conform to the use habit and the expectation of users so as to improve the user experience and the user viscosity of the dialog system.

The invention discloses a building construction safety monitoring method based on a convolutional neural network. The building construction safety monitoring method based on the convolutional neural network comprises the following steps: 1) obtaining safety management measure data of operators, operation environments and building construction sites, obtaining a safety evaluation grade, and forminga training set and a test set of a CNN convolutional neural network model through preprocessing; 2) taking data of operators, operation environments and safety management measures in the training setas input, taking safety evaluation levels as output, determining a safety monitoring index arrangement mode and a CNN convolution neural network kernel function through training, and storing a trained CNN convolution neural network model; and 3) predicting the safety of the operator and the operation environment. The building construction safety monitoring method overcomes the defects that an existing monitoring method is inaccurate in problem evaluation result, and the evaluation factor of the evaluation method is single, and the calculation cost is high, and real-time monitoring cannot be achieved.

The invention provides a soft-measuring method for density in concentrating process of salvia miltiorrhiza injection production, comprising the following steps: gathering the historical data of each sensor and densimeter in the concentration process of salvia miltiorrhiza, wherein the history data relate to concentrated solution density value and sensor data gathered on line in production process;selecting easily obtained procedure variables including gas phase temperature, liquid phase temperature, pressure and concentrated solution liquid level that have higher degree of correlation with density from each sensor data, and screening representative dataset; using a multivariate analysis method for building a soft-measuring model of density; and gathering the procedure variables on line, and using the soft-measuring model of density to carry out real-time estimation to control the concentration process. Aiming at the problem that the density in the concentration process of salvia miltiorrhiza injection production is difficultly monitored in real time, the invention provides a fast density soft-measuring method with high precision, which makes full use of the history data obtained by each sensor from the production process, and is beneficial to improvement of quality control on salvia miltiorrhiza injection production.

The invention discloses a portable intelligent plant physiological state detection system and method. The portable intelligent plant physiological state detection system comprises a lighting module and a camera unit used for collecting fluorescent images of a to-be-detected plant, the lighting module comprises an LED array providing excitation light for lighting and an optical thin film arranged on a light emitting side of the LED array, reflectivity, to fluorescent light of the to-be-detected plant, of the optical thin film is smaller than 5%, and reflectivity, to LED light, of the optical thin film is larger than 95%. By the portable intelligent plant physiological state detection system and method, a user can perform chlorophyll fluorescence detection on plants in field conditions in a portable manner, the fluorescent images acquired by measuring can be processed and analyzed in realtime, a measuring result can be matched with plant physiological states in a database of a remote cloud sharing server in realtime, and updating and machine learning of new samples of the database can be realized.

The invention discloses a method and a system for automatic operation of a data communication network (DCN), and relates to the field of DCNs. By the method and the system, an unnumbered Ethernet interface can be created and used. When routing equipment is powered for the first communication, by using a loopback interface automatically generated the routing equipment and operating an open shortest path first (OSPF) point-to-point network, the routing equipment automatically learns network routing and forms a DCN routing table so as to realize automatic operation of the DCN. According to the method and the system, the DCN automatic operation function is used, after the access layer routing equipment is powered, a management channel to a network management system can be broke through by the DCN routing table, and a series of maintenance and operation allocation can be performed by the routing equipment and the network management system. By the method and the system, a complete automation function can be achieved, the deployment, site commissioning and configuration operation of the routing equipment are relatively simple; not only the time is shortened, but also the workload is reduced; and the working efficiency is high, and the method and the system are convenient to use.

The invention discloses an electric automobile power battery SOC (State of Charge) detection system. The characteristics lie in that the detection system comprises a battery parameter acquisition platform and a battery SOC estimation system, the battery parameter acquisition platform is responsible for real-time parameter acquisition for the voltage, current and temperature of an automobile power battery pack, and the battery SOC estimation system can accurately estimate a battery SOC value through the acquired parameters; and the battery SOC is a non-linear, delayed, multivariable coupling and complex highly demanding real-time system. The detection system effectively solves a problem that the traditional automobile battery SOC estimation method is difficult to achieve an ideal effect.

A fault diagnosis method suitable for an electrical system of an unmanned aerial vehicle comprises the following steps of: establishing a fault diagnosis knowledge library and a fault symptom set of the electrical system of the unmanned aerial vehicle at first and then taking out a fault diagnosis rule completely containing all symptoms and all fault diagnosis decision tables; and outputting a fault reason with the highest diagnosis success frequency as a fault diagnosis prediction result, verifying the fault reason, and if the verification is passed, taking the fault reason as a final diagnosis result for positioning a fault mode to finish autonomous and quick dynamic reconfiguration, otherwise, reselecting the fault reason. The method greatly improves the reliability and fault-tolerant capability of the electrical system, thereby having relatively high engineering application values.

The invention discloses an intelligent SOC (State of Charge) prediction device for an electric vehicle power battery, which is characterized by comprising a battery parameter acquisition platform and a battery SOC prediction system, wherein the battery parameter acquisition platform is used to acquire real-time parameters of voltage, current, temperature, and ambient temperature of the vehicle power battery pack; and the battery SOC prediction system is used to predict the battery SOC value through the acquired real-time parameters. The battery SOC is a nonlinear, delayed, multivariable-coupling, and complex real-time system with extremely high real-time performance requirements. The problem that the conventional prediction device can not achieve ideal battery SOC prediction precision effects can be effectively solved.

The invention belongs to the field of power plant safety control systems, and particularly relates to an operation guidance system based on a typical and accident case base of a thermal power plant. The operation guidance system comprises a data collection system (DCS) and is characteriazed in that a memory internally saves a mentoring function module, and stores an accident case base and typicaloperations; a processor encodes and classifies fault messages of different components of the thermal power plant; based on an expert industry knowledge base and analysis on historical fault data, normal threshold ranges of operating indicators of a thermal power unit are defined, a change rule in an operating process of the indicators is mined, data sample sets in different states are constructed,and a fault trigger model is established using a machine learning algorithm. When triggering conditions are reached, a display visually shows the location of a fault and a data distribution rule, andinformation about the cause of the fault and treatment suggestions is displayed on the display; fault early warning is accurate and the operation guidance is accurate and in place.

The invention discloses a self-learning method of vehicle configuration information. The self-learning method comprises the steps that current configuration information of all electronic control unitsof the vehicle is obtained; the current configuration information of all electronic control units is compared with historical configuration information; if differences exist between the current configuration information of all electronic control units and the historical configuration information, configuration information of all electronic control units is updated according to the current configuration information. According to the method, a man-machine interactive system can learn the vehicle configuration information by itself, and consistency of a human-computer interaction interface and the actual hardware function is kept. The invention further discloses a self-learning system of the vehicle configuration information and the vehicle.

The invention provides an airborne flight parameter data health diagnosis method based on deep learning. The method is based on a deep learning algorithm, data collected by aviation airborne flight parameter equipment can be analyzed in real time, the health condition of an airplane in the flight process is obtained, the fault risk of the airplane is predicted in advance, and a decision for avoiding faults is given to be adopted by pilots. According to the method, specific acquired data is required to be trained, and particularly, analysis and learning of relevance of various characteristic parameters before an airplane has a fault are the key to effectively predict and avoid the fault. And the problems of data real-time analysis, fault diagnosis and risk decision in the field of aviationairborne flight parameter equipment are effectively solved.

The invention discloses an urban-level traffic video detection equipment quality detection method. The method comprises the following steps of a, constructing the network topology structure of each video detection device in a road network, and acquiring a network topology structure table; b, according to the network topology structure table, constructing the randomness model of road section strokeconfidence time; c, identifying correct vehicle passing data and error data in a total equipment detection amount N detected by the video detection device S and calculating the false detection rate of the video detection device S; and traversing all the video detection devices and acquiring the false detection rates of all the video detection devices; and d, according to the road section stroke confidence time, randomly simulating missing data, defining the stability index of a road section vehicle passing amount, and calculating the omission ratios of the video detection devices in a networktopology structure table. The method has characteristics that traffic data quality is improved; a workload is reduced; and data description accuracy is increased. And the method is the basis of intelligent traffic platform construction.

The invention discloses an intelligent gas cooker. The intelligent gas cooker is manufactured by electric technology and computer technology. A valve of the gas cooker is provided with an actuating mechanism controlled by a motor; the actuating mechanism is controlled by a gas cooker controller, and completes turn-on / off, firepower control and the like of the gas cooker by receiving the instructions of the gas cooker controller, so that the dependence of the gas cooker to a user is reduced; meanwhile, music and video can be played during use, so that the user works happily.

The embodiment of the invention discloses a method for monitoring harmonic waves of a power grid of a dredge boat. The method comprises the steps of obtaining original power signals in the power grid, extracting measurement quantity of the harmonic waves of the original power signals in the power grid each time, calculating the optimal estimation value (please see the formula in the specification) of the amplitude of the hth harmonic wave at the moment k according to measurement quantity (please see the formula in the specification) of the hth harmonic wave at the moment k, calculating a covariance matrix (please see the formula in the specification) of the optimal estimation value (please see the formula in the specification), according to the optimal estimation value (please see the formula in the specification), calculating the harmonic distortion rate (THD) k at the moment k, and according to the harmonic distortion rate of each time, judging the quality of the power signals of the power grid. According to the embodiment, on the basis of the conventional Kalman filtering algorithm, a correction factor gamma is introduced to express dredging operation environment, a robust Kalman filtering algorithm on the basis of a fuzzy self-adaptation optimizing reasoning model is built, one-step prediction and estimation of the harmonic data of the power grid are completed, robustness of a power grid harmonic monitoring system is obviously improved, the optimal estimation value of the harmonic distortion rate is made to be closer to the original power signals, finally, the quality of the power grid is judged according to the harmonic distortion rate, and the accuracy of judgment is improved.

The invention provides an event pre-warning and classifying method by an external safety pre-warning and positioning system of photoelectric composite cables, which comprises the following steps: obtaining event signals occurred within the range monitored by the system obtained through the continuous acquisition performed by the external safety pre-warning and positioning system of the photoelectric composite cables through high-speed data acquisition card; and performing pre-warning and classification of invasion events on the acquired event signals by using a loose coupling wavelet radial basis neural network algorithm described in the following text. The event pre-warning and classifying method by the external safety pre-warning and positioning system of the photoelectric composite cables provides a practical basis for the event occurrence position calculation and pre-warning so that the external safety pre-warning and positioning system of the photoelectric composite cables can conveniently detect the occurrence of warning events for the convenience that monitoring staffs can arrive at the occurrence positions of the events in time and perform corresponding protection and maintenance on the photoelectric composite cables; besides, the method saves the investigation time of event precaution and provides time allowance for the precautionary measures of the events.