224 results about "Online model" patented technology

A system and method for the identification of users and objects using biometric techniques is disclosed. This invention describes a biometric based identification and authorization systems which do not require the incorporation of an on-line database of stored complete biometrics for the security infrastructure. In order to remove the connectivity requirements, an off-line biometric system is achieved by incorporating an identity verification template (IVT) on a storage device / token (e.g., magnetic strip or smart-card) during the user's registration which provides for a reliable storage medium; however, there are no security requirements required of the token even to protect the privacy of the stored biometric. The IVT does not contain complete information of the user's biometric but allows for the verification of the user when that user later provides a biometric reading. To deal with errors that may be introduced into later scans of the biometric (for example at the time of verification) error correcting techniques, well known in the art of communication and error control systems, are incorporated into the system. The system is also usable in the online model. Moreover, it may also be used to enable cryptographic operations by being used to partially compose or encrypt private keys for cryptographic operation.

The invention relates to a fault monitor and diagnosis method of a continuous annealing unit, in particular to a PCA (Principle Component Analysis) model based furnace temperature and tension monitoring of a continuous annealing unit, mainly comprising the following steps of firstly, according to process variable data obtained in the field, and establishing a temperature and tension monitor modelby utilizing a principle component analysis PCA method; secondly, establishing an off-line model and calculating the T2 statistics quantity and the SPE statistics quantity as well as contributed control limits thereof by utilizing the data, obtained in step one, when process variable is in a normal work condition; thirdly, applying an on-line model, calculating the T2 statistics quantity and the SPE statistics quantity of current data, monitoring whether a current state is normal or not according to information supplied by the off-line model, and giving alarm signals if abnormal; fourthly, determining a leading variable which causes a fault by utilizing contribution of the T2 statistics quantity and contribution of the SPE statistics quantity. The invention monitors the furnace temperature and tension in real time in the production process and traces back a fault reason for leading to system abnormality when the abnormality occurs.

The invention relates to a modeling method of an on-line hydraulic model of a municipal drainage pipe network, belonging to the technical field of municipal engineering information; the modeling method can construct a municipal drainage pipe network on-line model and realize to input real-time rainfall data to the on-line model continuously when flood situation is forecasted, and analog computation is carried out to the model; furthermore, the model provides analog computation data of whole future water recession process for a drainage pipe network system continuously, so as to achieve real-time forecasting water regimen and timely warning functions. The method overcomes the disadvantages that the conventional drainage pipe network event model only can simulate the complete rainfall events in the city flood simulation and can not simulate real-time rainfall events, can effectively forecast the generating and changing process of the city flood in a real-time and on-line way, ensure long-term and stable operation of the on-line model system and provide technical support for city flood-prevention strategic decision.

A Model Predictive Control System, particularly useful in controlling gas turbine engines, formulates a problem of controlling the engine to achieve a desired dynamic response as a solution to a quadratic programming problem. The Model Predictive Control System also includes a quadratic programming problem solver solving the quadratic programming problem in real time using an interior point algorithm that searches for an optimal solution.

The invention discloses a real-time recommendation system and method based on Spark. A stacked recommendation system framework based on Spark is constructed, and comprises a data collecting module, an offline recommendation module, an online recommendation module and a recommendation module. The offline recommendation module is used for selecting a corresponding recommendation algorithm from an offline recommendation algorithm library according to user configuration parameters to train user behavior data, and a user feature model is obtained; the online recommendation module is used for sending the user behavior data to corresponding algorithms in an online recommendation algorithm library for training, and an increment user feature model is obtained; an online model training engine is used for adopting the user feature model obtained through training as a basic model, carrying out incremental updating through a current increment recommendation algorithm and the newly received user behavior data, and a newest user feature model is obtained. The recommendation module is used for updating a user recommendation list in combination with an inertia updating mechanism according to the user feature model. The accuracy and timeliness of the recommendation result can be effectively improved.

The invention relates to a comprehensive control method for the outlet temperatures of a heating furnace, belonging to the field of furnace temperature control technology. The invention is characterized in that branch balance control makes the outlets temperatures of a multi-branch heating furnace uniform, and automatic load lifting and dropping can be realized. By using state feedback prediction control of measured states and feedforward control, the method effectively improves the anti-interference ability. A state space model of the heating furnace is obtained by mechanism modeling, in no need of testing devices. An on-line model is self-adaptive, which effectively overcomes process characteristic change and improves the rate of putting into operation. Low-value fuel can realize idea valve location interval control, thus saving high-value fuel. The comprehensive control of the outlet temperatures ensures the uniform and stable outlet temperatures of the heating furnace, which is easy to operate and economical.

A method, medium, and apparatus detecting a real time event in a sports video. The method may include testing a confidence of an online model, calculated in a sports video stream, detecting an event by using an offline model in the sports video stream, when the confidence of the online model does not meet a threshold, training the online model through an event detected by using the offline model, and detecting an event by using the online model in the sports video stream, when the confidence of the online model meets the threshold.

A target tracking method based on semi-supervised learning and a random fern classifier relates to a target tracking method and solves the different problems that an existing target tracking method is large in calculated amount and high in complexity, cannot meet real time processing requirements, is poor in effects on large angle rotation and hyperplane rotation and the like. The target tracking method can solve the difficult problems target scale change, rotation, hyperplane rotation, noise, shading, bonding, camera vibration, dim and the like in scene to form a stable target track and a full-view online model of the target. The tracking method comprises the steps of initializing the online model to generate and standardize positive examples; initializing a detector; training a detection; tracking a short-term tracker; tracking validity evolution; detecting the detector; combining detection and track; and learning and upgrading training sets. The target tracking method is a complete target tracking solution and is widely applied to fields of actual video monitoring, behavior analysis, intelligent transportation, electronic police, precision guidance and the like.

The invention discloses a visual saliency model based automatic detecting and tracking method. The visual saliency model based automatic detecting and tracking method comprises the steps of: calculating a color, brightness and direction saliency graph of an input video image by using a visual saliency model, and defining a simple scene and a complex scene based on the weighted saliency graph; establishing a rectangular frame to serve as a tracking target to be tracked by using a saliency region when the simple scene is detected; correcting a manually selected tracking frame based on different weights when the complex scene is detected; tracking the tracking frame by utilizing a tracking studying and detecting algorithm, and detecting that the tracking is failed; detecting the image of each frame after the failure by using the visual saliency model, performing histogram matching on each region in the saliency graph and the online model before the tracking failure, and tracking a region with the highest similarity; and sending multiple regions with similar similarity into a target detector at the same time for detection, repeating tracking detection for the image target of the next frame, using a histogram comparison step until a target is detected again and tracking.

The invention relates to a main steam temperature predictive control method of an ultra supercritical unit of a thermal power plant. A layered control structure is arranged, the layered control structure consists of a bottom direct control layer and an upper guide optimization layer, and the upper guide optimization layer realizes model parameter identification, prediction optimization and PID (Proportion Integration Differentiation) parameter setting control; and the bottom direct control layer is in a cascade system structure and comprises main loop control and assistant loop control, and a PID controller orienting a DCS (distributed control system) is directly controlled. The control method can be used for replacing a control engineer to set PID parameters, on-line model identification and control parameter setting methods are directly adopted to achieve the purpose of main steam temperature self-adaptive control, the application of an advanced control technology and an intelligent optimization technology under a DCS is realized, the stability and the robustness are stronger, and a changing process of a controlled object model of the main steam temperature under a variable load can be adapted.

The invention discloses a batch process online-monitoring method based on a small number of batches. Aiming at the defect of insufficient modeling samples under the situation of a small number of batches, the method fully utilizes the information of a small number of modeling batches to construct a novel analytical unit generalization time slice to provide sufficient samples for analysis modeling. Through the analysis on the generalization time slice, the time-varying characteristics of potential characteristics of a process are effectively extracted, and meanwhile, automatic and ordered time frame partition is conducted on the process with the time sequence during a real process considered. A multi-period monitoring model is established based on the result of the time frame partition, and the number of models and complexity are reduced. The monitoring model has good monitoring performance and is perfected through online model updating, and then monitoring reliability of the model is guaranteed. The method is simple and easy to implement, an industrial engineer can make accurate judgment on the operation state of the process conveniently and find out failures in time, and then safe and reliable operation of actual production and high quality of products are guaranteed.

A lithographic projection apparatus includes a measurement system for measuring changes in projection system aberrations with time, and a predictive control system for predicting variation of projection system aberrations with time on the basis of model parameters and for generating a control signal for compensating a time-varying property of the apparatus, such as the OVL values (X-Y adjustment) and the FOC values (Z adjustment) of a lens of the projection system for example. An inline model identification system is provided for estimating model parameter errors on the basis of projection system aberration values provided by the predictive control system and measured projection system aberration values provided by the measurement system, and an updating system utilizes the model parameter errors for updating the model parameters of the predictive control system in order to maintain the time-varying property within acceptable performance criteria.

The invention relates to an extended Kalman particle filtering based SOC (State Of Charge) estimation method and system. The method comprises the steps of performing a discharging-standing test on a battery through a single battery test system so as to acquire an SOC-OCV (Open Circuit Voltage) function relational expression; building an equivalent circuit model of the battery; generating a particle set according to initial probability distribution; performing online parameter identification on the model according to the current data; and performing extended Kalman particle filtering on the particles by using the updated model; and updating a particle weight, and judging whether resampling is required or not according to the number of effective particles. In allusion to problems of low estimation accuracy, existence of an accumulative error, high requirements for model precision and the like of an existing SOC estimation method, the invention adopts the method for online model parameter identification to improve the model precision and adopts the method of combining Kalman filtering and particle filtering to improve the SOC estimation accuracy. The method provided by the invention can effectively estimate the SOC and suppress noises, has the advantage of high precision, and can be applied to the field of a battery management system.

The embodiment of the invention discloses an online model training method. The method comprises the steps of obtaining a training sample from streaming data, determining an objective function of the model according to the training sample, historical model parameters and non-convex regular terms, determining current model parameters enabling the objective function to be minimum, and updating the model according to the current model parameters. In the online training process, since the non-convex regular term is adopted to replace the L1 regular term for feature screening, the penalty deviationcan be reduced, effective features can be screened out, the sparsity is guaranteed, and the generalization performance of the model is improved. The invention further provides an information pushing method. The method comprises: obtaining user feature data and content feature data, based on the pushing model obtained by the online training model method, determining the probability that a target user is interested in target information according to the user feature data, the content feature data and the pushing model, and determining whether pushing is conducted or not according to the probability that the target user is interested in. The invention further provides an online model training device and an information pushing device.

Provided is a boiler combustion optimization air distribution method based on online model prediction. The method includes: constructing a boiler combustion optimization air distribution model by employing an improved BP neural network algorithm and model online update calculation; and performing optimization calculation on parameters such as the optimal oxygen amount, the primary air amount, each secondary air door and a burnout air door etc. with the combination of an adaptive genetic algorithm and dynamic optimization boundary so that a multi-target function constructed by the boiler efficiency and the pollutant discharge capacity reaches the optimal combustion range.

The invention discloses a video object tracking method based on a feature optical flow and online ensemble learning. The technical problem that tracking results of a tracking method of a specified object in an existing digital video are poor is solved. According to the technical scheme, the method comprises steps of inputting a tracking portion into a video sequence, tracking characteristic points of every frame through an iterative pyramid optical flow method by using functions of OpenCV, and obtaining positions of characteristic points of the next frame; selecting a positive sample or a negative sample to be subjected to adaboost algorithm processing for a detection portion; and conducting machine learning for possible object positions obtained by the tracking portion and the detection portion. A tracking feature extraction mode and a detection feature extraction mode are separated, the filtering for possible object position limitation is conducted during detection, possible objects which are far away from objects are removed, after tracking results and detection results are obtained again, the fisher discrimination ratio of objects and the online model is calculated in a self-adapting mode, the corresponding weight is determined, a fixed value is not used for fusing two results, and then the tracking effect is good.

The invention provides an object tracing method based on active scene learning, belonging to the technical field of computer graphics and image mode identification. The difficulty needing to be confronted all together is to not only guarantee the adaption ability of a tracing system to the target change, but also guarantee the learning accuracy to avoid shifting to result in tracing failure. According to the invention, the object tracing is achieved by the steps of online modeling, short-time tracing, overall detection, active scene learning, a constraint method, analysis and extraction of a motion area and the like. By learning a scene online, the background information can be found actively, meanwhile, the motion area analysis and extraction method based on light streams is provided, and the problem of target loss and difficulty in recovering the target, resulting from rapid movement of an objected or severe movement of the scene, can be effectively solved according to the structured constraint. The adaption ability to the target change can be effectively improved, so that the long-time stable rapid object tracing can be achieved, and the object tracing method is mainly applied to various object tracing occasions.

The invention discloses an urban and rural overall planning water supply pipe network hydraulic water quality simulation platform, and belongs to the technical field of pipe network operation monitoring intelligence, and comprises an online alarm early warning module, an optimal scheduling management module, an accident emergency disposal module, a leakage monitoring alarm module and a system configuration module. The on-line alarm early warning module monitors the hydraulic condition and the water quality condition at the current moment on line through the system and predicts the hydraulic and water quality trend in the future 24 hours. The optimization scheduling management module comprises three parts of hydraulic optimization scheduling management, water quality optimization schedulingmanagement and a scheduling expert database. The leakage monitoring and alarming module simulates the influence caused by an accident point and the range of an affected area through the operation online model, and the model is used for making various water plant pump station scheduling schemes. Research and analysis of various plans are provided for a decision maker to simulate a decision processand a scheme environment, and meanwhile various information sources and analysis tools are accessed to help the decision maker improve the decision level and quality.

The invention provides a target tracking method based on supporting online clustering in detection, and belongs to the technical field of a computer graphical image. The target tracking method comprises the following steps of: simultaneously storing characteristic vectors of a target and a background sample in a blade node of a random fern detector; exploring distribution characteristics through the online clustering; and carrying out type probability density estimating by taking the distribution characteristics as sampling data points of a core function. A target object to be tracked is selected and determined from an original image, and is added into an online model consisting of target image blocks; and a target selecting process can be automatically extracted through using a motion target detecting method. In a real-time treatment condition, a video image collected by a camera and stored in a storage area is extracted as an input image to be tracked. The target image block is extracted as a positive sample, and a background image selecting block is used as a negative sample, so that an online training set is generated and input to the detector. A plurality of target types exist according to a condition of a plurality of targets, and each target type corresponds to one target. The target tracking method is mainly used for various fields of target tracking.

The invention relates to a deep convolutional neural network-based stockyard smoke monitoring and online model updating method. The method comprises the following steps of a, converting a smoke video into an image sequence by using a tool, and conducting the smoke labeling operation; b, by using the deep convolutional neural network, obtaining the more abstract high-level features of the smoke, iteratively training and optimizing model parameters, evaluating an iterative model according to a loss function and selecting an optimal model; c, according to the false-information and misinformation condition, updating the model in the online manner. According to the technical scheme of the invention, the effective feature extraction for smoke images is conducted based on the deep convolutional neural network. Based on the method, the smoke monitoring and the model updating are conducted in real time in the video monitoring condition. Therefore, the detection accuracy is further improved. The method is simple in operation, fast, effective, and high in robustness.

The invention provides a video-based vehicle detecting and tracking method and system. The method comprises the steps of inputting a video containing a vehicle; detecting the vehicle in the video by means of a detector, obtaining a detection result of the vehicle, and building an online model of the vehicle according to a first-time detection result of the vehicle; tracking the vehicle in the video, and obtaining a tracking result of the vehicle; utilizing the online model of the vehicle for respectively conducting confidence coefficient calculation on the tracking result and the detecting result of the same fame, and obtaining a final detecting and tracking result according to integration of a confidence coefficient calculation result, the tracking result and the detecting result; utilizing the final detecting and tracking result for updating the online model of the vehicle and the detector through online learning. According to the video-based vehicle detecting and tracking method and system, the problems that the vehicle deforms in the tracking process and other objects shield the vehicle are solved.

The invention relates to an online model reasoning system. According to the embodiment of the invention, the method comprises the steps of: establishing a model warehouse and a container mirror imagewarehouse in which container mirror images required by an inference model are stored; when receiving an online reasoning request of a user, enabling a model micro-service engine to call an inference model required by the user from a model warehouse according to the configuration information of the user; and storing the container mirror image in the container mirror image warehouse, so that the situation that the container mirror image required by the training model is inconsistent with the actual container mirror image is avoided, the inference model can be packaged into inference service capable of being operated in a containerization manner, and online inference service is provided.

The invention discloses a branch coefficient calculation method based on a real-time data of a power grid. The method comprises the steps of: obtaining an online mode and real-time measurement data of the power grid from an EMS (energy management system); obtaining a real-time operation mode of the power grid by state estimation, or directly obtaining state estimation data from the EMS; obtaining a zero-sequence parameter of each power grid element from relay protection setting value setting software, and associating the zero-sequence parameter of the power grid into the online model data of the power grid through a name matching table; setting a fault position and a fault type; obtaining current vectors of a lower-level branch and an upper-level branch by short-circuit current calculation; and taking a mode value of the ratio of the current vector of the lower-level branch and the current vector of the upper-level branch, or taking the ratio of an amplitude of the current vector of the lower-level branch and the amplitude of the current vector of the upper-level branch, so as to obtain the branch coefficient based on the real-time data of the power grid. By adopting the method, the branch coefficient calculation method is simple in calculation process, and is suitable for setting and checking a protection setting value.

The present invention provides a data analysis processing system and an online model deployment method thereof. The method comprises the steps of: performing publishing of an algorithm model based ondetected user operation used for a deployment algorithm model, and displaying the published algorithm model in a user interface to allow preset users using the data analysis processing system to access and / or call the published algorithm model. The data analysis processing system and the online model deployment method thereof can provide the system or publish the algorithm model established by theusers to services capable of being accessed by other users so as to improve the cooperation of a data analysis processing system, facilitate user usage and improvement of the usage efficiency.

The invention relates to a flood prevention dispatching decision support system, which comprises an on-line model server, a database, a data interface and a WEB application server, wherein the on-line model server is used for receiving hydrologic remote measurement data, and performing flood prediction simulation according to the hydrologic remote measurement data; the database is connected with each piece of hydrologic remote measurement equipment, and is used for storing hydrologic remote measurement data collected by the hydrologic remote measurement equipment; and the data interface and the WEB application server are respectively connected with the database and the on-line model server, are used for pushing the hydrologic remote measurement data in the database to the on-line model server, receiving the prediction simulation result and issuing the prediction simulation result to a remote network terminal. Compared with the prior art, the flood prevention dispatching decision support system has the advantages that the on-line model server performs flood prediction simulation according to the hydrologic remote measurement data; the outward issuing of the prediction result is realized through the data interface and the WEB application server; the rainfall water accumulation condition is simulated in real time according to the rainfall condition; the water accumulation point positions, the water accumulation depth and the water subsiding time are predicted; and the decision support is provided for water discharge emergency rescue and pump station optimization operation.

The invention discloses a mixed learning model-based DSP real-time bidding advertisement system. The system comprises an RTB advertisement bidding engine module, an offline model training module and an online model training module, wherein the RTB advertisement bidding engine module is connected with the offline model training module and the online model training module, calls a CTR prediction engine to predict a CTR value of a corresponding advertisement in combination with a CRT model generated by one or two of the offline model training module and the online model training module, and performs corresponding advertisement release.

The invention discloses an effective wind speed estimation method of a high wind speed section of a wind turbine generator system based on SVR. The method comprises two steps of SVR model training and online model use. In the SVR model training process, a sensor is used for obtaining a training characteristic set and a target set; the characteristic set is normalized to obtain a SVR training set; a GA algorithm is used for selecting a penalty parameter and a nuclear function parameter to obtain a trained SVR model; and in the online model use process, output data of an unit is obtained for normalization to input in the trained SVR model to pass through a low-pass filter to obtain a final effective wind speed estimation value. The method reasonably uses the output data of the unit, can perform effective wind speed estimation for a wind turbine generator at the high wind speed section, is simple in design process and liable to actuate, and can replace a LIDAR wind measurement device; and the obtained effective wind speed estimation value can be used for providing feedforward control information and wind power field wind resource estimation for reduction of the mechanical load of the unit, so that the economic benefit of a wind power field is improved.

The invention discloses an online novel content similarity comparison method which comprises the following steps that a pre-processing step is carried out: a compared online novel is pre-processed, key words are extracted, and the key words are replaced by synonyms to form a standard online novel; a characteristic fingerprint is extracted: the online novel to be compared is divided into a plurality of groups of close order words, a Hash table which is formed by the Hash operation of each group of close order words serves as the characteristic fingerprint; and characteristic fingerprints are compared: the characteristic fingerprint is compared with the characteristic fingerprints of existing online novels which are stored in a characteristic comparison library, and the similarity of the online novel to be compared and the existing online novels is determined according to the quantity or proportion of the same characteristic fingerprints. The method is used for comparing the similarity of the online novels, solves the problems of low efficiency caused by high complicacy of the method by pre-processing and checking omission caused by low purposefulness, and has self-learning ability, and the repetition checking accuracy is improved.

The invention relates to a distribution network model incremental storage method based on the E language. According to the technical scheme, the method comprises a first step of receiving outer net models, conducting model transaction detection and transferring the models into an off-line bank, a second step of examining and approving model transaction, and a third step of conducting increment synchronization on the distribution network change models from the off-line bank to an on-line mode bank on the basis of the E language. The method is reasonable in design. According to efficient mechanisms, standardly achieving information sharing with external systems on the basis of the models and graphs up to the IEC 61970 standard, and achieving inside model analysis and model incremental storage on the basis of the E language, of the method, strategies of model detection, examination, approval and storage of feeder granularity are provided, transaction mechanisms of distribution network model equipment are completed, results of model transaction are effectively improved, and the friendliness degree of software and practicality of programs are promoted. The method is suitable for model data information access of power distribution automation systems, and provides data foundation for on-line monitoring and static and dynamic safety analysis of power distribution networks.

The invention provides a method for determining thermal crown of a hot rolled strip roller, and belongs to the field of metal rolling. The method comprises the steps of periodically forecasting a rolling temperature filed through an online model of a finishing mill, and determining the initial temperature and the rolling thermal crown of a roller which is machined again. The method is characterized by comprising the steps as follows: reading the rolling temperature field of the unloading time of the roller through an online computer model; calculating the dynamic variation of the temperature field according to a cooling condition of an unloaded roller; determining the machining temperature and the crown according to the temperature generated after unloading; processing / calculating such data, thus providing the accurate initial temperature of the roller for machining again; forecasting the temperature field and the thermal crown of a working roller; and improving the precision in forecasting of the thermal crown of the holt roller and the shape setting control. With the adoption of the method provided by the invention, the problem that the initial temperature of the roller which is machined again when the roller is not repeatedly worn cannot be accurately given during producing the hot rolled strip roller can be solved; and the method can be widely applied to the technical field of thermal crown forecasting and shape control of roller during producing the hot rolled strip roller.