760results about How to "Verify accuracy" patented technology

Systems and methods for automated segmentation of anatomical structures, such as the human heart. The systems and methods employ convolutional neural networks (CNNs) to autonomously segment various parts of an anatomical structure represented by image data, such as 3D MRI data. The convolutional neural network utilizes two paths, a contracting path which includes convolution / pooling layers, and an expanding path which includes upsampling / convolution layers. The loss function used to validate the CNN model may specifically account for missing data, which allows for use of a larger training set. The CNN model may utilize multi-dimensional kernels (e.g., 2D, 3D, 4D, 6D), and may include various channels which encode spatial data, time data, flow data, etc. The systems and methods of the present disclosure also utilize CNNs to provide automated detection and display of landmarks in images of anatomical structures.

The invention discloses a method and a system for discovering a network hot topic based on situation assessment. The method comprises the following steps of: acquiring internet forum data in real time; preprocessing corpus of the internet forum data and extracting a candidate word string; forming a characteristic set in a vector form to express each post based on the candidate word string; clustering the posts, acquiring the discovered hot topic according to the clustering result and quantizing; and performing situation assessment on the hot topic generated by clustering and correcting the discovered hot topic according to a situation assessment result. By comprehensively utilizing a topic discovering and tracking algorithm and a situation assessment idea, the organization form of algorithms of selectable and customable processing processes is realized to analyze text information and mine data so as to discover the hot topic. The experiment result verifies the validity and the accuracy of the method and the system and the method and the system have great practical value.

The invention relates to a method for monitoring the running state of an industrial glue spread robot based on a digital twin technology. Compared with the prior art, the method has the effect of resolving the defect that the running state is hard to predict for the industrial glue spread robot. The method comprises the following steps of obtaining the running state data, collected by a sensor, ofthe industrial glue spread robot; preprocessing the running state data of the industrial glue spread robot; establishing a driving model of digital twin equipment; evaluating the running state of theequipment preliminarily; correcting the driving model of the digital twin equipment; and monitoring the running state of the industrial glue spread robot. According to the method, by using the equipment running model driven by digital twin, the real-time evaluation of the running state of the industrial glue spread robot is achieved, so that more accurate operation suggestions and maintenance suggestions are obtained; and the effective management of the life cycle of equipment operation is achieved.

The embodiment of the invention discloses a method and device for detecting network hot topics found based on a maximal clique. The method comprises the following steps of: acquiring data of a network news website, a forum, a blog and a microblog in a real time; carrying out word segmentation, word frequency statistics and other processing on the acquired data to find all hot point word pairs and construct a hot point word pair set; expressing each hot-point word by using a unique serial number; viewing the hot point word pair set as an undirected graph and excavating the undirected graph to obtain all maximal cliques; and transforming each maximal clique into a word combination for expressing one hot point topic. The invention also discloses a device for detecting the network hot topics. According to the embodiment of the invention, the hot point topics in the network can be accurately found in a real time, the detection speed and precision of the hot point topics are improved and higher practical value is obtained.

A computer based method for controlling an industrial process that includes at least two unit processes, including the steps of directly controlling the process by a control model including one or more algorithms, and automatic diagnosis of the validity of the process data retrieved from the process for the purpose of preventing irrelevant process data from being used as an input in the control model.

The invention provides a method for precisely editing a genome DNA sequence. A TALEN, CRISPR / Cas9 and other gene targeting technologies and homologous recombination technologies are utilized in a combined mode for precisely editing the genome DNA sequence. According to the method, the technical problem that in the prior art, a specific sequence cannot be deleted is solved, the specific base sequence can be seamlessly inserted or deleted in a fixed point mode, and seamless free modification of the genome DNA sequence of eukaryocyte is achieved for the first time.

A system and method for performing testing of data at a data warehouse is provided. The methodology of the invention describes steps to develop and further invoke one or more data quality-accuracy test cases from a framework. The data quality-accuracy test cases check the sanity of the data stored at the data warehouse. The one or more data quality-accuracy test cases are developed based on at least one predefined strategy, which in turn are stored in the framework. The methodology further executes the developed one or more data quality-accuracy test cases as either batch or independently, based on the requirements of the test. Thereafter, the methodology maintains traceability of the executed test at the data warehouse, incorporating details from the development of the one or more data quality-accuracy test cases to the final output of the test.

Belonging to a rock mechanical determination method, the invention provides a determination method for mesoscopic parameters simulating mechanical properties of an intermittent crack rock. The method includes: firstly sampling an engineering rock mass to be tested, and making a complete and intermittent crack cuboid sample; placing the made sample in a sealed pressure cylinder of a servo test machine, firstly bringing a testing machine pressure head into close contact with the upper end and lower end of the rock sample; applying an axial pressure on the sample according to displacement loading, and conducting loading on the sample till fracture, thus obtaining a sample compression process stress-strain overall process curve, a crack propagation process and a final fracture mode; employing particle flow code in 2 dimensions (PFC2D) to establish a complete sample numerical model, and debugging the mesoscopic parameters continuously to acquire a set of mesoscopic parameters able to reflect complete rock sample test results; establishing an intermittent crack numerical model based on the mesoscopic parameters, carrying out compression process simulation, comparing the simulation results with the intermittent crack rock test results so as to further verify the accuracy of the mesoscopic parameters.

The invention discloses a system upgrade method, an OTA upgrade package encryption method, a terminal device and a vehicle. The system upgrade method comprises the following steps: obtaining description information of an over-the-air technology based OTA upgrade package, wherein the description information comprises a key, an OTA public key, an OTA public key signature, and a digest signature of the OTA upgrade package; downloading an OTA upgrade package encrypted by the key; decrypting the OTA upgrade package by the key; verifying the legality of the decrypted OTA upgrade package by using a pre-configured root public key, the OTA public key signature, and the digest signature of the OTA upgrade package; and storing the decrypted OTA upgrade package, and performing system upgrade accordingto the decrypted OTA upgrade package if the OTA upgrade package is judged to be legal. The system upgrade method can ensure that the OTA upgrade package is not easily maliciously changed during the transmission process, and can ensure the integrity of the upgrade package, so as to verify the accuracy, thereby improving the security of the system upgrade.

The invention provides a system and method for three-dimensional dose verification in radiosurgery. First a patient's image data and radiotherapy planning data are led in through a data management module, a patient's actual illuminated three-dimensional dose field distribution during the treatment is quickly reconstructed by utilization of a two-dimensional field size dosage collecting module and a three-dimensional dose inversion module; three-dimensional evaluation analysis is performed on a irradiation dose and a planning dose through a dose evaluation module; whether an error is in a permissible range is judged; if the error is in the permissible range, the patient is treated according to a radiotherapy plan, or else the radiotherapy plan is adjusted or a radiosurgery accelerator is subjected to quality control detection, irradiation is not performed on the patient until the dose verification passes, and the consistency between the irradiation dose and the planning dose is guaranteed finally. According to the system and method for three-dimensional dose verification in radiosurgery, the shortcoming of two-dimensional plane dose verification of using water instead of a person prior to existing treatment is overcome, and precise dose verification of three-dimensional space dose field distribution in the patient prior to fractionated treatment and during the fractionated treatment can be implemented.

The invention relates to an elevator equipment fault prediction method based on depth learning, belonging to the technical field of elevator fault prediction. The technical problem to be solved is toprovide a timely and accurate method to predict the type and time of elevator failure. In order to solve the technical problems, the concrete steps of the invention are summarized as follows: firstly,the elevator fault record information is collected and a real-time elevator fault information database is established; then the elevator fault information is processed into event sequence and time sequence; the event sequence and time sequence are used as the input data of double LSTM respectively, and the output embedding of the two sequences is obtained by iterative training of the loop neuralnetwork; then, the background knowledge and the nonlinear representation of the historical influence of the intensity function are obtained through joint player combined with two output embedding; finally, according to the characterization results of the strength function, the fault type and time of elevator are predicted. The invention can assist elevator maintenance personnel to take relevant preventive measures as early as possible to avoid occurrence of failure events.

A method for monitoring and diagnosing the combustion dynamics of a gas turbine engine system includes mounting at least one sensor on an external surface of at least one combustor can, receiving a signal from the sensor mounted to the combustor can, validating an accuracy of the signal from the sensors, determining the combustion dynamics of the can based on the received signals, and generating an indication when a combustion dynamic threshold has been exceeded.

The invention discloses an analysis method of high-speed compressible supercavity flow characteristics and belongs to the field of fluid mechanical engineering. The analysis method comprises the stepsas follows: step 1, establishing a geometric model; step 2, performing mesh division on the geometric model established in step 1 and a fluid computational domain; step 3, setting boundary conditionsand a computing method; step 4, performing secondary development through CFD software to realize compressible correction of physical attributes of water; step 5, conducting numerical computation analysis on the high-speed compressible supercavity flow characteristics; step 6, performing numerical simulation analysis and comparing a numerical simulation result with an experiment result to verify accuracy and feasibility of the analysis method. The high-speed compressible supercavity flow law and mechanism can be revealed, so that a theoretical foundation is laid for structural design of underwater high-speed navigation bodies, and the actual application engineering problems of the high-speed navigation bodies can be solved.

The invention discloses a double-spectrum night vision instrument vehicle-mounted system and a double-spectrum fusion design method. The system comprises an infrared double-spectrum imaging die set used for imaging of a visible spectrum and an infrared spectrum and for collection, processing, fusion and transmission of the image data, an image receiving system used for completing the receiving of the image data, a display terminal used for displaying the images for an observer, and a control terminal used for controlling the function and on-off of an imaging mode. Due to the adoption of the system and the method, the problem that coaxial imaging of the double spectrum image is slight bad is solved, the machining precision requirement of the imaging system is reduced, more clear images are provided for forward observation for a train; a fusion method based on a scene moving object provides a reference source for the double-spectrum imaging system, the computation load is reduced, the registering reliability is improved, the imaging on the hardware is convenient, and the real-time performance is possessed; a weight fusion method based on multi-scale bilateral filtering is simple and feasible, improves the imaging detail information and reduces the calculation ability.

The invention relates to a modular multilevel converter (MMC)-based electromechanical transient simulation method and system of a flexible DC and DC power grid. The method comprises the step of building a transient simulation model, wherein the transient simulation model comprises a controlled current source-based steady-state power flow model, an AC-side transient model and a DC-side transient model, and the controlled current source-based steady-state power flow model is used for generating initial power flow data required by transient simulation. The method involves at least one additive module based on an outer ring control structure for function expansion, the additive module is used for setting a shunting signal between an AC side and a DC side of the flexible DC and DC power grid to generate a logic and is connected by the outer ring control structure of the AC side and the DC side of the flexible DC and DC power grid, and the electromechanical transient running characteristic of the flexible DC and DC power grid under different working conditions is simulated by the transient simulation model. By the method, the calculation of an initial value of a DC network is not considered, the steady-state power flow model is simplified on the premise of no influence on transient simulation accuracy and stability, and the expansibility of the model is improved.

The invention discloses a width transfer learning network and a rolling bearing fault diagnosis method based on the same, and belongs to the technical field of bearing fault diagnosis. The invention provides a novel width transfer learning network and a rolling bearing intelligent diagnosis method based on the same, and aims to solve the problems of scarcity of vibration data with mark informationof a rolling bearing under a variable load, large distribution difference between source domain data and target domain data in the same state, unbalanced distribution of multi-state data and low diagnosis accuracy and model training efficiency. According to the invention, a width learning system is utilized to extract features of source domain data and target domain data and construct a sample set, and on the basis, a balanced distribution adaptation method in transfer learning is adopted to reduce the difference between a source domain and a target domain. A chicken swarm algorithm is introduced to optimize width transfer learning network parameters and establishing a width transfer learning network model. The proposed network model is applied to rolling bearing fault intelligent diagnosis under the variable load, and an experimental result verifies the high efficiency and accuracy of the proposed method.

The invention relates to a method for fast assessment on bearing capacity of a bridge based on a moving load test. The method is characterized by comprising the steps of 1 determining loaded vehicles and monitoring point arrangement, 2 arranging sensors in a field according to a monitoring point arrangement scheme, 3 driving the vehicles in the field and acquiring dynamic response of the bridge, 4 extracting the dynamic response of a bridge structure, 5 calculating static response and calculating static deflection and static strain according to maximum dynamic deflection, maximum dynamic strain and power increasing coefficient extracted according to a dynamic response curve of the bridge, and 6 assessing the bearing capacity of a tested bridge. The moving load test is conducted on the vehicles under the static load test condition at certain speed, so that acquired dynamic response is converted into static response by utilizing a formula, then the static response is compared with a theoretical calculating value and a relevant specification specified value, and further the bearing capacity of the bridge structure is assessed according to current specifications. The method for fast assessment on the bearing capacity of the bridge based on the moving load test has the advantages of saving time and labor and basically having no influence on traffic.

An electromagnetic positioning and navigation device for a transcranial magnetic stimulator is composed of an electromagnetic positioning and navigation module (1), a coil gesture control module (2), a three-dimensional seat (3), an image collection and processing module (4), an electromagnetic calculation module (5), a display module (6) and a transcranial magnetic stimulator coil (7). The electromagnetic positioning and navigation module (1) is connected with the coil gesture control module (2) and the three-dimensional seat (3), and the coil gesture control module (2) is installed on the transcranial magnetic stimulator coil (7). The image collection and processing module (4) can rebuild a three-dimensional head model of a patient in the image space to mark therapeutic targets. The electromagnetic positioning and navigation module (1) records space position and coil gesture of the head (8) of the patient and conducting alignment on head positions in the image space and the sensor space and the position of the transcranial magnetic stimulator coil (7). Induction field distribution of the head of the patient is displayed in the display module (6) by means of calculation of the electromagnetic calculation module (5).

The present invention provides taxonomy system for enterprise data management and analysis. Building a taxonomy is based on defining taxonomy themes to solve different aspects of enterprise data management. For each theme, a taxonomy is created by classifying data related to the theme. Taxonomy nodes include enterprise data management policies and data analysis methods to be applied to data stored at the nodes. The amount of data that companies need to access and manage today is complex and voluminous. Because of computational requirements and time constraints, it is very difficult to classify and process all data. In many cases, current solutions ignore a sizable amount of data and process a small subset of it. This leads to incomplete and inaccurate results as ignored data may contain valuable information. The present invention provides taxonomies that encompass all data. Voluminous data is processed first to eliminate irrelevant data that is not related to the theme of a taxonomy. This in effect reduces the amount of data used to create the taxonomy without impacting its accuracy.

The invention discloses an RGB-D salient object detection method based on foreground and background optimization. The method comprises the following steps: initial foreground modeling is performed based on low-level feature contrast, and a superpixel-level initial salient figure is obtained; a middle-level aggregation processing is performed on the superpixel-level initial salient figure, and a middle-level salient figure is obtained; a high-level prior is introduced in the middle-level salient figure to improve the detection effect, and a foreground probability is generated; edge connectivity mixing depth information is calculated, and the edge connectivity is converted into a background probability; the foreground probability and the background probability are optimized, and a objective function is obtained; the objective function is solved, a optimal salient figure is obtained, and the detection of a salient object is realized. According to the invention, a optimization framework based on foreground and background measurement and the depth information of a scene is fully utilized by the invention, a high recall rate can be obtained, and the accuracy is high; the method can accurately position the salient object in different scenes and different sizes of objects and can also obtain nearly equal salience values in the target object.

The invention relates to a method for measuring interfacial heat transfer coefficient of the rapid cooling process of steel. The method includes the steps: 1 spot-welding a thermoelectric couple on the surface of a workpiece and connecting the thermoelectric couple to a temperature acquisition module; transmitting tested temperature change information to the temperature acquisition module by the aid of the thermoelectric couple when the workpiece is cooled; 3 reading temperature change data of a surface point by the aid of the temperature acquisition module; 4 obtaining the interfacial heat transfer coefficient of the cooling process by means of an interfacial heat transfer coefficient verifying function of heat treatment software; 5 using the obtained interfacial heat transfer coefficient for simulating change of a temperature field of the workpiece in the cooling process, computing the temperature change of a surface testing point, and comparing the temperature change with actual measurement results until computed results are well matched with actual measurement results; and 6 using the interfacial heat transfer coefficient obtained in the step 5 for simulating deformation of the workpiece in the cooling process until computed results are well matched with deformation tendency of actual measurement results. Therefore, the problem of inaccuracy of interfacial heat transfer coefficient is solved.

The invention discloses a method for forecasting the noise value of a propeller noise value in uneven inflow. The method for forecasting the noise value of the propeller in the uneven inflow comprises the steps that firstly, a CFD computation grid is divided, steady iterative computation is conducted so that the water flowing performance parameter and the water speed of an inflow opening can be worked out and the accuracy of a model is verified; secondly, CFD unsteady computation is conducted, the period state of blade cavitation and the blade cavitation area change of the propeller are recoded; thirdly, through CFD flow field data, acoustic boundary element value computation is conducted, and load noise of the propeller is worked out; finally, the noise of the propeller is worked out through the blade cavitation monopole radiation noise of the propeller and the load noise of the propeller.

The invention discloses a real-time compression tracking method of multi-characteristic transfer learning. The method comprises the following steps: generating a training sample, and respectively extracting a positive sample and a negative sample for training a classifier from a current frame and a video frame which previously tracks an object; extracting the characteristics of an object and a background in the training sample, projecting an extracted high-dimension characteristic to a low-dimension characteristic of a compression domain by use of two complementary sparse mapping matrixes, and generating two balanced characteristics for indicating the object and the background; constructing and updating the classifier, and training a naive Bayes classifier for classifying samples to be detected by use of the characteristic of the compression domain; and employing a secondary object search strategy. According to the invention, by use of the coarse-to-fine secondary search strategy, the quantity of generated scanning windows is reduced, the quantity of the samples to be detected is reduced, the object search process is accelerated, the scanning window with the maximum response value is taken as a tracked object, and the training sample and the classifier are updated based on this.

Systems and methods for automatically matching in real-time an advertisement with a video desired to be viewed by a user are provided. A database is created that stores one or more attributes (e.g., visual metadata relating to objects, faces, scene classifications, pornography detection, scene segmentation, production quality, fingerprinting) associated with a plurality of videos. Supervised machine learning can be used to create signatures that uniquely identify particular attributes of interest, which can then be used to generate the attributes associated with the plurality of videos. When a user requests to view an on-line video having associated with it an advertisement, an advertisement can be selected for display with the video based on matching an advertiser's requirements or campaign parameters with the stored attributes associated with the requested video, with the user's information, or a combination thereof. The displayed advertisement can function as a hyperlink that allows a user to select to receive additional information about the advertisement. The performance or effectiveness of the selected advertisements can be measured and recorded.

The present invention provides an experimental apparatus and a method for petroleum pipe scaling and wax precipitation, the apparatus comprises a liquid storage tank, a variable frequency motor, a screw pump, a constant temperature water bath tank, a first test tube, a second test tube, a cover plate and a lacing film test sample. The liquid storage tank is used for containing oil water needed for the experiment, and is connected with a temperature control system and a heater to ensure oil water temperature is constant at a preset value; a predetermined flow rate of oil output can be achieved by the variable frequency motor and the screw pump, the constant temperature water bath tank is used for simulation of test tube section wall temperature to study changing laws and mechanisms of sediment characteristics along with pipe wall temperature differences; the first test tube is a full size pipe segment, and is used for fully simulating pipe scaling and wax precipitation, and the second test tube is a pipe segment with an opening, and is used for study on the laws and mechanisms of the sediment characteristics along with time. The experimental apparatus and method can provide new methods and supports for research and development and evaluation of measures scaling and wax removal and prevention, and can ensure the petroleum pipe safe, economic and efficient operation.

The invention discloses a coupling calculation method of deepwater gas-liquid two-phase flow circulating temperature and pressure. The coupling calculation method comprises the following steps: (1) calculating the vertical coordinate of a grid; (2) applying the initial condition; and (3) from the beginning of an original value, carrying out iterative calculation on node temperature and pressure data of drill string drilling fluid and annulus drilling fluid in sequence, until the temperature and the pressure reach the condition of convergence, finishing the iteration, and saving and outputting the last-time iterative calculation result as the final deep sea gas-liquid two-phase flow shaft temperature and pressure simulation result, wherein the iteration is named as a global iteration. The coupling calculation method improves the calculation precision.

The invention discloses a method for establishing an electromagnetic transient simulation model of a large-scale AC-DC (Alternating Current - Direct Current) power system. At first, an AC-DC power system network is simplified in an electromechanical transient mode by a power system dynamic equivalence method through the utilization of the electromechanical transient simulation data of the large-scale AC-DC power system; secondly, a conversion correspondence principle of an electromechanical transient element model and an electromagnetic transient element model is provided, so that the electromagnetic transient simulation model is established according to the simplified electromechanical transient simulation model of the AC-DC power system. By the adoption of the method provided by the invention, the electromagnetic transient simulation model can be established directly according to the electromechanical transient simulation data of the large-scale AC-DC power system; moreover, the dynamic characteristics of a prototype system are kept; therefore, an AC system and a DC system can be further studied in terms of detailed interaction, multi-direct-current commutation failure characteristics and recovery characteristics after direct current commutation failure.

The invention discloses a microblog similar account detection method based on graph analysis clustering and multidimensional similarity calculation. The method comprises the following specific contents that: S1: converting a malicious account identification problem into a user similarity calculation problem; S2: constructing a directed graph by user information through graph calculation; S3: usinga graph analysis algorithm to cluster users; S4: importing density weight d, and filtering users with sparse data; S5: imprinting an MDUS (Multi-Dimensional User Similarity) algorithm, and calculating a similarity on the basis of multidimensional information; S6: using an analytic hierarchy process to calculate the weight of each dimension to obtain a weighting similarity; and S7: obtaining the data of m users by a crawler, inputting target user information in a spark experiment, and obtaining a similar account set as a suspicious malicious account, wherein the accuracy of the MDUS algorithmcan be 80%. By use of the method, the graph analysis clustering and the multidimensional similarity calculation are combined to realize a purpose that exceptional accounts are quickly found, and the method has an important meaning for maintaining the stability of a social network site.