35results about How to "Demonstrate effectiveness" patented technology

Disclosed is a method and system for detecting malicious entities and malicious behavior in a time evolving network via a graph framework by modeling activity in a network graph representing associations between entities. The system utilizes classification methods to give score predictions indicative of a degree of suspected maliciousness, and presents a unified graph inference method for surfacing previously undetected malicious entities that utilizes both the structure and behavioral features to detect malicious entities.

The invention relates to an active distribution network energy-storage system dynamic planning method which studies active distribution networks which contain distributive photovoltaic and energy-storage systems. The method herein optimizes an output strategy of the energy-storage system on the basis of a bilevel programming model in combination with the shortest path theory so as to determine the installed capacity of the distributive photovoltaic and energy-storage system under the optimal output strategy of the energy-storage system. For the bilevel programming model, upper level programming determines the installed capacity of the distributive photovoltaic and energy-storage system with the minimum annual general fee of a power distribution network as the goal and generates a planning scheme, and the lower level planning optimizes the output strategy of the energy-storage system in one operation cycle and in one day with the minimum typical daily operation fee and load peak-valley difference of the power distribution network as the goal. Compared with prior art, according to the invention, the method herein can effectively improve operation characteristic of the active distribution network and optimization configuration of the distributive power supply and energy-storage system.

The invention discloses a depth video significance detection method based on motion and memory information; the method comprises the following steps: obtaining a detection data set formed by an imagesignificance object detection data set and a video eye portion setpoint detection data set; calculating a ground true graph of the video eye portion setpoint detection data set, and obtaining a finaleye portion location map of the current frame; building four models used for extracting local information and global information and with different anti-convolution layers; pre-training the four models on the image significance object detection data set, and fine-tuning the pre-trained four models on the video eye portion setpoint detection data set; using a significance moving object boundary detection algorithm to extract motion information between two video frames, using the detection result graph of the previous frame as memory information, and integrating the memory information and motioninformation into a depth model SGF (E), thus realizing point-to-point detection. The method can detect valid eye portion setpoints of a video.

The invention discloses a method for predicting and warning a soil heavy metal pollution area. The method comprises: collecting and preprocessing an original soil sample; carrying out sample separation and detection; carrying out data processing and analysis; and building a prediction model. According to the method disclosed by the invention, the soil-attached plants are received into the heavy metal detection range, so that the detection contents become comprehensive and accurate; and on the basis of combination of the heavy metal effective state theory with the GIS visualization technology,the prediction result has the great practical significance, so that the theoretical basis and scientific basis are provided for ecological risk evaluation, control, treatment, and repairing of the heavy metal pollution on soil.

The invention discloses a parametric modeling based stiffened wall plate structure vibration fatigue resistance optimization method and relates to an aircraftstiffened wall plate structure. The method comprises the following steps of 1 stiffened wall plate structure design, 2 optimization variable selection, 3 optimization flow and program development performing based on commercially used finite element software, 4 finite element parameter modeling, 5 random vibration response calculation under basic accelerated speedbroadbandstimulation, 6 extraction of a global stressroot-mean-square value, to serve as an optimization objective,in random vibration response, and optimization using structure optimization parameters and structural weight as constraints. The shortcomings of shape optimization and topological optimization can be overcome, and a selectable optimization parameter range is wider. The convenience and effectiveness of the parametric modeling based optimization method are embodied based on a commercially used finite element software platform, the optimization method can be suitable for a more complicated engineering structure. The optimization method is oriented to the global stressroot-mean-square value of a structure, aglobally optimal solution of the structure under a random load is obtained, and the vibration fatigue life of the structure can be effectively prolonged.

The invention discloses an administrative affair management system, which comprises a leave management system comprising leave types of official travel examined and approved leaves and national ruled leaves, a comprehensive management system which comprises examination and approval items such as official reception, use of official vehicles, object purchase, vehicle maintenance, consumable supply, rejection, breakage and the like, and circulation items, such as work instruction report, department work coordination, leadership task assignment, and is provided with a leadership mail box function for receiving different opinions and views at any time and accepting good advices, and a charge management system comprising China Taxation Administration Information System (CTAIS) authority adjustment, case transfer, 'four small ticket' checking, manager working paper and application examination and approval items. The invention provides a quick and effective platform, so that requirements of 'efficiency, convenience, standardization and conversation' on administrative affairs are met.

The invention discloses a fast calculation method for synchronous compression transformation and reconstruction, and the method comprises the steps: 1) selecting the standard deviation sigma of a Gaussian window, and calculating the length L of the window; 2) selecting a window function jumping point H; 3) performing overlapping segmentation on the original signal x [n], wherein the number of overlapping points is L-H, constructing a matrix X by taking each section of data as each column of the matrix; 4) multiplying the invariable discrete sequence g [n] corresponding to the window function by each column of the matrix X, and then performing FFT of Nf points on each column to obtain Sx [m, k]; 5) according to the selected frequency band, intercepting a sub-matrix for the Sx [m, k], and only rearranging along the frequency to obtain a time-frequency matrix Tx [m, l] after synchronous compression transformation; 6) performing inverse discrete Fourier transform on each column of Tx [m, 1] to obtain a double-variable function r [m, 1], and then multiplying the mth sequence r [m, 1] by a sliding window g [n-mH]; and adding the results to obtain a reconstructed signal. According to themethod, the operation efficiency is greatly improved while the excellent time-frequency aggregation of the synchronous compression transformation is ensured, and the reconstruction of the synchronouscompression inverse transformation can still be realized under a fast algorithm.

The invention discloses a color-based batching method for an automobile coating workshop, which is mainly characterized in that the physical position of a white automobile body in a sequence is kept unchanged, color attributes are adjusted among automobile bodies of the same type under the constraint that the original order demand in the sequence is not changed, and the purpose is to minimize theswitching frequency of coating colors in the sequence. The heuristic algorithm and the rolling solving method based on ant colony optimization can quickly and effectively solve the problem of color-based batching of different scales, and significantly reduce the number of times of color switching in an original sequence. Besides, according to the color-based batching method, hardware facilities such as an automatic access system or a linear buffer area do not need to be adopted, and color-based batching is realized by performing order adjustment between the same type of body-in-white, so thatthe occupied area of a workshop can be saved, and high equipment and facility construction and operation costs can be saved.

The invention discloses a multi-layer control adaptive micro-service system, which comprises: a micro-service system as a target system, wherein the micro-service system is composed of a series of micro-services, and the micro-services are organized into application programs through lightweight protocols; an infrastructure Control Layer (ICL), which is used to solve the adaptive problems at the infrastructure level of the adaptive micro-service system; and an Application Control Layer (ACL), as the upper layer of adaptive micro-service system, solving the adaptive problem of application layerof adaptive micro-service system. The invention has the advantages of simple principle, easy realization, meeting the requirements of dynamic change and the like.

The invention relates to the field of high-speed train simulation models, in particular to a locomotive multi-body structure dynamic model correction method. The locomotive multi-body structure dynamic model correction method comprises the steps of locomotive multi-body structure dynamic modeling and analysis, locomotive dynamic test modeling and analysis, correlation analysis between calculation and testing, locomotive model correction parameter selection, sensitivity and residual analysis and iteration correctio based on sensitivity analysis. The locomotive multi-body dynamic model correction is based on a commercial multi-body dynamic software platform, can obtain an accurate simulation model, embodies the convenience and effectiveness of a parametric modeling correction method, can be applicable to a complicated engineering structure, can be popularized to other multi-body dynamic model correction of a locomotive and meanwhile has very high practicability.

The invention discloses an administrative affair management system. The administrative affair management system comprises a leave asking management system, a total management system and a charge management system, wherein the leave asking management system comprises approval for going out for official business and leave asking regulated by the state, the total management system comprises examination and approval, such as official reception, outgoing of public cars, article purchasing, vehicle maintenance, getting and usage of articles, waste and breakage report, and dynamic items, such as work asking and reporting, desk work coordinating, and assigning of tasks of leaders, establishment of leader main boxes for receiving different opinions and views at any time, and wide receiving of good opinions, and the charge management system comprises CTAIS permission adjustment, case transfer, 'four receipt' check, administrator working paper and examination and approval of application. The administrative affair management system provides a fast, effective and good platform, and requirements for efficient, convenient, standard and saving management of administrative affairs are met.

The invention relates to a road side unit placement method based on an affinity propagation algorithm. The road side unit placement method includes the following steps: firstly, building a road network model according to traffic scenes; then building a mathematical model: referring to factors such as the traffic density, the vehicle speeds and the distances to define influence factors, converting a road side unit placement problem into a clustering problem, and defining a corresponding objective function; obtaining vehicle track information, and solving the influence factors according to the vehicle track information; finally, optimizing the objective function through the affinity propagation algorithm, and obtaining an optimal road side unit placement intersection set. According to the road side unit placement method, the traffic condition information is sufficiently used for finding the optimal road side unit placement positions through the affinity propagation algorithm, when road side units are placed at the positions, the delivery rate of network data packages can be remarkably improved, and the validity of the road side unit placement method is also shown through evaluation results. In addition, the road side unit placement method is low in time complexity and can be rapidly converged.

The invention discloses a TV set, comprising image input interface, user operation unit, image processing unit and image display unit, where the image input interface is connected to the image processing unit, the image processing unit is connected to the image display unit, the user operation unit is used to generate corresponding control instructions according to user operations, and the image processing unit is used to process image signals inputted from the image input interface according to the control instructions, and transmits the processed image signals to the image display unit, where the processing comprises dividing image and intensifying integral image display effect or partial image display effect after division. And the invention discloses an image displaying method for the TV set.

The invention provides a method and a device for measuring the performance of search engines. The method includes rating the correlation of at least one query result item obtained by at least one piece of Query by means of querying the to-be-measured search engines so as to obtain the first theoretical maximum score values of home page query results obtained by the various pieces of Query from the to-be-measured search engines; determining second theoretical maximum score values of the home page query results obtained by the various pieces of Query from the to-be-measured search engines according to rated values of the correlation of query result items obtained by the various pieces of Query from at least one reference search engine; measuring the performance of the to-be-measured search engines according to the first theoretical maximum score values and the second theoretical maximum score values. The method and the device in an embodiment of the invention have the advantages that in-depth and refined measurement results can be provided by the method and the device, and accordingly follow-up search engine optimization work can be effectively guided.

The invention belongs to the technical field of intelligent building equipment self-control alarm, and discloses an intelligent building equipment self-control alarm device which comprises a data acquisition module, an image acquisition module, an energy consumption acquisition module, a central control and processing module, an alarm module, a control execution module, a storage module, a power supply module, a display module and a signal transmission module. According to the invention, information acquired by the data acquisition module, the image acquisition module and the energy consumption acquisition module can be processed and analyzed through the central processing unit, and the environment information and the working state information of the building equipment are judged according to detection data, so that the abnormal state of the equipment is detected; the abnormal information is subjected to abnormal grade division, so that different types of alarm reminding can be carried out according to different abnormal grades; and the field controller and the execution mechanism are controlled to carry out corresponding control and adjustment processing, unmanned intelligent control and monitoring are achieved, and problems can be solved most effectively and reasonably.

The invention discloses an immune network theory based production scheduling optimization algorithm. According to an information processing mechanism in an immune system, an immune network theory in an artificial immune system is introduced into a production scheduling system to design an unique network based immune scheduling algorithm; a unique network for optimal scheduling rules is establishedaccording to the unique network theory, an effective antibody structure and an antigenic structure are defined, and an optimal scheduling rule and a current workshop state are defined as antibodies while a workshop state is defined as an antigen; a process of solving scheduling problems through the immune scheduling algorithm is explained through antibody and antigen matching, calculation of affinity of the antibodies, antibody concentration calculation and antibody selection. The immune network theory based production scheduling optimization algorithm is far superior to other single heuristic rule scheduling algorithms and shows efficiency and effectiveness of the immune scheduling algorithm in processing of multi-target dynamic scheduling problems.

The invention provides a demand response method considering user real-time comfort and a power grid load rate. Different from a passive control terminal in the traditional sense, the method considersa schedulable strategy of actively participating in response to improve the operation load rate of the power grid by integrating the will of a user. Firstly, household appliances of the user are classified, types of the intelligent household electrical appliances are determined, the comfort index of the user is calculated based on operation characteristics, a working state and the external environment information of the intelligent household electrical appliances, a power grid power scheduling limit value is set according to the power grid load rate improvement demand, and finally the comfortindex of the user is updated in real time to realize demand response control of the intelligent household electrical appliances. The method is advantaged in that the comfort level of the user is ensured, and the operation load rate of the power grid can be effectively improved.

The invention relates to the technical field of optical fiber temperature measurement, in particular to a fluorescent optical fiber temperature measurement system for biomedicine, which comprises a probe, a coupler and a driving circuit, the output end of the coupler is connected with a lens I through an optical fiber, the coupler is connected with the probe through a Y-shaped optical fiber, and the probe is connected with a lens III through a Y-shaped optical fiber. The sides, away from the optical fiber, of the second lens and the third lens are each provided with a laser sheet, the sides, away from the optical fiber, of the laser sheets are connected with a photoelectric detector, the output end of the photoelectric detector is connected with a signal amplification circuit, and the output end of the amplification circuit is connected with a filter. The filter output end connected with one end of the Y-shaped optical fiber is connected with a division circuit. According to the fluorescence optical fiber temperature measurement system for biomedicine, the reference signal related to the reference excitation signal is adopted, the influence of inherent excitation light leakage of a phase locking scheme on a fluorescence fading signal is successfully solved, and the measurement precision is remarkably improved.

The invention provides a network embedding learning method based on topology perception text representation, which comprises the following steps of: adaptively generating a topology perception filter by using local topology structure information of a node, and learning the text representation so as to obtain the topology perception text representation. Topological structure information is more effectively fused into mining of text representation; besides, the method can be combined with an existing network embedding model based on context awareness, the application range is wider, the performance of link prediction and node classification tasks is improved, and the effectiveness of network node representation learned by the method is reflected.

The present invention relates to a beverage with functions of nourishing the stomach, benefiting qi, promoting the secretion of saliva or body fluid and harmonizing viscus qi and blood and a preparation method thereof. The beverage is prepared from the following raw materials: 50-90 parts by weight of caulis spatholobi, 30-60 parts by weight of moringa oleifera and 20-50 parts by weight of red dates; and the moringa oleifera successively contains tender branches and stems, leaves and flowers, and seeds at a preferred mass ratio of 10:1:1. The beverage is prepared by selecting purely natural plants containing natural active ingredients as raw material components, scientific in component ratios, easy to obtain in the raw materials, relatively low in production cost, and also free of any pigments and preservatives, presents natural brownish red tea color, and is refreshing and sweet in mouthfeel, clear and transparent in appearance, and free of impurities. Specifically, the medicinal andedible plants of the caulis spatholobi, moringa oleifera stems, leaves and seeds, and red dates are selected as the main components. Besides, according to a compatibility and supplement principle of traditional Chinese medicine soup drink, the beverage effectively utilizes nutrients for promoting qi and blood, and effectively reflects scientificity, practicalness, safety and effectiveness of cold,hot, warm and cool natures and sour, bitter, sweet, pungent and salty tastes of natural herbaceous plants.

The invention provides a parking stall selection method and system of a parking lot. A vehicle owner starts software to search for parking lots nearby when arriving at a destination, one parking lot is selected according to willing of the vehicle owner, an entrance retrieval device uploads vehicle information when a vehicle enters the selected parking lot, a broadcast instruction to unlock a parking stall lock is issued automatically via a background service, the vehicle owner can lock the parking stall, and drives directly to the parking stall, and parking stall information is stored. When the vehicle owner takes the vehicle, the parking stall can be found according to the parking stall information in the mobile phone software directly, information of a background database is updated whenthe vehicle passes an exit retrieval device, and the parking stall selected by the vehicle owner is recovered to a vacant state automatically. Jamming near the parking lot and in the parking lot canbe avoided; fewer management staff are needed in the parking lot, and the management cost of the parking lot can be reduced; the vehicle owner spends less time searching for the parking stall; the vehicle owner can find the position of the vehicle conveniently; and the parking stall lock is mounted additionally so that the parking stall can be reserved more accurately.

The invention relates to a twin classifier certainty maximization method for a cross-domain complex visual task; the method comprises the following steps: 1, constructing a neural network frame, and initializing the neural network frame; 2, inputting the source domain sample into a feature generator G to obtain a corresponding feature representation; and 3, under the supervision of the label information of the source domain sample, calculating an empirical risk error between a model prediction output p and a real label y on the source domain sample by using a standard cross entropy loss function. The method has the advantages that a novel classifier deterministic difference measurement CDD is designed, wherein the difference of the classifiers is measured by means of category correlation between target prediction of the twin classifiers, and implicit constraints are applied to the identifiability of target features at the same time.

The invention discloses an information service system for engineering management, and the system comprises an investment control module, a progress control module, a quality control module, a contract management module, a construction management module, a safety management module, and an engineering data management module. According to the method and the system, the collected project information is classified according to the generation time node of the project information, and the classified project information is correspondingly stored by establishing an information storage unit in a computer, so that the collected project information is quickly queried, and the query efficiency is improved. According to the technical scheme of the invention, the engineering project information has higher standardization, so that each piece of engineering project information can effectively play a practical role, the regulation of engineering information management is perfected, the effectiveness and sharing of the engineering project information management are reflected, and the engineering information management is developed in a more convenient and normalized direction.

The invention relates to a large-scale edge machine learning training method based on probabilistic sampling. The training method mainly comprises the following steps: 1, setting a training model and training parameters in a server; 2, performing data preprocessing on each device and preparing a local training data set; and 3, uploading gradient parameters obtained by local training to each device, uploading the gradient parameters to a server and the like. The learning training method has the advantages that large-scale edge equipment can be effectively trained, through probabilistic sampling, the process of judging synchronization is not limited by the size of the scale any more, the expandability of a training system can be effectively improved, large-scale edge training is supported, conciseness and effectiveness of the learning training method are shown.

The invention relates to the technical field of power distribution network fault recovery, and particularly provides an SOP-considered active power distribution network fault recovery method and device, and the method comprises the steps: determining the SOP wheeling capability based on a preset SOP wheeling capability constraint condition; substituting the SOP transfer capacity into a pre-constructed active power distribution network fault recovery model, and solving the pre-constructed active power distribution network fault recovery model to obtain a power distribution network branch on-off adjustment vector for fault recovery; wherein the pre-constructed active power distribution network fault recovery model comprises a target function, a model optimization variable and a constraint condition configured for the pre-constructed active power distribution network fault recovery model. According to the technical scheme provided by the invention, limitation of the transfer power supply capacity of the SOP in the active power distribution network is considered, power supply recovery of the power distribution network is realized, and line loss generated in a fault recovery process is reduced.