3696results about How to "Clear structure" patented technology

The invention discloses a method for accelerating convolution neutral network hardware and an AXI bus IP core thereof. The method comprises the first step of performing operation and converting a convolution layer into matrix multiplication of a matrix A with m lines and K columns and a matrix B with K lines and n columns; the second step of dividing the matrix result into matrix subblocks with m lines and n columns; the third step of starting a matrix multiplier to prefetch the operation number of the matrix subblocks; and the fourth step of causing the matrix multiplier to execute the calculation of the matrix subblocks and writing the result back to a main memory. The IP core comprises an AXI bus interface module, a prefetching unit, a flow mapper and a matrix multiplier. The matrix multiplier comprises a chain type DMA and a processing unit array, the processing unit array is composed of a plurality of processing units through chain structure arrangement, and the processing unit of a chain head is connected with the chain type DMA. The method can support various convolution neutral network structures and has the advantages of high calculation efficiency and performance, less requirements for on-chip storage resources and off-chip storage bandwidth, small in communication overhead, convenience in unit component upgrading and improvement and good universality.

The invention discloses a clustering analysis and decision tree algorithm-based truck loading work time prediction model. A clustering analysis and decision tree mixed algorithm is introduced, factors influencing inventory control are abstracted out, related historical data serves as a training sample, and finally the truck loading work time can be effectively predicted by using a trained decision tree data model; and the historical data of truck loading is deeply mined by utilizing a data mining technology based on a demand, and an available, easy-to-use and high-accuracy data model is generated. The clustering analysis and the decision tree algorithm are combined and complement each other, so that the accuracy of the data model is improved; an optimization policy is adopted for an original decision tree algorithm under the condition of establishing a simple and accurate data model, so that the calculation amount is reduced and the algorithm efficiency is improved; and through the data model, a relatively accurate time interval of cargo loading can be predicted and used for better manual decision-making.

The invention discloses a code automatic generation method used for constructing a code automatic generation device, comprising a data layer, a component layer, a process layer, a control layer and a user interface layer, wherein, the five-layer structure is divided in accordance with functions, and every layer is independent and creates a corresponding processing model to complete data processing, component extraction, flow processing, function control and user interface processing. At the same time, a code template is created based on certain rules, and different processing codes are formed for different business objects. Based on the five-layer model and the code template, a code generating engine controls and calls the model of layers and the code template, which cooperate to complete the code automatic generation process. The invention also discloses a code automatic generation device and system; by adopting the invention, the development efficiency, the reusability and the standardization of software can be enhanced and code quality is improved.

The invention relates to a pallet machine which can automatically pallet bagged materials as bulked cement, chemical materials, rice or the like. The invention comprises a feeder, a bag rotation transmitter, a compression transmitter, a buffer transmitter, a grouped transmitter, a grouped bag pusher, a layering machine, a side face shape corrector, a vibration shape corrector, a lifter, a disc feeder, a pallet bearing outputting device, an output transmitter, a pallet disc storage transmitter, an electric control cabinet, a walk bench, and an arm rest lift. The invention has reasonable structure, low cost, and high speed, which combines a plurality of single machines, with clear and reasonable structure and simple maintenance.

Re-use and platform independence of application software is achieved by using component types comprising an interface and a body. This is followed by automatically generating components based on the developed component types. The development process steps are preceded by an adaptation process step, during which the functionality of all component types is defined for all the platforms.

The invention discloses an actually-enhanced driver-assisted early warning system, which comprises a vehicle camera unit, a central processor unit and a head-up display unit. The vehicle camera unit comprises a front camera for shooting circumstance of a road ahead and two face cameras for shooting head information of a driver, wherein the front camera is arranged in the front of a middle-rear view mirror; the face cameras are arranged in an instrument panel symmetrically; and the three cameras are connected with the central processor unit. The central processor unit is used for identifying the specific locations of external risk factors according to the specific image contents of the front camera, and marking the locations in the images. The head-up display unit is arranged in the front of a steering wheel and below a front windshield and is used for projecting modified identification information to the front windshield together with warning prompt patterns after receiving an early warning command. The actually-enhanced driver-assisted early warning system can effectively avoid the distraction of the driver so as to improve an early warning effect.

The invention discloses a fault diagnosis and assessment method of an intelligent substation. The fault diagnosis and assessment method is characterized by including the specific steps that a fault diagnosis and assessment information model of the intelligent substation is constructed; fault and assessment indexes of a primary device, a secondary device and a network device in the intelligent substation are calculated on the basis of the fault diagnosis and assessment information model; a Petri net is used to perform primary fault diagnosis and assessment to the intelligent substation; and final diagnosis and assessment are performed by means of a distributed type expert system based on a knowledge base. The fault diagnosis and assessment method has the advantages of being clear in the diagnosis process, fast and practical in diagnosis methods and accurate and reliable in diagnosis results.

A city emergency evacuation simulation system based on multi intelligent agent belongs to the complex system simulation in the technical field of computerized emulation and solves the problems of single function of emergency simulation systems of all kinds and lack of intelligence of code of conduct of emergency entity. The system comprise a database module, a communication module, a parameter configuration interface module, a data management module, an intelligent agent generation module, a program execution module, a simulation module, a commanding and decision-making module and an assessment module. The database module comprises a generalized database, a knowledge base and a model base; wherein, the generalized database comprises a dangerous source subdatabase, a resource subdatabase, a foundational geodata subdatabase. The system can establish an efficient intelligent model, describes the precise behavioral process of emergent intelligent agents of all kinds in the state of disasters and emergency from two levels of macroscopy and microcosmic, can custom-make a parameter and expanded intelligent model, has commonality and expansibility, and makes up the shortage and deficiency of the city emergency evacuation simulation study in regionality and comparatively large ranges.

The invention relates to a three-dimensional reconstruction algorithm introducing an interactive operation for a user, belonging to the technical field of a computer multimedia. The method is as followings: the user utilizes a plurality of cameras to obtain a two-dimensional image Ii of a plurality of viewpoints of an object and extracts a contour map of the object; the computer obtains a visual shell model of a real scene according to a geometric parameter of every camera and the contour map, and turns the visual shell model into a point cloud form; a two-dimensional characteristic in the image is extracted to optimize a part of a real surface model corresponding to the characteristics; the user divides the image Ii into a region of a smoothness and the region of a subsidence; a nonuniform weighted graph is set up according to the areas selected above to find out a minimal cut of the image to obtain the optimized scene surface model. Combined with an auxiliary operation of the user, the three-dimensional reconstruction algorithm is capable of reconstructing the high-quality three-dimensional model of the object relatively fast according to the multi-view image which is shot of the actual object and the camera parameter corresponding to each image.

The invention discloses a vehicle cooperative formation driving system, belonging to the field of vehicle network-connected automatic driving. The vehicle cooperative formation driving system comprises a network layer, a link layer, a coordination layer, a control layer and a physical layer from top to bottom; wherein a vehicular unit in the network layer is communicated with a wireless router inthe link layer through a wireless network; the wireless router is communicated with a central service unit CSU and a road side unit RSU respectively through Ethernet; and the ZeroMQ communication is adopted by the central service unit CSU and the road side unit RSU. The vehicle cooperative formation driving system adds low cost vehicular sensors such as vision and millimeter wave radars, and integrates perception fusion, path planning and decision control into a same software architecture, so as to integrate vehicle road environment information, and achieve reliable vehicle formation driving function through high precision controllers; the vehicle cooperative formation driving system is in line with the development trend of the network-connected vehicle formation driving, which is highly modular and easy to expand function; and the decentralized following strategy can reduce the impact of network communication uncertainty.

The invention belongs to the field of networking automatic driving of vehicles, and discloses a cooperative vehicle formation driving method and a cooperative vehicle formation driving system. Sensinginformation of vehicles and absolute position information of the vehicles are captured for sensing fusion, path planning and decision-making control, and road environment information of the vehiclesis integrated to implement formation converging, following, quitting, parking and cooperative lane changing co-driving strategies for the vehicles. According to the method and the system, connection of all modules and direct communication between the modules are implemented in a 5G (5th-generation) manner, meanwhile, low-cost vehicle-mounted sensors such as a visual sensor and a millimeter wave radar are additionally arranged, sensing fusion, path planning and decision-making control are integrated into the same software architecture, the road environment information of the vehicles is integrated, and a reliable vehicle formation driving function is realized by a high-accuracy controller. The system is consistent with the development trend of networking vehicle formation driving, high in modulation degree and easy to functionally extend, and by a decentralized following strategy, the influence of uncertainties of network communication can be reduced.

The invention provides an implement method of lightweight-class global multi-dimensional remote-sensing image network map service. The implement method comprises the following steps: a global geographic space pyramid model is constructed, and a Tile file encoding system is built according to the model; pre-treatment mainly comprising projection transformation, file format conversion and the like is performed on the remote-sensing data so as to cut the remote-sensing data into tiles, and file encoding is performed; according to a distribution strategy, Tile files are grouped to generate a corresponding directory structure, and different images are updated on different Web servers; the visible range of the map requested by a client side is analyzed to calculate the storage paths of all images within the visible range, and the images are loaded and fed back to the client side. According to the invention, the implement method of the light-weight-class global multi-dimensional remote-sensing image network map service can be provided for a remote-sensing data center (such as the China Center for Resources Satellite Data and Application) and can provide remote-sensing data service to users more conveniently.

The present invention discloses a micro service servitization application system. The system is constructed by using a micro service architecture and carries out unified configuration and management on micro-applications and micro services according to a unified application architecture. The system comprises: an infrastructure unit, for providing soft hardware resources and realizing virtual management of the network; a service unit, for realizing atomic services and carrying out classification management, carrying out service realization and service management by using the technology provided in an integrated platform, and providing unified service access support for the realization of the micro-applications; and an application unit, for carrying out groupware encapsulation of interactive interfaces on the applications and services, and realizing the micro-applications capable of directing operating according to actual needs, carrying out pool management on the micro-applications, clustering to form the application, and providing a unified access entry for various types of terminals. According to the system disclosed by the present invention, problems existing in the transformation of existing large-scale traditional enterprise information can be overcome, micro-service design of the information system is realized, and stability and reliability of the system are improved.

A method for collecting and issuing real time dynamic traffic information includes receiving source data collected by floating vehicle, carrying out treatment periodically on source data from various floating vehicles, obtaining road state information of each road section traveled by each floating vehicle through calculation, integrating and treating road state information of multiple traveled road section to calculate out road information of each unit element in traffic path then issuing relevant traffic information data according to requirement.

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 discloses a method for performing inter-process communication between a server and a client by service management. The method is characterized by comprising the following steps that: an operating system starts a server process and registers the server process by service management; a client process requests the service management for using the server process to transmit a message to the server; and the server processes the message and returns the processed message to the client process. By the method, the coupling of the operating system module is reduced, and great convenience is brought to the management on the inter-process communication.

The invention discloses a manufacturing method of paraffin sections of zostera marina embryo, comprising the following steps: (1) material drawing and fixation: zostera marina seeds are taken and the fruit skin is stripped, and after endosperm is removed, an integral zostera marina embryo is soaked in FAA (free amino acids) fixing solution and is fixed for more than 48 hours; (2) dehydration: the zostera marina embryo is taken out from the fixing solution and is soaked in alcohol for dehydration; (3) waxing and embedding: paraffin is added gradually in a vessel containing the zostera marina embryo and dimethylbenzene, and conventional paraffin embedding is carried out to obtain the wax blocks containing the zostera marina embryo; (4) slicing, spreading and drying: the wax blocks coated with the zostera marina embryo is fixed on a wheel rotation type slicing machine for carrying out continuous slicing of the paraffin, so as to obtain wax bands and enable the wax bands to be attached to an object slide; (5) dewaxing, rehydration and dyeing of hematoxylin dye solution; (6) slicing, dehydration and dyeing of eosin and alcohol; and (7) transparence and mounting are carried out to manufacture a permanent section. In the manufacturing method disclosed by the invention, the paraffin sections of the zostera marina embryo, which have clear dyeing and an integral texture structure, can be obtained.

The invention discloses a multi-head numerical control drilling machine, which comprises a base, a tool changing mechanism, a drill bit mechanism and a feeding mechanism, wherein the base is provided with a workbench, an upright post is fixedly arranged on the base, and the top of the upright post is provided with the tool changing mechanism; a supporting frame is fixedly arranged on the base; the drill bit mechanism is arranged on the supporting frame, the drill bit mechanism is connected with the feeding mechanism, and the feeding mechanism drives a main shaft sleeve of the drill bit mechanism to move along the vertical direction so that a pneumatic clamp arranged on the main shaft sleeve can move along the vertical direction; and the control ends of the tool changing mechanism, the drill bit mechanism and the feeding mechanism are connected with the signal output end of a numerical control system. The tool changing mechanism, the drill bit mechanism and the feeding mechanism are mutually independent, so the drilling machine has clear and compact structure and is easy to maintain; and the mechanisms are controlled uniformly by the numerical control system so as to increase the operation flexibility, improve the production efficiency and reduce the cost. The drilling machine can be widely applied in the machining field.

An axially substituted silicon-phthalocyanin match and its composition are disclosed, which have high optical spectrum characteristics and can be used as the photosensitizer used for photodynamic therapy to tumor and the diseases other than cancer, photodynamic diagnosis and photodynamic disinfecting.

The invention relates to a networking method of a wireless electricity consumption information acquisition ad hoc network. According to the technical scheme, the networking method of the wireless electricity consumption information acquisition ad hoc network includes the steps that first, the network is established; second, child nodes without access to the network search for a network center coordinator or child nodes with access to the network; third, the child nodes without access to the network which are searching for the network select the network center coordinator or the child nodes with access to the network as ancestor nodes and send network access requests to the ancestor nodes; fourth, after the ancestor nodes receive the network access requests, network access addresses are allocated to the child nodes applying for access to the network; fifth, the child nodes applying for access to the network send the network access addresses and fixed addresses to the network center coordinator; sixth, if the network center coordinator agrees with that the child nodes applying for access to the network have access to the network, the child nodes applying for access to the network successfully get access to the network, if the network center coordinator does not agree with that, reject information is returned to the child nodes applying for access to the network, and after receiving the information, the child nodes applying for access to the network execute the second step to the sixth step till successfully getting access to the network. The networking method of the wireless electricity consumption information acquisition ad hoc network is used in the field of wireless acquisition of electricity consumption information.

The present invention discloses a speech recognition text segmentation method and device. The method comprises: performing endpoint detection on speech data, to obtain speech segments and a starting frame serial number and an ending frame serial number of each speech segment; performing speech recognition on each speech segment, to obtain a recognition text corresponding to each speech segment; extracting a segmentation feature of the recognition text corresponding to each speech segment; by using the extracted segmentation feature and a pre-established segmentation model, performing segmented detection on the recognition text corresponding to the voice data, to determine a position where segmentation is needed; and segmenting the recognition text corresponding to the speech data according to a segmented detection result. According to the method and apparatus disclosed by the present invention, the recognition text can be segmented automatically, so that the structure of the recognition text can be clearer.

In the generator arrangement according to the invention at least one main bearing is arranged between the generator and the hub wherein the rotor of the generator is supported only on the front end of a rigid carrier in the form of a support disc (30). The support disc is a hollow chamber structure of low weight and high rigidity.

The invention discloses a method used for capturing carrier frequency offset and time delay in a single-carrier frequency domain balancing system, comprising the steps as follows: frame-arrival detection is carried out on the received data; when effective data frame is detected, the frame head position of the data frame is determined roughly; the determined frame head position is used as an initial position to carry out rough frequency offset estimation and compensation on the received data frame; the rough timing estimation is carried out so as to determine the initial position of the data block; the timing deviation is locked within one code element offset; the precise timing estimation is carried out by combining a maximum-likelihood searching method with an O&M timing estimation arithmetic, thus gaining precise code element timing information; precise frequency offset estimation and compensation are carried out; synchronous error lock judgment and correction are carried out. The method designs unique preamble code data form and structure, precisely carries out the estimation and compensation on carrier frequency offset and time delay during the synchronous capturing stage,gains precise carrier frequency offset and timing information, and determines the initial position of Fourier transform windows more exactly.

The invention discloses a testing system for a gas explosion characteristic size effect and relates to a testing system which studies on the gas explosion characteristic size effect. The device comprises a gas distribution device, an ignition device, an explosion device, a data collecting and processing device and a high-speed digital camera, wherein the gas distribution device comprises a combustible gas source, a distribution gas source and a gas distribution tank; the ignition device comprises a igniter, and the igniter is installed on the explosion device; the data collecting and processing device comprises a pressure sensor, a flame propagation rate sensor, a temperature sensor, a signal amplification circuit, a data collecting circuit and a computer, and the pressure sensor, the flame propagation rate sensor and the temperature sensor are installed on a spherical container, a cylindrical container and a pipeline; the explosion device is installed between the gas distribution device and the data collecting and processing device, and the explosion device comprises the spherical container, the cylindrical container and the pipeline; and the high-speed digital camera is used for shooting flames during explosion of the combustible gas inside the containers and the pipeline, and an output end of the high-speed digital camera is connected with the computer.

The invention provides a pear tree two-arm direct-motion type shed frame tree shape which comprises a tree trunk. The tree trunk has two boughs which grow in a direct-motion mode, and base angles of the two boughs are 45 degrees. When reaching a frame-climbing point, the boughs are fixed on a horizontal shed frame in a direct-motion mode, and extended long branches of the boughs are guided onto an elevation guide line in a perpendicular mode. The boughs have long placed fruit branches which are arranged in parallel and vertically extend into intervals. Before the two boughs climb a frame, two horizontal robust side position branches which are pulled along the reverse direction to be parallel to the boughs after the horizontal robust side position branches grow upward to the frame face. The side position branches have F-shaped single-shaft long placed type fruit branches which are arranged in parallel and vertically extend to the intervals. The tree shape only has the two boughs, and therefore the problems that an inclined crown appears easily in a three-bough tree shape and the main boughs intersect are solved. The two boughs oppositely extend in a direct-motion mode and do not influence each other. The two-bough tree shape is easy to reshape, and the two boughs are easy to form after the trunk is fixed. The phenomenon of lack of branches appears in a three-bough shed frame due to the fact that sprout force of a pear tree is weak. The three shape has the advantages that cultivation density is high, early-stage output is high, light conditions are good, the branches are easy to arrange in an ordered mode, uniformity of fruit quality is high, and standardization is easy to achieve.

Digital audio signal transmission is provided for digitally processing, for transmission in a digital VCR format, audio signals from a plurality of channels into a digitized audio data segmented into blocks having a data size, each block reserved for audio data digitized from one of the channels. The audio data is converted from the channels into the blocks of digitized audio data by varying the data size of the blocks. The audio data is composed by scaling the audio data in at least one of the blocks and combining the scaled audio data with the audio data of another of the blocks into a block reserved for composite audio data.

The invention discloses a PDF form extraction method. According to the technical scheme, the method comprises the steps of analyzing a PDF document by a page number to obtain all image data, first line data and character data, processing the image data in sequence according to the page number by adopting an image identification algorithm, and obtaining second line data corresponding to form data from the image data with the form data; processing the first line data and the second line data in sequence according to the page number by adopting a graphic algorithm, thereby obtaining form frame data with form row data and column data; performing clustering processing on the character data by adopting a clustering algorithm to obtain text data with a character string set; and obtaining all the form data in the PDF document through final all form frames and all text data. According to the method for extracting forms in the PDF document, the accuracy and efficiency of extracting the forms in the PDF document are improved, and more accurate form data can be obtained; and the method is suitable for the field with higher requirements on the accuracy and efficiency of extracting the form data.

The invention provides a coal mine production three-dimensional visualization information system based on an Internet of things, belonging to the coal mine production safety integration information monitoring research field. According to the system, starting from an application layer of the Internet of things, through computer technology such as system integration, three-dimensional visualization and the like, the three-dimensional visualization information system is constructed. According to the coal mine production three-dimensional visualization information system based on the Internet of things, computer Internet and sensor technology are utilized to connect the real world with the Internet, integration of coal mine production safety and safety risk avoidance six systems and dynamic control of a three-dimensional image are realized, and one-stop management of a coal mine enterprise is realized. The system has the characteristics of clear structure, a friendly interface, and high easiness of operation. A coal mine tunnel, ground detection information, machine running, underground personnel activities can be subjected to three-dimensional displaying conveniently, data is subjected to situation analysis, accident early warning is provided for operation monitoring personnel, and personnel security and production efficiency of a coal mine are raised.

The invention relates to a strip steel quality forecasting, furnace condition early-warning and fault diagnosis method, in particular to a strip steel quality forecasting, furnace condition early-warning and fault diagnosis method based on partial least square, comprising the following steps: model selection: multiple models are adopted to describe the process characteristics of corresponding steel types; data preprocessing: data alignment based on the model is carried out, synchronization relation of process input and quality output is built and data dimensionless treatment is carried out to eliminate effect of process data on modeling precision owning to non-unity of physical units; an off-line model building; a PLS (partial least square) model for strip steel quality and process variable is built by utilizing a great amount of historical data in normal working conditions; determining control limit of an monitor-control index; determining variable quantity control limit; and on-line forecasting and on-line detection and fault diagnosis. In the invention, the model for the strip steel quality and the process variable is built by PLS algorithm, so as to realize real-time quality forecasting, process monitoring and fault diagnosis.

The invention provides a microcontroller flush bonding on-line simulation debugging system, comprising a communication port control module, a state control module, a debugging command control module and a hardware breakpoint monitoring trigger module; the modules are collected into a chip through a special debugging command transmitted by a mainboard to control the work of the whole on-line simulation debugging system; complex boundary scan to the inside of the chip through a standard test boundary scan can be finished; the microcontroller chip is controlled according to an outer command to be converted into a debugging mode from a normal mode; a program can be downloaded and numerated from any position in a memorizer of the microcontroller through the debugging communication interface of the mainboard to support the target system single-step debugging and support software breakpoints and hardware breakpoints; the contents of a specific function register, a data memorizer and a program memorizer of the inside of the chip are checked and altered real-time; address bus and data bus are monitored, and the functions of real-time logical tracing, etc. are realized. The invention is of clear and simple structure, high-efficient operating effect and precise real-time monitoring.