231 results about "Smart computing" patented technology

A computer implemented system for the production of medical records, legal documents and other frequently produced semi-technical documents. This is accomplished by generating an intelligent computer-guided interview and the use of serialized scriptable objects. Major program elements include a knowledge base text file, a parse engine, and an execution module. The knowledge base uses a unique rule syntax. The parse engine converts the textual knowledge base file to a compiled binary representation which can then be interpreted by the execution module. The execution module leads the user through the interview by generating a series of questions and presents possible answers in the form of pick lists. The data is recorded with a computer pen and collated into a document file. This file is coded in binary format and can be written to and recalled from disk. If the file is recalled from disk the user can continue to answer questions or change answers previously given. The result is a mobile computing system whereby data is input in a structured format. When the program is executed the user is prompted for answers to questions and, based upon the user's response, the final document can change considerably. Depending upon each answer, the program may change the subsequent questions being asked, change the list associated with a question, change the text being generated, or change the entire structure of the document. The data collected may be also be stored in a database for analysis at a later date. Finally, the data collected is output in a narrative text format which can be tailored according to the traditions and expectations of the user's profession. The program will output the text via a printer or it can be transmitted via electronic means.

The present invention provides a vehicle with redundant systems to increase the overall safety of the vehicle. In other aspects, the present invention provides a method for learning control of non-linear motion systems through combined learning of state value and action-value functions.

The invention discloses a model training method, a server and a computer readable storage medium, and belongs to the field of artificial intelligence calculation. The model training method comprises the steps: after a training job is received, acquiring job information; copying a model to each GPU, and synchronizing the initial values of model parameters of each GPU; extracting a part of sample data each time of iteration, splitting the extracted sample data, distributing the split sample data to different GPUs for training, performing global protocol operation on parameter gradients obtainedby training of all the GPUs, and updating the model parameters on the GPUs according to the parameter gradients after protocol; and after iteration of the iteration times is completed, selecting the model parameters of any GPU and storing the model parameters as a model training result. According to the model training method, the bandwidth bottleneck and the computing power bottleneck between thecomputing nodes are eliminated by fully utilizing the high-speed data transmission bandwidth of the GPU-GPU, so that the synchronous training efficiency and the speed-up ratio of the model on the GPUcluster are improved.

The invention relates to a method for predicting ammonia process flue gas desulphurization efficiency based on multiple parameters. The method is characterized by comprising the following steps of: selecting four different artificial intelligent computation models and taking parameters acquired in an ammonia process desulphurization system operational process such as multiple groups of flue gas amounts, flow of a circulating pump, the flow of a concentration pump, the ammonia concentration, the concentration of absorption liquid, the liquid-gas ratio, the inlet flue gas temperature, the ammonia consumption, the density of spraying slurry, the pH value of slurry of a spraying tower and the pH value of the slurry of a pre-washing tower as input variables of the four models; respectively training each model, and establishing a non-linear function relationship between four desulphurization parameters and the desulphurization efficiency; then respectively transmitting parameters monitored in real time into the trained artificial intelligent model, and predicting the desulphurization efficiency; and taking the average value of two predicted values in the middle as a final predicted value... The method disclosed by the invention can be used for better predicting the ammonia process desulphurization efficiency and has the characteristics of higher stability and stronger prediction capability compared with single model prediction.

A system, a method, and a computer program for utilizing a WiFi beacon for low-latency communication with smart computing devices. Embodiments of the invention use beacon-stuffed network announcement broadcast signals to promptly convey urgent safety information to pedestrians and other vulnerable road users and to warn them of approaching vehicles. This information can then be used to generate visual or audible alerts that are presented on a smart computing device of the vulnerable road user. Similarly, beacon-stuffed network announcement broadcast signals transmitted by pedestrians' smart computing devices can be used to warn vehicle operators of pedestrians in their path.

The invention discloses a method for extracting a tower target in an unmanned aerial vehicle routing inspection power transmission line image, and belongs to the technical field of pattern recognition and intelligent computation technologies. The method for extracting the tower target comprises the steps of converting an image acquired by aerial photographing from an RGB color space into an HSI (hue/saturation/luminance) space, carrying out binarization processing on an I component image of the HIS space after being subjected to median filtering so as to acquire a binary image of an electric power tower profile; carrying out angular point detection on the binary image of the electric power tower profile by adopting an improved Susan angular point detection method, and marking positions of all angular points in the binary image of the electric power tower profile; and detecting the mass center of an angular point sequence based on a moment invariant mass center detection algorithm, and marking the position of the electric power tower in an original image by taking the mass center as a center. Through applying the technical scheme disclosed by the invention, the electric power tower can be extracted accurately in real time, and a guarantee is provided for fault diagnosis of electric power components such as insulators, hardware fittings and the like on the tower.

The present invention discloses a private conversation robot. The robot comprises a housing, wherein a driving apparatus, an intelligent controller, an intelligent calculation processing analysis system, a sensor apparatus and a database are arranged in the housing; the driving apparatus is connected to the intelligent controller; the intelligent controller is connected to the intelligent calculation processing analysis system; the intelligent calculation processing analysis system is connected to the sensor apparatus and the database; the intelligent calculation processing analysis system can determine an environment surrounding a user; classification options of relatives, friends and strangers are arranged; different secrets correspond to different levels of security hierarchies; and interactive content between the robot and the user is classified by security levels. The private conversation robot provided by the present invention is high in intelligence, has a powerful service function, keeps secret for the user maximally, solves problems for the user, talks and interacts with the user, help the user reduce mental pressure or overcome mental obstacles, facilitates more easy communication of the user, and enables interactive communication to be more user-friendly.

According to the industrial intelligent control system based on software definition provided by the invention, computing power optimization scheduling and dynamic reconstruction can be realized in a software definition manner, resources are allocated to different industrial control tasks and edge computing tasks as required, and an edge intelligent controller with differentiated control and computing capabilities is constructed; meanwhile, a real-time industrial control task and a non-real-time edge intelligent calculation task are supported, and the method has great significance and practicalvalue for improving the utilization rate of industrial edge equipment resources, improving the overall edge service capability and reducing energy consumption. Besides, real-time control and non-real-time computing task requirements of industrial intelligent control, edge computing, industrial AI and the like can be met, platform capabilities of an edge side and a cloud side are broken through, the method has the advantages of fusion, intelligence, real-time performance and flexible expansion, and a brand-new solution which is efficient, differentiated, low in cost, space-saving and easy to maintain is provided for industrial intelligent control.

The demand for automation design of power electronic circuits becomes higher and higher with the development of the power electronic technology. In the invention, a particle swarm algorithm is applied to the design and optimization of the power electronic circuits, and the invention mainly relates to the power electronics field and the intelligent computation field. In the optimization method, an optimization process is divided into two parts by a decoupling technology to respectively optimize the power transmission of the power electronic circuit and a feedback network. Meanwhile, a mutation operator is introduced into the particle swarm algorithm to increase diversity of the swarm and improve the optimization efficiency of the algorithm. The optimization design of a buck converter is taken as an example for testing, which proves that the optimization method is very effective.

The invention discloses a classroom teaching evaluation system based on artificial intelligence micro-expression analysis. The evaluation system consists of a video input system, a cloud processing system and an output module. The cloud processing system comprises a server, a GPU acceleration artificial intelligence calculation module, a micro-expression module, an attendance checking module, a seat position information module, an interpersonal relationship module, a learning assistant system, an educational administration system and an evaluation and estimation system. According to the invention, through classroom real-time monitoring, comprehensive information like expressions and micro-expressions, seat position information and interpersonal relationships of students can be dynamically captured in real time; through a data base connected with educational administration information, learning assistant information and a mental health center, classroom effects, recognition and mental states of students can be reflected through big data processing; and important scientific analysis and decision basis are provided for each of teachers, educational administrators, learning assistants, and parents.

Optimization for relay node deployment in a wireless sensor network is an important technology for prolonging service life of the wireless sensor network and reducing network construction cost. According to the invention, the multi-objective mother foraging algorithm is applied to the strategy of relay node deployment in the wireless sensor network, and the invention relates to two large fields of sensor networks and intelligent calculation. The method adopts the multi-objective mother foraging algorithm to optimize positions of all relay nodes in the wireless sensor network, so as to maximize the whole service life of the network and minimize the numbers of the relay nodes. According to the invention, three novel partial optimization operation related to the problems are introduced into the algorithm, which effectively improves the optimization efficiency of the algorithm. Taking a plurality of networks with different sensor scales as examples for simulation tests, the method is proved to be effective.

The invention discloses a quality processing method for a video conference. The quality processing method comprises the following steps of (1), establishing a human face database, a general mien database and a conference scene definition database; (2), collecting a video signal and an aural signal of the video conference; (3), image intelligent recognition: carrying out human face recognition and expression posture analysis on a conference participant from the video signal, so as to obtain the look data of the conference participant, and storing the data in a database; carrying out analysis on the aural signal, separating out a figure speech and a conference hall scene sound, using the figure speech as a weighting coefficient to correct the look data of the conference participant, carrying out analysis on conference scene audio frequency under the support of the conference scene definition database, so as to obtain conference atmosphere data, and storing the data in the database; (4), conference quality judgment: according to scores of a conference and the conference participant, carrying out intelligent calculation, so as to obtain the final grades of the conference and the conference participant, and giving out a conference evaluation and improvement report. The invention also discloses a quality processing system for the video conference.

The invention discloses a machine vision based parking lot management system. The system comprises a user intelligent terminal, a parking lot assembly and a server; the server comprises data obtaining, sending and processing devices; and the server obtains and process information from the user intelligent terminal, and transmits user indication information to the user intelligent terminal and theparking lot assembly. Thus, vehicle, user and parking lot information is obtained intelligently, and calculated and predicted intelligently by the server, so that the user is guided to using parking stalls reasonably, the parking efficiency is improved, and traffic load is alleviated.

The invention relates to the technical field of building construction. The invention particularly relates to a BIM method for automatically generating longitudinal bars and side structural steel barson the lower portion of a beam. The method mainly comprises the steps of establishing a beam structure model through a mainstream BIM software platform Revit in a BIM technology implementation process; intelligently calculating various parameters of the lower longitudinal bars and the side structural steel bars of the beam through visual programming, and finally generating multiple steps such as constructing rebars of lower longitudinal ribs and side ribs of beams, wherein the steps meet the design picture set in the national building standard entitled " drawing rules for an overall representation method of concrete structure construction maps and details of construction". Information, a core factor of BIM technology, is integrated from concrete engineering to steel engineering connection.Therefore, a most fundamental and important step in the application process of BIM technology. The invention quickly builds the steel bar model in BIM model to form structural information of steel bars and guide construction. The invention can be widely applied in the steel bar modeling process of the floor frame beam, greatly improving the modeling speed of the steel bar and enhancing the application effect of the BIM technology.

The invention discloses an intelligent computer remote interaction system, which comprises a man-machine operation module, a remote monitoring terminal, a security gateway connection module, a central processor, a startup and shutdown module, a timing module, a database format conversion module, a permission granting module, a timing inspection module and a defense decision generation module. All data and application programs are in an encrypted mode, and corresponding permission is granted only after biological authentication is conducted, so the data is well protected from leaking. Different kinds of permission are connected and transmitted through different service nodes, thereby further ensuring data security. Vocal print authentication, face authentication, palm vein authentication and random secret key authentication are combined as a whole, so the system is greatly improved in security. Each control command and each data calling command are traceable. Moreover, permission designs all need to be controlled by the remote monitoring terminal, so the security of a host is further improved.

The method for reading and writing the data in the network card comprises the following steps: decoding a received data packet; If a write command is obtained, continuing to receive the data packet, retransmitting the data requirement of an unrecoverable error, carrying out artificial intelligence and compression calculation on the data without the error according to user definition, carrying outerror correction coding on the data, and writing the data subjected to error correction coding processing into a storage chip of the network card according to a storage chip protocol; If a read command is obtained, reading the data from the storage chip of the network card according to the storage chip protocol, verifying the read data, performing artificial intelligence reasoning calculation anddecompression calculation on the correct data and the correct data obtained by using the error correction algorithm according to user definition, and packaging and sending the calculated and processeddata according to the network protocol. The network protocol processing function, the artificial intelligence calculation function and the storage control function can be completed in the network card, a host CPU and a memory are not needed, the delay is very short, and high performance is achieved.

The invention provides a system for dispatching large-occasion service vehicles, which comprises a vehicle-mounted terminal, a wireless communication network, a server system and a dispatching system, wherein the vehicle-mounted terminal is connected with the server system through the wireless communication network, and the dispatching system is connected with a communication server directly. The system mainly aims at occasions such as airports covered with wireless networks; positions of the vehicles can be displayed on an electronic map by a GPS (global position system) module in real time; for emergencies, the vehicles matching conditions are intelligently computed and screened out by a complete set of backgrounds of the dispatching system; a dispatching order is executed by selecting an optimal scheme; after the vehicle-mounted terminal receives dispatching information, sound-light alarm is prompted; position information of the vehicles can be judged in real time; service areas can be monitored; and the service vehicles can be dispatched temporarily under abnormal conditions to maintain the normal operation of the large-occasion service vehicles.

The invention discloses a method and system for configuring PTN service to solve a problem of high operational complexity of configuring PTN service in the prior art. The method comprises the following steps: a user can apply for PTN service through a terminal and obtain PTN service related information; then a specific access device uploads the PTN service related information to an authentication subsystem to complete the both-end authentication of the access device; after the both-end authentication, a control subsystem configures the corresponding PTN service for the access device. Thus, the method greatly simplifies the PTN service configuration steps and achieves intelligent PTN service path computation. An individual line subscriber can apply for and issue the PTN service without the assistance of professional network operation and maintenance workers, which greatly improves PTN service release efficiency.

The embodiments of the invention provide a system for dynamically adjusting short-term rental house prices based on a machine learning model. The system is characterized in that: the system includes: a data resource storage module which stores mass house data, house owner information and tenement requirement information, and transaction historical information in a distributed cluster; a user front end module which is intended for receiving house reservation information which is input by a user; an intelligent artificial computing module which is intended for estimating house prices through an intelligent artificial model; a house bidding module which is intended for pushing the estimated prices to a house owner and acquiring an on-sale price of a current house that is input by the house owner; and a sorting engine module and an output module which present house information to the user in a reasonable order according to user's requirement in combination with reservation resource information which is filled by the user. The invention also provides a corresponding method.

The invention discloses an all-in-view fire detection and artificial intelligence fire extinguishing system and method. The system comprises an all-in-view infrared thermal imaging fire detection device, an artificial intelligence computer and a fire extinguishing device which are sequentially connected; the all-in-view infrared thermal imaging fire detection device is arranged at the front end of the system and is used for probing hot spots through all-in-view infrared scanning, acquiring hot spot imaging data and carrying out quantitative analysis on hot spot imaging data and forming an artificial intelligence linkage strategy according to quantization analysis results; the artificial intelligence computer is used for calculating the association of all hot spots and the layout position of the fire extinguishing device according to the artificial intelligence linkage strategy and the layout position data of the fire extinguishing device, providing a matched fire extinguish plan and controlling the fire extinguishing device according to the fire extinguish plan to control the fire. The fire extinguishing system disclosed by the invention is beneficial to scientific strategy and humanization of fire rescue, and has a great technological advantage and a practical value on all-in-view accurate fire extinguishment and dynamic tracking fire extinguishment of early-stage fire and spreading and development process of the fire.