32results about How to "Reduce resource costs" patented technology

The invention discloses a pedestrian target movement track acquisition method and system based on multiple cameras. The method comprises the following steps: acquiring a video through calibrated cameras, and preprocessing the video to obtain a training sample set and a test sample set; constructing, training and testing a deep convolutional neural network model; matching the feature vector and theimage coordinate of the pedestrian target output by the model, converting the image coordinate into a space coordinate, and connecting the space coordinate of the pedestrian target in series according to the time sequence list to obtain a moving track of the pedestrian target under the single camera; in combination with calibration external parameters, realizing pedestrian target cross-camera association through feature matching and overlapping region position information, and connecting spatial coordinates of multiple pedestrian targets under multiple cameras in series to form a multi-pedestrian target moving track. According to the invention, the two-dimensional image coordinates of the pedestrian target are converted into the space coordinates in the actual scene, so that the positionand track of the pedestrian target can be accurately obtained; the accuracy of pedestrian target recognition and track acquisition is improved by using multi-angle video fusion analysis.

The invention discloses a user identity authentication system, belonging to the field of computers. The system comprises a server, a client and a databank, wherein the server respectively communicates with the client and the databank. Corresponding to the system, the invention discloses an authentication method of the user identity authentication system; the method comprises the following steps: (1) a user accesses to a Web server through the client; (2) the Web server queries user information through the databank; and (3) the user information queried in the databank is transmitted to a user authentication server to be authenticated, if the authentication is correct, the user is allowed to log in, relevant information of the user is displayed, and a unique and unrepeated authentication key is set up; or else input error is displayed. The system supports the lead-in of third-party user information and different authentication modes of the third-party, so that the situation that multiple systems use different user authentication systems is avoided, and thus the use complexity of users is reduced.

The invention provides a high-temperature multi-axis vibration fatigue test method, and relates to the technical field of material high-temperature vibration fatigue performance testing. Test instruments needed by the method comprise a vibration controller, a power amplifier, two vibration exciters, two sliding block guide rails, a square platform, a test piece, a clamp, a high-frequency magneticinduction heating machine, an acceleration sensor, an infrared temperature detector and a paperless recorder. The test method comprises the following steps: inputting vibration parameters into the vibration controller, and controlling the two vibration exciters through the power amplifier; starting the high-frequency magnetic induction heating machine to heat a test piece; setting a test heating temperature, detecting the temperature of the test piece through the infrared temperature detector, and controlling whether the high-frequency magnetic induction heating machine heats or not in a closed-loop manner; ending testing according to the failure criterion or the preset time; and deriving test data. The method can be used for measuring the high-temperature multi-axis vibration fatigue performance of a material, the design cost is reduced, and the design level is improved.

The invention provides a rapid focus measurement system and method for an actual sound field of a water immersion type focused ultrasonic transducer. The rapid focus measurement system for the actualsound field of the water immersion type focused ultrasonic transducer comprises an ultrasonic sound field simulation module and a measurement path planning module. According to the current sound fielddistribution condition of the water immersion type focused ultrasonic transducer and the instantaneous sound pressure value around a current measurement point, a path optimization algorithm is used in a sound field scanning range to calculate and obtain an optimal path for sound field scanning, and the optimal path is used as a planned measurement path to carry out rapid focus measurement of an actual sound field ultrasonic focus. Since the process of determining the focus in the real-time sound field is to perform scanning by planning the optical path of sound field scanning step by step based on the instantaneous sound pressure value around the current measurement point; compared with the traditional grid scanning method, the method has the advantages that the number of sound pressure measurement points can be reduced, the sound field test time is effectively saved, and the time cost and the resource cost are reduced.

An embodiment of the invention discloses a resource inquiry method and device. The method comprises steps as follows: receiving a resource consumption inquiry request which carries the quantity of to-be-processed tasks; determining grades of available resource quantities matched with the quantity of the to-be-processed tasks in a plurality of resource grades divided on the basis of extra resource consumption in advance; determining the estimated newly increased resource quantity of each grade; determining the total quantity of the estimated resources of each grade according to the estimated newly increased resource quantity; determining the grade matched with the quantity of the to-be-processed tasks according to the total quantity of the estimated resources and feeding back an inquiry result, that is, the resource consumption corresponding to each grade. By means of the resource inquiry method and device, the inquiry result fed back by a resource platform is closer to the practical extra resource consumption.

The invention discloses a generation method of time-division inter-satellite measurement communication network long multiple-address codes. The method comprises the steps that (1) the number of code elements of long spreading codes in one cycle is calculated; (2) the number of needed long multiple-address codes is determined; (3) the order of a high-order m sequence generator needing to be selected is determined; (4) a spreading code sequence based on a primitive polynomial is generated; (5) code segments with the balance characteristic are searched for; (6) orthogonal spreading code segments are paired and selected; (7) the paired spreading code segments are sequentially distributed to satellite constellation inter-satellite link network access nodes, and the initial states of shifting registers corresponding to the spreading code segments are calculated; (8) the time-division inter-satellite measurement communication network long multiple-address codes are calculated according to a primitive polynomial SSRG structure of the high-order m sequence generator. The method has the advantages that the principle is simple, the process is standard, the balance characteristic and correlated characteristics of code blocks are controllable, the number of taps and the number of registers in project implementation are small, and the resource cost is low.

The invention discloses a vehicle structure random vibration fatigue life calculation method considering welding residual stress, and belongs to the field of welding structure fatigue life calculation. The objective of the invention is to accurately calculate the random vibration fatigue life of a welding structure by considering the influence of welding residual stress on fatigue damage. The principle is that a time domain pavement irregularity calculation method is established based on an inverse Fourier transform method, and random pavement irregularity excitation power spectral density isdetermined; a thermoelastic-plastic theory and an ellipsoid heat source model are combined to carry out numerical simulation on the residual stress of the welded structure, excitation power spectral density is applied to the mechanical structure, random vibration analysis is performed and a structural stress power spectrum is calculated; and a stress amplitude probability density function is calculated according to the structure stress power spectrum, and the fatigue life of the vehicle structure is solved under the action of the random load according to the S-N curve of the weld joint. The method has the advantages that the initial welding residual stress and the working stress of the vehicle structure can be coupled, and the service life calculation precision under the weld random vibration load is improved.

The invention discloses a long-term joint optimization method for base station activation control and beamforming. The method comprises the following steps: initializing a central processing unit, a network control unit and an auxiliary processing unit in an MISO network; collecting and processing information such as a channel and a base station service state in the network by the network controlunit; performing corresponding analysis by the central processing unit according to the received current environment information; t opening or closing the specified base station by the network controlunit; then calculating and configuring a formed wave beam in combination with channel information, and feeding back a power consumption indication quantity; and receiving and storing historical dataof system operation by the auxiliary processing unit, assisting the central processing unit to update the neural network weight by utilizing the data stored by the auxiliary processing unit, and assist the system to normally operate. The invention provides a self-adaptive online solution, the network power is reduced, and the problems of low environmental adaptability, complex operation, poor practicability, high operation and maintenance cost and large capital investment in the prior art are solved.

The invention relates to a method for processing signal and relative device. Wherein, it is characterized in that: processing advanced detection on received baseband signal, to output detected effective multi-path information; based on said information, processing judgment, and according to the result of judgment, distributing the multi-path information that need demodulation; based on distributed multi-path information, selecting the signal of relative branch to be demodulated and decoded. The invention can directly process demodulation and filtration on received baseband signal, without storing long demodulate data to correct the signal deflection, to reduce the time delay of memory resource and data treatment. In addition, the invention processes information demodulation on the total branches with effective multi-path information, after the advanced detection, to avoid the decrease of demodulation property caused by mistake selection of demodulate branch, to confirm the access property of signal.

The invention discloses a production method for full-resource cyclic utilization of sulfuric-acid-method titanium dioxide production wastewater, which comprises the following steps of: neutralizing wastewater produced by sulfuric acid method titanium dioxide production by lime, separating gypsum by a filter press to obtain treated wastewater, and adding a recycled sodium carbonate solution to precipitate saturated calcium sulfate in the treated wastewater; clarifying the slurry to separate calcium carbonate precipitate and a sodium sulfate solution, and carrying out membrane separation on theseparated dilute sodium sulfate solution through a membrane filter; adding lime into a concentrated-phase sodium sulfate solution obtained after membrane separation for causticization reaction, takinga filtrate as a sodium hydroxide solution, employing carbon dioxide-containing tail gas generated in the titanium dioxide production process for carbonization, obtaining a sodium carbonate solution,and then returning the treated wastewater to precipitate saturated calcium sulfate. The method economically solves the technical problem that the lime neutralization treatment wastewater cannot be economically recycled in the sulfuric acid method titanium dioxide production process, eliminates the influence factors of outward discharge of the existing neutralization treatment wastewater on the environmental water body, and saves a large amount of raw water resources for production.

The invention provides a software disconnection monitoring method. The software disconnection monitoring method includes the following steps that software to be tested changes sent data with a first period T1 through a sending module; a receiving module obtains the sent data changed by the sending module with a second period T2, wherein N*T1<T2<(N+1)*T1 and N is a positive integer; whether the software to be tested undergoes disconnection or not is judged based on the obtained sent data. With the software disconnection monitoring method, the number of times in reading and writing the data is reduced, cost is saved, the error probability is greatly reduced, disconnection occurring is determined more effectively. The invention further provides a software disconnection monitoring system.

The invention discloses a position embedding interpretation method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring an original data set, and preprocessing the original data set; acquiring a plurality of positions corresponding to the plurality of pieces of position data; according to a plurality of position context sequences in the preprocessed original data set, performing learning by using a Skip-gram model to obtain a plurality of position embedding vectors; obtaining a plurality of position categories corresponding to the plurality of pieces of position category data; performing learning by using a Skip-gram model to obtain a plurality of position category embedding vectors, and enabling the plurality of position embedding vectors and the plurality of position category embedding vectors to be located in an original vector space; and according to a predetermined rule, converting each position embedding vector from an original vector space to a semantic vector space to obtain a position semantic representation corresponding to each position embedding vector. According to the method, more semantic information can be learned, so that the vector of each dimension has interpretability.

The invention provides a management method of multi-version software. The software comprises at least two versions, wherein the software is applied to embedded equipment, and the embedded equipment isconnected with an upper computer, and the embedded equipment comprises a microprocessor, and the microprocessor comprises code units corresponding to each version of the software. The management method comprises the following steps that S1, an upper computer issues a corresponding switching signal according to the version of software; and S2, the microprocessor switches to the corresponding codeunit according to the switching signal, and rewrites the control parameter and/or code in the code unit according to the switching signal, so that the rewritten control parameter and/or code meets thecorresponding function requirement of the version of the software, and the processing progress is sent to the upper computer. The management method has the beneficial effects that the management requirement of the multi-version software of the software product is met in one embedded device, so that the resource cost and the time cost are reduced, and the management efficiency of the multi-versionsoftware is further improved.

The invention discloses an English voice teaching system, which comprises a local acquisition device and a cloud server. The local acquisition device is used for acquiring a voice signal and uploadingthe voice signal to the cloud server. The cloud server comprises a receiving module, a conversion module, an analysis module and a storage module. The receiving module is used for acquiring a voice signal uploaded by the local acquisition device based on a binary stream. The conversion module is connected with the receiving module and is used for converting the voice signal uploaded to the cloudserver into an mp3 file and transmitting the mp3 file to the analysis module. The analysis module is used for reading the conversion module, carrying out data cleaning on the mp3 file based on a pre-stored formula, and outputting analysis information. The storage module is used for storing the address information of the mp3 file generated by the conversion module and the analysis information of the analysis module. According to the invention, the online oral level evaluation of students can be completed. The oral level of students is stably improved. The input time cost and the resource cost are reduced.

The invention discloses an energy consumption prediction method of a parallel BP neural network based on Hadoop, and the method combines a current mainstream ensemble learning related algorithm with a Hadoop distributed platform, improves the model building speed through a parallel calculation mode, and improves the generalization capability of the algorithm through a method of building a plurality of models or even combining a plurality of different models. According to the method, accurate prediction of future energy consumption of the metro can be realized, the processing capability, the calculation speed and the generalization capability of a prediction algorithm are improved, and reliable data support is provided for energy efficiency management and energy distribution of the metro by predicting the energy consumption of a future period and optimizing a control strategy and daily operation of the metro.

The invention provides a medical image brightness homogenization correction method. The medical image brightness homogenization correction method comprises the steps: constructing a brightness defectimage set of a target medical image; performing deep convolutional neural network training by taking the target medical image as a standard to obtain an image brightness distribution feature; constructing an image brightness adjustment model according to the image brightness distribution characteristics; and training a to-be-corrected medical image according to the image brightness adjustment model to obtain a medical image with uniform brightness. An image brightness adjustment model (namely, a generative adversarial neural network) is constructed through image brightness distribution characteristics obtained through deep convolutional neural network training, and medical image brightness homogenization correction is automatically performed by utilizing the image brightness adjustment model, and the medical image utilization rate is improved, and the time and resource cost caused by re-scanning is reduced.

The invention discloses a video CDN node instant capacity expansion method, a capacity expansion virtual machine room and a CND system, and the method comprises the steps: determining the capacity expansion virtual machine room which comprises a newly-added node machine and an existing node machine; when the scheduler receives a user request, selecting a node machine from the expansion virtual machine room; if the selected node machine is the existing node machine, directly using the existing node machine to provide service for the user; and when the new node machine is selected, the schedulerre-selecting the node machine from the to-be-expanded machine room to provide service for the user, and the new node machine selected by the scheduler acquiring the file from the existing node machine to provide service for the user. When the node is expanded, video migration is not needed, and the newly added node machine is available, so that the operation and maintenance cost and the total resource cost are greatly reduced, and the expansion efficiency is improved.

The embodiment of the invention discloses an object storage management method and device, electronic equipment and a computer storage medium. The method comprises the steps that a target object storage state query request is sent to an object storage service system; receiving a target object storage state query result returned by the object storage service system; and when the target object is a to-be-archived object and the storage state of the target object is a non-archived state, sending a target object archiving storage request to the object storage service system, so that the object storage service system executes an archiving storage operation on the target object according to the target object archiving storage request, and after the archiving storage operation is executed, returning the archived operation feedback information. Compared with the prior art, the technical scheme has the advantages that the working efficiency is improved, the time cost and the resource cost are greatly reduced, and the potential safety hazard can be effectively avoided.

The invention discloses an anti-extrusion storage device for transporting PLA straws, which belongs to the field of storage devices, and comprises a storage box main body. A No. 1 protective box cover and a No. 2 protective box cover are arranged on the storage box main body. The No. 1 protection box cover and a No. 2 protection box cover are left-right symmetrically arranged, and a strengthening device is installed inside the top of the No. 1 protection box cover. Swipe right. An anti-extrusion storage device for PLA straw transportation according to the present invention belongs to the field of storage devices. By setting a movable reinforcement device, the edges of the No. 1 protective box cover and the No. 2 protective box cover can be strengthened. The anti-extrusion performance of the edges of the No. 1 protective box cover and the No. 2 protective box cover is improved, thereby reducing the chance of the No. 1 protective box cover and the No. 2 protective box cover being pressed into the main body of the storage box during transportation. Ultimately reduces the chance of straws being squashed and reduces resource costs.

The invention discloses a joint optimization method for access control and beamforming in a downlink MISO network, which comprises the following steps: S1, initializing outer parameters and configuration in a downlink MISO network; S2, calculating and saving an outer-layer current user access status indication quantity matrix by assignment according to the outer parameters and configuration; S3, according to the outer parameters and the configuration, calculating an outer-layer next-generation user access status indication quantity matrix and an outer-layer next-generation beamforming matrix;S4, judging whether to meet the requirements according to the outer-layer current user access status indication quantity matrix, the outer-layer next-generation user access status indication quantitymatrix and the outer-layer iteration number indication quantity; if yes, outputting the outer-layer next-generation user access status indication quantity matrix and the outer-layer next-generation beamforming matrix to realize joint optimization of beamforming; otherwise, entering the step S2. The joint optimization method for access control and beamforming in a downlink MISO network in the invention solves the problems of complex operations, low practicability, high maintenance cost and large capital investment, which exist in the prior art.

An embodiment of the invention provides a complex network containment synchronization method and device, electronic equipment and a computer storage medium. The method comprises the following steps of: acquiring behavior description data for describing a random packet loss behavior of a complex network; and carrying out pinning control on the first l nodes of the complex network based on the behavior description data to realize state synchronization of all the nodes of the complex network, wherein l is an integer smaller than N, and N represents the number of all nodes of the complex network. Thus, for the complex network with random packet loss, the complex network containment synchronization method does not need to control all nodes in the complex network, only the first l nodes in the network are subjected to containment feedback control, the synchronization of the whole network is realized, and the resource cost and the time cost are effectively reduced; in addition, the packet loss of the network is described through the ideal output data and the transition probability of each node in the complex network, and the packet loss condition of the actual network is better met.