42results about How to "Shorten communication time" patented technology

The invention discloses an on-satellite full-autonomous task planning system for an agile remote sensing satellite. Comprising an imaging target database module, a high-precision orbit prediction and orbit calculation module, a target visible arc section calculation module, an imaging task preprocessing module, a subsystem constraint model module, an imaging task planning module, a data transmission task planning module, a solid memory autonomous erasing module, an instruction data block generation module and a satellite state real-time monitoring module. By carrying out constraint modeling on subsystems or single machines such as load, data transmission, control and the like, an on-satellite full-autonomous operation process from receiving of subsystem or single machine instruction blocks from user requirements is established, satellite control is effectively simplified, and remote control uploading efficiency and on-satellite autonomous operation capacity are improved. And through a satellite state real-time monitoring module, the execution condition of tasks on the satellite, satellite state changes and the like are monitored in real time, task execution stopping, whole satellite sun-facing or re-planning are further triggered, task requirements and environment changes are quickly responded, task plans are formulated and adjusted in real time, and the task execution efficiency is improved.

The utility model provides a high-precision diaphragm pump flow detection device which is characterized by comprising a shell, a pneumatic electromagnetic valve, a diaphragm pump, an air cavity, a throttle valve, a flow velocity sensor, a carbon dioxide sensor module, a temperature sensor module and a humidity sensor module which are connected in sequence through hoses, after a patient blows air into an air bag, an operator starts sampling, and the air bag is filled with air. At the moment, the diaphragm pump is started, the pneumatic electromagnetic valve is opened, and air in the air bag enters the equipment host; by controlling the rotating speed of a motor of the diaphragm pump, airflow in the sampling process is kept at a stable value, and sampling is automatically stopped according to set time; the storage device is used for recording and tracking, and the flow stabilizing device and the flow sensor are arranged, so that the sampling quality is high; the air bag is externally connected for collecting air during use, so that the waiting time of a patient is short.

The invention discloses an energy storage optimal configuration method considering new energy-power grid-user capacity sharing. The method comprises the following steps: step 1, constructing a new energy-power grid-user side energy storage configuration model; 2, constructing a new energy-power grid-user side energy storage sharing strategy; and step 3, based on the energy storage configuration model and the energy storage sharing strategy, establishing a new energy-power grid-user side energy storage optimization scheduling model. And the access number of reactive compensation equipment is reduced, so that unnecessary communication time, regulation and control cost and equipment access investment cost are reduced. The problem of resource utilization rate reduction due to new energy output reduction is relieved, and power consumer income is guaranteed; energy storage is configured on the power generation side, so that the problems of stabilizing output fluctuation, tracking a power generation plan, reducing prediction errors and reducing wind curtailment or light curtailment can be effectively solved.

The invention relates to a method and a device for realizing a distributed DHCP (Dynamic Host Configuration Protocol) service, and belongs to the technical field of DHCP services, the method comprises the following steps: creating a subnet on an Openstack cloud platform, and starting a DHCP service corresponding to the subnet; the method comprises the following steps: creating a virtual machine on a cloud platform, linking the virtual machine to a subnet, scheduling the virtual machine to a server, creating a virtual machine instance on the server, and sending a notification after creating the virtual machine instance; the SDN controller analyzes a server where the virtual machine is located, a switch connected with the server and a port of the switch connected with the server where the virtual machine is located according to the notification, obtains virtual machine information, and associates the virtual machine information with the switch connected with the server where the virtual machine is located; the SDN controller configures a DHCP service of a switch connected with a server where the virtual machine is located; and the virtual machine sends a DHCP request, and the switch connected with the server where the virtual machine is located sends the IP address allocated by the cloud platform back to the virtual machine according to the configured DHCP service.

The invention discloses a direct current voltage transformer frequency response test method and system based on wireless synchronization, and the method comprises the steps: configuring a test host and a wireless sub-machine, and carrying out the wireless synchronization of the test host and the wireless sub-machine; the test host collects a standard signal output by the standard source, and the wireless sub-machine collects an optical fiber digital signal output by the tested DC voltage transformer; the test host detects the fundamental frequency of the standard signal and sends the standard signal to the wireless sub-machine; the wireless sub-machine performs interpolation on the acquired optical fiber digital signal according to the fundamental wave frequency so as to adjust the sampling frequency; the test host performs frequency spectrum calculation on the standard signal, and sends the harmonic order needing to be tested to the wireless sub-machine; the wireless sub-machines return the corresponding actual spectrum parameters to the test host; and the test host executes a direct-current voltage transformer frequency response test according to the actual frequency spectrum parameters. The method has the advantages of simplicity in implementation and operation, low test cost, high test efficiency and precision, safety, reliability and the like.

The invention discloses a large model reasoning hardware accelerator based on data flow execution, and belongs to the technical field of large model reasoning hardware acceleration. Each node comprises a linear calculation kernel used for sequentially completing query / key / value projection and attention output projection in multi-head attention calculation, dimension raising projection and dimension reduction projection calculation in a feedforward neural network layer, and inverse quantization operation after each projection is output; the multi-head attention calculation kernel is used for completing loading and quantization of query / key / value vectors, multi-head division and cache management, attention score calculation, Softmax normalization, context vector generation and KV cache updating; the RLA kernel is used for completing residual addition, layer normalization, quantization / inverse quantization and nonlinear activation calculation of each layer; the three cores in the same node are located in the same SLR area, and data transmission is carried out through a hardware queue and an HLS blocking mechanism. According to the invention, high hardware utilization rate and low-delay reasoning can be realized.

The invention provides a communication device and method with read address bias. The communication device comprises an MIPI main control chip and at least two sub-chips connected with the MIPI main control chip through a bus, and each sub-chip is provided with an MIPI module. The MIPI modules of the sub-chips on the same bus are configured with register address offsets and register address response ranges, so that register address access ranges of the sub-chips connected to the same bus do not conflict with each other. The MIPI main control chip sends an MIPI reading instruction including register address information to the MIPI module, and the MIPI module controls the bus and outputs the value of the corresponding register if the instruction is judged to be responded according to the register address information, the register address offset and the register address response range. According to the invention, the problem of read-back competition of a plurality of MIPI modules under the same MIPI bus and USID can be well solved, the communication time can be reduced, the read-back capability of each sub-chip can be reserved, and comprehensive testing and debugging are more convenient.

The present invention discloses a clustered wireless communication method and system, comprising: step 1, a base station clusters all edge devices participating in the communication into multiple clusters according to the geographical location, and the edge devices in each cluster communicate using a D2D link; step 2, after each edge device updates the decentralized model parameters according to the local data, the edge device exchanges and updates the model parameters with the edge devices connected to it through the D2D link, and uses the parameter averaging algorithm to aggregate the parameters in the cluster to obtain the aggregated parameters; step 3, the base station selects an edge device from all edge devices in each cluster as a cluster head, and the cluster head uploads its aggregated parameters to the base station; step 4, the base station globally aggregates the aggregated parameters uploaded by all cluster heads, and broadcasts the obtained global model to all edge devices. The problem of central node communication traffic congestion in the federated learning scenario with large-scale terminal participation is solved, and the model convergence rate is improved.

The invention relates to a multi-authorization-party cooperation attribute and key management method with clear ownership based on RAFT consensus. In the system initialization stage, each attribute authorization agency (AA) registers and initializes an identity on a block chain, and then generates and broadcasts a global attribute public key and constructs an attribute block through multi-party cooperation. In a user registration process, a data user (DU) submits an identity and attribute request to a system, and after preliminary verification of a Leader AA, a consensus is achieved among the AA through an RAFT algorithm, so that the consistency and security of registration information are ensured. In the user attribute updating stage, a user or an organization needs to submit an attribute change request. And the system processes and verifies an update request through an RAFT consensus mechanism, so that the correctness and consistency of attribute change are ensured. And the new attribute key is generated after the update is confirmed, and the old key is replaced, so that the system security is ensured. According to the method, the RAFT consensus algorithm is utilized to improve the fault tolerance and data consistency of the system, and the block chain technology is combined to enhance the security and non-tampering property of the data.

The invention discloses a culture resource big data reference model construction method, and relates to the technical field related to culture big data. Comprising the following steps: setting a multi-source data integration module, setting a real-time data perception module, creating a data fusion engine, setting a user feedback analysis module, constructing a cultural heritage damage prediction calculation module, creating an emotion analysis feedback system, and setting a cultural resource digital display platform. The invention further discloses application of the culture resource big data reference model, specifically, the culture resource big data reference model is applied to an online and offline system, and a framework of public culture digital resources should grasp effective online and offline combination. By analyzing emotional feedback of visitors, not only can cultural understanding and resonance be enhanced and inheritance and innovation of cultural heritage be improved, but also cultural education and experience can be optimized, cultural policies and market decisions can be supported, cultural diversity and inclusiveness can be promoted, and development of the cultural industry can be promoted.

The present invention relates to the technical field of quantum key distribution, and proposes a data security transmission method and system based on quantum communication, including: obtaining the original keys of both parties in quantum communication; taking each verification of the original key as a matter, constructing an FP tree and obtaining a number of nodes; obtaining the node initial value of each node; obtaining the character security of each unverified position in the original key; obtaining a number of key segments to be verified and their comprehensive security; verifying the key segments to be verified based on the comprehensive security to obtain a cutoff key; replacing the characters in the cutoff key and obtaining a hash value to obtain a replacement hash value group for each position in the cutoff key; obtaining a number of added characters, combining the cutoff key to obtain a final key; and transmitting data based on the final key. The present invention aims to solve the problem that the key length is too short due to the iteration of the quantum key through the hash function, which affects the security.

The invention discloses an ecological data processing method, apparatus and device, a medium, a product and a system. The method comprises the steps of obtaining first ecological status data of a first object; establishing at least one second object according to the first ecological status data, and obtaining simulated ecological data corresponding to each second object; determining target data according to the first ecological status data and simulated ecological data; and obtaining second ecological current situation data, and comparing the second ecological current situation data with the target data to form a first result. The method has the beneficial effects that a three-action feedback adjustment mode formed by'simulation preferential selection, dynamic monitoring, integrated evaluation and operation and maintenance management 'is constructed, the limitation of a traditional landscape design one-way process is broken through, and links of full life cycle design are comprehensively considered.

This application provides a method, apparatus, device, and storage medium for recommending test environments. The method includes: receiving an input identifier of a business system to be tested; determining, based on the identifier, whether there is a business logic relationship between the business system to be tested and other business systems, indicating whether the testing of the business system to be tested depends on the test results of other business systems; if a business logic relationship exists, identifying the other business systems as related business systems; determining a target environment chain from multiple preset environment chains based on the identifier of the business system to be tested and the identifiers of the related business systems; and displaying test environment recommendation information for the business system to be tested, including information about the target environment chain. In this application, by matching the business system to be tested to determine related business systems with which it has a business logic relationship, and automatically recommending the target environment chain, communication time is saved.

The application provides a kind of dynamic selection method and system of ultra-wideband positioning base station, comprising: by mobile terminal broadcast query signal, collect the base station response message and corresponding RSSI reaching preset quantity;RSSI validity screening is carried out to the collected base station, and weak signal base station is eliminated;Effective base station is divided into odd-numbered base station group and even-numbered base station group according to the parity of its number;The strongest RSSI of the preset number of base stations is selected from the two groups respectively, and the final positioning base station combination is formed by merging, also includes smooth switching mechanism and GDOP check step.Through the staggered layout mode that odd-numbered base station is deployed on one side of scene and even-numbered base station is deployed on the other side, the application ensures that the selected base station is distributed on both sides of the tag from the mechanism, forms the positioning configuration of geometric distribution balance, effectively reduces GDOP value.At the same time, the cut-off mechanism of limiting the number of acquisition significantly shortens the positioning waiting time, improves the real-time performance of the system.

The invention relates to the technical field of part manufacturing, and particularly discloses a section parameter optimization method for a sealing strip, which comprises the following steps of: S1, determining a section design concept of the sealing strip according to a preset requirement, and finishing a primary section outer contour; s2, parameterizing the wall thickness of each part of the section, and establishing a compression load analysis model; s3, establishing a DOE matrix by taking the wall thickness of each part of the section as an input parameter and the compression load of different contact positions as an output parameter, generating permutation and combination of a plurality of input parameters in a numerical value setting range of the parameters, and calculating a result of each combination to obtain a complete DOE matrix; s4, performing response surface analysis by using the established DOE matrix; s5, according to a response surface analysis result, performing optimization by using a multi-optimization target genetic algorithm; and S6, deriving an optimal design point, and completing optimization analysis. By means of the technical scheme, the design efficiency of the sealing strip section parameters can be improved, and meanwhile it is ensured that the product performance is close to the design target value to the maximum extent.

The present invention belongs to the field of computer technology, and specifically relates to a dynamic scheduling method based on the longest path task insertion, including: obtaining the execution time and communication time of tasks in a workflow, converting the workflow into a directed acyclic graph; calculating the priority of each task in the directed acyclic graph, and calculating the key tasks; specifying a processor to only process the key tasks, and other processors to process the remaining tasks, to obtain a task scheduling table; traversing the key tasks from back to front, and scheduling qualified non-key tasks to the processor that processes the key tasks for execution. The present invention reduces the communication time or waiting time between tasks, thereby achieving the purpose of shortening the completion time of the processor that processes the key tasks, achieving the purpose of accelerating the total task completion time, and effectively reducing the completion time of the workflow.

The present invention provides a method and computing device for artificial intelligence computing optimization, the method comprising: the main processor calculates the amount of memory occupied by control information, the control information comprising the control information of the first operator and the control information of the second operator; the main processor determines that the amount of memory occupied by the control information is greater than the memory amount of a single tightly coupled cache, then the main processor copies the control information of the first operator to the first tightly coupled cache of the slave processor, so that the slave processor enables the intelligent computing engine to perform operations using the control information. According to the technical solution of the present invention, it is possible to reduce the memory space occupied by data, reduce communication time, and improve computing efficiency and hardware resource utilization.

The utility model discloses a novel EMS energy management system and a communication system, the EMS energy management system comprises a plurality of energy storage containers, a control unit and an EMS, each energy storage container is internally provided with an energy storage module, a plurality of energy storage converters, a first unit switch and a second unit switch, one energy storage container is defined as a first energy storage container, a first unit switch of the first energy storage container is connected with first unit switches of other energy storage containers, and a second unit switch of the first energy storage container is connected with second unit switches of other energy storage containers; the control unit comprises a first controller, a second controller, a main network switch and a standby network switch; and the EMS is connected with the main network switch and the standby network switch. Wiring is convenient, and communication is stable.

The present invention relates to the technical field of electronic detonators, and in particular to a remote detonation system of electronic detonators and a control method thereof. In the system, a remote control module and a signal conversion module are connected by cellular network communication, and a local wireless network communication is used between the signal conversion module and a plurality of detonation control modules. The plurality of detonation control modules are respectively connected by wire with each group of electronic detonators. The remote control module is used to generate a detonation instruction and transmit the detonation instruction to the signal conversion module; the signal conversion module is used to realize the conversion between the cellular signal and the local wireless signal, and forward the detonation instruction to the detonation control module; the detonation control module is used to verify the detonation instruction, and control the detonation of the electronic detonator based on the verified detonation instruction, which guarantees the requirements of the electronic detonator blasting for clock accuracy to a certain extent, and realizes the real "remote control" and the unified control of the overall blasting of the entire blasting range.

The invention discloses a gradient communication compression method and system for a shared memory heterogeneous processor, and the method comprises the steps: carrying out the Top-k precise sorting of a gradient obtained by carrying out the forward propagation and back propagation of a deep learning model in a specified round through all nodes based on a universal CPU (Central Processing Unit), and combining with a compression ratio, so as to obtain a latest threshold value through compression; gradient compression is carried out on the basis of cooperation of the heterogeneous acceleration core and the universal CPU in combination with the threshold value of the last round; each node performs gradient communication, sparse decoding and parameter updating based on the general CPU; carrying out step self-increasing on the training round; and according to the training round step, judging whether the training round ends or not, and deciding to continue iteration or exit. The invention aims to realize optimization of a communication algorithm for distributed deep learning training of a shared memory heterogeneous processor, solve the problem of low efficiency of a traditional gradient communication compression method, reduce the communication time, reduce the compression cost and improve the communication efficiency.

The embodiment of the invention provides a data verification method, an information processing device, a consumable and an image forming device. The method comprises the following steps: receiving a verification request sent by the image forming device; and sending original data and target verification information to the image forming device based on the verification request, wherein the target verification information is used for verifying the original data. In the embodiment of the invention, the information processing device can generate the shorter target verification information, and the image forming device verifies the original data based on the shorter target verification information. Because the data length of the target verification information is short, when the target verification information is transmitted between the image forming device and the consumable, the communication time between the image forming device and the consumable can be shortened, and the communication efficiency is improved. The problem that in the prior art, after an image forming device verifies data, the time for processing a next task or entering a next stage is long is solved.

The present invention provides a multi-parameter adaptive heterogeneous parallel computing method for image classification. First, the image data is copied into R copies, which are respectively input into R multi-parameter adaptive modules. The multi-parameter adaptive modules adopt heterogeneous parallelism. The multi-parameter adaptive modules adaptively select the optimal parallel mode according to the data type, data volume, video memory occupied by model parameters, batch size of model input and number of model layers, train the multi-parameter adaptive modules, and pass the prediction results of the trained multi-parameter adaptive modules into the main parameter server for voting, and output the optimal classification result. While maintaining the original deep learning training effect, the present invention can greatly improve the model training efficiency and improve the robustness of image classification.

The invention discloses a multi-service fusion system based on a cloud video. The system is composed of a background, a middle platform and a foreground, and comprehensive fusion of cloud video conferences is achieved. The background constructs an efficient and reliable video conference system based on sea oil cloud platform resources, and supports smooth expansion. The middle station provides abundant video conference services, covers basic, advanced, development, access and the like, and realizes seamless connection with a multi-manufacturer system. The foreground faces users, provides full-scene business applications such as video conferences, commanding and dispatching, cooperative work and the like, and improves the intelligent communication and cooperation capability by fusing various businesses. The whole system is safe and reliable, is easy to expand, and meets remote communication requirements under multiple scenes.

The present disclosure provides a sharded parallel model training method, apparatus, device and storage medium. The present disclosure relates to the field of artificial intelligence technology, specifically natural language processing, deep learning, computer vision, image processing and other technical fields. The specific scheme is: deploy the model to be trained on multiple computing devices; segment the sequence length variables in the input variables of the model to be trained to obtain the segmented data; perform sharded parallel training on multiple computing devices according to the segmented data to obtain the target model. According to the scheme of the present disclosure, a sharded parallel model training method is adopted to train long sequence models, which can reduce the amount of communication between devices, reduce the cross-device communication time, and improve the efficiency of model training.

The present invention relates to an event-triggered communication and control method for platoon driving, comprising the following steps: S1, setting a convoy consisting of N vehicles, installing an event-triggered communication module and an event-triggered control module on each vehicle; S2, checking whether the communication of the self-vehicle meets the triggering condition at a fixed sampling period, if so, executing step S3; if not, executing step S4; S3, the self-vehicle sends status information to the neighboring vehicle; S4, the neighboring vehicle uses the self-vehicle status information at the previous sending moment to predict the self-vehicle status at the current moment; S5, checking whether the self-vehicle control meets the triggering condition at a fixed sampling period, if so, executing step S6; if not, executing step S7; S6, the self-vehicle sends a new control signal to the engine and executes it; S7, the control signal of the self-vehicle remains unchanged. Compared with the prior art, the present invention can synchronize the speed and acceleration of all vehicles in different initial states and arrange them in sequence at a preset vehicle spacing.

The invention discloses a diabetic retinopathy lesion visualization method and device and a readable medium, and the method comprises the steps: extracting a region of interest from an original fundus image, and generating a region of interest image; determining a feature stroke code of a feature segmentation region in the region-of-interest image in the original fundus image, and determining a focus stroke code of a diabetic retinopathy focus segmentation region in the region-of-interest image in the original fundus image; based on the feature run-length codes, a reference area used for assisting in judging the diabetic retinopathy focus is obtained, and corresponding reference run-length codes are generated; and displaying the feature stroke code, the focus stroke code and the reference stroke code in a blank image differently by using different colors to generate a fundus focus visual image. Therefore, the original fundus image is simplified, and the simplified image is visually displayed, so that a doctor and a patient can check the image conveniently, the doctor-patient communication time is saved, and the utilization rate of medical resources is improved.

The invention discloses a forklift local end digital system which comprises a forklift body, a hole is formed in a side plate on one side of the forklift body, a side small door is hinged to the hole, a two-dimensional code fixing support is arranged on the inner side of the side small door, and a static two-dimensional code containing forklift information is fixed to the two-dimensional code fixing support. A camera is fixed to the inner side face of the side plate, the shooting direction of the camera faces the static two-dimensional code and obtains static two-dimensional code information, the camera is connected through a data line and outputs an acquisition signal to a display screen to be displayed, and the display screen is installed in a cab of the forklift body. According to the system, the forklift management efficiency and accuracy are improved, and the communication cost and time for a driver to solve problems are reduced. Therefore, the driver can conveniently and quickly obtain the required information, and the working convenience is improved. The informatization level in the forklift service process is improved, the problem of a driver can be better understood and solved, and the service quality is improved.

The invention discloses an ad hoc network wireless communication implementation method based on narrowband spread frequency hopping, and relates to the technical field of ad hoc network wireless communication. The invention aims to realize long-distance wireless communication, ensure the success rate of communication and enhance the anti-interference capability. Comprising the following steps: performing communication by adopting a spread spectrum mode, increasing the bandwidth of a radio frequency signal, adding a spread spectrum factor into the bandwidth, and increasing the transmission distance; a frequency hopping mechanism is realized in a frequency division multiplexing mode, and the transmission speed is improved through frequency division multiplexing. The bandwidth of the radio frequency signal is increased, the spread spectrum factor is added, the error-tolerant rate is effectively increased, the transmission distance is increased, the problem of short communication distance is solved, channels are randomly monitored through a random number generation algorithm, optimal frequency channels are optimized, the network is maintained by a main node and is directly communicated with equipment, and the communication efficiency is improved. Therefore, the network structure complexity is effectively reduced, the management difficulty is reduced, and the communication timeliness is improved.

The invention relates to a cable and a cable current-carrying capability identification method, device and system, the cable comprises a signal processing module, the voltage input end of the signal processing module is used for being connected with power supply equipment, and the output end of the signal processing module is used for being connected with to-be-charged equipment; and the signal processing module is configured to process a target electric signal into a target voltage signal according to the current-carrying capacity of the cable when the voltage input end receives the target electric signal output by the power supply equipment, so that the to-be-charged equipment identifies the current-carrying capacity of the cable according to the target voltage signal. Therefore, through the signal processing module, the current-carrying capability of the cable can be identified according to the hardware structure of the cable instead of identifying the current-carrying capability of the cable depending on a communication protocol, the communication time is shortened, and the communication efficiency is improved.

The present invention proposes a multi-channel RS485 data acquisition method and system, which is applied to an acquisition device, wherein the acquisition device includes multiple acquisition units, each acquisition unit includes multiple RS485 interfaces, each RS485 interface is connected to each sensor through an RS485 bus, and the multi-channel RS485 data acquisition method includes: for each acquisition unit, in accordance with a preset order, sequentially sending an acquisition instruction to each sensor through each RS485 interface; repeating the above steps until the acquisition stop condition is met; wherein, after sending an acquisition instruction to the current sensor through the current RS485 interface, in the process of sending an acquisition instruction to the next sensor through the next RS485 interface, receiving the response data returned by the current sensor through the current RS485 interface. The present invention can reduce the communication time required for one polling and effectively improve the data acquisition efficiency.