35results about How to "Guaranteed computing power" patented technology

The invention discloses a bilayer sliced high-efficiency block chain consensus mechanism and a working method thereof. The consensus mechanism is based on a bilayer gossip topology network, and comprises a full node, a local full node, a light node and a micro node, wherein the full node is in charge of node topology and slice maintenance and is elected and updated through a full node maintenancemechanism; and the local full node is elected and updated through a local full node maintenance mechanism and is in charge of transaction validation, transaction consensus, transaction storage and account synchronization. The invention also provides a working method for the consensus mechanism. A credible and incentive node joining and updating mechanism is provided, the stability, the bandwidth capability and the computing capability of the full node and the local full node are ensured, and the impacts on a transaction confirmation speed are reduced. The consensus mechanism proposed by the invention forms a divide-and-conquer distributed account system by means of intra-slice autonomy and inter-slice collaboration to ensure to consider decentralization, scalability and security.

The invention discloses a flexibly configurable neural network computing unit, a computing array and a construction method thereof. The neural network computing unit comprises a configurable storage module, a configurable control module and a time-division multiplexing multiplication and addition computing module. The configurable storage module comprises a feature map data buffer, a step-size data buffer and a weight data buffer. The configurable control module comprises a counter module and a state machine module; the multiplication and addition calculation module comprises a multiplier andan accumulator. The invention can support any type of convolution calculation and parallel calculation of multi-size convolution kernel, fully exploits the flexibility and data reusability of the convolution neural network calculation unit, greatly reduces the system power consumption caused by data moving, and improves the calculation efficiency of the system.

The invention discloses a method for controlling energy consumption of a cluster system based on changing a control mode, which carries out system level management and control under the operating condition of a computation task on a computation node in the cluster system through a control node and belongs to the field of computer cluster systems. In the method for controlling the energy consumption, the control node respectively chooses fixed, manual or semi-automatic control modes for the change of a computation task amount and the frequency of the change and controls the energy consumption through controlling the state of a computation node and the computation node enters different sleep states according to an idle rule so as to save the energy consumption of a processor. In the invention, the energy consumption of the cluster system is greatly reduced aiming at parallel computation of the cluster system under the premise of not greatly degrading computation performance and the balance between the performance and the energy consumption is obtained.

The invention provides an indoor intelligent mobile robot and a control method. The global advancing path of the robot includes a plurality of key point positions of the path; when an obstacle existsin the advancing path, an obstacle avoidance point is arranged out of the original path, a local path of the robot advancing to the obstacle avoidance point from the current key point is planned, andafter the robot advances to the obstacle avoidance point along the local path, a local path of the robot advancing to a next key point from the obstacle avoidance point is planned again; and if a newobstacle appears during the process of the robot advancing along the local path, another obstacle avoidance point is set. A three-dimensional lidar is adopted for laser oblique shooting, the scanningpoint of the laser avoids a dynamic obstacle, and thus, laser data reflecting the surrounding environment information are acquired for building a surrounding environment map. Trackless autonomous navigation in an indoor complex environment with human mobility can be realized.

The invention provides a cooperative computing unloading method based on a splittable task in a vehicle-mounted edge computing environment, and the method comprises the steps: sending a to-be-processed task unloading request when the calculation amount of a to-be-processed task cannot be borne; acquiring relevant information of callable computing resources according to feedback of the to-be-processed task unloading request; and obtaining an optimal splitting ratio for task segmentation according to the related information of the callable computing resources. Therefore, the optimal unloading decision is selected by comprehensively considering the computing power of the vehicle, and the minimum task completion time delay can be conveniently obtained. After the request vehicle receives the relevant information of the callable computing resources, a task splitting model is constructed to perform construction, the optimal splitting ratio is obtained, a distributed task splitting and unloading scheme is practiced, it is guaranteed that each request vehicle autonomously decides splitting and unloading decisions of a task only based on received information, and frequent information interaction in a centralized scheme is avoided.

The invention is applicable to the technical field of computers, and provides a coprocessor, and a method and a system for matrix operation acceleration. The coprocessor comprises a controller, a register bank, a multiplier-adder MAC array, and a memory. The controller parses a vector expansion instruction, and calls and sends data to be calculated to the register bank according to a parsing result. The register bank transmits the data to be calculated to the MAC array and transmits an operation result to the memory, the operation result being obtained by matrix operation of the MAC array on the data to be calculated, and the data to be calculated and/or the operation result are stored in the memory. In the process, the MAC array performs matrix operation according to the vector expansioninstruction, and the vector expansion instruction is generated by a main processor according to a preset function, ensuring computational power of a general microprocessor. The register bank transmitsand stores all kinds of data, and a data storage channel is independent of a receiving channel of the vector expansion instruction, improving data throughput of a general-purpose microprocessor in anoperation process.

The embodiment of the invention provides a power supply voltage control method and device of a digital currency mining machine and the digital currency mining machine. The method comprises the steps of determining a computing power ratio of the digital currency mining machine based on actual computing power and theoretical computing power of the digital currency mining machine; generating a powersupply control instruction based on a comparison result of the calculation force ratio and a preset threshold value; and controlling the output voltage of the mining machine power supply based on thepower supply control instruction. According to the embodiment of the invention, the output voltage of the power supply is controlled according to the computing power ratio, and a good compromise can be obtained between the power consumption loss and the computing power of the mining machine.

The invention relates to a water environment analysis method based on a geographic information system and a Delft3D-FLOW calculation engine. The water environment analysis method comprises the steps that a software system framework is built through a GIS platform; a Delft3D mesh file format is analyzed; a configuration file MDF file format of the Delft3D-FLOW and meanings of multiple parameters in the configuration file MDF file format are analyzed, and a human-computer interaction configuration interface is developed for configuring the multiple parameters under the GIS platform; a model is built, configuration tasks of a plurality of environmental elements are completed through GIS operation; a configuration result is output as a file format which can be identified by the mathematical calculation engine, and the calculation engine is called to conduct calculation; a calculation result file format is analyzed, a file is read under the GIS platform and processed into a raster image with space coordinates through the image processing technology, and the raster image is loaded into the GIS platform for display. The water environment analysis method combines the mathematical model with spatial and geographic information, thereby improving analysis accuracy. The invention further provides a water environment analysis system based on the geographic information system and the Delft3D-FLOW calculation engine.

The invention discloses an intelligent driving system, the intelligent driving system comprises a sensing and computing module, a decision making and planning module and an underlying control module,wherein the sensing and computing module comprises a first communication unit and a sensing and computing unit comprising a plurality of artificial intelligence chips; the decision making and planningmodule comprises a second communication unit and a decision making and planning unit; and the underlying control module comprises a third communication unit and an underlying control unit. The invention further discloses an intelligent driving method and a computer readable storage medium. By adopting the intelligent driving system disclosed by the invention, the cost of the intelligent driving system can be reduced, and meanwhile, the computing power of the intelligent driving system is ensured.

The invention provides an embedded platform-oriented dual-mode target detection method and system, and the method comprises the steps: carrying out the segmentation processing on a to-be-recognized image according to a preset segmentation algorithm, and obtaining at least one image box; performing feature extraction on the image box through depth separable convolution in the improved YOLOv4-Tiny network to obtain image features corresponding to the image box; determining an infrared image and a visible light image according to the image features, respectively inputting the same a predefined visible light target detection model and an infrared target detection model after registration, and outputting detection results of the visible light target detection model and the infrared target detection model; and performing decision-level image fusion on the detection results through a pre-established decision-level fusion detection model, so that infrared and visible-light multi-band information complementation is realized, and a dual-mode target detection result is obtained. According to the scheme, real-time detection of an embedded platform for a multi-source image is realized, and the method has more obvious advantages in detection precision compared with a single-band target detection model while the requirements of real-time performance and effectiveness are met.

The invention discloses a medical data calculation method and system based on a cloud platform. The method comprises the steps that medical feature data of a user is collected on the basis of a terminal application layer, and a calculation task is generated on the basis of the medical feature data; the terminal application layer sends a request to a service cluster of the cloud platform on the basis of a security protocol, and the calculation task is sent to the service cluster; when it is judged that medical service data does not exist in a mechache cache unit, the service cluster analyzes cloud calculation keywords to obtain medical feature data classification, and target service program is called for executing cloud calculation according to the cloud calculation keywords and the medical feature data classification, medical service data matched with the cloud calculation keywords is obtained, and the medical service data is sent to the terminal application layer. Accordingly, the problem of the rapid acquisition process of existing data is solved, and the whole platform experience is improved.

The present invention relates to an image processing method, an image acquisition device, a movable platform and a storage medium. The method comprises acquiring an image acquired by a first image acquisition device, and determining the current ambient light brightness according to the image (S101); acquiring the image target brightness corresponding to a second image acquisition device, and determining a target exposure parameter corresponding to the second image acquisition device according to the current ambient light brightness and the image target brightness (S102); and controlling the second image acquisition device to generate a target image of the image target brightness based on the target exposure parameter (S103). The method improves the image quality.

The invention provides a method and device for controlling a computing power board, digital certificate processing equipment, a storage medium and a computer program product. The computing power boardcomprises a plurality of voltage domains, wherein each voltage domain comprises a plurality of chips; the plurality of voltage domains adopt series power supply; and the plurality of chips in each voltage domain adopt parallel power supply. The method comprises the following steps: obtaining an effective frequency combination, wherein the effective frequency combination comprises the frequency ofeach chip of each voltage domain during effective operation; and controlling the chip of each voltage domain to operate at the corresponding frequency in the effective frequency combination. According to the method and device for controlling the computing power board and the digital certificate processing equipment, for the computing power board with the voltage domains connected in series, the probability of chip failure caused by insufficient voltage distribution of the voltage domains can be reduced, so that each voltage domain chip operates effectively, and the computing power and the energy efficiency ratio of the computing power board are improved.

The invention discloses a calculation system for evaluating parameters of comprehensive energy equipment, which comprises an evaluation system body, the evaluation system body internally comprises a global model, an evaluation model A, an evaluation model B and an evaluation model C, and the global model internally comprises a first algorithm module, a second algorithm module, a third algorithm module and a profit factor formula module. According to the calculation system for evaluating the parameters of the comprehensive energy equipment, the comprehensive energy scene, the profit factor, the calculation process, the required parameter profit factor application scene, the leveling unit energy cost and the effect of the profit factor in production are combined with a learning framework, so that the parameters of the comprehensive energy equipment are conveniently and fully calculated; the three algorithms unify energy output and facilitate calculation, so that operation of workers is facilitated, guidance is provided for follow-up construction of a new comprehensive energy station through collection of a large amount of comprehensive energy equipment profit factor data, data privacy protection is achieved, and the model popularization degree is improved.

The embodiment of the invention discloses a storage space estimation method and device, electronic equipment and a readable storage medium. The storage space estimation method comprises the following steps: calculating and estimating a first storage space occupied by result data after data calculation based on sampling sum of data; writing the result data into the estimated first storage space; and writing the residual result data into a second storage space based on the fact that the residual result data is not written after the first storage space is fully written, so that the overall calculation performance is ensured while the space use efficiency is ensured.