28results about How to "Speed up execution" patented technology

The invention relates to the technical field of integrated circuits, in particular to a processing system of a single-precision configurable matrix operation unit based on RISC-V instruction expansion, which comprises an RISC-V processor used for executing a software program and identifying a user-defined expansion instruction, and the RISC-V processor is connected with the single-precision configurable matrix operation unit through a user-defined coprocessor interface; the single-precision configurable matrix operation unit comprises an instruction response and decoding module and an instruction retirement and write-back module, and the instruction response and decoding module is used for receiving an instruction function code, source operation data and a destination register number and decoding the instruction function code; after decoding is completed, different execution units are selected according to a decoding result to send corresponding control signals; and after the execution unit completes the corresponding task, the instruction retirement and write-back module returns an execution result and a completion signal to the RISC-V processor, so that execution of a self-defined extension instruction is completed. According to the system, the matrix operation speed can be accelerated through bottom hardware under limited hardware resource consumption.

The invention relates to the technical field of path planning, in particular to an emergency evacuation path planning method and system and a storage medium. The method comprises the following steps: obtaining building layout data; performing building structure topology analysis on the building layout data to generate building structure layout data; arranging monitoring equipment according to the building structure layout data to obtain monitoring equipment deployment data; performing environment perception on the monitoring equipment based on the monitoring equipment deployment data so as to generate environment perception data; performing static image acquisition on the environment sensing data to obtain a static environment image; performing channel connection on the static environment image to generate building channel distribution data; and performing potential dangerous channel area identification on the building channel distribution data through the static environment image to generate potential dangerous channel area data. According to the invention, by integrating environment perception, dynamic video stream analysis and intelligent path scheduling, the flexibility and accuracy of emergency evacuation path planning are improved.

The invention discloses an explosive and powder safety information management system, and belongs to the technical field of explosive and powder management. The system comprises an information acquisition module, an explosive and powder state monitoring module, an explosive and powder safety evaluation module, a data processing module, an early warning and control module and an emergency plan storage module, the information acquisition module acquires safety information of explosives and powders; the explosive and powder state monitoring module predicts a potential explosion risk; the explosive and powder safety evaluation module monitors and analyzes stress distribution conditions of explosives and powders in storage and transportation processes; the data processing module is used for comprehensively analyzing potential explosion risks and stress distribution conditions of explosives and powders; the early warning and control module sends out an early warning signal; and the emergency plan storage module stores and calls a corresponding emergency plan. According to the invention, the monitoring precision and response speed of the system are improved, and the safety and reliability of explosive and powder management are ensured.

The invention discloses an NPU time division multiplexing real-time reasoning system and a scheduling method, belongs to the technical field of edge intelligent computing, and solves the problem that the reasoning real-time performance of multiple reasoning tasks is difficult to guarantee by traditional edge intelligent computing service equipment and a reasoning scheduling method in the prior art. According to the method, an intelligent reasoning model is converted through a preparation state pre-divider in a preprocessing stage, namely, the intelligent reasoning model is divided into blocks with different granularities in combination with unit granularities, and operation attribute information of the model and the blocks of the model is obtained; in the execution stage, a task request called by a remote procedure is received through a runtime planner, then the optimal scheduling granularity is determined through nonlinear optimization, a corresponding task scheduling sequence is generated according to a low-segmentation NPU real-time scheduling algorithm, and model blocks needed by tasks are obtained through a runtime executor according to a job sequence and executed. According to the method, the real-time performance of time division multiplexing of NPU resources of the edge intelligent reasoning calculation task in a multi-task scene is effectively improved.

The invention is suitable for the technical field of big data mining, and particularly relates to an entity resolution parallel processing method and device, equipment and a medium. According to the method, entity tuples in a to-be-resolved data set are divided into non-intersecting partitions, the entity tuples in any partition are equally divided into N intervals, M continuous intervals form a window, all the windows are subjected to serial processing, a thread block is distributed to each interval in the window for any window, and a thread block is distributed to each interval in the window. The thread blocks are used for matching target tuples in the corresponding intervals with entity tuples except the target tuples in the partitions to obtain matching results, and the thread blocks in the window are processed in parallel; and if not, processing the interval in the next window of the window by using the executed thread block until all the intervals are processed by the thread block to obtain all matching results. The entity data is effectively segmented, resource utilization and parallelization are optimized, and the entity resolution efficiency is improved.

The invention provides a heat sink preparation optimization method and device and a readable storage medium, and the method enables the process parameters of process control models corresponding to at least two adjacent deposition layers in the same heat sink to be different. The process parameters of the process control models corresponding to the deposition layers formed in the heat sink generation process and the performance parameters of the heat sink are collected to generate multiple sets of samples, and the models corresponding to the different deposition layers are trained through the samples and qualification judgment of the samples, so that the process control models are improved to optimize the performance of the deposition layers. And finally, the purpose of optimizing the comprehensive performance of the heat sink is achieved. The scheme provided by the invention is applicable to preparation of heat sinks with various thicknesses, the problems of defects and non-uniform performance in the production of the heat sinks can be effectively avoided, diversified and differentiated effects can be embodied on each deposition layer of the heat sinks, the different deposition layers can have the advantage of different directionality, and the production efficiency of the heat sinks is improved. And thus, the finally generated heat sink is ensured to have more excellent comprehensive performance.

The invention discloses a Tranform architecture operator hybrid optimization method. The Tranform architecture operator hybrid optimization method comprises the following steps: step 1, operator mapping analysis and optimization demand identification; 2, the operation process of the operators is disassembled, and the calculation kernel of each operator is constructed by using OpenCL; and 3, taking an OpenCL kernel as a computational graph node, and constructing a computational graph by using OpenVX. According to the method, the OpenVX provides a powerful graphic processing optimization capability, a computational graph can be optimized into an efficient execution form, and the OpenCL accelerates the execution of operators by utilizing a multi-core parallel computing capability, so that the overall reasoning efficiency is improved.

The invention relates to the technical field of RISC-V instruction set processors, in particular to an RISC-V-based AI processor design method for small convolution kernel acceleration, which comprises the following steps: determining the acceleration demand of a 3 * 3 convolution kernel in a convolutional neural network (CNN), selecting an open architecture RV32I of RISC-V as a basis, defining a core architecture of an AI processor, comprising a calculation unit, a register file and a control unit, the processing of the small convolution kernel is optimized; designing and realizing a vector arithmetic logic unit (VALU) and a convolution operation unit (CUU); the VALU has parallel computing capability and can execute addition, subtraction and multiplication vector operations to support instruction set extension; a convolution arithmetic unit CUU processes 3 * 3 convolution operations, including dedicated hardware logic to perform element-by-element multiplication and accumulation operations, thereby accelerating the convolution calculation process.

A method for facilitating playing a legacy game is described. The method includes receiving user input during the playing of the legacy game, determining whether one or more code blocks for servicing the user input are cached, and accessing one or more instructions of legacy game code upon determining that the one or more code blocks are not cached. The method also includes compiling the one or more code blocks from the one or more instructions of the legacy game code, caching the one or more code blocks, and executing the one or more code blocks to display a virtual environment.

The utility model discloses a remote control smoke exhaust fire damper actuating mechanism which comprises an actuating mechanism body, a circuit board and a first relay, the power input end of the circuit board is connected with a power source, and an electromagnet is respectively connected with the reset end of the circuit board and an action switch through lines. The first relay is connected with the motor, the reset switch and the action end of the circuit board through lines, the reset switch is connected with the common end of the circuit board and the binding post through lines, the action switch is connected with the reset switch and the binding post through lines, and a signal receiving module is arranged on the circuit board. The signal receiving module receives a remote control signal to remotely control opening or closing of the valve. The valve is reasonable in structural design, the signal receiving module receives remote control signals, opening and closing of the valve can be remotely controlled, potential safety hazards are effectively reduced, and safety performance is enhanced.

The invention discloses an autism spectrum disorder language rehabilitation training device with interaction ability, which comprises a data processor, loudspeaker modules symmetrically distributed on two sides of an image display screen arranged on the outer wall of the data processor and used for playing audios, and microphone modules arranged below the loudspeaker modules and used for collecting external voice information; a storage box is arranged below the microphone module and used for installing keys. A monitoring probe, a liquid storage tank and a snack box are arranged at the top of the data processor. According to the device, the language understanding and language expression ability of children can be improved through voice, picture and video communication strategies of different levels, forms and scenes, man-machine back-and-forth dialogue training and a way for providing guardians or doctors to remotely observe the states of the children and interactively communicate with the children; and on the basis of language training, the capabilities of concentration, hand-eye coordination, music perception and execution and the like of the children can be further improved, the mood of the children is relaxed, and the learning interest of the children is improved by performing real-time positive feedback on training items of the children.

The specification provides a transaction execution method, a node and a blockchain system. The method is applied to a first node in a blockchain system, the first node comprising a control process and a plurality of computing processes, and the method comprises: the control process acquiring a first transaction to be executed in a first block; the control process, in a case where it is determined that there is a second block in an execution stage, sending the first transaction to a first computing process in the plurality of computing processes in response to the first transaction not depending on a second transaction in the second block; and the first computing process executing the received first transaction.

The invention relates to a communication system, a communication method and a cleaning robot. The communication system is suitable for a cleaning robot of a photovoltaic panel array, the cleaning robot comprises a main adsorption control module, a first cleaning control module and a second cleaning control module, and the communication system comprises a first serial port communication module and a second serial port communication module. The first serial port communication module is arranged between the first cleaning control module and the main adsorption control module and used for achieving serial port communication, and the second serial port communication module is arranged between the second cleaning control module and the main adsorption control module and used for achieving serial port communication. Through the special serial port communication module, each control module of the cleaning robot can quickly and accurately exchange information, so that the control response speed and the operation efficiency of the whole cleaning robot are improved.

The invention belongs to the technical field of power systems, and more specifically relates to a large-scale power grid multi-agent reinforcement learning rapid scheduling method, and the method comprises the steps: obtaining power grid data in real time based on a data collection system, and improving the change trend of a prediction load through a load prediction model; a heuristic algorithm and a parallel computing architecture are adopted, so that the generation and execution process of a scheduling scheme can be accelerated, and the problems of efficiency and expandability of a scheduling algorithm are solved; therefore, the problem of limitation in the aspect of real-time data processing in a large-scale power grid multi-agent reinforcement learning rapid scheduling method is effectively solved, and the power grid scheduling efficiency and response speed are improved.

The invention discloses a voice interaction graphene AI intelligent tea table control method, and relates to the technical field of home voice interaction, and the method comprises the control steps: S1, monitoring a surrounding user instruction based on a built-in microphone of a tea table, and after the tea table receives the voice instruction, synchronously carrying out the preliminary noise reduction processing, and retaining the voice segment of the instruction; s2, training a data set containing a noise scene, enabling a built-in model to learn features for distinguishing voice and noise, and filtering non-voice signals in real time; and S3, for dialects and accents in different regions, collecting a scale data set, training a dialect recognition model, and inputting the obtained voice signal into the dialect recognition model to recognize and judge the current dialect and accents. Through preliminary noise reduction and microphone array layout, the voice acquisition quality is improved, and instruction omission is reduced; secondly, by training a multi-noise scene data set, the adaptability of the model to a complex environment is enhanced, and reliable recognition is guaranteed; and aiming at dialects and accent optimization, the product universality is improved.

The present invention relates to the field of microgrid technology, and specifically to a photovoltaic storage AC / DC hybrid microgrid system. The photovoltaic storage AC / DC hybrid microgrid system provided by the embodiment of the present invention is a highly integrated comprehensive energy system, which integrates photovoltaic power generation, energy storage technology, AC / DC power conversion and intelligent management technology. The system realizes unified coordination and management of photovoltaic power generation, energy storage, energy conversion, energy management, monitoring and protection, and load modules through a microgrid controller. The introduction of a wireless transmission module ensures the real-time transmission of information between modules within the system and the rapid execution of instructions, thereby improving the response speed and reliability of the system. In addition, the modular design enables the system to be flexibly expanded and upgraded according to actual needs, enhancing the flexibility, adaptability and scalability of the system. Overall, the system efficiently coordinates energy utilization, has strong real-time communication capabilities, and is highly flexible and scalable.

The utility model relates to the technical field of unmanned aerial vehicle machining, and discloses an unmanned aerial vehicle wing machining mold which comprises a bottom plate, four threaded rods are movably installed at the top of the bottom plate, a lower mold is arranged among the four threaded rods, and a groove is formed in the lower mold and extends to the bottom of the bottom plate; the mounting plate is movably mounted at the bottom of the bottom plate, an electric push rod is fixedly mounted at the top of the mounting plate, and a movable plate is fixedly mounted at the top of the electric push rod; and a sleeve. By using the electric push rod, the push-out action can be accurately and quickly executed, the production efficiency is greatly improved, the requirement of manual operation is reduced, the labor cost is reduced, the possibility of human errors is reduced, the position and the speed of a push-out part can be accurately controlled, and the production efficiency is improved. This ensures that the parts in the mold can be accurately and completely pushed out, avoiding damage or residue.

The present invention provides a heat sink preparation optimization method, device and readable storage medium. The method makes the process parameters of the process control model corresponding to at least two adjacent deposition layers in the same heat sink different. The process parameters of the process control model corresponding to each deposition layer formed during the heat sink generation process and the performance parameters of the heat sink are collected to generate multiple groups of samples. The models corresponding to different deposition layers are trained through samples and qualified judgment of the samples, so as to improve each process control model to optimize the performance of each deposition layer, and finally achieve the purpose of optimizing the comprehensive performance of the heat sink. The scheme proposed by the present invention is suitable for the preparation of heat sinks of various thicknesses, which can effectively avoid the defects and uneven performance problems in the production of heat sinks, and can reflect the effects of diversification and differentiation in each deposition layer of the heat sink, so that different deposition layers can have advantages of different directions, thereby ensuring that the final generated heat sink has better comprehensive performance.

The invention discloses a model training method and a task execution time prediction method and device, and the method specifically comprises the steps: screening similar historical model training tasks according to historical resource occupation data of historical model training tasks; and inputting the historical resource occupation data of the historical model training task and the similar historical model training task into a prediction model, so that resource occupation feature data corresponding to the historical model training task is determined, and the prediction execution duration corresponding to the historical model training task is determined. And training according to the predicted execution duration and the historical task execution duration. And the trained prediction model determines the prediction execution duration of the target model training task according to the resource occupation data of the target model training task. The method in the specification is higher in prediction efficiency and more accurate. Therefore, the utilization rate of training resources is effectively improved during subsequent resource allocation, and the training efficiency of the whole training process is improved to a great extent while resource waste and shelving are avoided.

The invention relates to the technical field of remote control, in particular to a remote control method and device based on a low-power-consumption Hall sensor. The method comprises the following steps: acquiring remote control equipment type data; performing Hall sensor deployment based on the remote control equipment type data to generate Hall sensor deployment data; performing sensor working magnetic field direction calibration on the Hall sensor deployment data to generate Hall sensor deployment correction data; performing real-time signal acquisition on the Hall sensor deployment correction data to obtain a standard remote control equipment magnetic field detection signal; performing lightweight signal decoding on the standard remote control equipment magnetic field detection signal to generate remote control equipment lightweight decoding data; and performing remote control mode identification on the remote control equipment lightweight decoding data to generate remote control intention data. By optimizing Hall sensor deployment, signal decoding, response delay and multi-device cooperative control, the precision and efficiency of remote control are improved.

The invention discloses a transaction positioning acquisition method, a transaction method and electronic equipment. Creating a main thread, wherein the main thread is used for executing a transaction process; the main thread is utilized to create a sub-thread, the sub-thread is parallel to the main thread, and the sub-thread is used for obtaining positioning information of the transaction terminal and storing the positioning information in a cache; and in the process of executing the transaction process by the main thread, when a positioning data demand is triggered, obtaining positioning information from the cache. In the process of acquiring the positioning information by the sub-thread, the main thread can normally execute the transaction process without suspending a result waiting for the acquisition of the positioning information, so that the execution process is accelerated, and the sub-thread directly stores the positioning information in the cache for the main thread to acquire after acquiring the positioning information. According to the technical scheme, the main thread can obtain the positioning information from the cache when the positioning data requirement is triggered, if the positioning data meet the use condition, the sub-thread cannot be triggered again to obtain the positioning information, and consumption of positioning resources is reduced.

The present invention relates to an intelligent search method for power data based on resource graph, comprising the following steps: S1: integrating power data from different systems, cleaning the data, and unifying the data format; S2: using Stanford NLP and BERT to identify entities, and using a relationship extraction algorithm to establish relationship mapping between entities; S3: using a graph database to store knowledge graphs, mapping entities to nodes, mapping relationships to edges, and establishing multiple types of indexes; S4: converting natural language queries into graph database query languages ​​such as Cypher, and using the index of the graph database to cooperate with a cache strategy to accelerate query execution; S5: based on the established multiple types of indexes, using a query pattern prediction model to predict query patterns and dynamically adjust indexes; S6: adjusting result sorting based on user history and context information, and performing semantic reasoning to expand query results according to the semantic concepts and hierarchical structures of the knowledge graph. The present invention effectively improves the query efficiency on large-scale power multi-source data sets.

The invention relates to an intelligent fire alarm response method and system for a nuclear power plant, a storage medium and electronic equipment. The method comprises the following steps: receiving a fire alarm detection signal, associating field fire-fighting information and starting a fire-fighting isolation measure; fire information is determined based on the fire alarm detection signal, and large-screen display and fire alarm prompt output are carried out in a nuclear power plant fire alarm control center; obtaining process guidance data of the fire-fighting action card; on-site fire information is reported according to the process guidance data; performing verification and confirmation based on the on-site fire information and transmitting the on-site fire information; fire alarm response work is executed according to the on-site fire behavior information; and tracking and recording all fire-fighting response data. According to the invention, structuring and digitization of the fire-fighting action card can be realized, automatic fire-fighting action card file automatic identification process and scheduling can be realized, an operator can carry out visual interaction operation through a mobile terminal or a PC terminal, and efficient fire-fighting action support is provided for the operator; and the corresponding fire-fighting response task can be quickly and accurately executed under the emergency condition.

The present invention is a method for visually grading the appearance quality of pears and a grading production line. The method can reflect the fruit shape, color, defect type and defect area of ​​the sample through characteristic values, and select a suitable machine learning model as a classifier according to the characteristics of the features to make accurate predictions for the above-mentioned various types of information. Finally, the information is input into a decision tree for decision-making, simulating the decision-making classification process of people to obtain the grading results. The whole grading process uses a multi-model fusion method to process various types of information in a targeted manner to avoid the occurrence of overfitting, so that the processing results are more accurate, the detection efficiency and detection accuracy are improved, and the requirements for fast and accurate detection on the assembly line are met. The method is used for the rapid grading of pears on a circular production line. The pears are graded by feature fusion, and the recognition rate is expected to reach more than 95%. It is suitable for grading the appearance quality of different types of pears.

The invention discloses a training examination simulation method and system based on language and action synchronization. The method comprises the following steps: constructing a plurality of police and people communication teaching plan libraries and an expert library comprising standard police and people communication characteristics; according to the police and people communication teaching plan library, establishing a voice database, including a virtual alarm condition model of a face model and an action model, so that the virtual alarm condition model reproduces alarm processing performance; constructing a virtual alarm processing scene based on the voice database and the virtual alarm condition model; acquiring interaction information when the trained police interacts with the voice database and the virtual alarm condition model in the virtual alarm processing scene through acquisition equipment; and scoring the similarity between the interaction information and the communication standard characteristics in the expert database to obtain a training examination simulation result of the trained policeman. According to the invention, through systematic training and assessment, it is ensured that traffic police can rapidly and accurately execute tasks in a complex and changeable traffic environment.

This invention discloses a smart construction site management system, relating to the field of construction site management technology. The system monitors and analyzes the working status of crane operators in their control rooms and the operational status of the cranes, thereby analyzing the corresponding working status of the crane operators. Simultaneously, it monitors and analyzes the operational status of hoist operators and the safety of hoist operations, thus analyzing the corresponding working status of hoist operators. This achieves intelligent monitoring of both crane and hoist operators, improves the dissemination and implementation of safe operating knowledge among crane and hoist operators, provides timely alerts to operators exhibiting improper behavior or operations during work, enhances the focus of crane and hoist operators, and reduces the probability of dangerous accidents.

The invention provides a function calling method, electronic equipment, a readable storage medium and a computer program product, and the method comprises the steps: determining a parameter type of a passing parameter of a first function in response to a calling operation of the first function on a second function in an application program; generating an access constraint instruction matched with the parameter type, executing the access constraint instruction, and recording an access strategy of the passing parameter; when the second function is executed, the passing parameters are accessed according to the access strategy, and the access constraint instruction comprises pointers and immediate operands of the passing parameters. According to the embodiment of the invention, a relatively isolated secure execution environment can be provided for execution of the application program by taking function call as granularity, so that an isolation means with finer granularity is realized. In addition, the security isolation means is quick to execute and does not need to fall into a kernel mode, so that the performance is relatively high, and the response is relatively quick.

The invention provides a processor. The processor comprises an execution module and a memory checking module, in response to a calling operation of a first function on a second function in the application program, the execution module is used for executing the access constraint instruction generated by the compiler and determining a security execution strategy; when the second function is executed, the memory checking module is used for analyzing the access address of the second function according to the security execution strategy to obtain an analysis result and controlling execution of the second function according to the analysis result. Therefore, an isolation means with finer granularity is realized. In addition, the security isolation means is quick to execute and does not need to fall into a kernel mode, so that the performance is relatively high, and the response is relatively quick.