161 results about "Parallel running" patented technology

The invention provides an intelligent system development method and device based on a large model. Receiving an original demand input by a user, and carrying out demand clarification and improvement through a multi-round dialogue reinforcement learning mechanism; synchronously generating a user story, an application architecture model, an executable code and a test case based on the improved original demand; constructing a three-dimensional task scheduling model and dynamic optimization of a retrograde development plan according to the generated development elements; running static risk mode detection and dynamic progress risk prediction in parallel in a development full cycle; and when the project is finished, automatically executing a knowledge distillation process, and carrying out vectorization storage and model iteration of development experience. According to the embodiment of the invention, the intelligent level of software development is improved, and the development efficiency is improved, the quality risk is controllable and the knowledge is continuously evolved by establishing the full-link digital twinning of demand-design-development-test-operation and maintenance.

The invention discloses an in-memory computing architecture simulation method, tool and system, and belongs to the technical field of computers. The method comprises the steps that a plurality of instruction sequences input in parallel are distributed to processing modules running in parallel in an in-memory computing architecture, each processing module comprises a plurality of execution units, at least one of the execution units has a non-linear data computing function, and the non-linear data computing function corresponds to the non-linear data computing function; at least two execution units in the plurality of execution units run in parallel; each processing module is scheduled to parse each distributed instruction sequence into a plurality of operation instructions, and the operation instructions comprise at least one nonlinear calculation instruction; and by scheduling the execution unit in each processing module to process the corresponding operation instruction, simulating the instruction processing operation of the execution unit.

The invention is suitable for the technical field of virtual screening, and provides a molecular virtual screening method based on a parallel depth map Bayesian optimization framework. According to the method, a depth map Bayesian optimization framework is used for performing a virtual screening task; a parallel screening strategy is designed under a complex condition, that is, a parallel depth map Bayesian optimization framework is used for parallel operation on a plurality of computing nodes. In order to further reduce the time cost, the agent model is trained by using an incremental learning method, and meanwhile, the problem of disastrous forgetting is relieved by using an EMA strategy and an OCS sampling method. Compared with a serial depth map Bayesian optimization framework, the parallel depth map Bayesian optimization framework provided by the invention has the advantages that when the same screening task is carried out, the time cost is reduced by 75.1%, the training time of an agent model is reduced from 13.97 h to 4.49 h, and meanwhile, the screening performance is basically kept stable.

The invention relates to a hot update control method and system for an industrial PLC (Programmable Logic Controller) of a microkernel operating system, belonging to the field of industrial controllers. A shadow instance is created by an isolated address space and is bound with a shadow I / O (Input / Output) mirror image and a shadow communication stack; capturing a containment point in the scanning period, wherein the containment point is located between input sampling completion and output submission; acquiring running state information from the current network instance, converting the running state information, and transmitting the converted running state information to the shadow instance; pre-operating the shadow instance in the isolation channel in an accelerated or constant-speed manner, and switching the I / O channel from the current network instance to a shadow I / O mirror image of the shadow instance at the next restraint point if the verification is passed; and continuing to run in parallel for a certain period, unloading the old instance if the continuous running is qualified, and executing rollback if an exception occurs. Depending on a microkernel operating system, shadow instances are drilled and switched under the controlled priority, the real-time performance and certainty are met, the time delay defect of periodic gap switching is overcome, and updating and switching are completed in a single scanning period or a plurality of scanning periods in an atomic mode.

The invention discloses a robot application secondary development method supporting user behavior pattern self-learning, and relates to the field of robot application development, and the method specifically comprises the following steps: S1, data collection and treatment; s2, performing behavior abstraction; s3, mode learning; s4, performing double-domain scheduling during operation; s5, performing mode capitalization; and S6, low code reuse. According to the robot application secondary development method supporting user behavior mode self-learning, during operation, a double-domain isolation framework physically isolates a safety domain from a learning domain, and the situation that key operation of a robot is disturbed in the learning process is avoided; the scheduling algorithm combining the SMP and the EDF and a real-time monitoring mechanism can accurately control core indexes such as thread scheduling delay and interrupt response time, ensure that key tasks such as robot motion control and emergency fault processing are executed preferentially, meanwhile, parallel operation of behavior learning and execution is achieved, the strict requirement for real-time performance of an industrial scene is met, and the real-time performance of the industrial scene is improved. And the operation safety of the robot is ensured.

The invention discloses an intelligent main plan scheduling system oriented to the chemical industry, and the system comprises a data center layer which is used for aggregating order, equipment, materials, inventory and process constraint data; the MPS optimizer operates the tabu search module, the step-counting hill-climbing module and the flood algorithm module in sequence or in parallel in the same optimization period, and automatically switches algorithms under the following conditions: when the continuous iteration improvement rate is lower than a set threshold value, the tabu search is switched to step-counting hill-climbing, and the step-counting hill-climbing is switched to step-counting hill-climbing; when the local search step number is not improved, starting the flood algorithm to re-initialize the candidate solution; and the process scheduler is used for splitting the coarse-grained plan output by the MPS optimizer into minute-level process schedules, and transmitting the execution deviation back to the data center layer in real time through a bidirectional feedback loop so as to trigger incremental rearrangement. According to the intelligent main plan scheduling system for the chemical industry, through a hybrid algorithm dynamic switching mode, the thousand-level order rearrangement time is greatly shortened, and the plan response speed is remarkably improved.

The invention discloses a DCS system IO module channel testing device, and relates to the technical field of industrial automation equipment detection. The device comprises an IO module point inspection instrument and an upper computer, the IO module point inspection instrument is in communication connection with the upper computer, a plurality of sets of independent test modules capable of running in parallel are arranged in the IO module point inspection instrument, and each test module comprises a controller, an excitation source circuit, a signal acquisition circuit and a matrix switch circuit and corresponds to one slot. And the test module establishes a signal path with the IO module connected to the slot through the matrix switch circuit. Wherein a controller in the test module can automatically identify the type of the IO module to be tested, dynamically configure an excitation source circuit and a signal acquisition circuit which are matched with the IO module to be tested, and realize automatic and high-precision detection of various types of IO modules. According to the device, the point inspection efficiency of the IO module of the DCS system can be greatly improved by integrating a multi-channel signal output function and a signal detection function.

The embodiment of the invention provides an operating system and a signal system cloud platform, and the operating system comprises an abstraction layer which is used for packaging the architecture-level drive of a computing node, architecture-level virtualization and the core of a CPU, and providing a packaging interface irrelevant to a hardware system for other layers; the microkernel layer is used for constructing various components of the microkernel by using the interface provided by the abstraction layer; and the application layer comprises a plurality of virtual machines running on a micro kernel, the virtual machines comprise real-time virtual machines and non-real-time virtual machines, and each real-time virtual machine is configured with resources isolated from other virtual machines. According to the technical scheme of the embodiment of the invention, the computing node hardware is abstracted, and the microkernel constructed based on the abstraction supports real-time and non-real-time virtual machines, so that applications with different security levels run in parallel on the same physical computer, and the utilization efficiency of hardware resources is improved.

The invention discloses a multi-agent parallel planning and execution-oriented dynamic task cooperation method and system. The method comprises the following steps of: constructing agents containing planning threads and execution threads which are independent and run in parallel, configuring a callable large language model and a readable and writable centralized memory system for the planning threads, and configuring a callable tool library for the execution threads; a planning thread and an execution thread of each agent establish an exclusive communication link, and a centralized memory system configures a file lock mechanism; the planning thread obtains latest environment information, team communication logs and execution history from the centralized memory system at each time step, calls a large language model to generate a formatted tool library API instruction as a next plan, independently judges whether an interrupt mechanism is triggered or not by the large language model, generates an interrupt signal if the interrupt mechanism is triggered, and sends the interrupt signal to the centralized memory system; writing the tool library API instruction into an action buffer area; the execution thread continuously obtains the latest tool library API instruction from the action buffer area and calls a tool library to execute a corresponding task; if an interrupt signal sent by the planning thread is received through the exclusive communication link, the current task is terminated immediately, and after occupied resources are released, the latest instruction of the buffer area is loaded for execution; and the centralized memory system updates contents at each time step, including an environment observation result returned by the tool library, a communication log between the intelligent agents and a task execution history of each intelligent agent, for all planning threads to access.

The invention provides a BMC optimization method and system of an AI server, electronic equipment and a storage medium, and relates to the technical field of computers. The BMC optimization method of the AI server is applied to a BMC optimization system of the AI server and comprises the steps that a first kernel and a second kernel are started in response to a starting instruction for the AI server, after the first kernel is started, a basic monitoring subsystem of the AI server is taken over, and after the second kernel is taken over, a BMC optimization subsystem of the AI server is started; continuously collecting operation data of a fan and a power supply in the AI server and controlling the fan and the power supply; after the second kernel is started, taking over a global monitoring subsystem of the AI server, performing global monitoring on the AI server, and performing strategy optimization based on a monitoring result; in the parallel running process of the first kernel and the second kernel, if the target kernel detects that the other kernel is abnormal, the target kernel takes over the monitoring task of the abnormal kernel. Therefore, the starting window period of a traditional single-kernel AI server can be eliminated.

The invention provides a data processing method, device and equipment for a large model and a storage medium, and the method comprises the steps: generating an input feature matrix according to prompt words input by a user; according to the processing type of each input feature sequence in the input feature matrix, determining at least one thread block corresponding to the input feature matrix, the operation type of each thread block and a target basic calculation unit to which each thread block belongs; according to the operation type of each thread block and the processing type of each input feature sequence, distributing each input feature sequence to each thread block; running each thread block in parallel to obtain an operation result corresponding to the input feature matrix; and obtaining an output result of the large model based on an operation result corresponding to each input feature matrix. According to the method, the dynamic scheduling of resource allocation can be realized at the kernel function level of the GPU, so that one SM unit can process pre-filling and decoding tasks at the same time, idle or repeated calculation is avoided, and the reasoning performance of a large model is remarkably improved.

A computer program product provides program instructions executable by a processor to cause the processor to perform operations. The operations include identifying a first inference workload instance using a first inference model and a second inference workload instance using a second inference model and identifying whether the first and second inference models have affinity for being run concurrently on a core processing unit. The first and second inference models have affinity if the first and second inference workload instances can run concurrently on the core processing unit without causing either of the first and second inference workload instances to experience latency above a predetermined limit. The operations further include causing the first and second inference workload instances to be run concurrently on the core processing unit if the first and second inference models have been identified to have affinity for being run concurrently on the core processing unit.

The invention discloses an operator arrangement method and device for spatio-temporal data analysis and calculation and a storage medium, and relates to the technical field of data processing.The method comprises the steps that after an operator arrangement diagram is generated, the operator arrangement diagram is analyzed, and to-be-executed operators and the dependency relationship between the to-be-executed operators are determined; generating an executable file according to the code snippets corresponding to the operator to be executed and the dependency relationship; dividing a running task corresponding to the executable file into a plurality of sub-tasks according to partitions of a distributed data set of the input data; and distributing the sub-tasks to respective corresponding computing nodes in the distributed cluster, performing parallel execution, and obtaining an execution result after the execution of each sub-task is completed. According to the method, the subtasks which run in parallel are created and allocated to the computing nodes for computing according to the partition condition of the distributed data set of the input data, so that a single operator node runs and computes on a plurality of servers, and the computing efficiency of spatial data under a large data volume is improved.

The invention discloses a front-end system reconfiguration version control method and device, equipment, a storage medium and a product, and relates to the technical field of system reconfiguration, and the method comprises the steps: receiving a first access request of a user for a target menu in a parallel operation stage of an old system and a new system, and obtaining the identity information of the user; judging whether an access system corresponding to the first access request is the new system or not based on the identity information and a version mapping table; when it is judged that the access system is the new system, a micro-front-end frame is called to load sub-application static resources, corresponding to the target menu, in the new system to a browser for rendering, and a special area page is obtained, and forwarding an interface calling request triggered by the user in the special area page to a new back-end service through a new system gateway. According to the invention, the technical risk and integration complexity of large-scale front-end system reconstruction can be reduced.

According to the electroencephalogram acquisition and signal processing system and the information processing method provided by the invention, the control module is arranged, so that an upper computer is prevented from directly interacting with massive electroencephalogram acquisition data, and a data processing transfer station and a temporary control core are provided, so that the processing efficiency of electroencephalogram signals is improved. Furthermore, the control module comprises a processing unit and a programmable logic unit, the programmable logic unit can control a plurality of acquisition front ends in parallel to realize high-sampling-rate and high-passband signal acquisition, and the processing unit plays a role in connection in the whole electroencephalogram acquisition and signal processing system and can help the upper computer to control and manage the programmable logic unit, so that the accuracy of the electroencephalogram acquisition and signal processing system is improved. And meanwhile, a part of functions which can be realized without continuous data transmission are shared. The control module is provided with a programmable logic unit and a processing unit to realize parallel operation of two independent threads, namely a control flow thread and a data flow thread, so that the control flow and the data flow of the whole system are separated, and the data processing efficiency and the stability of the system are improved.

Systems and methods for artificial intelligence (“AI”) bidirectional monitoring with a quantum-computing-powered system as a flexible guardrail to AI are provided. The systems and methods may include a quantum processor and a classical processor. The systems and methods may include requesting data elements pertaining to boundaries. The systems and methods may include controlling boundary rules and creating classical boundary rules via a classical processor. The systems and methods may include interfacing classical boundary rules with a quantum processor. The systems and methods may include running Grover's conversions in parallel over the boundary rules. The systems and methods may include pulling dynamic market data through a legacy transformation platform including dynamically derived data values and a machine learning model (“MLM”) thereby monitoring and controlling an AI and machine learning (“ML”) processor.

The embodiment of the invention provides an agent task processing method and system. According to the method, an execution module and a verification module are arranged in an intelligent agent. Wherein the execution module runs in the main execution link so as to reasone and execute an operation sequence according to an instruction input by a user, and the verification module runs in parallel in a bypass monitoring thread or process independent of the main execution link so as to monitor interface state data in the process that the execution module executes the operation sequence in real time. In addition, the verification module performs verification evaluation based on the interface state data and the user input instruction to verify whether the evaluation operation sequence is correctly evolved along the target task direction corresponding to the user input instruction. And when the verification evaluation indication is passed, the execution module performs normal reasoning and executes the operation sequence. And when the verification evaluation instruction is not passed, the verification module generates a correction instruction and triggers the execution module to correct the operation sequence based on the correction instruction.

The invention relates to a high code coverage rate fuzzy test method based on a parallel strategy. The method comprises the following steps: deploying a plurality of fuzzy test examples running in parallel; acquiring an execution path and state data of the target program in real time through a dynamic instrumentation technology; dynamically evaluating the seed priority based on the central management system; distributing a high-priority seed to each fuzzy test instance through a feedback driving mechanism, and optimizing file operation and multi-thread synchronization; and dynamically adjusting the seed priority and the test strategy according to the real-time test data. The method has the advantages that system calling and resource contention are effectively reduced, the test efficiency and coverage rate are improved, and efficient cooperation and intelligent test strategy adjustment are achieved through the design of a multi-thread synchronization mechanism and a central management system.

The invention provides a heterogeneous distributed debugging system which comprises a main control screen, at least one power distribution terminal debugging screen and at least one intelligent gateway debugging screen. The main control screen is used for acquiring a test instance; when it is detected that the test operation is started, running the test software to start a test thread for each to-be-tested debugging screen in the test instance; the power distribution terminal debugging screen is used for executing a test case for the power distribution terminal according to the test thread to determine a first test report; and the intelligent gateway debugging screen is used for executing a test case for the corresponding intelligent gateway according to the test thread to determine a second test report. Thus, the main control screen in the heterogeneous distributed debugging system manages the debugging screens corresponding to the different test instruments by obtaining the test examples, then parallel operation of the power distribution terminal debugging screen and the intelligent gateway debugging screen is achieved, each debugging screen works independently, the debugging efficiency is improved, and the overall debugging time is shortened.

The invention provides a power grid micro-front-end construction method and architecture based on a qiankun framework. The method comprises the following steps: starting a main application of the qiankun micro-front-end framework; registering a global state in the main application, and monitoring a global state change event; when one sub-application loaded through routing matching receives a service request corresponding to the other sub-application, triggering a global state change event, and transmitting parameters about calling the service request of the other sub-application into the global state change event; after monitoring the global state change event, the main application judges whether the other sub-application is mounted in a hidden container or not, and if not, manual mounting is carried out; and the other sub-application acquires the global state change event from the main application and then executes a corresponding service function. According to the method, parallel operation of the sub-applications and cross-sub-application communication can be realized in the power grid micro-front end based on the qiankun framework, and service multiplexing is realized while the independence of the sub-applications is kept.

The system and methods described herein allow users to give their applicant information to at least one entity when seeking to submit an inquiry associated with an item, and have various lender microservices run in parallel, segregated by entity, in a jailed environment. The result of these microservices may be returned as a response to the inquiry, being determined autonomously for each respective entity based on one or more respective rule sets or executable logic for each respective entity. Payloads for multiple entities may be combined in a single output from the jailed environment due to outputs from the environment being encrypted in a universal format.

The present application provides a simulation method, apparatus, and electronic device for a multi-core mechanism, including: during the running of a test case, identifying whether there are multiple target test sequences that can be parallelized; in response to the existence of multiple target test sequences that can be parallelized, configuring a matching target simulation core for each target test sequence; running each target test sequence in parallel based on the target simulation core until the end; continuing to run the next test sequence in the test case until the test case runs to completion. In the present application, the parallel running of test sequences is realized based on the multi-core mechanism, shortening the running time of the test case and improving the efficiency of chip verification.

A computer-implemented method comprising: obtaining real-time text data relating to a matter, the text data comprising a plurality of portions of information; performing a categorisation process, wherein the categorisation process is configured to run a plurality of threads in parallel, wherein each thread of the plurality of threads acts on one portion of information at a time, wherein each thread performs the following steps: (i) obtaining a sentiment score based on the portion of information using a Sentiment Analysis machine learning, ML, model; (ii) assigning a category to the matter based on the sentiment score using a classification ML model trained on historical data; and (iii) updating a live category based on the category assigned to the matter; and finally, outputting the live category to a user in real-time.

The invention belongs to the technical field of spiking neural networks. The invention provides a construction method of a reversible pulse neural network supporting parallel computing. According to the embodiment of the invention, the reversible calculation structure is introduced, the neuron state of an intermediate layer does not need to be explicitly stored in the forward propagation process of the network, and the reconstruction of the activation value can be realized through the reversible mapping relation in the backward propagation stage. In the design stage of the reversible structure, hardware mapping characteristics are fully considered, and forward calculation and reverse reconstruction processes of reversible mapping are simultaneously incorporated into a parallel execution framework of a configurable calculation array. Through collaborative design of a data path and control logic, forward propagation and reverse reversible reconstruction can be efficiently and parallelly operated in the same array, and inter-stage calculation idle time and data waiting time are remarkably shortened. On the premise of ensuring reversibility, streamlined and parallel execution of forward and reverse calculation is realized, and the throughput rate and the training speed of the system are integrally improved.

A RAN system includes Capsules that operate in parallel to cooperatively perform RAN stack operations, each capsule utilizing a processor and a memory configured to perform a component of the RAN stack operations, wherein a first subset of plurality of capsules is implemented by a first module under control of a first liquidity controller, wherein a second subset of capsules is implemented by a second module under control of a second liquidity controller, and wherein the first subset of capsules are mutually exclusive of the second subset of capsules. The RAN system migrates, via the first liquidity controller and the second liquidity controller, a first capsule from the first subset of capsules to the second subset of capsules as part of a security mechanism in response to a security attack on the RAN.

The embodiment of the invention discloses a vehicle and an upgrading method thereof, a server and an upgrading file processing method thereof, a medium and a program product, and the method comprises the steps: obtaining an upgrading file package under the condition of receiving an upgrading instruction, and building a first thread and at least one second thread based on grouping information; upgrading the vehicle based on the first compressed upgrade package in the first thread; the vehicle is subjected to serial upgrading based on the second compressed upgrading package in the corresponding second group in each second thread, and the first thread and at least one second thread run in parallel. According to the embodiment of the invention, the longest upgrading task in the system is ensured not to block the execution of other upgrading tasks, the longest upgrading task is continuously operated as an independent parallel unit, and other threads can synchronously promote intra-group upgrading by fully utilizing calculation and I / O (Input / Output) resources, so that the total time consumption of remote online upgrading is greatly shortened globally, and the upgrading efficiency is improved. And the upgrading efficiency is improved.

The invention relates to a hydrogen production system 4E multi-objective optimization method based on an NSGA-II-MOGWO parallel mixing mechanism. According to the method, a primary population is generated firstly, and global operator parameters and corresponding strategies of NSGA-II and MOGWO are set. System design and operation decision variables are formalized into vectors, and a feasible region is defined by inequality / equality constraints. Then simulating each candidate solution by calling the system hourly to obtain a 4E fitness function value; in the optimization process, NSGA-II and MOGWO are operated in parallel in each generation, and excellent individuals are exchanged according to a preset period to give consideration to exploration and utilization; and after the parent and the two branch filial generations are combined, a GPR is adopted to establish a proxy model for batch screening of candidates for real evaluation and proxy retraining. And finally, through rapid non-dominated sorting and congestion degree selection and external file maintenance, outputting a high-quality Pareto frontier solution and a plurality of representative configuration schemes, and visualizing the high-quality Pareto frontier solution and the representative configuration schemes.

Enhancing data processing is provided. A plurality of derivative application instances is generated to run a plurality of parallel jobs based on an image of an instance of an application providing a service corresponding to a data processing request. One derivative application instance is generated for each respective job of the plurality of parallel jobs to run the plurality of parallel jobs to meet defined data processing performance objectives. The plurality of parallel jobs is run on the plurality of derivative application instances at the same time in parallel to increase data processing throughput and decrease an amount of time and resources needed to fulfill the data processing request. Each job of the plurality of parallel jobs retrieves a particular chunk of a dataset corresponding to the data processing request from a database to process that particular chunk of the dataset to generate a sub-result of the data processing request.

The invention discloses a container trusted measurement and verification method in a confidential environment, and belongs to the technical field of information security. According to the method, a dual-system architecture in which calculation and protection components run in parallel is designed, TCMTPCM and vTPCM management modules and a trusted software base TSB are constructed and deployed in the protection components, protection and hardware trusted roots are provided for running containers in a confidential virtual machine in a confidential environment, and integrity measurement judgment of key content is carried out in the starting and running processes of the containers. Confidentiality, integrity and credibility of a user service container provided in the cloud service are realized. And finally, a dual verification mechanism for providing a service container is introduced, and a remote verification platform not only needs to verify the identity and integrity of a confidential virtual machine where the container is located, but also needs to verify the integrity of the container and the integrity of the container during operation based on the vTPCM in the virtual machine. The method can get rid of absolute dependence of a user on single hardware trust of a hardware trust provider, and integrity measurement and verification are carried out during container operation in a confidential environment.