223 results about "Toolchain" patented technology

The invention provides a dynamic tool integration method and system for an intention-driven AI dialogue system, and the method comprises the steps: S5, extracting resource occupation information from a transaction log according to a state persistence record, carrying out the resource release and temporary context cleaning of each heterogeneous component, and obtaining a released resource list; s6, searching a breakpoint state snapshot and a transaction log from the distributed database according to the released resource list, and reconstructing a tool chain execution flow historical state to obtain a fault context analysis result; and S8, according to the context recovery state and the semantic feature vector, updating a tool chain execution stream, and reallocating computing resources through a heterogeneous component coordination mechanism to obtain a task continuity confirmation result. According to the method, the recovery efficiency and the resource utilization rate of the tool chain in an abnormal scene are remarkably improved, and the task continuity and the system robustness in a high-concurrency environment are guaranteed.

A simulation system based on a large language model comprises a large language model module, a scenario generation tool module, a meta-scene database module, a simulation deduction engine module, a simulation model module, a simulation front-end module and a simulation tool set module. The large language model module analyzes an instruction and a situation description input by a user, splits semantic information and meta-scene information in the instruction and the situation description, sends the split semantic information and meta-scene information to the scenario generation tool module, and receives a scenario file output by the scenario generation tool module; further sending the received scenario file to a simulation deduction engine module, and receiving a simulation process and a simulation result generated by the simulation deduction engine module; and meanwhile, the large language model module can call the simulation tool set module by using a tool chain instruction through a simulation deduction engine, so that the simulation tool set module generates a situation processing result. According to the method, man-machine interaction based on the large language model is realized, and a tool set is called by utilizing the large language model to assist a user in completing simulation and analysis.

The invention discloses a container mirror image construction and use method oriented to embedded cross compiling, relates to the field of compiling methods of operating systems, and solves the compiling problem of a software source program under an embedded Linux operating system. The invention relates to an embedded cross compiling method based on a container mechanism, which comprises the following steps of: integrating tool chain software and an operating system header file, a library file and the like required by compiling to form a new mirror image file based on a basic container mirror image, starting a container through the mirror image when a program is compiled, and after a source code catalog is mounted and a catalog is constructed, starting the container through the mirror image; and calling the tool chain through the script to complete cross compiling. According to the method, the construction of a cross compiling environment can be separated from the shackle of a host machine operating system, the method has the characteristics of high environmental adaptability, flexible and convenient deployment, good platform safety and the like, compilation of engineering source codes such as application programs, library files and drivers is supported, and powerful support can be provided for cross compiling of embedded programs.

The invention discloses a large model calling system and method based on an MCP tool chain, and the method comprises the steps: receiving user input, and carrying out the preliminary matching of tools: querying whether the user input contains an entity in an entity database or not through a fuzzy matching technology, and carrying out the preliminary screening of a plurality of candidate tools based on a matching result, so as to obtain limited to-be-selected tools; user input is converted into semantic vectors, semantic matching is conducted in a vector database, and semantic similarities of all tools to be selected are ranked through a similarity calculation method; extracting key information in user input; abstracting the user input into a specified format; and matching the specified format with the historical workflow template, and if the historical workflow template is matched, replacing the key information to a corresponding position of the historical workflow template by adopting a structure-preserving parameter replacement algorithm to realize workflow reuse. According to the method and the device, a complete technical link from accurate tool screening to intelligent workflow multiplexing can be realized.

The invention discloses a building construction carbon emission dynamic optimization method and system based on Revit and AI linkage. The method comprises the following steps: step 1, data integration; step 2, model training; step 3, logistics simulation; step 4, collaborative optimization; 5, dynamic early warning is carried out; step 6, generating a thermodynamic diagram; step 7, integrating paths; 8, iterative optimization is carried out; and 9, multi-scene analysis is carried out. According to the method, real-time variables are dynamically captured through an Azure AutoML time sequence model, and the prediction error is reduced; automatic management is realized by integrating a code-free tool chain, and the cost is reduced; double objective functions are innovated to synchronously optimize transportation and environmental protection benefits; a full-life-cycle closed-loop system is constructed to improve the dynamic adjustment capability; and the generation of visual decision support efficiency is improved, and revolutionary technical support is provided for green construction.

The invention discloses a lightweight cross-platform PLC (Programmable Logic Controller) multi-language code generation method, which relates to the field of industrial process control, and firstly provides a unified intermediate data model, and on the basis, a module for converting each PLC language into the unified intermediate data model is realized; on the basis of a standardized JSON data structure, a modularized unified intermediate language conversion chain is further supported, an analyzer and a generator communicate through a JSON interface, a cross-platform conversion path is supported, and unified intermediate language conversion from PLC to JSON to XML to ST to C is completed; the method comprises the following steps of: integrating an open source compiler and runtime, constructing an open and pluggable automatic execution framework, compiling a generated C code into a platform executable file by using an adaptive GCC tool chain aiming at different target platforms, and deploying a program generated by compiling to a soft PLC runtime system supporting multiple platforms. According to the method, a JSON intermediate data structure which is light in weight, unrelated to languages and uniform in structure and a matched conversion framework of the JSON intermediate data structure are constructed.

The invention provides a script language-oriented seismic batch processing job editing method and device, and the method comprises the steps: obtaining a to-be-processed seismic job demand, and determining a seismic job project and database information; initializing each module object in a pre-established seismic batch processing operation framework, wherein the framework is a scripting language framework comprising a preset core library, a module library and a tool chain; determining a standard type parameter corresponding to each module object; initializing an operation topological structure of the seismic batch processing operation framework; selecting a target module object and adding the target module object to the target operation based on the earthquake operation project and the database information; and outputting the topological graph of the module according to the debugging requirement, and generating a target operation corresponding to the seismic operation requirement by calling the XML format file. According to the method, seamless connection of job editing and code editing is realized through a script language-oriented seismic batch processing job framework, so that the method is more flexible and efficient; and diversified requirements of a user for job editing by using a script programming language are met.

The invention provides a micro-front-end multi-application architecture, a micro-front-end multi-application system and an integration method of the micro-front-end multi-application system. The method comprises the following steps: constructing a main application based on a Vite tool chain; a multi-application universal code warehouse is constructed based on Monorepo; based on the main application and the multi-application universal code warehouse, introducing a Wujie micro-front-end, and constructing a micro-front-end multi-application architecture; performing technical agreement based on a micro-front-end multi-application architecture; based on the technical agreement, transforming a main application and a sub-application which are integrated and accessed into the micro-front-end multi-application architecture, and constructing a multi-application integrated system; according to the dependency relationship of different application resources in the multi-application integrated system, public resources are planned, a CDN content distribution network is constructed, and a distribution strategy of the resources and a distribution service of the resources are constructed; performing configuration setting and construction deployment on the multi-application integrated system and the resource distribution service; according to the method, the problem of'giant application 'of enterprise-level integrated software is solved.

A method and a system for compiling a neural network, a computer storage medium, and a compilation device are provided. The method for compiling the neural network comprises: translating a network file into an intermediate representation file; optimizing the intermediate representation file to obtain an optimized intermediate representation file, based on a performance analysis, single-node optimization, and collaborated optimization; generating a network template file based on hardware interfaces through the optimized intermediate representation file; compiling the network template file into an executable inference application. The present disclosure aims to design and implement an automated compilation toolchain framework. This framework adjusts parameters, generates code, creates intermediate representations (IRs), and applies optimization algorithms based on software and hardware information. When this compilation toolchain operates on a target chip, it ensures consistent network output results, achieves higher computation rates within shorter optimization times, reduces computation delays, and facilitates user debugging and tuning.

The invention provides a method for searching structured data based on natural language driven by a large model, which comprises the following steps of: preprocessing interface parameters through a data interface of an openAPI (Application Program Interface) to generate an interface file, and adding the preprocessed interface parameters into an agent tool chain; the large model extracts keyword parameters from a question input by a user through a natural language and performs semantic analysis on the extracted keyword parameters, so that the keyword parameters are converted into interface fields or table fields in an interface file; the large model adapts an analyzed result to a tool name in the agent tool chain; calling an adaptive interface name by the large model, and calling a corresponding tool to query a data result; the large model outputs the result according to the form of the user demand, standardization processing is performed on the interface of the traditional data, the agent tool is added, the large model is used, different data search requirements are met, and the personalized data query requirement is met.

The invention discloses a smart home control method and system based on a multi-modal large model, and the method comprises the steps: carrying out the cross-modal feature extraction and semantic alignment processing of voice, image and sensor data through an improved multi-modal large model, and achieving the semantic space mapping of multi-source data based on a context sharing mechanism and reasoning logic; the method comprises the following steps: abstracting local smart home equipment into a standardized resource entity by using an MCP technology, defining an equipment function feature set through a protocol interface, and packaging the equipment function feature set into a callable tool chain; constructing a real-time data fusion model, and generating a control instruction sequence through the multi-modal large model based on the real-time data fusion model; and calling a corresponding tool in the tool chain according to the control instruction sequence to execute equipment control. The problems that an existing system depends on single-mode data and is easily interfered by the environment, the access cost of new equipment is high due to the fact that a unified data fusion framework does not exist, a control strategy is difficult to adapt to user behavior changes, equipment communication protocols are different, and the user requirement matching degree is low are solved.

The invention discloses an Alliance tool chain logic synthesis method based on multiple logic domains. The method comprises the following steps that an intermediate representation structure of a target circuit is obtained; mapping a Boolean expression in the intermediate representation structure into an XMG network equivalent to the logic function of the target circuit; logic optimization with the optimization logic hierarchy as the target is executed on the XMG network obtained through mapping; converting the XMG network after logic hierarchy optimization into an AIG network; logic optimization with optimization of the number of nodes as a target is executed on the AIG network obtained through conversion, and the optimized AIG network is output as an aig file; and reading the aig file, analyzing the aig file, outputting the analyzed aig file as a VHDL file, inputting the VHDL file to the Alliance tool chain, and generating a gate-level netlist corresponding to the optimized target circuit. According to the method, the logic optimization efficiency and expandability can be greatly improved, the performance bottleneck of an existing Alliance tool chain is effectively solved, and the large-scale circuit design processing capacity of the Alliance tool chain is remarkably improved.

The invention relates to a cross-platform software development method and related equipment. The method is applied to the electronic equipment providing a low-code tool chain, and the low-code tool chain comprises a low-code development platform and a compiler. Based on the configuration of the meta model of the vehicle-mounted resource, the low-code development platform generates an automobile domain specific language VDSL file for describing the vehicle-mounted software function; the compiler obtains the VDSL file, compiles the VDSL file into running program codes running in different controllers and sends the running program codes to a cloud server, and the cloud server is used for sending the running program codes to the different controllers. According to the method and the device, the running program codes corresponding to different controllers are generated through the unified VDSL, so that cross-platform running software programs can be conveniently developed in the same platform, the software development efficiency is improved, and the development cost is reduced.

The invention discloses an agent and tool chain collaborative data processing method and device and a medium, and relates to the technical field of agent application. The method comprises the steps of receiving a data processing task target and related configuration parameters input by a user, and initializing task context data based on the task target and the configuration parameters; executing a circular processing process by an agent based on a ReAct framework, and dynamically adjusting a task strategy by iteratively executing a thinking stage, an action stage and an observation stage until a task completion condition is met; in the circular processing process, based on standardized tool calling instruction data generated by the intelligent agent, cooperatively calling the MCP tool chain and a plurality of tools in the external tool adaptation layer to execute a data processing task; performing aggregation processing and data verification on the multi-source result data returned by the tool execution to generate integrated analysis result data; and displaying task execution process data and a final analysis result in real time through a front-end interaction interface, and dynamically adjusting a task execution process based on user intervention instruction data.

The invention relates to a container application migration method and system and a storage medium, and the method comprises the steps: extracting a dependency list and call information through a script, compiling a source code of a current architecture of a container application into an executable file under a target architecture, replacing a dependency library file under the current architecture of the container application with the dependency library file under the target architecture, and carrying out the migration of the container application. Based on the dependency list and the calling information, using the basic mirror image of the target architecture to reconstruct the container mirror image so as to ensure that the container application can run on the target architecture, intercepting and redirecting the system calling through an LDPRELOAD mechanism so as to enable the system calling to adapt to the target architecture, and using a cross compilation tool chain so as to improve the system calling efficiency. And replacing the instruction set of the current architecture with the instruction set of the target architecture, and carrying out container application migration operation. The dependency list and the calling information are extracted through the script, so that the possibility of omitting the dependency list and the calling information required by operation of the container application is reduced, and efficient and stable operation of the container application on the target architecture is ensured.

The invention provides a method for achieving automatic interface testing by integrating a software CICD process with MeterSphere. The method comprises the following steps that S1, environment preparation is conducted, and a MeterSphere platform is built; s2, a Jenkins server and a MeterSphere platform are integrated, and the Jenkins server and the MeterSphere platform are S3, generating a pipeline script; s4, automatic testing is carried out; s5, sending a test result notification; through seamless integration of Jenkins and MeterSphere, research and development personnel do not need to manually transmit deployment information to test personnel, and the test personnel can obtain latest deployment and test feedback in time, so that delay of manual intervention and information transmission is reduced, and the overall communication efficiency is improved; the interface test is automatically executed, manual operation is reduced, and the test execution speed is increased; by integrating a test report and tool chain automation, the accuracy, transparency and traceability of a test result are ensured.

The invention belongs to the technical field of computers, and provides a macro instruction execution method, a macro instruction compiling method, an upper computer, a touch screen device and a storage medium. The upper computer can provide a platform for compiling a macro instruction through a Windows operating system for a user, the programming language threshold of the user is lowered, then a cross compiling tool chain is utilized, and therefore the programming speed of the user is improved. Compiling the macro instruction into a target dynamic library file capable of running in an embedded operating system; and the touch screen device downloads the target dynamic library file and calls the target dynamic library file to execute the macro instruction. According to the method and the device, the macro instruction written in the upper computer of the Windows platform can be compiled to generate the dynamic library capable of running in the embedded operating system platform through a cross compiling mode, so that the function of executing the macro instruction in a cross-platform manner is realized, and the macro instruction can more intelligently and efficiently execute complex operation or flow defined by a user.

The invention discloses an AI dialogue flow control method, system and device supporting real-time interruption of a user, a medium and a program product. The method comprises the steps that a user request is received, and a flow session is established; based on the stream session, requesting LLM increment generation, receiving fragments generated by increment, returning the fragments, and pushing the fragments to the client; when a click stop request of a user is received, an STOP signal is sent; when the STOP signal is monitored, sending a mark termination request; after a mark termination request is received, a subsequent generation termination request is sent to the LLM, a STREAMTERMINATED semantic event is published and broadcasted, and generated content and a termination mark are returned to the client; terminating subsequent generation comprises stopping untriggered subsequent reasoning and tool chain calling; discarding the tail cache which is not sent; and recording the output token count, and terminating the offset. The AI dialogue flow control method supports real-time interruption of the user, is low in delay, clear in semantics and recoverable, and further has the advantages of saving resources, being high in stability, capable of being universally expanded, gaining behavior analysis and the like.

The invention relates to the technical field of artificial intelligence, in particular to a brain-equipped robot end-side cloud collaborative management system, which comprises a brain-equipped robot unit, a man-machine interaction client, an APP / applet, a resource management platform and an application market platform, the brain-equipped robot unit comprises a multi-mode sensing module which is used for understanding natural language instructions, screen contents and voice interaction; a large model decision module; the execution control module is used for simulating a keyboard and a mouse to operate computer software; the man-machine interaction client and the APP / applet serve as an interface for interaction; the resource management platform comprises equipment and a user management module; the resource matching algorithm module is used for evaluating and distributing optimal computing resources; the block chain registration module is used for recording ownership and use right information of the task; the sandbox execution environment module is used for creating a running environment; the application market platform is used for providing a developer tool chain and a revenue distribution mechanism. The application efficiency of the agent system can be improved.

The invention provides a macro-cell layout evaluation optimization method based on a large language model, and belongs to the field of electronic design automation. According to the method, a large language model is introduced to assist layout generation, a current initial layout is improved in combination with connection weight analysis, on the other hand, an improved result is judged and continuously iterated according to natural language feedback of an experience manual or an expert, and closed-loop optimization is formed; carrying out weight quantization on direct and indirect signal paths among macro units, constructing a logic association strength matrix, and realizing accurate identification and priority ranking of key time sequence paths; the availability score of each candidate layout scheme is calculated through a layout availability scoring module, dynamic measurement is carried out on surrounding standard unit resources, the resource distribution state is evaluated in real time, the position of a macro unit is actively adjusted, and resource congestion or waste is avoided; the method is deeply integrated with an open source EDA tool, the dependence on a commercial tool is reduced, the tool chain deployment and iteration period is shortened, and the transportability and usability of the method are guaranteed.

A framework that brings together application development, resource management, and accelerator design into a single compilation and runtime toolchain. In one example aspect, a method for enabling execution of multiple software applications on an embedded system includes receiving a first invocation of an Application Programing Interface (API) function by a first software application and a second invocation of the API function by a second software application, and determining available computing resources corresponding to the API function. The method includes selecting, based on one or more characteristics associated with the available computing resources, a first set of computing resources and a second set of computing resources. The method also includes scheduling the first invocation of the API function using the first set of computing resources and scheduling the second invocation of the API function using the second set of computing resources.

A system and method for a protocol parser that has native support for parsing TLVs and flag-fields and allow users to code in a language convenient to them and leverage standard tool chains and tools. The same parser source can be compiled to arbitrary software and hardware targets without code change and provide the highest performance possible given its capabilities.

The present invention discloses a method for driving BI analysis scenarios by an intelligent agent, the analysis method comprising: step S1: initializing a tool chain object library, a tool library and a BI scenario library; step S2: the front-end module interacts with the user, obtains the user input and sends it to the Web service; step S3: the Web service receives the request, records it in the database, and calls the intelligent agent; step S4: the intelligent agent parses the task and matches the tool chain according to the scenario; step S5: the intelligent agent executes the tool chain using the scenario parameters; step S6: the intelligent agent calls the BI scenario tool; step S7: the intelligent agent completes the execution of the remaining steps of the tool chain; step S8: the Web service returns the received output to the front end. A hierarchical software architecture is adopted to enable the BI scenario to be flexibly expanded to support the ever-changing analysis needs, and a fine-tuning technology is adopted to make the large model more suitable for the BI analysis scenario.

A system and method for designing and optimizing a hardware accelerator of a neural network is provided. In a pre-design phase, rules are extracted from compilation features that describe a transformational relationship between a neural network operator, a coarse-grained operator, and a fine-grained data stream. A fast mapper and a data flow mapper are generated for converting the neural network model into coarse-grained operator descriptions. In the coarse-grained design stage, a plurality of preliminary hardware accelerator designs are generated using an architecture optimizer. The coarseness operator descriptions are simulated using a coarseness simulator to obtain performance metrics for each preliminary accelerator design. In the fine-grained design stage, a data stream mapper and a fine-grained simulator are adopted to complete preliminary hardware accelerator design. A hardware accelerator is generated from the finally determined hardware accelerator design and a corresponding software tool chain is created, including a compiler and a software development kit for programming, debugging and deploying the hardware accelerator design.

The invention relates to the technical field of embedded system software development and continuous integration, in particular to a docker-based AGV (automatic guided vehicle) cross compiling method, which comprises the following steps of: constructing an ARM (advanced RISC machines) root file system on an x86 development host by using a virtualization tool, configuring a header file and a dependency library, and packaging and compressing to obtain an ARM root file system mirror image; pulling the basic mirror image from the public warehouse to create an x86 Docker container, copying the ARM mirror image to a specific path of the container and mounting the ARM mirror image to prepare a Docker mirror image special for AGV cross compiling; writing a CMake cross tool chain file in the container, introducing an ARM mounting path, setting a dependency path and appointing a tool chain; then integrating the commands, automatically updating the container through mirror image ID comparison, and constructing operation commands such as Makefile packaging and compiling and the like at the same time; and finally, calling an integration command to start cross compiling, and generating a program capable of running on an ARM board card of the AGV. According to the method, unified and efficient multiplexing of the AGV cross compiling environment can be realized, the development efficiency is improved, the maintenance cost is reduced, and the method is adaptive to an AGV project needing cross compiling.

The invention discloses a resource optimization method and system based on demand driving, belongs to the technical field of software product development, and aims to solve the technical problems that resource allocation in software product life cycle management depends on artificial experience, is low in efficiency, subjective in allocation standard and difficult to dynamically adapt to project changes. Comprising the following steps: acquiring demand characteristics, and endowing the demand characteristics with weights; the method comprises the following steps: modeling available resources of a project to form a resource constraint model, defining attributes and constraint conditions of the resources through the resource constraint model, and collecting state information of the resources in real time; on the basis of the weights of the demand features, the state information of the resources and the attributes and constraint conditions of the resources, the resources are allocated through a multi-objective optimization algorithm, and an optimal resource allocation scheme is formed; issuing the resource allocation scheme to an execution end through tool chain integration; and optimizing the weight of the demand feature and the optimal resource allocation scheme based on a machine learning method.

The invention provides a collaborative management system and method for an EDA all-in-one machine, the collaborative management system comprises a business layer, an application layer, a platform layer and a hardware layer, the hardware layer is used for providing hardware resources and deploying container instances; the platform layer is used for performing instance management, resource quota allocation and instance life cycle control of container instances on the hardware resources provided by the hardware layer; the application layer is used for running the EDA tool of the corresponding version in the container instance; and the business layer is used for carrying out resource planning, template configuration, task scheduling and result management on the chip design whole process of the EDA all-in-one machine. According to the scheme, the whole process of chip design is integrated through the business layer, scattered project information, resource requirements and result management are incorporated into a unified framework, the tool chain information barrier is broken, and the manual coordination cost is reduced. And meanwhile, the predictability and traceability of task execution are improved, and an efficient and ordered collaborative management scheme is provided for the EDA all-in-one machine.

A system and method is provided for designing and optimizing hardware accelerators for neural networks. During a pre-design phase, rules are extracted from compilation patterns that describe conversion between neural network operators, coarse-grained operators, and fine-grained dataflow. A fast mapper for converting neural network models to coarse-grained operator descriptions and a dataflow mapper are generated. A coarse-grained design phase employs an architecture optimizer to generate plural provisional hardware accelerator designs The coarse-grained operator descriptions are simulated using a coarse-grained simulator to obtain performance metrics of each provisional accelerator design. A fine-grained design phase employs a dataflow mapper and fine-grained simulator to finalize provisional hardware accelerator designs. A hardware accelerator is generated from a finalized hardware accelerator design and a corresponding software toolchain is created including a compiler and software development kit (SDK) for programming, debugging, and deploying the hardware accelerator design.

The invention belongs to the field of aircraft fault diagnosis and maintenance guarantee, and particularly relates to a flight parameter diagnosis rule dynamic compiling method. According to the method, codes required by the criterion, the blank-leaving criterion formula and the associated parameters are designed in advance according to the criterion function, the criterion formula and the associated parameters are filled after the criterion formula is analyzed, the to-be-compiled criterion file is generated, and finally the to-be-compiled file is compiled into the executable program capable of being linked to the main program through the compiling tool chain. And thus, the criterion is dynamically compiled into an executable program. The time for converting the criterion into the executable program is greatly shortened, the software development process is guaranteed to a great extent, and the guarantee efficiency is greatly improved.

This invention discloses a method and system for constructing an AI-native operating system based on ecosystem collaboration, belonging to the field of operating system technology. The invention first acquires and decomposes target requirements to form a set of functional and non-functional requirements, constraints, and priority lists. Based on this, it generates an operating system architecture description including kernel configuration and subsystem division. Then, it generates and completes multi-dimensional checks on module code and configuration files, and compiles, links, and packages them through a toolchain to generate an operating system image. This invention employs a three-layer AI model to realize requirement reasoning, system construction, and test optimization, forming a complete automated closed loop from requirement input to runnable image output. Simultaneously, it establishes a unified ecosystem collaboration mechanism to support multi-vendor collaborative development. This invention effectively reduces errors and repetitive work caused by manual intervention, shortens the development cycle, improves system stability and adaptability, and promotes the collaborative evolution and improvement of the AI-native operating system ecosystem.