54results about How to "Improve iteration efficiency" patented technology

The invention provides a speech synthesis model training method, a speech synthesis method, a speech synthesis device and a storage medium, and the speech synthesis model training method comprises the steps: obtaining the sample text data of a plurality of speakers and the acoustic features of the sample audio data of the plurality of speakers according to the sample audio data of the plurality of speakers, feature extraction is carried out on the sample text data of each speaker to obtain sample text features of each speaker, a speech synthesis sample of each speaker is constructed according to the sample text features of each speaker and acoustic features of the sample audio data of each speaker, and speech synthesis is carried out according to the speech synthesis samples of multiple speakers. And respectively carrying out model training to obtain speech synthesis models of the plurality of speakers. Automatic training of the speech synthesis model is carried out in a one-button mode, and the iteration efficiency of the speech synthesis technology is improved.

The invention discloses a reaction spectrum reverse iteration algorithm of an additional damping ratio, relates to the technical field of constructional engineering, and aims to reversely calculate reasonable parameters of a damper and a connecting component from the final demand (namely, the additional damping ratio demand and the rigidity demand of a damping target) of a main body structure, and merge and iterate according to the model of a common damper product to obtain a response spectrum reverse iteration algorithm of the additional damping ratio. And compared with the traditional forward iteration, the efficiency of the reverse iteration algorithm is higher.

The invention relates to the technical field of research of reactor fuel irradiation, in particular to a fuel power calibration method for a loop irradiation test, which comprises the following steps: carrying out an out-of-reactor hydraulic test, and establishing a high-precision CFD (computational fluid dynamics) model of a fuel loop irradiation device based on a hydraulic test result; calculating the gamma heat release rate ratio of each material and the total heat release in the test section, and preliminarily determining the nuclear power value of the fuel assembly; based on the heat conduction differential equation and the convective heat transfer equation, the gamma heat release rate of the material and the nuclear power value of the fuel assembly are corrected; repeating the calculation process of S3 until the nuclear power value of the fuel assembly converges, and obtaining the final theoretical calculation value of the gamma heat release rate of each material and the nuclear power of the fuel assembly; and further verifying the calculation result to finally obtain the nuclear power of the fuel core body, and obtaining the burnup of the fuel assembly in the irradiation test through time integration. The method aims to solve the technical problem that the nuclear power of the fuel assembly calculated based on thermotechnical measurement data in a loop irradiation test has deviation compared with an actual value.

The invention provides a dynamic route replacement method, system and device based on a micro front end and a medium, which are applied to a micro front end container running in a browser, and the method comprises the following steps: receiving an access request sent by a user; the access request is a uniform resource path; acquiring a set routing mapping table in the micro front-end container; the set routing mapping table comprises first routing information and second routing information; determining a corresponding accessible route through a logic matching rule according to the access request and the set route mapping table; and loading and displaying the corresponding sub-application according to the accessible route. Compared with the traditional technical scheme, the method has the advantages that the route dynamic replacement logic is moved forward into the web container, so that the complexity of operation and maintenance deployment can be greatly reduced, and the problems of tedious maintenance and high maintenance cost are solved.

The invention discloses a to-do list collection and processing method and system based on Java, and belongs to the technical field of computer software, collection and processing of to-do lists are achieved through feature dictionary matching, and the method comprises the steps that multi-source data are collected, a to-do feature dictionary is constructed, and to-do information is recognized through a multi-dimensional matching algorithm; the rear end constructs a layered architecture based on the SpringBoot and the SpringMVC, and the SpringMVC is used for providing a standard RESTfulAPI (Application Program Interface) and supporting to-do query and state updating functions; the front end adopts Vue3 + ElementPlus to realize a response type Web interface, and a back-end interface is called through Axios to realize efficient collaboration of the front end and the back end. According to the method, automatic access of multi-source heterogeneous data and extraction of the information to be handled are realized, the accuracy of identification of the backlogs is improved, and automatic classification, priority ranking, reminding triggering and state updating of the backlogs are realized.

The invention discloses a software testing device for an escalator printing plate. The software testing device comprises a relay output module, an IO input board card, an IO output board card, a pulse output board card, a CAN communication board card, an RS232 communication board card, a relay and a testing system. The test system is connected with the relay output module, the IO input board card, the IO output board card, the pulse output board card, the CAN communication board card and the RS232 communication board card. The IO input board card is connected with the relay, and a level output signal of an escalator printing plate is transferred by the relay and then is input into the IO input board card; the IO output board card and the relay output module are connected with the relay, and level output signals of the IO output board card and the relay output module are transferred through the relay and then input into an escalator printing plate level input interface. And the pulse output board card is connected with a pulse input interface of the escalator printing plate. All external interfaces of an escalator hardware system can be replaced, more functions and scenes are covered, and the automation degree of software testing is improved.

This invention provides an approximate Izhikevich neuron model and calculation method based on an adaptive Cordic algorithm, belonging to the field of neuron model technology. The method includes: obtaining the current residual angle to determine residual angle update parameters and membrane electrical displacement; updating the current residual angle based on the residual angle update parameters to obtain the target residual angle, and continuing to iterate on the target residual angle if it does not meet a preset accuracy; obtaining the membrane potential input, and calculating the membrane electrical displacement value based on the membrane potential input and the membrane electrical displacement; accumulating the membrane electrical displacement values ​​after multiple iterations to obtain the target voltage value, and outputting the target voltage value when the target residual angle meets the preset accuracy. The above method dynamically adjusts the iteration step size according to the residual angle size, quickly approximating the target value to reduce the number of iterations, ensuring the accuracy, reliability, and precision of the final output. The process of residual angle update and shift operation is parallelized, improving the efficiency of iteration.

The invention discloses a wind tunnel model path planning method based on multiple motion mechanisms, is used for designing operation paths of an airplane model and an external store model in a grid force measurement test of the multiple motion mechanisms, realizes that the airplane model and the external store model efficiently and safely complete all grid points, and belongs to the technical field of grid force measurement wind tunnel tests. Comprising the following steps: (1) establishing a motion simulation model of a motion mechanism, a test model and a wind tunnel; (2) designing a plurality of paths with different motion sequences according to the distribution of test grid points of each test model; (3) setting the movement mode of the movement mechanism to be simultaneous movement at the maximum running speed, then carrying out analogue simulation on a plurality of paths, selecting a path A and a path B which consume less time in the first two as to-be-selected paths, and assuming that the path A is the path which consumes the shortest time; and (4) carrying out collision detection simulation on the path A and the path B, and determining a final test implementation path.

This invention discloses a software code optimization method and system based on big data. The method includes the following steps: constructing a code optimization knowledge base; acquiring the target code to be optimized and its corresponding estimated runtime data features; parsing the target code, extracting its structural features, and fusing the structural features with the estimated runtime data features to generate a target code fingerprint; performing similarity matching between the target code fingerprint and historical code fingerprints in the code optimization knowledge base to determine target historical code fingerprints that match the target code fingerprint; and determining a target optimization strategy for the target code based on the optimization strategy corresponding to the target historical code fingerprint. This invention can improve development efficiency, reducing the performance tuning work of senior architects, which can take hours or even days, to automated recommendations at the minute level, greatly improving software iteration efficiency.

The present disclosure relates to a vehicle control method, device, computer equipment, storage medium and computer program product. The method comprises: obtaining to-be-predicted vehicle data of a vehicle; inputting the to-be-predicted vehicle data into a target perception model, and outputting a perception prediction result through the target perception model, wherein the target perception model is obtained after an initial perception model is optimized based on labeled vehicle data samples and is verified to pass, the labeled vehicle data samples include vehicle data samples labeled with a perception result label, the vehicle data samples are obtained when an initial perception result of the initial perception model has an error, and the perception result label has a correlation with a problem type corresponding to the error; and controlling the vehicle in a control mode matched with the perception prediction result. The present method can improve the accuracy of vehicle control.

The invention relates to the field of polishing machine tools, and discloses a polishing head and a polishing device.The polishing head comprises a main rod, a second rod, a third rod and a fourth rod, the piezoelectric actuator is connected into the mounting groove, the piezoelectric actuator is provided with a moving end capable of moving up and down, and the moving end is arranged downwards; the transmission part is connected to the moving end and downwards extends out of the mounting groove; the piezoelectric actuator is used for performing micro-motion control on the polishing disc, the movement of the polishing disc can be quickly and accurately controlled, so that the machining efficiency and quality of the polishing head are improved, the polishing disc can be well attached to the surface of a workpiece in the polishing machining process for the surface of the workpiece which is not in rotational symmetry, and the machining precision of the polishing head is improved. And in combination with a smooth track and dwell time controlled by a computer, the surface shape iteration efficiency is improved, so that the machining universality of the polishing head on different workpieces is further improved.

The invention discloses a cooperative monitoring and early warning method and system for landslide in a high-risk area of ground disaster, and is applied to the technical field of intelligent monitoring and early warning. The method comprises the following steps: acquiring landslide surface data and landslide geologic structure data of a target landslide; setting a plurality of monitoring devices; acquiring landslide motion type data and landslide deformation type data; obtaining predicted displacement data; calculating a first correction weight, and correcting the first correction weight to form a second correction weight; updating the second neural network model; and judging whether the target landslide is safe or not according to the predicted displacement data output by the updated second neural network model. According to the method, different landslide motion types and landslide deformation types are analyzed, and the analyzed result is applied to the updating process of the neural network model, so that the model prediction precision is effectively improved, the iteration efficiency is improved, the prediction model can be updated to the latest state in time after various emergency situations, and the prediction efficiency is improved. And the safety and the applicability are effectively improved.

This invention provides a method, system, device, and medium for splitting and automatically distributing in-vehicle communication matrices, belonging to the field of vehicle technology. The method includes: acquiring a target vehicle communication matrix; acquiring a network segment and node configuration table and a gateway node configuration table; acquiring custom configuration splitting rules and determining splitting dimensions based on these rules; performing multi-dimensional splitting of the target vehicle communication matrix based on the splitting dimensions and the network segment and node configuration table and gateway node configuration table, obtaining split sub-matrix files; and converting the split sub-matrix files into target format files and distributing them. This invention improves iteration efficiency by dynamically splitting according to custom configuration splitting rules and configuration tables, automatically converting multiple formats, intelligently distributing based on roles, and supporting incremental splitting based on change history, thereby achieving automated delivery of results from the final version of the target vehicle communication matrix to various ECU suppliers.

The invention provides an MBSE-based full-electric chassis design modeling method and device, and solves the technical problems that the design work energy consumption is high in demand tracing and influence analysis, and the simulation calculation of each specialty is dispersed and cannot be integrated and verified. The method comprises the following steps: a demand quantification process: quantifying an overall demand to form a demand model; an architecture forming process: designing functional compositions and physical compositions according to the demand model, sequentially forming organizational structures of a subsystem, a communication architecture and a power supply and distribution architecture, and forming organizational structure optimization according to the demand model; in the demand parameterization process, demand parameterization of the population and the subsystems is formed according to the demand model and the organization structure, and parameter attributes of the population and the subsystems are optimized through inter-system joint simulation; and in the scheme comparison process, multi-scheme comparison is established according to parameter constraints between the overall system and the subsystems, and a UI is provided for data display of the comparison schemes. The method is suitable for design and development of a vehicle chassis scheme, and has the characteristics of high efficiency, universality and reusability.

The invention discloses a human body use state dynamic evaluation method based on artificial intelligence. The method comprises the steps that state information used for reflecting the overall human body use state is acquired; analyzing the state information based on an artificial intelligence model to judge whether the human body use state is suitable for continuing or not; outputting a judgment conclusion according to an analysis result, wherein the judgment conclusion is used for indicating whether the human body use state is allowed to continue; in the continuous use process of the human body, the steps are periodically repeated to keep the consistency and stability of judgment. The invention further discloses a human body use state dynamic evaluation system based on artificial intelligence. By using the method or the system, the overall state in the continuous use process of the human body can be dynamically tracked and evaluated, the long-term consistency and stability of judgment are ensured, and running state judgment parameters of the human body use state lacked in an existing diagnosis system are complemented; the judgment result has a blocking type control characteristic; and the judgment layer and the execution scheme are structurally decoupled.

The present application relates to the technical field of optical regulation, in particular to a deep medium polarization recovery method and device based on polarization photoacoustic signal feedback, comprising: pulsed laser is incident into a first SLM, and the absolute phase of light is updated by photoacoustic signal to realize the regulation of absolute phase; a half-wave plate, a second SLM and a quarter-wave plate constitute a polarization regulation channel, and the polarization state of light is updated by linear polarization degree to realize the polarization regulation of relative phase; and the additional absolute phase caused in polarization regulation is calculated and reversely applied on the first SLM for compensation, and finally the polarization regulation is realized. The method proposed in the present application overcomes the limitation that the traditional optical imaging is difficult to realize the simultaneous regulation of phase-polarization, and through accurate phase-polarization regulation, different feedback signals are continuously iterated to realize the light energy enhancement and polarization state recovery of deep scattering medium, which can be applied to the polarization detection field of biomedicine.

The present invention relates to a method for obtaining the convective heat transfer coefficient of solid three-dimensional thermal conductivity, comprising the following steps: placing a solid to be tested in a uniformly flowing fluid, obtaining the fluid temperature, the initial surface temperature of the solid to be tested, the thermal conductivity and the thermal diffusion coefficient, measuring a certain temperature reached by the surface of the solid to be tested and the time consumed to reach the temperature, and recording them as the measured temperature and the measured time respectively; calculating the surface temperature of the solid to be tested at a certain moment within the measurement time, constructing a temperature cost function and a Lagrangian function for measuring the convective heat transfer coefficient; estimating the initial convective heat transfer coefficient, respectively obtaining the temperature distribution and the adjoint operator of the solid surface at all moments within the measurement time, and obtaining the gradient of the convective heat transfer coefficient change; obtaining the optimal convective heat transfer coefficient by the Newton iteration method. Compared with the prior art, the present invention can accurately obtain the heat transfer coefficient considering the influence of the three-dimensional thermal conductivity of the tested firmware, and has no restrictions on the shape of the surface of the solid to be tested.

This invention provides a multi-objective optimization device based on dynamic elite learning and dynamic environment selection, comprising: an input module for user input of m objective functions and n decision variable bounding range functions constructed according to a specified multi-objective optimization task; an initialization module for constructing and initializing a population P consisting of N individuals, and setting the iteration count G to 1; a population storage module for storing the population P; a counting storage module for storing the iteration count G and a pre-set maximum number of iterations; and a population iteration module for iteratively updating the population P based on the iteration count G, the maximum number of iterations, the decision variable bounding range functions, and the objective functions. F The output module is used to display the population P to the user. F As a preferred solution for multi-objective optimization tasks, this method can more quickly and accurately obtain the preferred solution for a given multi-objective task.

This application provides a target detection method, apparatus, and electronic device. This embodiment uses weakly supervised learning to determine the labels of target candidate regions containing target objects. Weakly supervised learning does not require manual, precise labeling of the type corresponding to each sample; it only needs to record the user's operation information on the target candidate region. This process does not involve manual labeling of the target candidate region, thereby reducing the labor and time costs of labeling and improving the iterative efficiency of the detection model.

The present disclosure relates to a gameplay parameter management method, device, system, medium and computer program product. In at least one embodiment of the present disclosure, by providing a gameplay parameter editing page, it is convenient for business personnel to directly edit the gameplay parameters on the page, and then generate a gameplay parameter form to achieve business personalized customization of the gameplay parameters; after generating the gameplay parameter form, the business personnel can bind the gameplay parameter form with the activity configuration item; by providing an activity editing page, the various activity configuration items involved in an activity are organized into an activity configuration item list, which is convenient for business personnel to assign parameters to the gameplay parameter form bound to each activity configuration item, and then generate an activity, thereby achieving management of gameplay parameters in units of activities, and business personnel can directly participate in management, avoiding the problem of insufficient number of R&D personnel or long development cycle, and not affecting the efficiency of business iteration.

The present application relates to an image testing method, device, computer equipment and storage medium. The method comprises: obtaining at least one benchmark test atlas; wherein each benchmark test atlas comprises multiple task atlases, and each task atlas comprises multiple base library fingerprint maps and / or multiple unlocking fingerprint maps; obtaining the task size of each task atlas; running the target number of target processes according to the number of task atlases and the maximum number of processes allowed to run; according to the task size of each task atlas, assigning each task atlas as a task to be tested to the corresponding target process according to a preset allocation strategy and performing a running operation to obtain a test record; wherein the running operation comprises an entry operation for multiple base library fingerprint maps and / or an unlocking operation for multiple unlocking fingerprint maps. The use of this method can improve the efficiency of image testing.

The application discloses a virtual model rendering method and device, a storage medium and an electronic device. The method comprises the following steps: obtaining an art resource of a model to be rendered; dividing the art resource to obtain model information, material information and association information, wherein the model information is used to represent attribute information of the model to be rendered, the material information is used to represent material performance of the model to be rendered, and the association information at least comprises a corresponding relationship between the model information and the material information; generating a plurality of material layers based on the material information, wherein each material layer corresponds to different types of materials; and rendering the model to be rendered based on the plurality of material layers, the model information and the association information to obtain a first target model. The application solves the technical problem that, in the prior art, due to the mutual coupling of materials and models, the material iteration efficiency is low when the virtual model is rendered.

The application provides an optimization method and device of a machine learning model and an electronic device, relates to the technical field of cloud computing and intelligent operation and maintenance, and according to the embodiment of the application, when the machine learning model is changed, the configuration file and the training sample data of the machine learning model can be acquired, the machine learning model, the training sample features of the machine learning model and the running environment of the machine learning model are determined, the machine learning model can automatically train in the running environment, and the result related data is provided to a user, so as to provide a reference for optimization of the machine learning model. Artificial modification of various process related source codes is not needed according to the change of the machine learning model, the human and time costs of iteration of the machine learning model in the intelligent operation and maintenance scene are reduced, and the iteration efficiency is improved.

The present invention provides a method and application for simulating a large deformation process of an impact based on a meshless interpolation technique, which belongs to the field of computational mechanics and numerical simulation technology, including model parameter setting, discretization of the problem domain and the time domain, iterative updating of the variables to be determined in the problem domain using the regularized Newton method, solving the local deformation gradient at the next moment, solving the node acceleration, updating the node coordinates, updating the material point coordinates and physical information, completing the simulation of the impact process, and the like. The present invention uses the regularized Newton method for the objective function of the local maximum entropy, and by constructing a simple shape near the calculation point, provides a method for quickly, accurately, and efficiently determining the initial point, thereby ensuring the solution efficiency, and using the Nelder‑Mead algorithm to solve the extreme case iteration point, thereby ensuring the stability of the solution, and is used for the complete simulation of the large deformation process of materials under ultra-high-speed impact, and has the characteristics of high calculation accuracy and strong stability.

The embodiment of the invention provides a programmable logic cluster and a programmable logic device. The programmable logic cluster comprises a plurality of programmable logic units (CLB) arranged in an array form and a signal routing structure, wherein the signal routing structure comprises a transverse signal routing structure and a longitudinal signal routing structure, and the transverse signal routing structure is used for cascade connection between CLBs in the same column or cascade connection between the PLC and other PLCs in the transverse direction; and the longitudinal signal routing structure is used for cascade connection between the CLBs in different columns or cascade connection between the PLC and other PLCs in the longitudinal direction. Through the embodiment of the invention, the problem that the PLC wiring algorithm of the current programmable logic chip is complicated in the related technology is solved, and the effect of improving the logic function iteration efficiency is achieved.

The invention provides an automatic regression test method and system for process step files of battery test software. The method comprises the following steps: establishing a golden sample library containing a process step file set passing historical verification and corresponding standard output time sequence data; starting to-be-tested software and loading a virtual battery simulation engine, wherein the virtual battery simulation engine is configured to simulate battery response characteristics and respond to a hardware control instruction of the to-be-tested software; obtaining a to-be-tested step file associated with the to-be-tested software, loading the to-be-tested step file in batches, driving the virtual battery simulation engine to execute a test, and collecting actual measurement output time sequence data; and comparing the actual measurement output time sequence data corresponding to the part, contained in the gold sample library, in the to-be-tested work step file with standard output time sequence data to judge a regression test result and output a report. According to the method, the automatic regression test of the battery test software working step file which runs at an ultra-real-time speed and automatically performs high-precision data comparison is realized.

This invention relates to the field of intelligent question answering technology, and in particular to a context-based iterative question answering method, electronic device, and storage medium. The method includes: determining the current question to be processed from a constructed question queue; generating prompt information and inputting it into a preset large language model based on the current shared task context and the current question to be processed to obtain the answer to be output and evaluating the quality of the answer to be output; if the quality evaluation result is satisfactory, selecting to directly output the final answer or continue execution depending on whether it is the last question in the question queue; if the quality evaluation result is unsatisfactory, determining the operation to be executed and updating the execution result information to the shared task context to iteratively generate prompt information until the final answer is output. This invention can achieve targeted correction and iterative processing of question defects, improving the accuracy, reliability, and comprehensiveness of the final output answer.

The utility model discloses a digital twin data management device, and belongs to the technical field of industrial Internet of Things data management. A multi-protocol conversion chip is compatible with an RS485 interface, an Ethernet interface and an optical fiber interface, a PCIe interface is adopted to realize high-speed data interaction among a processing module, a security module and a storage module, a dual-redundancy circuit is utilized to guarantee verification signal transmission reliability, and hierarchical data management is realized in combination with a hierarchical storage architecture (an SSD cache layer, an RAID6 disk array layer and a distributed node layer). And the data integrity is ensured through a model checking unit (double-precision floating point residual calculation) and a self-checking unit (CRC checking and heartbeat packet detection) of the synchronous checking module. The method has the advantages that the protocol conversion delay is smaller than or equal to 2 microseconds, the encryption throughput of the SM4 algorithm reaches 5 Gbps, RAID6 double-disk fault tolerance and parameter updating batch processing efficiency are improved by two times, the system availability is larger than or equal to 99.999%, and the high-real-time, high-safety and high-reliability digital twin data management requirements in industrial scenes are met.

The invention discloses an application transformation method, device and equipment based on a micro front end, and a storage medium. The method comprises the following steps: modifying a sub-application configuration file in a micro front end sub-application framework according to a sub-application name and a sub-application port in a script starting command; according to a sub-application source code path in the script starting command, obtaining a to-be-migrated service catalogue under an original front-end framework of the webpage application program, and copying the to-be-migrated service catalogue into a micro-front-end sub-application catalogue of a micro-front-end sub-application framework layer by layer; obtaining a routing file in the micro front end sub-application framework, and modifying each routing path in the routing file according to the sub-application name; according to the project dependent file under the original front-end framework, modifying a dependent file to be verified in the micro front-end application framework; and integrating the transformed micro-front-end sub-applications according to the modified sub-application configuration file to obtain a webpage application program after micro-front-end transformation, so that front-end code decoupling can be realized, and the response speed of the application program is improved.

The present invention relates to the field of data processing, and discloses a method for analyzing the automatic performance anomaly of a distributed storage system, comprising: presetting performance collection points in each process in the distributed system, and inputting the parent-child relationship of the performance collection points; collecting the performance indicators of each performance collection point, and obtaining the rate of change of the performance indicators of the performance collection point with the time interval according to the performance indicators of the performance collection point; when the performance collection point generates anomalies or jitters within the set time period, the performance collection point is the target performance collection point, and the correlation between the target performance collection point and each descendant performance collection point is obtained; the correlation between the target performance collection point and each descendant performance collection point is sorted, and the descendant performance collection point and leaf node with the greatest correlation are obtained; according to the descendant performance collection point and leaf node with the greatest correlation, the performance anomaly propagation path is obtained. The present invention can realize the rapid and automatic positioning of the submodule causing the performance jitter.