11results about How to "Accurate call" patented technology

The invention provides an application access control method based on system call limitation, relates to the technical field of network security, and provides an application behavior learning module and an application authority control module. The application behavior learning module executes the following steps: A1, creating an application behavior list; a2, dynamic training monitoring: training and monitoring each item of content in the application behavior list; a3, generating a mapping table, and establishing a mapping relation between each item of content in the application behavior list and system calling; the application authority control module executes the following steps: B1, user authorization application: initiating an application access request to an authorization center; b2, authorization approval: after receiving the application access request, the authorization center authorizes the application as required; and B3, issuing the authority. According to the method, the granularity of authority management is refined, the exposure surface of the application is reduced, enterprises are guaranteed, the malicious behavior risk is reduced, the system security and data privacy are protected, and the overall security is improved.

This invention relates to the field of exercise prescription generation and recommendation technology, and discloses an LLM-based generative recommendation method for exercise prescriptions. The method involves acquiring and preprocessing data, constructing a knowledge graph based on the preprocessed data (S1), pre-training the LLM model, inputting the constructed knowledge graph into the LLM model, and optimizing it through reinforcement learning. The PPO algorithm is used to optimize the LLM model's output strategy. Based on the optimized LLM model, the exercise prescription is derived and generated in natural language using FITT parameters. This method combines user characteristics and scenario information to generate high-quality, personalized exercise prescriptions. The invention utilizes a multi-dimensional reward function framework to quantify prescription quality and optimize the generation strategy, ensuring that the generated exercise prescriptions are not only scientifically sound but also conform to the user's actual situation and preferences. Synthetic data is generated through a user simulator, achieving an upgrade from static recommendation to dynamic adaptation, improving user experience and the level of intelligence in exercise and health management.

The invention provides an FPGA (Field Programmable Gate Array) parallel password operation method, device and equipment based on simple scheduling, a medium and a program product. The method comprises the following steps: receiving and caching a to-be-processed packet; based on protocol unpacking and table lookup, extracting a payload to be processed from the packet to be processed and completing algorithm scheduling and kernel scheduling; according to algorithm scheduling and kernel scheduling results, corresponding algorithms and algorithm kernels in the FPGA are adopted to complete cryptographic operation on the payload to be processed; according to algorithm scheduling and kernel scheduling results, protocol packaging is carried out on the cryptographic operation result; and sending and caching the protocol packaging result. According to the method, the password performance can be quickly improved at low cost under an application layer, the password types are expanded, and meanwhile, the parallel scheduling complexity is reduced.

The application provides a large model fine-tuning method and device for a railway design intelligent agent tool call, the method comprising: constructing a supervised fine-tuning data set in the field of railway design; constructing a direct preference optimization data set based on the success and failure feedback of the large model intelligent agent call; using the constructed supervised fine-tuning data set and direct preference optimization data set to pre-train a large language model; using the supervised fine-tuning data set to perform single supervised fine-tuning on the pre-trained large language model to obtain a large language model after supervised fine-tuning; using the direct preference optimization data set to perform multi-round preference optimization fine-tuning with feedback on the large language model after supervised fine-tuning to obtain a large language model after optimization fine-tuning. By constructing high-quality data and fine-tuning the large language model, the application significantly reduces the risk of the large language model fabricating tools or misadjusting interfaces in complex design tasks, so that the intelligent agent has stable and controllable tool collaboration capability in the real railway design software environment.

The invention relates to the technical field of environmental assessment text processing, and discloses an intelligent auditing system for an environmental assessment report based on dynamic classification and knowledge mapping, which enables a classification system to automatically identify a novel pollution type and dynamically adjust a type boundary through multi-source data fusion and an incremental learning algorithm. The method ensures that the classification result always meets the newest environmental policy requirement, and ensures that the knowledge base can synchronize the newest standard specification content in real time by establishing a classification system and a trigger updating mechanism of the standard specification, thereby achieving the effect of comprehensively improving the professionalism and timeliness of the environmental assessment system. The system can automatically adapt to environmental policy changes and technical standard updating, can maintain scientificity of a classification system and integrity of a knowledge base without manual intervention, ensures that newest and most relevant standard specifications can be called for professional judgment when various environmental assessment reports are audited, and remarkably reduces audit deviation caused by knowledge lag.

The invention discloses an electric drive system rotary transformer zero position self-learning and curing system, which comprises a mechanical alignment module for driving a motor rotor to rotate to a preset mechanical zero position and generating a zero position offset value; after the motor reaches the rated rotating speed, the control module sends a rotary transformer zero-position self-learning enable signal, so that a motor controller executes current closed-loop control based on a zero-position offset value, a d-axis current instruction in a rotor synchronous rotating coordinate system is set to be a negative value and a q-axis current instruction is set to be zero, and an actual torque value is obtained; the zero calibration module obtains an average torque value through the actual torque value, calculates the deviation between the average torque value and a zero reference torque value, and when the deviation is larger than a positive deviation threshold value or smaller than a negative deviation threshold value, adjusts a zero offset value and triggers the control module to recollect the actual torque value; and obtaining a final zero offset value until the deviation is smaller than or equal to the positive deviation threshold value and larger than or equal to the negative deviation threshold value. The torque control precision and the operation reliability of the electric drive system are improved.

The invention relates to the technical field of computers, and provides a cloud desktop graphic processing method, an acceleration service component and computer device.The method is applied to the acceleration service component installed on a cloud desktop and comprises the steps that a first graphic acceleration processing request initiated by an application program in the cloud desktop is intercepted; determining a rewriting mode of the first graph acceleration processing request based on a driving interface mapping rule, and rewriting the first graph acceleration processing request based on the rewriting mode to obtain a second graph acceleration processing request; sending the second graphic acceleration processing request to a virtual graphics card driving layer through the acceleration interface; and receiving the first processing result returned by the virtual graphics card driving layer, and returning the first processing result to the application program to improve the compatibility of the virtual graphics card of the cloud desktop.

The invention discloses a carrier rocket controller parameter intelligent design system and method, and belongs to the technical field of carrier rocket attitude control. Aiming at the defects of long manual design time, low efficiency, dependence on empirical values, high error rate and the like of a traditional attitude control system, an expert system is adopted for carrying out carrier rocket controller parameter intelligent design, and the purpose of intelligently designing control parameters can be quickly and effectively achieved. By using the method, the efficiency of manually adjusting attitude control design parameters can be greatly improved, and the attitude control design efficiency under the dense emission background is improved by more than 80%. According to the method, successful model design parameters can be rapidly used, or combined parameters can be adapted to the successful model design parameters, and even new design parameters adapted to design requirements can be rapidly and intelligently generated according to an autonomous parameter selection method, and the parameters have an effect of good stability, so that the method has a relatively high practical value for the design of an attitude control system of a carrier rocket in active service.

This invention relates to a closed-loop coin detection module based on a novel eddy current magnetic core. The main control unit retrieves temperature drift curves for n temperatures from a storage unit based on temperatures obtained from a temperature sensor. It obtains sampled values ​​based on the sampling points corresponding to the temperatures in the drift curves, and outputs the coin detection result by comparing the sampled values ​​with a set range, thus achieving closed-loop coin detection. The storage unit stores temperature drift curves obtained from temperature data sampling within a set temperature range for different coin types. The voltage stabilizing unit stabilizes the voltage of the central control unit's operating units. This invention eliminates deviations caused by coin identification speed, calculation time, and data communication between modules, enabling precise curve retrieval within a limited time, thereby achieving closed-loop control for coin detection.

The invention discloses a wind-light-water complementary system real-time scheduling method based on multi-task deep reinforcement learning. A t-SNE algorithm is adopted to carry out dimension reduction on operation data of a wind-light-water complementary system; clustering the low-dimensional feature data by adopting a DBSCAN algorithm to obtain operation scene data sets of different scene categories; constructing a scene identification model based on a Stacking ensemble learning algorithm; training a scene identification model by adopting the obtained scene data set; taking the minimum wind power and photovoltaic output reduction amount, the minimum power shortage amount and the maximum energy storage amount of the reservoir at the end of the scheduling period as comprehensive optimization targets, and constructing an optimal scheduling model of the wind-light-water complementary system; a wind-light-water complementary system real-time scheduling model based on a multi-task SAC algorithm is constructed and trained; and performing real-time scheduling on the wind-light-water complementary system based on the trained scene identification model and the real-time scheduling model of the wind-light-water complementary system. The method can quickly make decisions, is high in adaptability to wind-light uncertainty, and improves the reservoir energy storage of the wind-light-water complementary system.

The invention relates to a method for constructing a patent large model question-answering system based on map enhancement generation, which adopts an incremental pre-training strategy, integrates a large number of patent abstract texts into an LLM learning process, and can more accurately call and generate related contents in subsequent question-answering interaction. In order to further transform the text continuous writing type LLM into a form suitable for executing an accurate question and answer task and improve the understanding and following degree of the LLM on a specific instruction, an instruction fine tuning technology is applied, targeted optimization is carried out on the model through a massive instruction data set, and the adjusted LLM can accurately meet the knowledge answering requirement needed by a user.