11results about How to "Real-time response" patented technology

The invention discloses a laser diameter measuring instrument lens self-cleaning device and method based on multi-mode cooperation, and belongs to the technical field of cigarette manufacturing. The device comprises a pollution sensing unit for acquiring the pollution area P, the oiliness index Q and the particle density G of the surface of the lens; the annular ion air curtain unit is used for generating ion airflow flowing by being attached to the surface of the lens; the piezoelectric ultrasonic unit transmits high-frequency vibration to the surface of the lens; the ultraviolet laser unit is used for emitting pulse type ultraviolet laser; and the controller is electrically connected with each unit. The method comprises the following steps: acquiring pollution characteristic parameters in real time; determining a pollution level; generating a time sequence control instruction; and all the cleaning units are cooperatively controlled to work. According to the invention, self-adaptive, non-contact, non-stop and efficient cleaning of composite pollutants such as tobacco dust, tar and sugar is realized, the real-time cleaning requirement of a high-speed production line is met, and the measurement precision and the operation stability are guaranteed.

The invention discloses a battery maintenance method based on a dynamic sparse neural network and a large language model, and belongs to the technical field of battery management. The method comprises the following steps: acquiring sensor time sequence data and an operation and maintenance text maintenance log of a battery pack, and respectively carrying out standardized preprocessing on the sensor time sequence data and the operation and maintenance text maintenance log; extracting time domain and frequency domain features of the preprocessed sensor time sequence data to obtain sensor features; carrying out semantic coding on the preprocessed text maintenance log by adopting a large language model subjected to battery domain knowledge fine adjustment to obtain text semantic features; performing feature fusion on the sensor features and the text semantic features based on a cross attention mechanism to generate fusion features; and inputting the fusion features into a dynamic sparse neural network which adaptively adjusts the sparseness of each layer of network according to the real-time health state of the battery, and outputting a battery fault probability and a fault root cause weight.

The invention provides a power grid dispatching method, a control device and a storage medium, and belongs to the technical field of power grids. The method comprises the following steps: inputting acquired operation data into a multi-objective optimization model to output an optimization scheduling variable; performing constraint condition verification on the optimal scheduling variable, when the optimal scheduling variable meets a preset constraint condition, generating a scheduling scheme of the target power grid, and issuing a scheduling scheme instruction; in the process that the target power grid executes the scheduling scheme instruction, monitoring an operation index of the target power grid; and calculating the deviation between the monitored operation index and the optimization target, and when the deviation exceeds a preset threshold value, adjusting the scheduling scheme of the target power grid. Through a technical path of offline training and online reasoning of the neural network, the generation time of the scheduling scheme is greatly shortened, and real-time response to dynamic change of the power grid is realized.

This invention relates to the field of expense reimbursement management technology, and in particular to a centralized management system and method for expense reimbursement documents based on financial sharing. The management system includes a financial sharing module, an interaction management module, and an expense reimbursement document interaction module. The financial sharing module pushes expense reimbursement document data to the interaction management module in real time. The interaction management module controls and monitors the entire interaction process of expense reimbursement documents. The expense reimbursement document interaction module implements the entire interaction process of expense reimbursement documents. A first machine learning model is established in the financial sharing module as its generator network; a second machine learning model is established in the interaction management module as its generator network; a discriminant network for the interaction management module is established in the financial sharing module; and a discriminant network for the financial sharing module is established in the interaction management module. This management system and method effectively solve the problems of risk and time-consuming / labor-intensive processes in expense reimbursement document interaction.

The invention relates to the technical field of power equipment monitoring and early warning, and particularly discloses a substation three-dimensional operation crane external damage prevention early warning system which comprises an environment sensing module, a dynamic early warning and collision prevention module, a data fusion and decision module, a man-machine interaction operation module and a system self-diagnosis and maintenance module. The environment sensing module is responsible for sensing the surrounding environment of the crane and collecting and analyzing surrounding obstacle information. According to the invention, a mode of fusing various sensor data is adopted, the system can accurately sense the surrounding environment and make a response in a three-dimensional space, especially in a complex and narrow environment such as a transformer substation, the sensor fusion can significantly improve the comprehensiveness and precision of environment sensing, the system can carry out danger prediction and intelligent decision making, and the system has a wide application prospect. And the system not only can make a response in real time, but also can carry out self-learning and optimization according to historical data and environmental changes, so that an optimal early warning strategy is provided under different operation conditions.

This invention relates to the field of electric motor control technology and discloses an electric motor control method for compressor drive, comprising: acquiring stator current signal and DC bus voltage signal; extracting voltage ripple characteristics based on DC bus voltage signal and determining initial value of quadrature axis current compensation; calculating the amplitude of stator current signal; when the amplitude crosses the linear magnetization current threshold, retrieving the inductor magnetization characteristic table to determine the inductor drop ratio; correcting the initial value of quadrature axis current compensation using the inductor drop ratio; and injecting the corrected compensation command into the current regulating loop. This invention, by integrating voltage ripple sensing and magnetic saturation correction mechanisms, synergistically cancels the pulse load interference of the compressor's single-cycle, fills the torque output gap under magnetic circuit saturation, suppresses stator current oscillation under low-frequency operation, and maintains torque response stiffness under all operating conditions.

The application provides an agent collaborative decision optimization method and system based on a multi-source heterogeneous knowledge graph, and relates to the technical field of knowledge graph intelligent decision-making. The application extracts a time-effect feature set from entities of a multi-source heterogeneous knowledge graph, generates a knowledge new-old evaluation matrix containing time decay characteristics and credibility weights, analyzes policy document version numbers, academic literature citation relationships and news event transmission chains to obtain a dynamic time-effect vector, labels the time-effect level of the knowledge graph nodes according to the dynamic time-effect vector, iterates through nodes with different time-effect levels to identify conflict node pairs, generates a directional retrieval rule set according to the contradiction type and the dynamic time-effect vector characteristics, then schedules three types of intelligent agents for parallel retrieval, exchanges intermediate results to optimize the strategy, aggregates the retrieval results to form a cross-platform evidence chain, and further generates a knowledge node update recommendation report. Based on the multi-source heterogeneous knowledge graph, the multi-agent collaborative retrieval and analysis can optimize the knowledge node decision and generate an update recommendation.

The application discloses an energy storage control method and device and an energy storage system. The energy storage system comprises a first energy storage unit and a second energy storage unit, and the first energy storage unit and the second energy storage unit are connected to a load. The second energy storage unit is a lithium ion battery. The energy storage control method comprises the following steps: detecting a load current in real time; determining an energy storage unit for supplying power to the load and controlling an output current of the energy storage unit according to the load current, an upper limit value of an output current of the second energy storage unit and a load current change rate, so that the second energy storage unit limits power output and the first energy storage unit responds to load power change in real time. The first energy storage unit and the second energy storage unit are combined, the second energy storage unit constantly outputs power according to requirements, the first energy storage unit changes the output power to respond to the load power change in real time, the real-time response to the load power change is realized, the economy of energy storage is realized, the two energy storage units are used to supply power, and the overall operation life of the energy storage system is prolonged.

This application discloses a multi-parameter fusion-based dynamic optimization method for SDC industrial control parameters, belonging to the field of parameter optimization technology. The method includes: predicting the predicted coal seam dip angle sequence and the predicted coal mining speed sequence within a future preset time zone; predicting the roof pressure of the working face and obtaining the predicted roof pressure field sequence; analyzing and determining the suitable safe time zone and the suitable efficiency time zone; optimizing the hydraulic support parameters of the hydraulic support group, outputting a queue of adapted support parameter commands, and executing hydraulic support within the preset time zone. This solves the technical problems of existing hydraulic support parameters being unable to respond in real time to complex and changing underground environments, leading to insufficient roof pressure prediction accuracy, untimely support, insufficient initial support force, or excessive pressure relief.

The application relates to the field of silicon-based optoelectronic integration, and particularly discloses an on-chip adaptive combining method and a feedback system for polarization diversity MZM, which comprises the following steps: optical coupling and mode conversion, polarization diversity electro-optical modulation, optical power collection and photoelectric conversion, signal processing and feedback control, and a tunable combining module; after the coupling conversion and modulation, two-way optical power is collected and photoelectrically converted, a power difference voltage is obtained through an analog circuit, a control voltage is generated by superimposing a bias voltage, and a tunable combiner is driven to dynamically adjust a beam splitting ratio to complete combining. The application solves the problems that the output power of the polarization diversity MZM randomly fluctuates with the polarization state, the existing polarization control scheme has slow response, a complex structure and is difficult to integrate with a silicon-based process, realizes real-time and accurate compensation of power fluctuation, guarantees stable combining output power, and has a simple system structure, low power consumption, is easy to monolithically integrate with a silicon-based chip, and improves the performance and compatibility of an integrated photonic link.

The invention provides a pre-training language model retrieval method and system based on knowledge graph enhancement, and the method comprises the steps: carrying out subgraph division processing on a domain knowledge graph, and generating a subgraph index matched with a user retrieval type; updating the domain knowledge graph and the sub-graph index by adopting an incremental training method of a graph neural network; user query is decomposed into an entity set and a relation path through entity recognition, entity disambiguation, entity linking and relation extraction operation; positioning a corresponding sub-graph index based on the entity set and the relation path, and extracting a related sub-graph from the domain knowledge graph; and respectively coding the head entity, the relationship and the tail entity in the related sub-graph triad into basic vectors, and performing semantic fusion processing on the basic vectors to generate knowledge prompt vectors. The method has the advantages of accurately capturing domain structured association logic and responding to domain dynamic iteration knowledge in real time.