41 results about "Multi field" patented technology

The invention relates to an abnormity alarm method and system for a feed production control system, and the method comprises the following steps: collecting multi-field data related in a feed production process in real time, the multi-field data at least comprising production equipment operation parameters, material characteristic parameters, environmental parameters and process control parameters; preprocessing the collected multi-field data, and fusing to generate a multi-field description feature vector representing the current production state; inputting the multi-field description feature vector into a pre-trained fault prediction model, wherein the fault prediction model respectively calculates the confidence of the fault type prediction result and the confidence of the fault point prediction result; wherein the confidence coefficient is in positive correlation with the distance from the feature vector to a known fault mode clustering center or the distance from the feature vector to a decision boundary; and when the fault prediction result indicates that an abnormal risk exists and the corresponding confidence exceeds a preset threshold, triggering a corresponding alarm signal.

The invention discloses a transformer substation global digital twinborn fusion system, which comprises a data fusion module, a model fusion module and an artificial intelligence fusion module, and is characterized in that the data fusion module is used for collecting multi-source data of a transformer substation, preprocessing the data and performing data fusion on the preprocessed data; the model fusion module is used for constructing a bidirectional dynamic mapping model of a transformer substation physical entity and a digital virtual body, constructing a multi-field model and constructing a model increment updating mechanism, and the artificial intelligence fusion module is used for correcting the deviation of an AI model, performing simulation verification on the output of the AI model and generating a visual alarm. According to the method, the technical problems of incomplete data fusion, poor model fusion collaboration and insufficient AI and digital twinning fusion depth in the existing transformer substation digital twinning system are solved, real-time bidirectional accurate mapping of the transformer substation physical entity and the digital virtual body is realized, and the operation and maintenance efficiency and the decision-making precision are improved.

In recent years, reinforcement learning has a significant effect in multiple fields, but a traditional method has limited adaptability in a dynamic environment or a multi-task situation. The invention provides a multi-task reinforcement learning method named as a priority weighted soft modularization (SM-PHT) method, which is used for enhancing the adaptability and the generalization ability of an intelligent agent, and relates to a multi-task reinforcement learning method named as SM-PHT (SM-PHT), which is used for enhancing the adaptability and the generalization ability of the intelligent agent. SM-PHT integrates three key mechanisms: priority weighted knowledge distillation, and student network robustness is improved through weighted fusion of multi-model knowledge; the hierarchical cache is used for respectively storing low-level experience and high-level parameter optimization efficiency; and task embedding: capturing environment characteristics to enrich task representation so as to promote knowledge transfer. In the Meta-World MT10 challenge, the success rate of SM-PHT is twice that of the existing optimal method, and the average reward is improved by 30%; in the MT50 challenge, the success rate and the average reward are both improved by about 10%. Results prove that the stability and generalization of the mechanical arm in complex tasks highlight the application potential in the field of actual multi-task mechanical arm operation.

The invention provides a multi-domain task processing method, system and device based on knowledge fusion and Agent cooperation and a medium, and belongs to the technical field of cross-domain knowledge fusion. Tasks input by a user are decomposed, and a subtask set is generated; building a specialized Agent set according to a subtask calling skill mapping table and a role template library; executing cross-domain knowledge retrieval and fusion on the Agent set, monitoring output consistency and classifying collaborative conflicts; triggering a corresponding extension mechanism according to different conflicts, and updating the knowledge graph or generating a new Agent; meanwhile, field knowledge updating is monitored, and knowledge weights are adjusted and synchronized to related modules. Precise disassembly and specialized processing of complex tasks are realized, and the cross-domain knowledge fusion efficiency is improved; cooperation conflicts are solved, and system adaptability is enhanced; the knowledge is updated in real time, so that the processing timeliness is ensured, and the task processing quality and efficiency of the multi-Agent system are improved.

The invention discloses a multi-field die cutting part intelligent manufacturing management system and method based on the industrial internet, and relates to the technical field of data analysis, and the method specifically comprises the steps: building a product feature set based on the product demands of die-casting die cutting parts, and calculating the product passing rate of each die-casting die; extracting die-casting dies greater than a preset first qualified threshold value, obtaining pouring features, recording the pouring features as a die feature set, obtaining die features influencing die-casting die cutting piece production based on a correlation algorithm, and recording the die features as key influence die features; and for any die feature in the die feature set, based on the number of times of taking the die feature as the key influence die feature, obtaining a product feature of taking the die feature as the key influence die feature, and counting a constraint set of the die feature, thereby obtaining an optimal feature value range, and selecting an optimal die-casting die according to the optimal feature value range. The key influence mold features are identified based on the correlation algorithm, and the limitation that the quality of the mold is judged only depending on experience or a single parameter traditionally is broken through.

The application discloses a multi-field fake news detection method based on hierarchical potential field modeling, and belongs to the technical field of fake news detection. The method comprises the following steps: obtaining multi-angle multi-modal feature embedding of to-be-detected news by using a pre-training feature extractor; enhancing original field labels based on multi-modal features by using a field label enhanced memory network; constructing a single-modal and multi-modal processing flow by using a layer-stacking Transformer block, thereby modeling the field dependency relationship in news features in a hierarchical manner while adaptively fusing multi-modal features; introducing a variational mutual information decoupling module to supervise each layer of expert mixed network; obtaining a fake judgment feature by using a field perception Transformer, and outputting a fake judgment result through a classification head. By means of hierarchical field modeling and multi-modal feature fusion, the application effectively solves the distribution difference challenge existing in the multi-field adaptation of social media news.

The invention relates to the technical field of airborne system design, and discloses an airborne system digital forward design method, device, equipment and medium, a design process from demand analysis, architecture design, architecture tradeoff to simulation verification is established, integration and optimization of the airborne system design process are effectively realized, machine learning and multidisciplinary design optimization are combined, and the design efficiency is improved. According to the method, knowledge in multiple fields such as machinery, electronics and software is fused, design rules are mined from historical design cases through deep learning, a multi-target tradeoff analysis model is established, automatic generation and optimization of an architecture scheme are achieved, dependence on artificial experience is remarkably reduced, airborne system design efficiency is improved, design change cost is reduced, and the method is suitable for large-scale popularization and application. And the reliability of the airborne system is improved.

The application provides a dynamic modeling method and system for multi-field skill AI automatic semantic labeling, and relates to information retrieval, medical health and cross-field skill fusion scene. Through nonlinear dynamic modeling and biomedical stability control, the semantic label automatic generation and real-time matching of talent skill and health intervention effectiveness are realized. The application innovatively integrates the sound wave frequency of fetal brain promoting music method, maternal action instruction and blood type nutrition scheme into the "skill-time" frequency domain space, combines the Gevrey smoothing operator and fractional derivative evolution equation, accurately separates the temporary fetal movement fluctuation and low-frequency core characteristics, and breaks through the characterization limitation of traditional models on nonlinear biological coupling effect. By introducing the "intervention effect-time" frequency domain component and local dependency constraint, the short-term effectiveness and long-term value of music intervention on fetal neurodevelopment are quantified, and the model distortion problem caused by extreme data or cross-field parameter resonance in traditional methods is solved.

The utility model discloses a multi-field wear-resistant scraper conveyor which mainly solves the technical problem that transportation is affected due to the fact that the bottom of a groove body of an existing scraper conveyor is prone to being abraded in the transportation process. The middle part is provided with a middle groove, and the machine head and the machine tail are correspondingly connected to the two ends of the middle part. The middle groove comprises a groove body, limiting plates, a plurality of supporting rollers and chains, wherein the limiting plates are symmetrically and internally buckled on the tops of the two sides of the groove body and correspondingly form limiting cavities with the groove walls of the two sides of the groove body, the supporting rollers are perpendicular to the axial direction of the groove body and arrayed in the groove body at equal intervals, and the chains circularly run on the upper sides and the lower sides of the supporting rollers along the axis of the groove body. The chain plates are fixed to the chain at equal intervals and are perpendicular to the axis of the groove body, the belts are connected between the adjacent chain plates, and the bottom faces of the belts are in sliding contact with the supporting rollers when materials are conveyed. Limiting parts correspondingly extending into the limiting cavities are arranged on the two sides of the chain plate. The scraper conveyor can avoid direct contact with the bottom of the material tank body, so that the relative integrity of the tank body is ensured.

1. The name of the design product: intelligent AI edge gateway. 2. The use of the design product: provide data encryption security for inter-device communication, smart home, Internet of Vehicles, industrial Internet of Things and other multi-field applications. 3. The design points of the design product: in shape. 4. The picture or photo that best indicates the design points: perspective view 1.

The application discloses a kind of adaptive control parameter optimization methods for multi-field industrial control system, belong to industrial control and intelligent control technical field;By collecting the multi-source heterogeneous time series data in multi-field industrial control system, a unified system state representation is constructed;The non-invasive virtual state evolution of multiple candidate control parameters is deduced, and the corresponding system response trajectory is generated;Introduce the preferred parameter verification mechanism based on physical residual and parameter risk potential, and screen out the parameter combination that violates physical constraints or has high risk;Based on spectral entropy and spectral geometric center, the frequency domain topology gating analysis is carried out, the stability and safety of the control parameters are constrained in frequency domain, and the application of the control parameters is triggered only when the frequency domain stability condition is met. Realize the safe, robust adaptive optimization of multi-field industrial control system control parameters without interfering with the existing production process, improve the reliability and engineering practicability of complex industrial control system parameter setting.

The invention discloses a multi-field time series data prediction method based on language guidance and near-end strategy optimization fine tuning. A used deep learning model comprises a time series encoder based on a Transform structure, a pre-trained large language model and five linear layers arranged in parallel; the method mainly comprises the following steps: carrying out preprocessing and normalization operation on multi-field time series data, dividing the time series data into single-channel input, and carrying out sliding window division processing; designing a structured template, inputting the structured template into a large language model in combination with language background information and a time sequence, and performing a pre-training stage, wherein two tasks of pre-training comprise reconstruction of an input time sequence and prediction of a future time sequence; and a near-end strategy optimization algorithm is adopted to carry out fine adjustment in a specified field, and finally a prediction model is constructed. According to the method, the large language model and the reinforcement learning technology are creatively fused, and the prediction accuracy of the model in multi-field time series data is remarkably improved.

The invention discloses a coordination control phase difference transition decision-making and distribution method, which comprises the following steps of: correlation adjustment amount calculation: taking any specified time point as a unified clock reference point, and calculating an extended transition adjustment amount required by a period switching moment and a shortened transition adjustment amount corresponding to the extended transition adjustment amount; according to the transition decision model establishment, through association adjustment amount and period phase boundary constraint, through accurate phase difference transition control, differentiated green time ratio distribution, real-time dynamic feedback and multi-field applicability, the defects in dynamic adaptability, multi-objective optimization and engineering economy in the prior art are systematically overcome, and the method has the advantages of high efficiency, high reliability and the like. And a core technical support is provided for intelligent upgrading of urban traffic.

The application provides a multi-field attention mechanism-based electric energy prediction method and system, wherein the method comprises the following steps: acquiring electric energy data and corresponding general features and field features; using a continuous feature encoder and a discrete feature encoder to encode continuous features and discrete features respectively; performing feature fusion on various encoded feature representations, taking the field feature representation as a query in the attention mechanism, and taking the general features as key values and key values in the attention mechanism; calculating the similarity between the query and the key value as the attention weight of the corresponding key value; performing weighted summation on the general features according to the attention weight, splicing the field features, and obtaining mixed features; and using a feedforward neural network to map the mixed features to the electric energy data to obtain an electric energy prediction value. The application can predict the electric energy of various cities and power consumption industries based on a single model, and can fuse electric energy data in different fields to realize accurate electric energy prediction.

The invention discloses a self-adaptive control parameter optimization method for a multi-field industrial control system, and belongs to the technical field of industrial control and intelligent control. Uniform system state representation is constructed by collecting multi-source heterogeneous time series data in a multi-field industrial control system; performing non-intrusive virtual state evolution deduction on the multiple groups of candidate control parameters to generate a corresponding system response track; an optimal parameter verification mechanism based on physical residual errors and parameter risk potential energy is introduced, and parameter combinations which violate physical constraints or have high risks are screened out; based on frequency domain topology gating analysis of spectrum entropy and spectrum geometric gravity center, frequency domain constraint is carried out on stability and safety of control parameters, and application of the control parameters is triggered only when the frequency domain stability condition is met. On the premise that an existing production process is not interfered, safe and steady self-adaptive optimization is carried out on the control parameters of the multi-field industrial control system, and the parameter setting reliability and engineering practicability of the complex industrial control system are improved.

The application discloses a cloud-computing-based distributed simulation platform elastic scheduling method and system, and relates to the technical field of computer systems.The cloud-computing-based distributed simulation platform elastic scheduling method and system comprise a cross-field dynamic dependency processing module, a real-time task ultrafast scheduling module, a hybrid cloud resource adaptation module and an elastic scheduling core engine; the cross-field dynamic dependency processing module is used for realizing dynamic processing of dependency relationships in multi-field collaborative simulation, cross-domain collaborative allocation of resources and breakpoint self-recovery of task chains.The cloud-computing-based distributed simulation platform elastic scheduling method and system realize dynamic analysis of multi-field dependency relationships, cross-domain collaborative allocation of resources and breakpoint self-recovery of task chains through the cross-field dynamic dependency processing module, solve the problem of rigid cross-field dependency processing mechanism in the prior art, reduce cross-field data synchronization delay, and improve resource utilization and stability of task chains.

The present application relates to the technical field of large language model, especially refers to a complex multi-field guided teaching question and answer generation method, device, equipment and computer storage medium.The complex multi-field guided teaching question and answer generation method disclosed by the present application solves the problem of single field by using multi-field data source to process corresponding multi-turn question and answer pair data set, fine-tuning the model based on the multi-turn question and answer pair data set, enhances the applicability and practicality of the teaching question and answer model;For the problem of lacking guiding ability of the model, the present application further processes the input of the teaching question and answer model by problem extraction and disassembly based on thought chain and processing based on prompt engineering and context learning mechanism, and improves the problem guiding ability of the model;For the problem of single evaluation index of teaching effect, the present application proposes a new evaluation system, and multiple evaluation indexes can more comprehensively and accurately evaluate and optimize the teaching effect of the model.

The invention discloses a multi-field and multi-scene-oriented arbitrary multi-frequency pseudo-random signal electrical stimulation device, which is characterized in that the device provides a signal generation module, a signal receiving module, a signal amplification module, a signal output module, a power supply module and the like which are connected in sequence. The multi-frequency pseudo-random signal used by the device integrates the advantages of low-order and high-order pseudo-random signals, is wide in frequency coverage, can flexibly customize and switch the low-order and high-order pseudo-random signals, is realized through specific coding, is high in practical operability, can accurately match the inherent frequency range of a target area in different application scenes, and is high in practicability. And multi-frequency combined signals are selected as required to realize synchronous stimulation. No matter in the fields of rehabilitation physiotherapy, tissue function adjustment, special biological effect regulation and control and the like, the device can adapt to differences of different groups and individuals, the personalized stimulation requirements of diversified scenes are met, and the adaptation range is far beyond that of a traditional single-field device.

The invention belongs to the technical field of forestry engineering maintenance, and particularly relates to a multi-field self-adjusting execution rocker arm which comprises a girder, a working arm, a connecting frame and a mounting seat. According to the invention, the motor I, the motor II and the motor III are controlled through the external integrated console. The mounting base mounting bolt is assembled on the movable carrier, then the positioning ring makes contact with the working face for use, and the positioning ring achieves the spraying effect on trunks or branches through the spraying holes in the inner side. When a working face is changed, for example, the diameter of a trunk is changed, a third motor drives a worm, when the worm rotates, a thread of the worm makes contact with a third shifting wheel, the third shifting wheel can be made to rotate, then a second shifting wheel is driven to rotate, when the second shifting wheel rotates, a shifting cylinder is shifted, a first shifting wheel is driven, then the first shifting wheel drives a second lead screw, and the second lead screw is driven to rotate. When the second lead screw rotates, the connecting rods on the two sides drive the positioning rings to move synchronously, and therefore the distance between the positioning rings is adjusted.

The application relates to a three-dimensional structure unit cell configuration multi-physical field performance coupling regulation method and system, and belongs to the field of electric digital data processing. A multi-scale self-similarity microstructure unit cell is constructed through a fractal geometry theory. A box-counting dimension method is used to quantitatively represent a fractal dimension, a mapping relationship model of the fractal dimension and electromagnetic, mechanical and optical physical performances is established, and directional regulation of macroscopic physical characteristics is realized. The material is adapted to electromagnetic wave absorption, vibration suppression, light field regulation and other multi-field requirements through parameterized design, the limitations of traditional material cross-physical field regulation are solved, and a universal framework is provided for mathematical design of functional materials.

The application relates to a Modelica-based multi-field simulation system of a three-axis turntable and a construction method, and belongs to the technical field of modeling simulation, and solves the problems of compatibility, precision and solving efficiency of modeling simulation of a three-axis turntable in the prior art, including: a modeling field of the three-axis turntable constructed based on Modelica, including a mechanical field, an electrical field and a control field; the mechanical field includes a three-axis turntable mechanical model, a reducer model for providing a reduced rotation angle signal and an amplified torque signal, and a shaft coupling model; the electrical field includes a motor model for providing a torque signal, a driving circuit model for providing a power voltage source, and a conversion circuit model for providing a converted voltage signal to the motor model; the control field includes a control instruction model for providing a theoretical rotation angle signal instruction, a PID controller model for calculating and providing a control instruction required by the conversion circuit model, and a sensor model for collecting and feeding back a signal required by the PID controller model.

The invention belongs to the technical field of detection pretreatment equipment in multiple fields of food, non-food and the like, and provides an efficient and accurate multi-field sample preparation and separation device and a use method thereof.The efficient and accurate multi-field sample preparation and separation device comprises a treatment cylinder, a top cover is arranged at the top end of the treatment cylinder, and a smashing and homogenizing mechanism is arranged above the top cover; a discharging assembly is arranged below the processing barrel, supporting seats are fixedly connected to the bottom face of the processing barrel, and a weighing assembly is fixedly installed on the side faces, close to each other, of the two supporting seats. The variable frequency motor drives the driving rod, the driving disc and the smashing cutter to efficiently rotate, a homogenizing structure can be formed, one-stop rapid sample homogenizing can be achieved in cooperation with auxiliary liquid supply of the water injection electromagnetic valve and the water injection pipe, subsequent cryopreservation and repeated freezing and thawing are not needed, it is avoided that the biological activity of a sample induces drug metabolic conversion or structural isomerization, and the efficiency is high. And the detection result reflects the original residual state.

Disclosed by the present disclosure is a method for obtaining key parameters for pollutant transportations under coupling effects of temperature and hydrodynamics based on a site scale. The method is to obtain the permeability parameters, thermal conductivity, and dispersion coefficient of the contaminated sites based on multi field coupling numerical simulations of hydrodynamic field-temperature field-concentration field. The transportation parameters for pollutants on the site are key indicators for the investigations and treatments of organic pollution on the site. The method as illustrated in the present disclosure includes establishing a water-thermal-salt coupling model for organic contaminated sites according to site drilling data and site physical test results in the early stage, performing multiple field coupling numerical inversion according to the site model in combination with multi-party data, adjusting the model parameters through manual trial and error method to improve model accuracy and obtaining the optimal model.

The invention discloses a multi-field machine translation method based on a large language model, and relates to natural language processing. The method comprises the following steps: constructing a multi-field monolingual data set, and dynamically generating a field term dictionary in combination with an expert term dictionary and a large language model self-reflection mechanism; on the basis of sentence length, syntactic complexity of SpaCy calculation and domain vocabulary proportion, screening single statements with different complexities, and prompting and guiding model learning by taking a term dictionary as a feature; a mixed course learning strategy is adopted, the difficulty degree of a data set is divided, monolingual continuous writing pre-adaptation and bilingual alignment fine tuning are alternately executed, and the multi-field adaptation capability of the model is improved in combination with progressive, multi-field mixed and monolingual and bilingual mixed training. According to the method, the BLEU and COMET indexes are improved by 10%-18%, the training time is shortened by 6-12 h compared with sequential fine tuning, small parameter quantity models such as Llama2-7B and the like are supported, and the method is suitable for multi-field machine translation scenes.

The invention discloses a temporary wharf special construction scheme intelligent generation method based on a multi-field knowledge base. The method comprises the steps that 1, the specific name of a temporary wharf special construction scheme is input; 2, a qwen-32B large model, a standard specification knowledge base and a design file knowledge base are utilized to automatically generate compilation basis chapters; step 3, generating project general situation chapters by using the qwen-32B large model and the design file knowledge base; and 4, generating a construction plan chapter by using the qwen-32B large model and the design file knowledge base. 5, designing a file knowledge base by utilizing the qwen-32B large model and the construction process knowledge base, and generating construction process chapters; step 6, generating guarantee measure chapters by utilizing the qwen-32B large model, the construction process knowledge base, the design file knowledge base and the basic guarantee knowledge base; and step 7, generating acceptance requirement chapters by using the qwen-32B large model and the design file knowledge base. And the problems in a special construction scheme compilation process are solved.

The invention discloses a product design method and a formulating system thereof, particularly relates to the technical field of product design, and aims to generate association records through'demand-design 'two-way mapping, capture multi-field design changes and quantify cross-field conflicts, evaluate change influences through a full-cycle cost-performance coupling model, and perform intelligent rollback verification on unqualified versions. And finally generating a log containing a'demand-design-change 'traceability chain. The system comprises a first draft and demand mapping module and a cross-domain deviation collaboration module, and the method is correspondingly executed. According to the product design method and the formulation system thereof, the problems of demand disjunction, conflict control difficulty, one-sided evaluation and the like in traditional design are effectively solved, full-period global balance is realized, the design efficiency is improved, and the cost and the risk are reduced.

The application discloses a large-range automatic precise control liquid dispensing method of a pipette, and belongs to the technical field of the pipette. In the application, a single device is combined with a double peristaltic pump mode, so that the wide-range liquid dispensing demand is covered, a plurality of pipettes are not needed to be matched, cost and space are saved, integer dispensing plus a high-precision iteration of a residual amount is adopted, the minimum precision reaches 0.01 muL, non-integer volume demand is accurately adapted, three modes of automatic, high-precision and manual are provided, different use scenes are adapted, and the operation threshold is reduced, manual contact is reduced, liquid pollution and volatilization risk are reduced, and the application is suitable for precise liquid dispensing in multiple fields.

The invention discloses a large-range automatic accurate control pipette liquid preparation method, and belongs to the technical field of pipettes. According to the invention, a single device depends on a double-peristaltic pump combined mode, the wide-range liquid preparation requirement is met, a plurality of pipettes do not need to be matched, and the cost and the space are saved; through integer allocation and margin high-precision iteration, the minimum precision reaches 0.01 [mu] L, and the non-integer volume requirement is accurately met; an automatic mode, a high-precision mode and a manual mode are provided, different use scenes are adapted, and the operation threshold is lowered; manual contact is reduced, liquid pollution and volatilization risks are reduced, and the device is suitable for accurate liquid preparation in multiple fields.

The application provides a multi-field task processing method, system, device and medium based on knowledge fusion and Agent cooperation, and belongs to the technical field of cross-field knowledge fusion. The application generates a sub-task set by decomposing a user input task; constructs a specialized Agent set according to a skill mapping table and a role template library; performs cross-field knowledge retrieval and fusion on the Agent set, monitors output consistency and classifies collaboration conflicts; triggers corresponding expansion mechanisms for different conflicts, updates a knowledge graph or generates a new Agent; and simultaneously monitors field knowledge updates, adjusts knowledge weights and synchronizes to related modules. The application realizes accurate decomposition and specialized processing of complex tasks, improves cross-field knowledge fusion efficiency, solves collaboration conflicts, enhances system adaptability, ensures processing timeliness through real-time knowledge updates, and improves the task processing quality and efficiency of a multi-Agent system.

The invention relates to the technical field of distributed artificial intelligence and collaborative perception, and discloses an intelligent perception and decision-making system for distributed multi-modal data fusion, which comprises the following steps: acquiring multi-field real-time data streams in a live broadcast process, aligning the multi-field real-time data streams according to a unified time granularity, and storing the aligned data streams in a sliding window buffer area; carrying out abnormal point detection on the deviation degree sequence of each field by applying an accumulation and detection algorithm, and outputting abnormal alarm information when the accumulation deviation exceeds a dynamic threshold value; outputting the type pre-judgment of the current exception in the field of triggering the exception alarm; when a trend type abnormal event is detected, calculating the occurrence probability and urgency degree of each downstream event, and generating early warning information and intervention suggestions; according to the invention, the pre-judgment from the current domain exception to the future associated domain event is realized. The problem of pre-judgment lag caused by single-field detection is solved, and a time window is won for operation intervention.