216 results about "Performance model" patented technology

The invention relates to the technical field of bridge engineering detection and repair, in particular to a bridge crack repair effect prediction method and system based on performance simulation, and the method comprises the following steps: obtaining basic data and environmental load data of a target bridge, and matching at least two candidate repair schemes from a preset repair scheme library; constructing an initial performance model before repair, and calculating an initial bearing capacity loss rate and a crack propagation rate; quantifying the interaction influence of the multi-modal data through a coupling factor matrix, and iteratively calculating the repair condition of a set time period after repair to obtain an initial simulation result; and collecting real-time monitoring data in a set time period in the repairing process and after repairing, and inputting the real-time monitoring data into a preset LSTM adaptive correction model to obtain a corrected simulation result. According to the method, the long-term repair condition after repair can be accurately calculated, the accuracy of a prediction result and the environmental adaptability are remarkably improved, and a scientific basis is provided for long-term service performance pre-judgment.

The invention discloses a local reasoning acceleration system and method for a large-scale MoE model under limited resources, and the method comprises the steps: carrying out the MoE model performance modeling on large language model reasoning equipment, and obtaining a module calculation performance model; executing a sliding window type greedy prediction-scheduling strategy: performing expert module routing prediction by utilizing a gating network in the MoE model to obtain activated experts of each layer and an input token of each expert, and dynamically determining the number of layers of the MoE model contained in a single scheduling window by utilizing the module calculation performance model in combination with a minimum sliding window algorithm; and according to the predicted expert module route, performing optimal execution strategy search and MoE model layer scheduling on the activated expert module in the window with the single scheduling window length. According to the method, efficient and stable local large model reasoning deployment is realized through an expert activation prediction mechanism without training cost.

The invention discloses a composite fault diagnosis method for motor overload and pressure abnormity of a wind power water cooling system. The method comprises the following steps: S1, data acquisition; the collected original signal data is preprocessed; s2, constructing a lightweight system performance model, and calculating a pressure residual error; s3, extracting time domain and frequency domain features from the original signal data and the residual signals to form feature vectors; s4, performing fault diagnosis based on a machine learning model, and inputting the feature vector into a trained model; the model outputs a diagnosis result and gives a fault confidence coefficient; and S5, outputting a result, and performing early warning. Predictive maintenance is achieved, early warning can be given out at the early stage of a fault when shutdown or secondary damage is not caused, a maintenance plan can be made in advance, passive first-aid repair is changed into active planning, the non-planned shutdown time is shortened, and the availability of a wind field is improved; the unnecessary on-site inspection frequency is reduced, the excessive dependence on the qualification operation and maintenance experts is reduced, and the troubleshooting time is shortened.

The invention relates to a hybrid vertical take-off and landing aircraft energy management method based on a model predictive control algorithm, which belongs to the technical field of aircraft energy management, and comprises the following steps: establishing a size model and a performance model of a multi-rotor configuration based on a top demand, a hybrid architecture and a propulsion configuration by using an iterative scheme of size design optimization; forming constraint of size iteration, and optimizing required overall size parameters of the aircraft through an fmincon function; a component-level turboshaft engine model and a battery internal resistance model are provided, a theoretical model is established based on gas-heat dynamic characteristics of each component of the turboshaft engine, the battery internal resistance model is established based on battery testing, and a battery peak current is obtained; and an energy management strategy based on an MPC algorithm, DP optimization information, supplementary CS and a charge strategy.

The invention relates to the technical field of forklift safety analysis, in particular to a forklift safety analysis system fusing multi-modal data, which comprises a multi-modal data acquisition module, an environment-level safety analysis module, a vehicle-level safety analysis module, a behavior-level safety analysis module, a dynamic decision engine, a safety intervention module and a model optimization module. By collecting multi-modal data, constructing a dynamic environment risk model, a forklift comprehensive performance model and a sub-scene safety evaluation model and fusing multi-dimensional analysis results to generate a multi-level instruction, early warning and active control are achieved, and robustness is improved through self-optimization. The problems that a traditional technology depends on single data, a fixed threshold value fails, and closed-loop intervention is lacked are solved, and the forklift operation safety and efficiency can be comprehensively improved.

Medical devices and related systems and methods are provided. A method of controlling medication delivery based on sensor input involves obtaining a measurement parameter representing an electrical response of a first instance of a sensing element to a physiological condition of a person. The measurement parameter is converted into a calibrated measurement parameter using calibration data specific to the first instance of the sensing element. The method further involves determining a measurement value using the calibrated measurement parameter as input to a performance model. The performance model is derived from historical calibrated measurement parameters and corresponding reference values. The historical calibrated measurement parameters are from other instances of the sensing element. A command is then determined based on the measurement value and sent to a medical device. The command causes the medical device to deliver a dose of medication influencing the physiological condition of the person.

The invention provides a performance prediction method and device based on a refrigerating machine, electronic equipment and a storage medium, and relates to the technical field of refrigeration, and the method comprises the steps that the condensation saturation temperature, the evaporation saturation temperature and the compression exhaust temperature of a refrigerating system are collected in the running test operation process, so that the saturated liquid phase enthalpy and the saturated gas phase enthalpy are determined; according to the condensation saturation pressure calculated according to the condensation saturation temperature and the compression exhaust temperature, the overheat gas specific enthalpy is determined, system refrigeration energy efficiency prediction is conducted according to the saturated liquid specific enthalpy, the saturated gas specific enthalpy and the overheat gas specific enthalpy, and the target system refrigeration energy efficiency is obtained; and performing unit performance prediction on the target refrigerator by using the target refrigeration performance model, the target system refrigeration energy efficiency, the frequency converter efficiency, the motor efficiency and the bearing transmission efficiency. According to the method, the actual refrigerating performance calculation of the refrigerating machine unit from the refrigerating cycle theory can be realized, the problem of performance prediction deviation of the refrigerating machine is effectively solved, and the performance prediction effect of the refrigerating machine is improved.

The invention provides a multi-source heterogeneous data integrated management method based on a digital main line, a medium and a system, and belongs to the technical field of industrial digital main lines. A digital main line platform is used for collecting data of various formats and conducting standardization processing, a deep ontology fusion model is adopted for achieving cross-domain semantic understanding, and the data of various formats is obtained. Semantic understanding efficiency is improved through adaptive pooling feature dimensionality reduction and separable attention calculation, a storage space is optimized by adopting similarity aggregation de-duplication or a distributed independent storage strategy according to a data overlap ratio, and a digital principal line data consanguinity tracing relation is established to realize full-life-cycle data link construction. Based on the semantic mapping deviation value, a fine fine tuning or coarse tuning mode is adopted to optimize semantic understanding parameters, longitudinal integrated mapping from a function model to a performance model and a physical model and a transverse integrated framework of multi-domain comprehensive simulation are constructed, and the technical problem that cross-domain model semantic understanding accuracy is insufficient is solved.

The invention discloses a power quality disturbance identification method based on a Markov transition field and a lightweight dense connection network, and the method comprises the steps: constructing a PQDs standard signal set, converting a complex PQDs one-dimensional signal set into a two-dimensional feature image data set based on an MTF visual conversion method, and dividing the two-dimensional feature image data set into a training set, a verification set and a test set; the method comprises the following steps: developing a DenseNet-L lightweight identification network model based on a dense connection mechanism; a CBAM attention mechanism model combining CAM and SAM is constructed; the DenseNet-L model and the CBAM model are combined through module fusion, and a DenseNet-LC lightweight identification model is constructed and obtained; the DenseNet-LC model is trained through the training set, and an optimal performance model is stored; and performing identification performance test on the DenseNet-LC optimal performance model by using the test set to obtain a final complex PQDs identification result. According to the method, the classification precision and noise immunity of the complex PQDs can be effectively improved, meanwhile, real-time performance is achieved in the efficiency performance aspect, and technical support can be provided for rapid sensing of the operation health situation of a power distribution network system.

The embodiment of the invention provides a humanoid robot joint motor driving performance model training method, a humanoid robot joint motor driving performance model prediction method and humanoid robot joint motor driving performance model equipment, and belongs to the technical field of robots. The method comprises the following steps: acquiring multi-source data; the multi-source data comprises static parameter sample data, dynamic parameter sample data, driving performance sample data and joint motion mode parameter sample data of the joint motor; establishing a multi-dimensional data set associated with the joint movement mode according to the multi-source data; setting a regression model; training the regression model according to the multi-dimensional data set to obtain driving performance prediction data; and adjusting parameters of the regression model according to the driving performance sample data and the driving performance prediction data to obtain a target model. According to the embodiment of the invention, the problems of low joint motor driving performance measurement efficiency, poor dynamic adaptability and high maintenance cost of the humanoid robot can be solved, and accurate performance prediction and optimal control based on data driving are realized.

The invention discloses a water pump energy efficiency optimization method and system, relates to the field of water pump energy efficiency optimization, is used for optimizing electric energy consumption of a water pump and further improving the energy-saving effect, and comprises the following steps: collecting operation parameters of the water pump; based on the operation parameters, an actual performance model of the water pump is periodically reconstructed; the actual performance model reflects a lift-flow relation and a power-flow relation of the water pump at different rotating speeds; according to the current target working condition and the actual performance model, optimal operation parameters of the water pump are calculated; and adjusting the operation of the water pump according to the optimal operation parameters.

The invention provides a zero-trust gateway encrypted traffic non-inductive filtering method and device based on an eBPF multi-level mapping architecture and a computer program product. The method comprises the steps that a data packet is intercepted in a network driving layer through an XDP mode of an eBPF, and quintuple and encryption protocol visible meta information are extracted; performing hierarchical strategy matching on the meta-information by adopting a multi-stage mapping architecture, and dynamically optimizing a search path through a performance model; and executing real-time verification and access decisions based on a zero-trust policy engine to realize non-perceptual filtering of encrypted traffic. According to the method, the problems of weak detection capability and high matching delay of a traditional scheme under encrypted traffic are solved, the method has the advantages of high speed, low delay and high precision, and meanwhile, data privacy and compliance are guaranteed.

The invention discloses an edge AI accurate real-time reasoning method based on heterogeneous calculation, and the method comprises the following steps: S1, collecting perception data on an edge platform, constructing an input tensor, and distributing the input tensor to each calculation unit; s2, establishing a performance model for the calculation unit, and extracting calculation capability, load, communication delay and power consumption parameters; s3, fusing the input tensor and the resource state vector to generate a task state vector; s4, constructing an improved TinyML model, forming a reasoning path, and binding a calculation unit; s5, constructing a scoring function, and selecting an optimal reasoning path to execute a reasoning task; s6, collecting reasoning process data, and generating a feedback vector; and S7, optimizing the scoring function and the model structure according to the feedback vector, and realizing path strategy updating. According to the invention, accurate real-time reasoning and dynamic scheduling under heterogeneous resources are realized.

The invention discloses a multi-source time sequence decomposition and fusion hydraulic structure behavior prediction method, which comprises the steps of 1, acquiring historical time sequence monitoring data of a hydraulic structure, including behavior parameters and environmental quantity parameters, the behavior parameters including dam body settlement and seepage flow, and the environmental quantity parameters including dam body water level and rainfall; 2, preprocessing the acquired monitoring data, and constructing a hydraulic structure behavior model input data set; 3, determining an optimal decomposition period of an input data set of the hydraulic structure performance model; 4, according to the determined optimal decomposition period, decomposing the preprocessed hydraulic structure performance model input data set into three component parameters according to an original time sequence, wherein the three component parameters are respectively a trend component parameter, a periodic component parameter and a residual component parameter; and 5, superposing the predicted values of the three component parameters according to time points to obtain a final performance predicted value of the hydraulic structure. According to the method, the prediction precision is remarkably improved, and the stability and reliability of long-term prediction are remarkably improved.

The invention discloses an HPC curtain wall installation safety protection intelligent monitoring system, and relates to the technical field of intelligent monitoring, and the system comprises a risk rehearsal module which constructs a digital neural rehearsal model, compares installation attitude data with historical accident features through a reinforcement learning algorithm, generates a safety risk thermodynamic diagram, and sends the safety risk thermodynamic diagram to an intelligent monitoring module; performing spatial superposition on the safety risk thermodynamic diagram and a preset installation scene to generate a safety risk index; the risk early warning module is used for generating an instruction type after the voice instruction signal is subjected to natural language analysis, performing contradiction value calculation on the instruction type and a safety risk index, and activating a risk protection protocol of a digital neural rehearsal model and sending out an alarm signal when the contradiction value exceeds a preset safety threshold value; and the communication module transmits the alarm signal and the safety risk index to a remote monitoring terminal. According to the invention, through dynamic visual presentation of the digital nerve rehearsal model and the security risk thermodynamic diagram, intelligent quantitative analysis and spatial visual expression of the security risk are realized.

The invention discloses a GPU BOX energy consumption regulation and control method and system and a computer device. The method comprises the following steps: performing feature analysis on a received to-be-processed task, and identifying task feature information; obtaining historical performance data of all GPU BOXs of the same node corresponding to the task type, constructing a GPU BOX performance model according to state information, obtained in real time, of all GPU BOXs, and obtaining a GPU BOX distribution strategy used for executing the task in the same node according to the model; determining a target GPU BOX according to the GPU BOX distribution strategy, and executing the to-be-processed task; and monitoring the operation information of the target GPU BOX in real time, obtaining the dynamic regulation and control strategy of the energy consumption of the GPU BOX according to the operation information, and executing the dynamic regulation and control strategy of the energy consumption of the GPU BOX. According to the method, dynamic energy consumption regulation and control of the GPU BOX in different working modes can be achieved, efficient utilization of energy consumption is achieved, the service life of equipment is prolonged, energy is saved, and cost is reduced.

Medical devices and related systems and methods are provided. A method of controlling medication delivery based on sensor input involves obtaining a measurement parameter representing an electrical response of a first instance of a sensing element to a physiological condition of a person. The measurement parameter is converted into a calibrated measurement parameter using calibration data specific to the first instance of the sensing element. The method further involves determining a measurement value using the calibrated measurement parameter as input to a performance model. The performance model is derived from historical calibrated measurement parameters and corresponding reference values. The historical calibrated measurement parameters are from other instances of the sensing element. A command is then determined based on the measurement value and sent to a medical device. The command causes the medical device to deliver a dose of medication influencing the physiological condition of the person.

The invention provides a low-altitude unmanned system air node optimal deployment method based on communication-link-energy consumption joint optimization. The method comprises the following steps: initializing an unmanned aerial vehicle group; the system constructs a performance model according to the real-time channel information, constructs a multi-objective collaborative optimization problem, and solves the multi-objective optimization problem by adopting an MOPSO-HGD algorithm to obtain an optimal result; deploying the unmanned aerial vehicle group to the optimal position according to the optimal result to form a virtual distributed MIMO array structure; setting an unmanned aerial vehicle electric quantity threshold value, judging the size of each unmanned aerial vehicle remaining electric quantity and the unmanned aerial vehicle electric quantity threshold value, and if the unmanned aerial vehicle remaining electric quantity is smaller than the unmanned aerial vehicle electric quantity threshold value, recycling the unmanned aerial vehicle; otherwise, re-determining the position of the unmanned aerial vehicle group; through a closed-loop control mechanism of channel perception-multi-target optimization-dynamic adjustment, the unmanned aerial vehicle relay system can still keep a high signal-to-noise ratio and strong robustness in a complex low-altitude environment, is suitable for scenes such as low-altitude broadband communication, emergency rescue and urban information networks, and has a good engineering application prospect.

The invention provides a method for determining surge margin loss of a compression part caused by production and assembly deviation. Key geometric parameters are determined; key geometric parameters are obtained through measurement; engine performance parameters and working condition parameters are obtained; correcting and confirming key geometric parameters by using the overall performance model, the performance parameters and the working condition parameters of the engine of the model; using a plurality of sub-engines to obtain probability distribution characteristics of each key geometric parameter; a Monte Carlo method is adopted, each key geometric parameter is valued according to the determined probability distribution, the engine overall performance model of the model is used for carrying out engine overall performance calculation, and a compression part common working line is obtained; determining a surge boundary line of the compression component when the key geometric parameters change; and a surge boundary line is determined, and surge margin loss caused by engine production deviation is obtained. According to the method, the surge margin loss of the compression part caused by the production and assembly deviation is determined according to a small number of engine production, assembly and test run results in the development stage.

A cognitive assessment system includes a computing device, a server coupled to the computing device, and an eye tracking device coupled to the computing device. The computing device includes a cognitive assessment program configured to perform at least one cognitive assessment task to assess cognitive function of a user. The server includes a database. The database stores historical performance data of the user, a historical performance model of the user, task performance data of a healthy population, and task performance data of a patient population. The eye tracking device is configured to capture eye movement information of the user.

The invention provides a performance analysis method and device for source address verification, equipment and a storage medium, and relates to the technical field of network security. The method comprises the following steps: receiving a source address verification processing program written through a description language; loading an algorithm configuration file; wherein the algorithm configuration file at least comprises a performance model configured for a plurality of preset algorithms, and the performance model is used for calculating the consumed time length of the operation statement based on the operand length of each operation statement; based on the source address verification processing program and the performance model, calculating the consumed time length of each operation statement; and generating a performance analysis report of the source address verification processing program according to the time-consuming duration of all the operation statements in the source address verification processing program. According to the embodiment of the invention, automatic and theoretical performance analysis can be carried out on the source address verification processing flow based on the formalized description language and the predefined algorithm configuration file under the condition of separating from the dependence of specific hardware.

The invention provides an aero-engine complete machine digital test method which comprises the following steps: firstly, establishing an engine main combustion chamber fuel flow neural network PID control model and an open-loop control rule of a guide vane angle, a channel area and a thrust augmentation fuel flow, and establishing an engine control system digital test model according to the PID control model and the open-loop control rule; and an engine performance digital twin model is constructed and used for predicting and outputting real-time performance parameters according to the engine control parameters and the test run environment parameters. In the test process, the real-time rotating speed of the engine serves as a middle bridge, a closed-loop control loop is formed between the control system model and the performance model, the whole machine dynamic test process of the aero-engine is simulated, and the aero-engine can be driven to be subjected to a high-fidelity simulation test in a digital space only through environment parameters and the angle of a throttle lever. The real physical test requirement is reduced, the engine development efficiency is improved, and compared with a traditional method, the high coincidence between the digital test process and the real test is ensured.

The application discloses a beyond-visual-range air combat decision-making method based on dynamic fire field and artificial potential field fusion, and relates to the field of aerospace. The method comprises the following steps: establishing a beyond-visual-range air combat simulation environment, modeling enemy and friendly camps in the air combat; grouping the enemy camp in real time; constructing a time-varying killing performance model of a missile to determine a three-view dynamic fire field of an aircraft-missile-target so as to establish an artificial potential field of combat groups, and determining the potential field gradient generated by each combat group in the grouping result; improving a particle updating formula in a particle swarm optimization algorithm by using the potential field gradient generated by the combat groups; solving a target function to determine a maneuvering decision of a friendly aircraft. The improved particle swarm optimization algorithm is obtained by improving the particle updating formula in the particle swarm optimization algorithm by using the potential field gradient generated by all the combat groups in the grouping result. The application can improve the coping ability, autonomous decision-making ability and coping ability in a complex battlefield environment of the beyond-visual-range air combat.

The application discloses a large language model parameter fine-tuning method and device based on manifold constraint multi-track adaptation and a medium. In view of the problems of unstable training and limited expression capacity of an existing low-rank adaptation technology, a plurality of parallel low-rank tracks are constructed, and a double random matrix is introduced to constrain the information flow between the tracks, so that the stability of gradient propagation is ensured. Meanwhile, the expression capacity of the adaptation module is enhanced under a limited parameter budget by dynamically fusing the outputs of the tracks through a dynamic gating mechanism. The method can realize stable, efficient and high-performance model fine-tuning by fine-tuning a small number of parameters, and reasoning has no additional overhead, and is particularly suitable for application scenarios with limited resources and high stability requirements, and has wide practical value.

A solar tracking system comprises a plurality of solar panel modules, each independently orientatable relative to a solar source. A control system determines a topography associated with the solar panel modules and generates, for each module, a performance model based at least in part on the topography and weather data. The control system independently orients each solar panel module to the solar source based on the respective performance model to optimize energy output from the array. The topography may indicate relative positions or heights of the modules and can be determined using laser site surveys, energy readings from photovoltaics, or aerial imaging. The weather data may include direct normal irradiance, global horizontal irradiance, or diffuse horizontal irradiance. The control system may periodically update the performance model for each module and account for interactions between neighboring modules due to shading.

The application provides an intelligent training acceleration method based on data dense characteristics in a neural network, which collects relevant data information and performance model information by pre-training a model, and dynamically reduces the data amount required to be transmitted between a GPU and a CPU by using the dense characteristics of a deep learning network in combination with compression and decompression algorithms, so as to improve the overall performance of the system. The application also provides an intelligent training acceleration system based on data dense characteristics in a neural network, which comprises a data acquisition module, a dense data compression rate decision module, a selective compression decision module and a training module; and the neural network training is carried out by using the system, so that the training speed and the overall performance of the system can be effectively improved.

This invention provides a task performance evaluation method based on a combination of hierarchical KAN and RBF networks. The method involves: constructing multiple KAN networks corresponding one-to-one with categories; each KAN network performs hierarchical nonlinear transformation and fusion of multidimensional feature variables of the same category, outputting low-dimensional feature variables; constructing an RBF network and inputting the low-dimensional feature variables output by each KAN network into the RBF network; the RBF network uses all low-dimensional feature variables as input, performs nonlinear mapping, and outputs the task performance evaluation result; finally, the transformation from a task performance evaluation index system to a task performance evaluation result is completed. This invention effectively fuses features from multiple categories using KAN networks to achieve feature dimensionality reduction, and uses the RBF method to quickly fit the performance model between input and output, improving the accuracy and efficiency of the evaluation. It addresses the shortcomings of existing technologies in modeling multi-category, multi-feature variable data, improving evaluation accuracy, computational efficiency, and interpretability.

The invention discloses a double-classification head voiceprint recognition training method based on federal learning. The process of the training method comprises the steps that a client provides local speaker category information, a server integrates global speaker categories, the server initializes a global classifier and a global encoder, the server distributes the global classifier and the global encoder, and the client receives the global classifier and the global encoder and then carries out training together with a local classifier. In the local training process, global classifier parameters are updated through a moving average algorithm, cooperative training of the global classifier and a client local classifier is provided, and the model performance is remarkably improved. In the model fusion process, in order to solve the influence caused by data distribution deviation, a global classifier carries out weighted averaging according to categories contained in each client. According to the method, the local model with better performance can be obtained, so that the global model performance loss during model aggregation is reduced, and a forward feedback mechanism is formed for subsequent model training.

The application discloses a lightweight adaptive network learning method for multi-modal multi-task learning, comprising the following steps: 1, constructing a downstream task dataset, 2, constructing a deep self-attention network model, 3, pre-training weight pruning, 4, constructing a task adapter, 5, adapting a pre-training model, 6, designing a progressive guided distillation training and training a model. The method prunes part of the weights of the pre-training model and adapts it with an efficient task adapter. The application proposes a progressive guided distillation training algorithm to better fill the gap between the pre-training task and the downstream task, ensuring the performance of the model on the downstream task. The application can be combined with any existing pre-training model based on a deep self-attention network, and through training, an adapter model with superior performance in the downstream task, total storage overhead during model deployment, calculation overhead during model inference and model configuration flexibility is obtained.

The invention provides a new energy automobile hardware-in-the-loop performance calibration method and system, and the system comprises a virtual function test module which is used for determining each calibration item and a corresponding calibration value range by taking the function set by a virtual whole automobile as a target; the virtual calibration module is used for determining a calibration value of each calibration item and performance data corresponding to the calibration item by taking the calibration item and the calibration value range as constraint conditions and taking the performance set by the virtual whole vehicle as a target; the calibration correction module is used for acquiring calibration values and performance data corresponding to all calibration items of a real vehicle corresponding to the virtual whole vehicle, and performing difference calculation on the calibration values and the performance data corresponding to all the calibration items of the real vehicle and the calibration values and the performance data corresponding to the calibration items determined by calibration of the virtual whole vehicle; and when the calibration difference value or the performance difference value exceeds a set threshold value, modifying the basic function model combination and the performance model combination used in the whole vehicle simulation virtual calibration to perform simulation virtual calibration again until the set threshold value is met.