268 results about "Physical modelling" patented technology

The invention provides a multi-mode ultrasonic fusion pressure vessel weld defect nondestructive testing method and system, and relates to the technical field of nondestructive testing. According to the method, geometric parameters of a welding seam are obtained through three-dimensional laser scanning, and an optimal scanning parameter set is generated; driving ultrasonic phased array equipment to scan for one time and synchronously acquire shear wave full-matrix capture and longitudinal wave linear scanning data; performing energy flow angular spectrum analysis and envelope analysis on the bimodal data, extracting defect feature parameters and constructing a three-dimensional feature tensor; carrying out multi-dimensional feature fusion by adopting Tucker decomposition, and enhancing a core tensor through physical modeling; generating three types of defect indication diagrams including a defect existence possibility diagram, a defect relative scale diagram and a defect space orientation diagram from the enhanced feature tensor; and the three types of indication diagrams are visually presented for comprehensive interpretation of detection personnel. Through multi-modal data fusion and physical modeling enhancement, the defect identification accuracy and detection efficiency are remarkably improved, the false alarm rate is reduced, and reliable technical support is provided for pressure vessel welding seam safety detection.

The invention belongs to the technical field of automatic control, and relates to a self-adaptive control method based on multi-physical modeling. According to the method, by collecting multi-source data of a controlled object, a thermal, electric and force coupling relation used for control analysis is established so as to describe dynamic responses under different operation conditions. And calculating stress, motor power, energy consumption and temperature rise change in the operation process based on a coupling relation to obtain system performance data, verifying stability and safety of different control parameter combinations in a simulation environment, and obtaining performance indexes including operation retardation risk, overload safety margin and safety response time limit. And according to a simulation result, under the condition of meeting safety constraints, taking energy consumption and temperature rise as optimization targets, adjusting control parameters, generating optimized control parameter configuration data, and feeding back the optimized control parameter configuration data to a control unit, so that closed-loop adaptive control and performance optimization are realized. According to the invention, through multi-physical coupling modeling and simulation optimization, the adaptability and reliability of the automatic control system are improved.

The invention relates to the technical field of data processing, and discloses a multi-scene-oriented unmanned aerial vehicle operation resource dynamic configuration and scheduling method and a multi-scene-oriented unmanned aerial vehicle operation resource dynamic configuration and scheduling system. The method comprises the steps of generating multi-scene task constraint data through semantic analysis; constructing a dynamic situation field matrix based on a Gaussian function; establishing a continuous motion velocity field containing pressure gradient and viscosity diffusion; the vortex velocity components are fused to form a self-adaptive control velocity field; and constructing a task allocation vector field to generate an operation area configuration scheme and converting the operation area configuration scheme into a flight control instruction. According to the method, the core problems that multi-scene adaptability is insufficient and real-time performance and optimality are difficult to balance in the prior art are solved, and the intelligent scheduling capability and the collaborative operation efficiency of the unmanned aerial vehicle group in a complex and changeable environment are improved through an innovative semantic understanding and physical modeling combined method.

The invention relates to a lithium ion battery life prediction method, and belongs to the field of battery life prediction and intelligent maintenance. The method comprises the steps that S1, a battery capacity degradation sequence is collected, and integrity is checked and normalized; s2, decomposing the sequence by using an improved complete set empirical mode decomposition algorithm, and dividing the sequence into a high-frequency component and a low-frequency component according to a zero-crossing rate; s3, modeling the high-frequency component: fusing multi-scale channel interactive attention, a time sequence convolutional network and a hybrid expert model, and extracting short-term fluctuation and capacity recovery features; s4, modeling a low-frequency component: introducing a two-way gating circulation unit network constrained by a double-index degradation model, and simulating a long-term trend; and S5, constructing a high-frequency migration module through tensor decomposition, improving cross-battery generalization, and fusing high and low frequency results to output a residual life prediction value. According to the method, a dual-channel framework combining signal decomposition, deep learning and physical modeling is combined, the prediction precision and adaptability under complex degradation are improved, and the method is suitable for various battery systems.

The invention discloses a method for establishing a high-voltage circuit breaker fault detection model, and the method comprises the following steps: collecting current, voltage, mechanical response, temperature and other multi-dimensional signals of a circuit breaker under different working conditions, and unifying the signals into standardized time sequence data; a nonlinear dynamic sparse identification method is utilized to establish a dynamic model for describing equipment state evolution, and sparse coefficients reflecting physical change rules are extracted from the dynamic model to serve as health features. And the features are fused with current monitoring data to generate a joint feature input vector, and a health prediction model based on a TabPFN architecture is introduced for training and discrimination. And finally, accurate prediction of the current health state or the potential fault of the circuit breaker is realized, and the model self-adaptive updating capability is realized. According to the method, physical modeling and data analysis are combined, so that the accuracy and interpretability of fault prediction are improved.

The invention discloses an unsteady flow field dimension reduction and prediction method fusing physical modeling and deep learning, and belongs to the technical field of computer-aided fluid mechanics analysis. According to the method, firstly, a modal coefficient reflecting global dynamics is extracted from an unsteady flow field by using a DMD, and meanwhile, low-dimensional feature representation of a potential space is learned from a flow field snapshot through CVAE; the two types of features have complementarity in physical and statistical meanings, and the complex dynamic evolution law of the unsteady flow field is more effectively represented through the low-dimensional features constructed in a combined mode. On the basis, an LSTM model is used for carrying out time sequence modeling on the joint features, and high-precision prediction of future evolution of the flow field is achieved. The hybrid modeling method provided by the invention improves the dimensionality reduction efficiency and prediction precision of a high-dimensional nonlinear unsteady flow field while keeping physical consistency, and is suitable for intelligent simulation and rapid prediction tasks in a complex flow scene.

The invention relates to an error compensation method of a birefringence self-calibration laser level meter, in particular to the field of laser level meters, and aims to improve the precision of the laser level meter by combining a birefringence effect, physical modeling and neural network optimization. Firstly, temperature data are collected in real time, Kalman filtering is used for noise reduction, and accurate data are provided for physical modeling; then, through physical constraints such as a heat conduction equation and a Jones matrix, a coupling relation between the temperature gradient and optical parameter changes is established; then, dynamically optimizing the refractive index correction by using a physical information neural network to avoid an overfitting problem; and finally, calculating optical path difference compensation by using a hardware accelerator, correcting a system error in real time, and feeding back and adjusting a network weight. According to the method, the precision and the stability of the laser level meter in an environment with relatively large temperature change are effectively improved.

The invention discloses a large model driven automobile model modeling method based on a Modelica language, and belongs to the technical field of intelligent modeling and automobile simulation. The method comprises the following steps: firstly, accurately analyzing a natural language demand into a structured triple by adopting a BERT-CRF (domain knowledge enhanced) multi-task model; matching an optimal component combination through a multi-objective optimization algorithm driven by a graph neural network, and cooperatively predicting an interdisciplinary parameter feasible region in combination with symbolic mathematical derivation and machine learning; a topological connection matrix is innovatively optimized by using a graph attention network, and intelligent generation and dynamic verification of simulation codes are realized by fusing a template engine and syntax tree analysis; and finally, constructing a multi-target reward function optimization control strategy through reinforcement learning, and establishing a closed-loop knowledge iteration mechanism. Compared with a traditional modeling method, through deep combination of the large model and Modelica, the technical difficulty of automobile system modeling is remarkably reduced while the preciseness of physical modeling is kept, and the method is particularly suitable for complex scenes such as new energy vehicle model development and intelligent driving system integration.

The invention discloses an extended-range hybrid propulsion double-source dynamic coupling energy management method, which comprises the following steps: carrying out global physical modeling on a double-source power system and a flight scene, and establishing a double-source dynamic coupling model; designing a reinforcement learning physical constraint reward function, performing optimization training on each coefficient of the reward function by adopting a QMPSO algorithm, and outputting an optimized reward function coefficient; a qualified double-source dynamic coupling model is verified, and a power distribution coefficient is optimized; outputting the optimal power distribution coefficient of the battery and the range extender; the superiority of the dual-source power cooperative control strategy in the aspects of flight economy, operation stability and system life guarantee is verified through multi-dimensional comparative analysis of each performance index. Cooperative power distribution of the battery and the range extender is achieved through dynamic coupling modeling and reinforcement learning, the flight scene load requirement is met, the system energy efficiency is improved, and the service life of parts is prolonged.

The invention relates to the field of battery thermal management, in particular to an AUV lithium ion battery thermal state prediction method. Comprising the following steps: constructing an electrothermal coupling reduced-order thermal model, and generating initial temperature estimation with physical consistency; a physical guidance space-time dynamic graph convolutional network PG-STDGCN is constructed as an error correction model, the model constructs a static and dynamic fused adjacency matrix by embedding physical priori such as a battery topological structure and circuit characteristics into dynamic graph learning, and a correction value of initial temperature estimation is output; and adding the initial temperature estimation and the correction value to obtain a final battery thermal state prediction result. According to the method, organic fusion from physical modeling to data-driven correction is realized, interpretability, precision and adaptability are considered under the dynamic working condition of the AUV, and the battery pack-level multi-cell temperature prediction performance is remarkably improved.

The invention discloses a magnetic source positioning method based on a spin-exchange relaxation-free atom magnetometer, and aims to provide a high-precision brand-new solution for magnetic source positioning in a near-zero magnetic environment by performing physical modeling and least square estimation on space magnetic field distribution of a target magnetic source. The magnetometer is enabled to be simultaneously sensitive to a three-axis magnetic field, a light absorption detection method is utilized to ensure that complete three-axis magnetic field information is obtained, and the sensitivity of the three axes reaches fT / Hz1 / 2 (1fT = 10 <-15 > T) magnitude. In the positioning process, the target magnetic source is abstracted into a physical model of the magnetic dipole, and the space magnetic field information generated by the target magnetic source is obtained through the atom magnetometer, so that the long-term stability of the magnetic source positioning process and the high precision of the positioning result can be ensured in the zero magnetic environment; therefore, a high-precision magnetic positioning technical means with great potential is provided for various application scenes such as paleomagnetic sample testing and magnetic source positioning in medical research, and technical progress and application expansion in related fields are expected to be promoted.

The invention discloses a building concrete waste treatment system, which belongs to the technical field of building waste treatment, and is characterized in that a data acquisition module acquires waste physical attributes and inputs the waste physical attributes into a model generation and optimization module; the model generation and optimization module establishes a waste physical characteristic model, dynamically updates the waste physical characteristic model and outputs a dynamic waste physical characteristic model to the strategy generation module; the strategy generation module outputs equipment operation parameters; the strategy mapping module maps the equipment operation parameters into physical execution signals through an equipment instruction mapping model to obtain execution results, and the execution results are input into the strategy feedback optimization module; and the strategy feedback optimization module identifies the execution deviation in the execution result and feeds the execution deviation back to the strategy generation module to optimize the equipment operation parameters. According to the scheme, an intelligent algorithm and physical modeling are combined, the waste treatment process is optimized through real-time data collection and feedback, equipment parameters are adjusted based on different characteristics of the waste, and the problems of excessive treatment and resource waste are avoided.

The invention provides a CFD optimization-based air volume matching and linkage control method and system for a smoke prevention and exhaust system of an underground powerhouse, and belongs to the technical field of underground smoke exhaust systems. The method comprises the following steps: performing physical modeling according to an actual underground powerhouse, and performing grid division, solver selection, boundary condition setting and airflow field change simulation to obtain an initial airflow field; cFD is adopted to simulate temperature changes, smoke concentration changes and airflow changes in different fire scenes, wind flow and smoke flow in the underground powerhouse are analyzed, the defects of an existing system are evaluated, and smoke exhaust pipeline arrangement and fan configuration are optimized according to the defects; the airflow change in the fire environment is dynamically calculated by combining the real-time monitoring system and the CFD, and the air volume of the pressurized air supply system, the smoke exhaust system and the air supplement system is adjusted. Efficient linkage of the system is achieved, while personnel evacuation safety is guaranteed, the smoke exhaust efficiency and the emergency response capacity are improved, energy consumption is reduced, and the dual goals of energy saving and efficiency improvement are achieved.

The invention discloses a semiconductor device etching structure prediction method and equipment, and a storage medium. The method comprises the following steps: acquiring three-dimensional etching evolution simulation data of a semiconductor structure to be etched under different etching process conditions; extracting a two-dimensional section image from the three-dimensional etching evolution simulation data; generating an etching morphology evolution sequence under high time resolution based on the two-dimensional cross section image; and obtaining an etched target contour image of the semiconductor structure according to the etching morphology evolution sequence under the high time resolution. Constructing a contour prediction model based on deep learning, and training the contour prediction model by using the data set; the input of the contour prediction model based on deep learning comprises etching process conditions and an initial contour image of the semiconductor structure before etching, and the output of the contour prediction model based on deep learning is a predicted contour image after etching evolving along with etching time. According to the method, physical modeling and a data driving method are combined, high-precision prediction and intelligent optimization control of the etching process are achieved, and the process stability and the product yield are improved.

The invention discloses an unmanned aerial vehicle aerial image imaging optimization method and device fusing a deep learning perception mechanism and physical modeling. The method comprises the following steps: acquiring an original image frame obtained in a flight process of an unmanned aerial vehicle; inputting the image into a MobileViT illumination estimation network, extracting local convolution perception and multi-scale global semantic features, and outputting a scene illumination intensity estimation value; constructing a differentiable imaging parameter reasoning module based on an illumination physical modeling relationship, reversely deducing an optimal exposure parameter combination of a current frame, and constructing a parameter optimization module based on a perceptual error; combining the difference between the reconstructed image and the target image in the semantic perception space to construct a multi-loss function joint training model, and optimizing an exposure combination; deploying an edge computing platform for the trained network model to complete parameter prediction, control feedback and image acquisition link closed loop; according to the method, exposure optimization is realized before imaging, image gamma decoding and target enhancement are realized after imaging, and the image quality in low-light and backlight scenes is improved.

The invention belongs to the technical field of textile, and discloses a reinforcement learning and physical modeling combined broken end dynamic capturing method and system. The method comprises the following steps: constructing a yarn microstructure evolution model; inputting real-time environment parameters of the textile workshop into the microstructure evolution model, and predicting to obtain real-time microstructure parameters of the yarn; performing nonlinear coupling feature extraction on the real-time environment parameters, and calculating an environment coupling feature vector; inputting the real-time microstructure parameters as material attributes into a yarn tension dynamical equation, and solving to obtain a predicted yarn macroscopic stress state; splicing the macroscopic stress state of the yarn and the environment coupling feature vector to form a state observation value; the reinforcement learning agent outputs a broken end risk probability value according to the state observation value; and when the broken end risk probability value exceeds a preset threshold value, generating and outputting a broken end early warning signal. According to the invention, early and accurate early warning of yarn breakage in a complex dynamic environment can be realized.

The invention relates to the field of unmanned aerial vehicle motor service life prediction, and discloses an unmanned aerial vehicle motor service life test method, which comprises the following steps: collecting motor data in real time through temperature, vibration, current and voltage sensors, carrying out cleaning, denoising and standardization processing, analyzing motor characteristics by adopting physical modeling, extracting parameters such as damage rate, temperature and load, and calculating the service life of the unmanned aerial vehicle motor. According to the method, historical data and physical modeling features are combined, an LSTM deep learning model is constructed for time sequence training, the remaining service life RUL is predicted, feature fusion is carried out, the prediction precision is improved, the health state of the motor is evaluated according to a prediction result, maintenance suggestions are provided, the model is optimized and fed back, and the prediction method is dynamically adjusted. According to the technical scheme of fusing physical modeling and the deep learning LSTM network, high-precision prediction of the service life of the motor is realized, the intelligent degree of data processing is improved, and the adaptability of the model to a complex operation environment is enhanced.

The invention particularly relates to a multi-level self-cognition system based on a large model, and relates to the technical field of large models. A neural symbol world model module; a large language model cognition core module; and a hierarchical decision planning system module. According to the method, deep integration of perception, cognition and decision making is achieved through the hierarchical fusion architecture, and compared with the prior art, the method has remarkable advantages; the multi-modal perception encoder adopts layered encoding and a cross-modal attention mechanism, so that the semantic alignment problem of multi-source perception data is effectively solved, and the understanding ability of the system to a complex scene is greatly improved; according to the neural symbol world model, the neural network and symbol reasoning are combined, the limitation of a pure neural network method in physical modeling is overcome, meanwhile, the calculation complexity of a pure symbol system is avoided, and efficient and accurate environment characterization and prediction are achieved.

The invention discloses a side plate type heat exchanger performance testing system and method, and relates to the technical field of heat exchanger testing, and the system comprises a computer which is used for constructing a digital twinborn model, generating multiple groups of dynamic working condition data, inputting the multiple groups of dynamic working condition data into the digital twinborn model to obtain dynamic performance data, screening according to the sensitivity coefficient of each piece of dynamic performance data to obtain sensitive working condition data; the test equipment executes the sensitive working condition data and collects actual measurement state data of the side plate type heat exchanger; and the computer judges whether the performance of the side plate heat exchanger meets the design requirement according to the actually measured state data. Physical modeling is carried out on the heat exchanger in a digital twinning mode, heat exchange simulation is carried out on the heat exchanger, invalid test working conditions are deleted from a parameter set, only sensitive working conditions are reserved, and time, manpower and energy cost needed by testing is greatly saved.

The invention provides an electric power facility dynamic safety evaluation system based on blasting vibration propagation characteristics, which relates to the field of electric digital data processing and comprises a sensing and data acquisition module, a physical modeling and parameter library module, a real-time evaluation and risk inference module and a feedback execution and self-adaption module. The sensing and data acquisition module is used for converting on-site physical signals into high-quality data streams, the physical modeling and parameter library module is used for providing and building physical modeling and managing parameter information, and the real-time evaluation and risk inference module maps sensing data into system states and risk indexes through observation-model assimilation. The feedback execution and self-adaption module is responsible for driving decision execution, collecting feedback and continuously improving a system model; the system can monitor the influence of blasting vibration on electric power facilities in real time, accurately evaluate the structural damage degree and the system operation risk, and provide a scientific basis for safety protection and emergency decision-making of the electric power facilities.

The invention discloses a key part remaining service life prediction method based on a physical information neural Wiener process, and the method employs a Wiener physical information neural network model to predict the remaining service life of a key part, and integrates data driving and physical modeling. By combining the feature extraction capability of the feature extraction prediction sub-network and the physical degradation modeling capability of the Wiener degradation modeling solution sub-network, the complex nonlinear degradation features of the key parts can be captured more accurately, and high-precision prediction of the remaining service life of the parts is realized. Specifically, the feature extraction prediction sub-network extracts time sequence degradation features through a neural network, and the Wiener degradation modeling solving sub-network uses a Wiener process to model degradation dynamic states of key parts; a physical information joint loss function is designed, and a dynamic weight adjustment strategy is adopted, so that the relationship between data-driven learning and physical modeling is effectively balanced, a prediction result is ensured to accord with a real degradation law, and the physical interpretability of the model and the credibility of the prediction result are enhanced.

The invention relates to the technical field of medical image processing, and discloses a multi-mode ultrasonic mammary gland and multi-mode fusion method. The method comprises the following steps: S1, acquiring medical images of at least two modalities of a mammary gland area, wherein the medical images comprise a first modal image and a second modal image; s2, performing standardization preprocessing on the first modal image and the second modal image; and S3, constructing a physical modeling model of the breast tissue based on the second modal image, and generating a tissue deformation displacement field to obtain tissue response information. By adopting a multi-modal ultrasonic mammary gland image registration and fusion technology and combining feature fusion and deformation field optimization based on an image neural network, the image registration precision is remarkably improved. Compared with a traditional image registration method in the prior art, the multi-modal image registration method not only optimizes the alignment effect among the multi-modal images, but also effectively reduces registration errors caused by modal differences, and particularly provides more accurate alignment on detail structures of breast tissues.

The invention discloses a standard modular energy storage thermal runaway early warning method and system based on digital twinning, and relates to the technical field of energy storage system safety monitoring. The method comprises the following steps: deploying multiple types of sensors in an energy storage unit to form a network, and collecting multi-dimensional data such as temperature, voltage and gas concentration; transmitting the data to the digital twin platform, and performing format conversion on the data for thermal runaway prediction; the platform is combined with the LSTM-GBDT hybrid model through physical modeling, and the risk probability is output; and comparing the risk threshold to trigger early warning. The system comprises a sensing module, a data acquisition and transmission module, a digital twinborn prediction module and an early warning execution module. According to the method, physical simulation and historical data training are fused through the digital twin platform, the heat transfer and out-of-control propagation process can be accurately simulated, virtual simulation optimization is supported, the early warning accuracy and reliability are improved, and the operation and maintenance burden and accident potential are reduced.

The invention discloses a composite material defect modeling method based on physical modeling generation and meta-learning migration, which comprises the following steps: S10, starting from a forming mechanism of a fatigue crack of a composite material, establishing a morphological function of a crack morphological simulation model based on a mechanical principle for simulating a spatial evolution process of a defect; s20, embedding the morphological function as a regular term into a generator loss function of a generative adversarial network, and generating a high-simulation pseudo-defect image; s30, introducing a transfer learning mechanism, constructing a model-irrelevant meta learning training strategy, and realizing rapid learning of the model on a new task through a two-layer nested optimization strategy; s40, introducing a domain adaptive residual module in the model training process to enhance the feature alignment capability of a cross-batch image domain; and S50, training the model by using a real sample and a synthetic defect sample in a mixed manner.

The invention discloses an offshore wind turbine multi-source heterogeneous data fusion and state evaluation method and device and a medium. The method comprises the following steps: constructing an offshore wind turbine multi-source heterogeneous database; data preprocessing is carried out; mapping the multi-source heterogeneous data subjected to data preprocessing to a unified embedding representation space by utilizing physical modeling and deep learning technologies; in the unified embedded representation space, constructing a feature interaction operator based on a multi-head cross attention mechanism, executing feature decoupling, and extracting a panoramic service state vector; and introducing the panoramic service state vector as a global observation item into a variational Bayesian calibration framework of an integrated physical information neural network, realizing dynamic calibration of digital twin physical parameters, and generating an offshore wind turbine service state evaluation report with uncertainty quantitative support. According to the method, the problems of data island, evaluation conservative, decision lag and the like are effectively solved, and a service state evaluation report with physical interpretability and uncertainty support is generated.

The invention provides a passenger compartment air conditioner control method based on model predictive control. The method comprises the steps that a passenger compartment model is constructed, and an original data set containing environment parameters, air conditioner control parameters and passenger compartment response parameters is obtained through experimental design; preprocessing the data; a neural network model is trained through the preprocessed data and serves as a prediction model to learn the nonlinear mapping relation among the air conditioner operation parameters, the environment parameters and the passenger compartment states; calling the prediction model to execute MPC rolling prediction, and generating a future state variable sequence; constructing a target function containing a state tracking error and control energy cost, obtaining an optimal control sequence through optimization solution, and acting the first control quantity on an actuator; and dynamically adjusting a prediction model or a control parameter according to real-time feedback. The technical problems that physical modeling is difficult due to high nonlinearity of a passenger compartment air conditioning system, and a traditional control strategy cannot give consideration to high-precision prediction and multi-working-condition adaptability can be solved.

The invention provides a wind turbine power prediction method based on multi-modal feature fusion, and relates to the technical field of wind turbine power prediction.The method comprises the steps that multiple types of sensors are used for obtaining multi-source sensor data, and a standardized time sequence data set is generated; establishing a first observation matrix based on the standardized time sequence data set, and constructing a weight sensing fractional order adaptive genetic algorithm to optimize the first observation matrix to obtain a second observation matrix; performing feature fusion and noise reduction on the second observation matrix by applying extended Kalman filtering and combining the working state and physical modeling of the wind turbine; constructing a bidirectional long-short-term memory network, and performing power time sequence modeling based on the multi-dimensional noise reduction feature vector to obtain an output power predicted value; and feeding back an error between an output power prediction value and a true value to a weight sensing fractional order adaptive genetic algorithm to carry out parameter iterative optimization to obtain a third observation matrix, and obtaining a power prediction value of the wind turbine based on the third observation matrix.

The embodiment of the invention provides a broadband oscillation source positioning method and device, electronic equipment, a computer readable storage medium and a program product, and relates to the technical field of electric power. The method adopts a data driving strategy, does not need to depend on accurate physical modeling of a power system, and is suitable for a modern power system with a complex operation state and a changeable topological structure. The causal influence intensity between generator sets is analyzed by introducing transfer entropy, an adjacent matrix is constructed, and reasonable graph structure modeling under the condition of no topological prior is realized. Deep feature extraction is carried out on active power time sequence data of the generator set through the gating circulation unit, dynamic evolution characteristics in the broadband oscillation process are fully captured, and the representation capacity of nodes is enhanced. A causal graph structure and enhanced features are fused in a graph neural network model, space-time joint modeling is realized, the description capability of a disturbance energy propagation path is effectively improved, the spatial positioning precision of a forced oscillation disturbance source is greatly improved, and the method has good engineering applicability and popularization value.

The invention provides an escalator running path sensing method based on sensor data. The escalator running path sensing method comprises the following steps: acquiring multi-source nine-axis sensor data; carrying out fusion and feature enhancement on the multi-source nine-axis sensor data; step 2, static state detection and scene classification; 3, escalator running direction judgment and cross-layer quantification are carried out; and 4, outputting a calculation result. According to the escalator running path sensing method based on the sensor data, high-precision and low-power-consumption vertical traffic scene recognition and motion trail reconstruction are achieved through a technical path of staged physical modeling and deep learning collaborative optimization.

The invention discloses an intelligent LED illumination energy-saving optimization method and system based on an AI algorithm, and relates to the field of energy-saving optimization, and the method comprises the steps: inputting a multi-modal data set into a multi-task neural network model, carrying out the abnormal energy consumption detection, recognizing the abnormal energy consumption, generating a photon space distribution optimization instruction through fault root cause reasoning, and carrying out the optimization of the photon space distribution. Based on the photon space distribution optimization instruction, a U-Net generator in the generative adversarial network is adopted to obtain a photon space distribution probability graph, physical simulation is carried out through differentiable ray tracing, an optimal lens curvature parameter is obtained, and multi-target lighting effect optimization and thermal management cooperative control is carried out through a dynamic light field regulation and control algorithm; according to the method, through closed-loop linkage of data fusion, physical modeling and intelligent optimization, the energy efficiency utilization rate of an LED system is improved, the service life of a lamp is prolonged, and meanwhile the personalized illumination requirement and the thermal management requirement in a complex scene are met.