180 results about "Back propagation algorithm" patented technology

The invention relates to the technical field of intelligent communication, and discloses an end-to-end optimization method, a signal constellation geometry and probability joint shaping optimization method in an end-to-end intelligent communication system, and a communication device based on a neural network. According to the method, a transmitter is constructed by using a neural network, and a received signal passes through a two-stage equalization processing system to form an end-to-end optimization framework without assistance of a channel model. The transmitter adopts a modular design, can realize joint optimization of constellation probability shaping (PS) and geometric shaping (GS), and performs joint updating on all trainable parameters of the system based on a total loss function containing an equilibrium error and a decoding error through a back propagation algorithm. According to the invention, computing resources are effectively saved; the two-stage equalization system can accurately compensate channel damage, the realized PS and GS joint optimization can significantly improve the performance of the communication system and reduce the bit error rate, and the method has good performance advantages and practical value in an actual optical fiber communication system.

The invention provides a PCB manufacturability intelligent analysis and early warning method and system based on artificial intelligence, and the method comprises the steps: collecting and marking the multi-source time sequence process parameter data in the PCB design and manufacturing process under working conditions, building a dynamic causal graph model with direction and time lag marks through a sliding window and standardization processing by applying a causal discovery algorithm, and carrying out the calculation of the dynamic causal graph model. Dynamic expression of causal relationships among process variables is realized; when manufacturing abnormity is detected, abnormity attribution is carried out by combining a Bayesian back propagation algorithm, high-contribution-degree root dependent variables are screened, the effectiveness of root causes is verified through virtual intervention simulation and statistical test, and finally verification results and a causal mode are stored in a knowledge base to support subsequent rapid matching and reasoning. According to the method, the accuracy, efficiency and interpretability of PCB manufacturing abnormity attribution are improved, and process optimization and preventive intervention are facilitated.

The invention relates to the technical field of ship design optimization, and discloses a hull shape optimization method based on neural network modeling. In the data acquisition stage of the method, initial appearance parameters and hydrodynamic performance data of a ship body are obtained, the appearance parameters comprise geometric dimensions and shape features, and the performance data comprise resistance coefficients and wave-making resistance values. In the neural network construction stage, a neural network model with a multi-layer perceptron structure is trained by using collected data, weights are updated through a back propagation algorithm, and a nonlinear mapping relation between appearance parameters and hydrodynamic performance indexes is established. In the shape optimization stage, the trained neural network model is used for carrying out iterative adjustment on the shape of the ship body, fluid dynamic performance indexes are recalculated through the model after each adjustment until preset convergence conditions are met, and finally optimized ship body shape data are output. According to the method, partial complex calculation is replaced by the neural network, and intelligent optimization of the hull appearance is realized.

The invention relates to a four-body coupling vehicle base vibration noise prediction optimization method and system, and relates to the technical field of rail transit operation and maintenance optimization, and the method comprises the steps: obtaining a vehicle base multi-physics field correlation monitoring data set; carrying out system level assembly in a multi-physics field co-simulation environment to generate a coupled system state parameter matrix; generating sound ray bending trajectory data; loading the vehicle operation condition time sequence data to a wave equation solver, and outputting vibration noise propagation characteristic field distribution data; building a network topology structure to form a prediction model, and executing a back propagation algorithm to train network parameters to a convergence state; designing a vehicle scheduling scheme chromosome coding structure to generate a candidate solution population; and performing selection, crossover and variation genetic operations on the primary solution set, and outputting a vehicle base operation and maintenance scheduling real-time optimization scheme. The method has the beneficial effects that the precision and generalization ability of vibration noise prediction in the future time period are greatly improved, and the fundamental transformation from passive vibration isolation treatment to active prediction optimization is realized.

The invention relates to the technical field of laser control, in particular to an intelligent control method and system of an MOPA laser. The method comprises the following steps: constructing a target training data set and a one-dimensional convolutional neural network model; training a one-dimensional convolutional neural network model based on the target training data set in combination with a target loss function and a back propagation algorithm to obtain a target model; acquiring current operation parameters of the MOPA laser; inputting the current operation parameters into a target model, and outputting target regulation and control parameters by the target model; the target regulation and control parameters comprise the working current of a pumping module in the MOPA laser and the working temperature of a doped fiber; based on the target regulation and control parameters, the working current of a pumping module in the MOPA laser and the working temperature of a doped optical fiber are adjusted, the working temperature of the doped optical fiber comprises a plurality of different target temperatures, and each target temperature corresponds to a different position on the doped optical fiber. In this way, the problems caused by the nonlinear effect generated under high-power pumping can be reduced.

The invention discloses a flood peak enhanced physical base flow and residual error correction collaborative runoff prediction method, which belongs to the field of hydrological prediction, and comprises the following steps of: dividing a training set and a verification set according to a proportion, performing oversampling processing on flood peak samples, and constructing a time sequence window; a Xinanjiang model is discretized and expressed by adopting an ordinary differential equation, rainfall and potential evaporation data are input, and intermediate variables are obtained. A physical base flow and residual error correction dual-channel module is constructed, and physical base flow and residual error correction is calculated through two full-connection networks. And calculating a final runoff predicted value by adopting a residual connection structure, taking basic NSE loss as a core, superposing a flood peak sample error weighted item, strengthening flood peak fitting precision, and updating physical parameters and neural network weight through a back propagation algorithm. And verifying the model, and respectively calculating prediction indexes of the training set and the verification set. According to the method, fusion of a traditional hydrological model and a deep learning method is realized, the physical interpretation of the model is enhanced, and the basin runoff prediction precision is improved.

The invention discloses a myopia image deep learning recognition model training method, particularly relates to the technical field of medical image processing and deep learning, and is used for solving the problem that an existing deep learning model lacks anatomical structure priori knowledge guidance in myopia eye bottom image analysis. The method comprises the following steps: acquiring a myopia eye bottom image and anatomical structure priori knowledge data, extracting a multi-scale feature map by using a deep learning model, analyzing the geometric morphology of a key anatomical component based on standard spatial relationship information, and generating a spatial constraint loss item; according to the method, key anatomical path topology coherence is evaluated based on topology connection information, topology constraint loss items are generated, a loss item fusion strategy is dynamically adjusted according to a training stage, finally, a model is iteratively trained to convergence through a gradient back propagation algorithm, and organic combination of medical priori knowledge and a deep learning model is realized. And the clinical rationality and reliability of model output are improved.

The invention provides a vertical furnace body temperature layout recommendation method, and belongs to the technical field of intelligent recommendation, and the method comprises the steps: collecting the internal temperature of a furnace body and the current production process parameters in real time through sensors distributed at different positions of the furnace body, and carrying out the cleaning and preprocessing of the collected data, dividing the preprocessed data into a training set and a test set according to a preset proportion; in combination with a back propagation algorithm and a gradient descent optimization algorithm, enabling the model to learn a complex mapping relation between each factor in the furnace body and the temperature layout; performing feature extraction on test set data by using the trained model, and obtaining key features of temperature distribution in the furnace body through multi-layer convolution and pooling operation of the model; and according to the key features and a preset temperature layout optimization target, an intelligent optimization algorithm is adopted to carry out optimization calculation on the furnace body temperature layout, and a recommended temperature layout scheme is generated. And the temperature control precision of the furnace body is obviously improved.

The invention discloses a satellite signal authentication method and device based on a complex valued neural network and a storage medium. The satellite signal authentication method comprises the following steps: acquiring a complex valued data sequence of a satellite signal to be authenticated; inputting the complex-valued data sequence of the to-be-authenticated satellite signal into a trained complex-valued neural network model to enable the trained complex-valued neural network model to output the radio frequency feature vector of the to-be-authenticated satellite signal, the trained complex-valued neural network model being obtained by training based on a preset triple loss function and a back propagation algorithm; calculating a target similarity score between the radio frequency feature vector of the satellite signal to be authenticated and the corresponding target anchor point sample, wherein the score is used for representing an average angular distance between the radio frequency feature vector and the target anchor point sample; and classifying and authenticating the satellite signal to be authenticated based on a preset score threshold and the target similarity score. According to the method, the recognition complexity can be reduced while the recognition precision of the satellite signal is improved.

The invention discloses a three-dimensional electromagnetic inverse scattering imaging method of a diffusion model based on physical constraint embedding. The method comprises the following steps: 1, in a specific three-dimensional electromagnetic inverse scattering imaging environment, defining a dielectric constant of an imaging area, arranging a transmitting and receiving antenna array at the periphery, and obtaining scattering field data of a target area for constructing an input and output sample pair of a diffusion model; 2, constructing a diffusion model on the three-dimensional grid space; 3, introducing a physical consistency constraint term based on a Maxwell equation set in the training stage of the diffusion model; and 4, synthesizing the generation loss and the physical constraint loss of the diffusion model, constructing a joint optimization objective function, and carrying out iterative updating on network parameters by adopting an end-to-end back propagation algorithm until the model converges. According to the method, the spatial resolution and continuity of three-dimensional imaging are remarkably improved, the stability and robustness of inverse scattering reconstruction are improved, and the physical consistency is enhanced. And high-precision and interpretable three-dimensional electromagnetic inverse scattering imaging can be realized.

The invention discloses an end-to-end scheduling method and system for a multi-source weather-driven urban integrated energy system. The method comprises the following steps: constructing an integrated energy system model; constructing a source load prediction model based on multi-source numerical weather forecast data in combination with historical photovoltaic output data and power load and thermal load data; the method comprises the following steps: establishing a comprehensive energy system optimization scheduling model with minimization of system operation cost as an optimization target, designing a differentiable optimization layer, and reversely transmitting the gradient of the optimization target in the scheduling model to a source load prediction model parameter to the source load prediction model through a back propagation algorithm by the differentiable optimization layer, the parameters of the driving source load prediction model are updated, and end-to-end linkage optimization from prediction to scheduling is achieved; and periodically obtaining updated multi-source numerical weather forecast data and source load data, readjusting prediction model parameters, and executing optimization solution of the integrated energy system optimization scheduling model. According to the invention, cooperative training and iterative optimization of the prediction model and the scheduling decision are realized.

The invention relates to the field of water turbines, and provides a water turbine speed regulation system PID parameter optimization method and device, and the method comprises the steps: evaluating whether the system performance under the current operation condition meets the performance requirements or not based on a digital twin model of a water turbine speed regulation system; if not, optimizing the PID parameter corresponding to the current operation condition on the digital twin model by applying a trained BP neural network to obtain an optimized PID parameter; the BP neural network is obtained by training through an error back propagation algorithm based on the sample signal and the corresponding expected output. According to the PID parameter optimization method and device for the water turbine speed regulation system, an online optimization mechanism combining the digital twin model and the BP neural network is introduced, a group of adaptive new parameters can be rapidly and automatically output according to the current working condition information, calculation depending on artificial experience or a fixed formula is replaced, and the calculation efficiency is improved. And real-time and accurate tracking optimization of the complex time-varying system is realized.

The invention discloses an arch bridge tensioning method and system based on a physical constraint space-time diagram attention network, and the method comprises the steps: constructing a dynamic topological graph of an arch bridge cantilever construction structure, taking a structure initial state parameter, a control force system parameter and a real-time environment parameter as node input features, and taking a structure state response parameter as an output label, generating a training data set based on finite element model simulation; designing a deep learning network model combined with space diagram attention and time sequence causal convolution, and modeling a space-time dependency relationship; defining a mixed loss function of the data fidelity item and the physical residual loss item of the structural mechanical control equation; and training the model through a back propagation algorithm and deploying the model to a field decision control system. According to the method, through graph structure abstraction, spatio-temporal joint modeling and physical constraint embedding, multi-source monitoring data driven structure state online prediction and tension control strategy intelligent optimization are achieved.

The invention discloses a double-branch lithium battery health estimation method based on multi-source feature fusion. The method comprises the steps that key operation data in the charging and discharging process of a lithium battery are collected through a battery management system or a test platform; carrying out multi-source health factor extraction on the collected key operation data of the lithium battery; performing data preprocessing on the extracted multi-source health factors; searching an optimal section in a reasonable range of the charging voltage by using a particle swarm optimization algorithm improved based on a quantum mechanics principle; in the determined optimal voltage section, the corresponding multi-source characteristic quantity is trained, a parallel double-branch network (CMNN) is established, parameters are optimized through a back propagation algorithm, and a QPSO-CMNN lithium battery state estimation model is obtained; according to the method, the optimal voltage section is searched by using the QPSO, the method adapts to actual data missing and actual use characteristics of a user battery, and the method has higher SOH prediction precision and robustness.

The invention relates to a lower limb function rehabilitation evaluation method based on a multi-dimensional feature fusion network, and belongs to the technical field of intelligent evaluation of lower limb function rehabilitation conditions. The method comprises the following steps: acquiring multichannel surface electromyogram signals of a stroke patient and preprocessing the multichannel surface electromyogram signals to construct a data set; constructing a lower limb function rehabilitation evaluation model which comprises a time domain feature extraction branch, a frequency domain feature extraction branch, a first cross attention module, a second cross attention module and a classifier; inputting samples in the data set into a lower limb function rehabilitation evaluation model, and training the model; optimizing the model by adopting a loss function, updating parameters by using an AdamW optimizer, and minimizing loss through a back propagation algorithm until the model converges to obtain a trained model; and preprocessing the to-be-evaluated surface electromyogram signal, and inputting the preprocessed to-be-evaluated surface electromyogram signal into the trained model to obtain an evaluation result The accuracy of lower limb function rehabilitation evaluation can be improved.

The embodiment of the invention provides a calibration data acquisition method, a model quantification method, equipment, a storage medium and a computer program product, and is applied to the technical field of model quantification. Specifically, firstly, a Gaussian noise image conforming to Gaussian distribution is generated; then, segmenting the Gaussian noise image into a plurality of image blocks, inputting the plurality of image blocks into a visual Transform model, and obtaining an attention map output by an attention module included in an encoder in the visual Transform model; determining an attention alignment loss value based on the attention map and a predetermined attention priori map; and updating the Gaussian noise image according to the attention alignment loss value and a back propagation algorithm. According to the method, Gaussian noise image updating operation is executed through iteration for multiple times, the Gaussian noise image updated through last iteration serves as a forged image, the forged image serves as calibration data and is added into a calibration data set, and therefore possibility is provided for the calibration data set needed by visual Transform model quantification.

Methods and systems for training a neural language model. Clean sequence pairs are received including clean source and target sequences. For each clean sequence pair, a noisy version is sampled with an adversarial generator to generate a noisy sequence pair. Parameters of the neural language model are optimized on the clean and noisy sequence pairs. Parameters of the adversarial generator are optimized to minimize a modeling loss of the adversarial generator and maximize a neural language loss of the neural language model using backpropagation.

The application discloses a kind of double-branch lithium battery health estimation methods of multi-source feature fusion, including the key operation data of lithium battery in the process of charging and discharging is collected by battery management system or test platform;The key operation data of lithium battery collected is extracted to multi-source health factor;The data preprocessing is carried out to the multi-source health factor extracted;Using the particle swarm optimization algorithm based on the principle of quantum mechanics is improved, the optimal section is searched in the reasonable range of charging voltage;In the determined optimal voltage section, the corresponding multi-source characteristic quantity is trained, the parallel double-branch network CMNN is established, the parameters are optimized by back propagation algorithm, and the QPSO-CMNN lithium battery state estimation model is obtained;The final model is used to estimate test set lithium ion battery SOH, the application uses QPSO to search optimal voltage section, adapts actual data missing and user battery actual use characteristics, with higher SOH prediction accuracy and robustness.

The invention discloses a parameter optimization method and system based on an agent model and multi-algorithm gradient collaboration, and relates to the technical field of nuclear engineering and nuclear security, and the method comprises the steps: firstly determining a nuclear facility parameter combination, carrying out the high-fidelity calculation through a Monte Carlo program to obtain a Keff value, and constructing a training set; training a micro-proxy model capable of predicting a Keff value and gradient information through a feedforward neural network in combination with a back propagation algorithm, taking the micro-proxy model as a target function, taking the minimum absolute error of a Keff predicted value and a target value as an optimization target, and generating candidate parameter solutions through global exploration of a hybrid optimization algorithm; and an optimal parameter combination candidate solution is obtained through gradient-assisted local optimization, and finally, through Monte Carlo program verification, if the error reaches the standard, the optimal parameter combination candidate solution is used as a final parameter combination solution. The problem that a parameter optimization method in the prior art cannot give consideration to calculation efficiency and precision is solved.

The invention relates to the technical field of data storage, and discloses a method and system for optimizing test data collected by a memory chip based on machine learning, and the method comprises the steps: collecting original test data from a test platform of a to-be-tested memory chip; constructing a convolutional neural network model, learning a mapping relationship between test data features and memory defects through the convolutional neural network model, training the convolutional neural network model through the collected original test data, optimizing parameters of the convolutional neural network model according to a training result through a back propagation algorithm, and obtaining memory defects through the optimized parameters. The performance of the convolutional neural network model tends to be stable; converting a parameter vector output by the convolutional neural network model into a test vector executable by ATE; and taking the generated test vector as a new training sample, feeding back the new training sample to an original test data updating data set, and iteratively training and optimizing the convolutional neural network model. According to the invention, intelligentization of the test process is realized, and hidden defects sensitive to voltage and time sequence can be effectively captured.

The invention discloses a pressurized water reactor steam generator liquid level control method and system, and belongs to the technical field of nuclear power station control. The method comprises the steps that a liquid level set value and a measured value of the steam generator are obtained, and deviation is calculated; inputting the set value, the measured value, the deviation and the bias constant into a preset BP neural network; the network dynamically outputs three coefficients of proportion, integral and differential through forward propagation to serve as real-time parameters of a PID (Proportion Integration Differentiation) controller; the PID controller calculates and outputs a water supply flow demand value, and a water supply valve is adjusted after conversion; and calculating a root-mean-square error of liquid level control as a loss function, and updating the weight of the neural network on line by using the new liquid level deviation through a back propagation algorithm to realize adaptive optimization of PID parameters. According to the invention, the self-learning ability of the BP neural network is combined with the reliability of PID control, the control quality and stability of the liquid level of the steam generator under various power levels are significantly improved, the system can be deployed in a power plant DCS through modular packaging, and the engineering practicability is high.

The invention provides a permanent magnet synchronous motor multi-parameter identification method and system based on a physical information neural network. The method comprises the following steps: constructing a physical information neural network model; constructing a composite loss function of the physical information neural network model, wherein the composite loss function comprises a data fitting item and a physical constraint item; training the physical information neural network model according to the operation data of the permanent magnet synchronous motor under different working conditions, and minimizing a composite loss function through an optimization algorithm; in the training process, the stator resistance, the d-axis inductance, the q-axis inductance and the permanent magnet flux linkage of the permanent magnet synchronous motor serve as multiple parameters of the motor to be identified and are merged into calculation of a physical constraint term, the multiple parameters of the motor to be identified are synchronously updated through a back propagation algorithm till the composite loss function is converged, and the identification result is obtained. And determining the motor multi-parameter to be identified at the moment as a motor multi-parameter identification result. The method is supported by the Zhejiang Scientific and Technology Project (2025C01199 (SD2)) of Zhejiang Province 'Scientific and Technical Project' of 'Scientific and Technical Project

The application discloses a kind of multi-camera automatic layout method in multi-vision measurement system, it includes: step S1: initialization camera layout parameter;According to the CAD model information of workpiece to be detected, the initial position of camera is obtained, and then the initial parameters of all cameras are obtained;Step S2: optimization camera initial layout based on simulated annealing algorithm;After obtaining the initial parameters of all cameras, iterative optimization is carried out using simulated annealing algorithm, and the optimal solution satisfying the constraint condition is obtained;Step S3: based on the differentiable adjustment promotion, the measurement accuracy and robustness of three-dimensional reconstruction under the above simulated annealing camera layout are improved.Based on the optimal layout obtained based on simulated annealing, the error back propagation algorithm based on stochastic gradient descent is used, and the Euclidean distance between reconstructed three-dimensional coordinates and CAD model coordinates is used as a loss function to optimize, to obtain a camera layout with better reconstruction robustness.The application has the advantages of high automation, strong layout practicality, good accuracy and robustness of three-dimensional reconstruction, etc.

The application discloses a lining cloth dyeing control method and system based on artificial intelligence, and is used for the control field, and the method comprises the following steps: real-time monitoring of multi-source data in the lining cloth dye vat, capturing the video stream inside the lining cloth dye vat; using a Gaussian mixture variational autoencoder model to extract features from the multi-source data, obtaining multi-source feature data; obtaining the changing moment of the fabric state in the video stream, and extracting the fabric state change feature data according to the observed fabric color change and the heterogeneity of the dye distribution in the video stream; fusion of multi-source feature data and fabric state change feature data; introducing a model predictive control layer, predicting the result to adjust the dyeing parameters in real time; establishing an intelligent feedback mechanism. The Gaussian mixture variational autoencoder model is trained and optimized through the small-batch stochastic gradient descent and back propagation algorithm, important features are learned and captured from the data.

The invention discloses a passive underwater maneuvering target state estimation method based on a long short-term memory network, a program, equipment and a storage medium, and belongs to the technical field of underwater acoustics. The method comprises the following steps: simulating a maneuvering track of a target by adopting a coordinated turning state space model to generate a motion state; performing normalization and time step sampling on the passive acoustic measurement data acquired by the multiple observers, and constructing a time sequence input sequence; inputting the time sequence data into a multi-layer LSTM network to capture a time dependency relationship and nonlinear characteristics of target motion; and training the network through a time back propagation algorithm, and minimizing a mean square error between a state estimation value and a true value. According to the method, verification is carried out in a complex maneuvering scene, and comparison with an interactive multi-model extended Kalman filter and an interactive multi-model unscented Kalman filter is carried out. Results show that the time sequence learning capability of the tracking system can be improved, and the robustness and accuracy of state estimation are enhanced in noise interference and dynamic change environments.

The application provides a method and device for training and post-quantization of an image super-resolution model based on corrector routing, comprising: training a corrector group by using a gradient back propagation algorithm according to the weight increment, pre-training weight and quantization function of each corrector in the corrector group, obtaining optimized corrector parameters and quantizer parameters; based on the optimized corrector parameters and quantizer parameters, performing static weighting coefficient optimization on each quantization module by using a calibration image dataset, calculating an optimal weight increment according to the optimal static weighting coefficient, and generating a static routing table; enhancing the pre-training weight of the image super-resolution model by using the optimal weight increment, and calling each quantization module to pre-quantize the enhanced weight; and inputting a to-be-processed image into the model based on the pre-quantized weight for processing, and restoring a quality-enhanced image. In this way, the information loss caused by low-bit quantization is effectively reduced, and the ability of the quantized image super-resolution model to restore fine image details is significantly improved.

The application discloses an intelligent traffic flow prediction method combined with an improved BWO-BP neural network algorithm, which comprises data preprocessing, construction of a multidimensional input feature set, optimization of the BWO algorithm by using multiple strategies, improvement of the global search capability of the model, initialization of the connection weight and threshold value of the BP neural network by using the improved white whale optimization algorithm, improvement of the initial solution quality of the model, construction of a feedforward neural network model, network training by using an error back propagation algorithm, future traffic flow prediction based on the trained network model, and realization of efficient and intelligent traffic management application. The application has stronger adaptability and higher prediction stability when dealing with complex and variable traffic scenes, can more comprehensively improve the grasping capability of the overall system for traffic trends according to the flow characteristics of different types of vehicles, has higher calculation efficiency, fast system response speed, and good practicability and deployment flexibility.

The application discloses a kind of based on adaptive noise reduction tool damage gray image intelligent classification method, mainly including a network consisting of adaptive noise reduction module and a tool damage image classification module;Noise reduction module and classification module are simultaneously accepted end-to-end joint training, share and optimize network parameters, while increasing a balance parameter in loss function, the parameter is according to the classification result of feedback of back propagation algorithm adaptive optimization noise reduction level, finally reach the optimal classification performance.The application can automatically identify whether the image contains noise and batch solve the classification problem of image noise for tool damage gray image, improve the prediction ability of model for difficult samples through adaptive noise reduction mode, can effectively remove image noise, while reducing the amplification effect of corresponding loss function to anti-noise, finally improve the intelligent classification ability of tool damage gray image from image processing quantity, image quality and prediction efficiency.

This invention discloses a multimodal collaborative denoising product recommendation method based on modality balance. The method first constructs behavior-aligned multimodal semantic encoding and projects it onto the behavior semantic space to obtain a multimodal feature product sequence. Next, for the multimodal feature product sequence, a multimodal-aware collaborative denoising module is constructed to collaboratively filter noise from each modality, resulting in a denoised intermediate product sequence. By introducing positional encoding in conjunction with the collaborative denoising module, an enhanced product sequence is obtained. Finally, based on the enhanced product sequence, cross-modal fusion weights are generated, prediction scores are calculated, and the product with the highest score is recommended. A joint loss function is constructed, and the global parameters are iteratively updated using a backpropagation algorithm. This invention effectively solves the problems of noise interference and modality learning imbalance in multimodal recommendation, suppresses the excessive dominance of strong modalities in the early stages of training, and improves the accuracy and robustness of product recommendations.

The invention discloses a gas turbine residual life prediction method and system based on an improved TTAO algorithm, and relates to the technical field of gas turbine residual life prediction, and the method comprises the steps: evaluating a sensor data contribution value through a random forest algorithm, screening key features, building an improved Autoformer model fusing a parallel multi-scale convolution kernel and an autocorrelation mechanism, and carrying out the prediction of the residual life of a gas turbine. A weighted loss function of JS divergence and a mean square error is constructed, an improved TTAO algorithm based on an adaptive disturbance balance strategy and a nonlinear tangential flight strategy is adopted to adjust model hyper-parameters, a prediction error is minimized through a back propagation algorithm, and finally high-precision residual life prediction is achieved. According to the method, redundant noise is reduced through feature screening, a dynamic convolution kernel is utilized to adapt to nonlinear degradation features, prediction precision and distribution consistency are balanced through a mixed loss function, convergence is accelerated by adopting a self-adaptive optimization algorithm, and the health management level of the gas turbine under complex working conditions is remarkably improved.