176results about How to "Simplify the training process" patented technology

The invention provides a frequency converter IGBT junction temperature prediction method based on a neural network, and relates to the technical field of neural networks, and the method comprises the steps: obtaining a circuit structure and operation parameters of a subway traction transmission system, and constructing a dynamic thermal network model; establishing a simulation working condition set, performing simulation to acquire a real-time parameter set, and establishing a sample data set after preprocessing; a neural network with multiple input layers is established, training is executed to construct an IGBT junction temperature prediction model, each input layer is mapped with a feature extraction sub-network, feature extraction is performed by using the feature extraction sub-networks, and neural network training is performed according to a feature extraction result; and performing data prediction by using the IGBT junction temperature prediction model, and establishing a junction temperature time sequence. The technical problems that dynamic changes of the junction temperature under complex operation conditions are difficult to capture in the prior art, and particularly when load fluctuation is large or multiple working conditions are switched, prediction errors of a traditional model are remarkably increased, and consequently the accuracy of junction temperature prediction is insufficient are solved.

The invention discloses an AE-GCN-based unmanned aerial vehicle system fault diagnosis method and equipment, belongs to the technical field of unmanned aerial vehicle fault diagnosis, and is used for solving the technical problems that an existing unmanned aerial vehicle fault diagnosis method depends on expert experience and feature engineering and has the defects of low efficiency, poor generalization ability and the like. The method comprises the following steps: constructing a corresponding structure topological graph based on a topological structure of a target unmanned aerial vehicle, and obtaining a topological matrix; performing node feature extraction on the structure topological graph to obtain a node feature vector; inputting the topological matrix and the node feature vector into a graph convolution neural network for graph convolution operation to obtain a node state feature vector; inputting the data into an attention mechanism module to obtain attention features, and mapping the attention features into node state updating parameters; comparing the node state updating parameter with the node state historical parameter, and positioning a fault node and a fault type; and predicting a fault propagation path based on other nodes directly connected with the fault node in the structural topological graph.

The invention provides a multi-degraded image restoration method based on self-adaptive prompt. The method comprises a two-stage hidden space diffusion model, and various degraded images can be flexibly processed. Specifically, the model generates a degradation cue by learning the difference between a degraded image and a clean image, and encodes the cue information in a hidden space. In the first stage, a degradation prompt is obtained through a degradation adaptive prompt learning mechanism as a key feature for guiding the diffusion model in the second stage to carry out image restoration. Finally, prompt information is integrated into a Fourier transform restoration network, and accurate restoration of different types of degraded images can be realized. The method not only improves the accuracy of image restoration, but also significantly reduces the calculation complexity, and is suitable for image restoration tasks in various real world scenes, including denoising, deblurring, defogging, rain removing, snow removing, raindrop removing and low illumination enhancement.

The invention provides a brain tumor image accelerated diffusion network segmentation method based on fuzzy learning, and belongs to the technical field of medical image intelligent segmentation. The technical problem of low segmentation accuracy caused by fuzzy texture and boundary disappearance of the brain tumor image is solved. According to the technical scheme, the method comprises the following steps of S1, performing data preprocessing on a brain tumor image data set; s2, performing pre-segmentation on the brain tumor image by using a 3D U-Net network, and performing iterative noise addition on the obtained segmented image to enable the segmented image to become a pure noise image; s3, performing iterative denoising on the pure noise image, and learning a denoising process by using a fuzzy U-Net network; and S4, stopping the denoising process in advance, and segmenting the segmented image by using a 3D U-Net network to obtain a final segmentation result. The method has the advantages that the segmentation accuracy is high, decision support is provided for segmentation of the brain tumor image, and convenience is provided for diagnosis of doctors.

The invention provides a vehicle track prediction method and device suitable for different driving styles, and relates to the technical field of automatic driving. The vehicle trajectory prediction method suitable for different driving styles comprises the following steps: acquiring a sample data set of an original vehicle trajectory; filtering the sample data set to obtain a filtered sample data set; performing dimension reduction processing by adopting a principal component analysis method to obtain dimension-reduced data; carrying out soft clustering processing by adopting a fuzzy C-means clustering algorithm to obtain a hard tag; constructing an initial global-local attention mechanism-based residual bidirectional long-short term memory network; training the initial network to obtain a trained bidirectional long-short term memory network based on the global-local attention mechanism residual error; acquiring state information of the target vehicle in the observation time domain; and inputting the state information into the trained network to obtain the future trajectory of the target vehicle in the prediction time domain. According to the invention, the accuracy of vehicle trajectory prediction can be improved.

The invention discloses a visual detection optimization method and system based on data set classification. The method comprises the steps of introducing a scene classification label and carrying out standardization processing on scene features in a data set labeling and preprocessing stage, and carrying out scene classification multi-task learning in a model selection and training stage, so that a model dynamically adjusts feature learning in different scenes, and the multi-scene adaptive capacity is improved. And in the model evaluation and verification stage, scene-based hierarchical evaluation is introduced and corresponding optimization is carried out. In the model deployment and reasoning stage, a dynamic scene is introduced to adapt to reasoning, and the reasoning path of the model is dynamically adjusted to adapt to the characteristics of the scene, so that the reasoning accuracy is improved. In the later optimization, scene feature clustering and transfer learning are used, and the multi-scene performance of the model is optimized. According to the method, under the condition that the model is not jointly trained through multiple data sets, the multi-scene adaptive capacity of the visual detection model is improved through data set classification.

The invention discloses a drug molecule optimization method based on a hidden space diffusion model. The method comprises the steps of 1, constructing drug molecule data sets for different tasks; 2, constructing and pre-training a molecular SMILES auto-encoder; 4, constructing and training a molecular optimization model based on the hidden space diffusion model; according to the method, new molecules which have relatively high similarity with original molecules, have better attributes and are more diversified can be generated, so that the optimization efficiency of multiple attributes of drug molecules is improved.

The invention relates to a speech recognition method, device and equipment based on an end-to-end cross-language large model. The method comprises the following steps: constructing a cross-language large model; obtaining a training data set formed by multiple groups of voice text label pairs, and inputting the training data set into the cross-language large model for end-to-end training until a trained cross-language large model is obtained; a to-be-recognized voice signal is input into the trained cross-language model, and filter feature and wav2vec2 feature extraction and splicing, text embedded feature representation, dimension alignment, splicing and decoding translation are carried out through a voice coding module, a text coding module, a feature alignment module and a decoding module in the model in sequence. And outputting and obtaining the speech recognition text of the target language. By adopting the method, the accuracy and robustness of speech recognition can be improved, and cross-language speech recognition is realized.

The invention belongs to the technical field of medical images and deep learning, and particularly relates to a non-paired cross-modal medical image conversion method based on contour guide path regularization, which comprises the following specific steps: constructing a conversion path from a source domain to a target domain; a path regularization method is applied to smooth path transition and ensure the stability in the image conversion process; according to a path regularization technology based on contour guidance, the model can be focused on a key anatomical region, the structural consistency and accuracy of the generated CT image are further improved, and the training time can be remarkably shortened by reducing additional training steps. The invention aims to simplify the unsupervised training normal form and overcome the defects of the traditional GAN-based single-step generation method in the aspect of capturing the complex anatomical structure of the medical image. While the training time and the sampling time are considered, the problem that the quality of the generated medical image is limited is overcome.

The invention discloses a pathological image self-supervised dyeing normalization method and device, and belongs to the technical field of combination of digital pathology and image processing, and the method comprises the steps: constructing a pathological image database; a dyeing normalization model is designed, the dyeing normalization model comprises a public feature domain extraction module, a content perceptron, a scene perceptron and a fusion module, and the dyeing normalization model can perform color conversion on the pathological image to obtain a dyed normalized image; a self-supervised loss function is constructed based on the pixel level difference of the input original pathological image and the dyed and normalized image, and a dyeing and normalization model is trained. The model can perform adaptive dyeing style normalization on any image of an unknown pathological data center towards a selected pathological data center image.

The invention discloses a drill jumbo working hour prediction method and device, electronic equipment and a storage medium, and belongs to the technical field of working hour prediction, and the method comprises the steps: updating an original particle swarm position formula of an SABO algorithm according to Cubic chaotic mapping, and generating a first SABO algorithm; updating an initial iteration formula of the first SABO algorithm according to Piecewise chaotic mapping, and generating a second SABO algorithm; updating a position iteration formula of the second SABO algorithm according to a golden sine algorithm, and generating an optimal GSABO algorithm; performing iterative training on an initial neural network model based on the optimal GSABO algorithm and the index data to obtain an operation man-hour prediction model; and performing drill jumbo working hour prediction on the target operation based on the working hour prediction model. According to the method, the model parameters are determined by optimizing the SABO algorithm through the two chaotic mapping and golden sine algorithms, and the accuracy and efficiency of predicting the working hours of the drill jumbo are improved.

The invention discloses an air jet loom auxiliary jet air consumption prediction method and device, equipment and a medium, and belongs to the technical field of industrial automation. The method comprises the following steps: obtaining to-be-measured structure parameters of a to-be-measured nozzle in the air jet loom, wherein the to-be-measured structure parameters comprise an inlet diameter, a straight pipe diameter, a nozzle taper, an elliptical outlet long axis and an elliptical outlet short axis; the to-be-measured structure parameters are input into an air flow prediction model, the nozzle air consumption, output by the air flow prediction model, of the to-be-measured nozzle is obtained, and the air flow prediction model is obtained based on sample structure parameters of the sample nozzle and air consumption label training. According to the method, the gas consumption of the nozzle can be accurately predicted in a complex industrial environment.

The invention provides a method and a device for generating a high-order light beam through physical model driven deep learning. The device comprises the following steps: firstly, establishing a non-training phase imaging physical model, loading phase modulation through a plane wave, carrying out Fourier transform by using a Fourier lens, and collecting light field intensity distribution on a rear focal plane through a detector; phase distribution and light field intensity distribution meet the Fourier transform relation, and the Fourier transform relation can be regarded as the process from phase forward coding to diffraction intensity diagram. And then, inputting the acquired light field intensity distribution I into the neural network, and outputting a phase result. And calculating the intensity distribution corresponding to the phase distribution by using the Fourier transform theory, and measuring the error between the input intensity and the calculated intensity. In the whole process, training of the neural network can be completed only by collecting a single light field propagation intensity graph, dependence on a large amount of training data is remarkably reduced, and reconstruction precision and generalization ability of phase imaging are improved.

The invention belongs to the technical field of voice signal processing, and discloses a method and system for driving a virtual human by a voice signal, and the method comprises the steps: carrying out the resampling processing of original audio signal data, obtaining a voice signal, solving a linear prediction coefficient based on a dynamic framing strategy and a linear prediction analysis technology, and obtaining a linear prediction result; obtaining a voice feature sequence of the voice signal; performing enhancement processing on the voice feature sequence by using a dynamic feature enhancement technology, obtaining an enhanced feature sequence based on a processing result, and constructing an animation sequence prediction model by taking the enhanced feature sequence as input; and obtaining animation sequence parameters based on an output result of the animation sequence prediction model, and combining the animation sequence parameters with the virtual human role model to perform facial action synthesis and rendering operation. According to the invention, the earlier stage processing flow is greatly simplified, and the usability and applicability in the driving process are improved.

The invention provides an ocean state variable forecasting method based on thermocline adaptive loss, and belongs to the technical field of ocean meteorology. The self-adaptive thermocline loss function designed by the invention can dynamically adjust the weight of the loss function according to the intensity and complexity of the temperature change of the thermocline region, and the self-adaptive mechanism enables the model to pay more attention to the temperature gradient change characteristics of the thermocline region in the training process; therefore, the model is guided to learn and simulate the complex evolution process of the thermocline more accurately. Compared with the prior art, the traditional loss function gives the same weight to the temperature change of all regions, and the importance of the thermocline region cannot be highlighted, so that the model has a poor effect when learning and simulating the complex evolution process of the thermocline; according to actual verification, the accuracy and stability of thermocline forecasting are remarkably improved, meanwhile, the model training process is simplified, and a more efficient and reliable thermocline forecasting tool is provided for marine scientific research and practical application.

The invention discloses a zero-sample food image detection method based on visual semantic bidirectional guidance, which belongs to the technical field of image data processing and generation, takes DETR based on Transform as a detection framework, and fuses a food semantic enhancement network and a cross-modal bidirectional guidance module. Firstly, an image is input to an image backbone network branch to extract visual features, a category label is input to a food semantic enhancement network branch to extract semantic features, then the two obtained features are sent to a cross-modal two-way guide fusion module, the module comprises a two-way guide encoder, a two-way guide query vector and a two-way guide decoder, and the cross-modal two-way guide fusion module comprises a two-way guide query vector and a two-way guide decoder. Visual and semantic features are aligned and fused through coding and decoding processes, and finally, regression and classification tasks of zero sample food image detection are completed in a full connection layer. According to the method, detailed food attributes and visual features can be captured in the detection process, effective interaction of visual semantic modals is achieved, and therefore the detection precision of unseen food categories is improved.

The invention discloses an adjusting and optimizing method and system based on a large power dispatching model, and belongs to the technical field of power dispatching, and the method comprises the following steps: obtaining historical power dispatching knowledge, and dividing the historical power dispatching knowledge into regional knowledge and global knowledge based on regional characteristics; constructing a first teacher model based on the difference between the global knowledge and the regional knowledge, and constructing a student model based on the lowest fluctuation cost; obtaining regional scheduling correlation according to historical power scheduling knowledge, and constructing a second teacher model according to the regional scheduling correlation; associating the first teacher model, the second teacher model and the student model based on knowledge distillation, and constructing a large power dispatching model; and outputting a tuning strategy according to the large power dispatching model and the current power data. The beneficial effects of the invention are that the output of the power tuning strategy gives consideration to pertinence and efficiency.

The invention discloses a multi-mode brain-computer interaction method and system. The multi-mode brain-computer interaction system is composed of a signal acquisition module, a control module and a human-computer interaction interface module. The signal acquisition module is used for acquiring various signals, which can be used for man-machine interaction, of a user, including electroencephalogram, myoelectricity, electrooculogram, voice, posture, somatosensory human body signals, position, motion posture and temperature human body surrounding environment signals, and combining the channels to form a multi-channel parallel input information group; the information is transmitted to the control module in a wireless communication mode; and the control module performs digital filtering, signal processing, classification detection and fusion algorithm identification according to the acquired multi-mode signal, and outputs a corresponding instruction according to an identification result for corresponding interactive operation.

The invention discloses a homestead building change detection method, device and equipment. The method comprises the following steps: determining a target area; acquiring two remote sensing images of the target area; marking the homestead areas in the two remote sensing images to obtain a homestead vector file; the two remote sensing image maps are subjected to registration, the two remote sensing image maps after registration are cut based on the homestead vector file, a plurality of image blocks are obtained, and each image block only comprises a building of one homestead; generating a pairing table according to the plurality of image blocks, and randomly distributing a training set and a verification set in proportion based on the pairing table; and carrying out model training on a twin neural network by adopting the training set to obtain a homestead building change detection model, and realizing homestead building change detection. According to the method, the change condition of the buildings of the known homestead can be monitored, and the method can be well matched with an existing geographic database.

The invention relates to a model deployment method, a fault diagnosis method, a system, equipment and a medium. The method comprises the following steps: acquiring historical working data of the intelligent cabinet; preprocessing the historical working data; providing an initial fault diagnosis model; training the initial fault diagnosis model based on the training set data; performing verification optimization on the trained initial fault diagnosis model based on verification set data; testing the verified and optimized initial fault diagnosis model based on the test set data to obtain a to-be-optimized fault diagnosis model; performing compression optimization on the to-be-optimized fault diagnosis model to obtain a fault diagnosis model; configuring a cloud server; and deploying the fault diagnosis model in a cloud server. The training and deployment process of the fault diagnosis model can be greatly simplified, manual intervention is reduced, and when the deployed fault diagnosis model is used for fault diagnosis of the intelligent cabinet, the accuracy and timeliness of fault diagnosis can be improved, and the operation and maintenance efficiency of the intelligent cabinet is improved.

The invention relates to the technical field of wind turbine generator running state monitoring, in particular to a wind turbine generator input torque intelligent identification method and system and a storage medium, and the method comprises the steps: S1, constructing a training data set of state data of a wind turbine generator; s2, constructing an intelligent recognition model based on a radial basis function neural network; s3, inputting the training data set into the intelligent identification model to obtain a trained input torque intelligent identification model; s4, constructing a real-time data set of the state data of the wind turbine generator; and S5, inputting the real-time data set into the trained intelligent identification model to obtain an intelligent identification result of the input torque. According to the intelligent identification method for the input torque of the wind turbine generator, long-term identification and monitoring of the input torque of the wind turbine generator are achieved through an intelligent algorithm, strain gauges do not need to be installed and maintained, and the cost of long-term monitoring of the input torque of the wind turbine generator is reduced.

The invention provides a photovoltaic output prediction method considering haze / dust factors. The method comprises the following steps: firstly, preprocessing data, screening out abnormal data and invalid data, and complementing missing data; the method comprises the following steps: firstly, analyzing haze / sand dust data, then classifying weather data by using a decision tree algorithm, then deeply understanding an internal rule and an operation mechanism by analyzing a haze / sand dust formation mechanism and a photovoltaic power generation principle, determining main characteristics, and then firstly establishing a haze / sand dust influence model 1 and a photovoltaic power generation power prediction model 2 for the haze / sand dust data; the model 1 obtains influence parameters of haze / dust factors on the photovoltaic power generation power, and the model 2 outputs a predicted value of the photovoltaic power generation power. The advantages of mechanism analysis and a data driving model are combined, and the accuracy and reliability of the model are improved.

The invention belongs to the field of computer vision, and discloses a multi-target detection method, a multi-target detection model training method, an edge computing device and a medium, and the multi-target detection model training method comprises the steps: extracting a first-class feature map; respectively inputting the first type of feature map into a first target detection module, a multi-layer feature fusion module and a second target detection module; calculating a target loss function value based on the obtained target detection result and the target labeling result; and adjusting network parameters of the to-be-trained model until a preset model training convergence condition is met, and obtaining a multi-target detection model. According to the method, the generalization ability of the model is improved, and the driving safety is enhanced.

The DASAC-based micro-grid fault recovery method comprises the following steps: establishing a micro-grid fault recovery model considering DG output and load demand time variation; the method comprises the following steps: converting a fault recovery problem of a micro-grid into a constrained nonlinear programming problem, constructing a double-agent reinforcement learning model by taking the minimum fault recovery cost of the micro-grid as a target, and distributing discrete and continuous double agents for different action types in an action space for control; a DASAC method is provided, efficient self-adaptive training learning is carried out on the double-agent reinforcement learning model, and the fault recovery efficiency is improved to meet the operation requirements of the micro-grid. According to the DASAC-based microgrid fault recovery method provided by the invention, discrete and continuous actions are processed respectively, so that the fault recovery efficiency is improved, the training process is simplified, and the robustness and adaptability of the system are enhanced.

The present invention belongs to the technical field of tank fermentation, and specifically is a biological fermentation device for Chinese herbal medicine, comprising a tank cover, wherein a control unit is arranged at the axial position of the tank cover, a main unit is arranged on the control unit, and a cleaning unit is arranged on the main unit; the present invention breaks the limitations of the single form and solidification of traditional scrapers through the coordinated cooperation between multiple rubber shovels, that is, the single scraper is made segmented and flexible, and the multiple rubber shovels can make corresponding shape adjustments due to changes in the external environment when moving, so as to adapt to the same-position cleaning of the suddenly changed area on the inner wall of the tank body, which helps to improve the cleaning integrity and thoroughness of the residues on the inner wall of the tank body, and at the same time, through the telescopicity of the telescopic gas rod, the contact consistency between each rubber shovel and the inner wall of the tank body is adjusted autonomously to avoid damage to the inner wall of the tank body during the operation of the rubber shovel.

The invention provides a device, a method and equipment for small target image detection and classification. The device for small target image detection and classification comprises an image input module, a feature extraction module and a detection or classification module. The image input module is used for processing an acquired image and inputting the image to the feature extraction module; the feature extraction module is used for extracting the features of the image through a feature extraction model to obtain the coordinate information of the small target; and the detection or classification module is used for cutting out local feature blocks on the feature map according to the coordinate information of the small targets and inputting the local feature blocks into a decision head to obtain the feature information of all the small targets in the image. The method can avoid the repeated calculation of the same features, thereby reducing the calculation amount, and improving the utilization rate of calculation resources.

The invention belongs to the field of bioengineering, and particularly relates to an in-vitro culture method of CD8 < + > dry memory T cells, a product and application of the product. The method comprises the following specific steps: (1) coating, stimulating and activating T cells by using an anti-CD3 antibody and RetroNectin; (2) performing resting culture on the activated T cells in a resting culture solution for 0-60 hours; and (3) adding cell factors into the resting culture solution to perform large-scale multiplication culture on the T cells after resting. According to the method, the efficiency of obtaining the CD8 + TSCM cells through in-vitro culture is improved, the culture process is simplified, and the risk of microbial contamination in the culture process and the cost of obtaining the CD8 + TSCM cells are reduced.

The present application discloses a method, device, equipment and medium for predicting the auxiliary jet air consumption of an air jet loom, and belongs to the field of industrial automation technology. The method comprises: obtaining the structural parameters to be measured of the nozzle to be measured in the air jet loom, the structural parameters to be measured include the inlet diameter, the straight tube diameter, the nozzle taper, the elliptical outlet major axis and the elliptical outlet minor axis; inputting the structural parameters to be measured into the air flow prediction model, and obtaining the nozzle air consumption of the nozzle to be measured output by the air flow prediction model, and the air flow prediction model is obtained based on the sample structural parameters and air consumption label training of the sample nozzle. The method can accurately predict the air consumption of the nozzle in a complex industrial environment.

The utility model discloses a perioral muscle function training device, which relates to the technical field of medical instruments, and comprises a support ring, an upper lip accommodating part and a lower lip accommodating part, the support ring is an elastic support ring, and the support ring is used for being placed between teeth and lips; the upper lip containing part is fixedly connected with the supporting ring, the upper lip containing part is provided with an upper lip supporting face, the upper lip supporting face can make contact with the upper lip when the supporting ring is placed between the teeth and the lip, and the upper lip can apply first external force to the upper lip containing part through the upper lip supporting face; the lower lip containing part is fixedly connected with the supporting ring, the lower lip containing part is provided with a lower lip supporting face, the lower lip supporting face can make contact with the lower lip when the supporting ring is placed between the teeth and the lip, and the lower lip can apply second external force to the lower lip containing part through the lower lip supporting face; the supporting ring can be bent under the action of the first external force and the second external force, and the upper lip containing part and the lower lip containing part are made to be close to each other. The training effect of the perioral muscle function training device is easy to achieve.

This invention provides a method for the germination and regeneration of mature rubber tree embryos into seedlings outside of test tubes, belonging to the field of seedling cultivation technology. The method uses somatic embryos derived from anther tissue culture as raw materials, and involves three stages of cultivation in a growth medium: dark culture, low-light culture, and strong-light culture. Then, the mature embryos that meet the requirements undergo closed-bottle / open-bottle acclimatization. After this acclimatization, they can be rooted and grown into seedlings in river sand outside the test tube. This method allows mature embryos to germinate and regenerate directly into seedlings outside the test tube, eliminating the need for a sterile culture room and specialized rooting medium, thus requiring less specialized knowledge from the operator. Furthermore, by combining the embryo germination and seedling regeneration with the in-test-tube and sand-bed acclimatization in one step, the cultivation process is simplified, the cultivation cycle is shortened by two months, and cultivation efficiency is improved while production costs are reduced.