16results about How to "Overcome dependence" patented technology

The invention provides a wind power wind speed prediction and optimization arrangement method, system and device. The wind power wind speed prediction and optimization arrangement method comprises the steps of obtaining wind power wind speed basic data of a target area; constructing a neural network model based on physical information based on the wind power wind speed basic data; acquiring a wind speed at a specific position in a speed field from a neural network model based on physical information, and creating a training data set; training the neural network model based on the physical information based on the training data set; reconstructing wind speed distribution in a speed field through the trained neural network model based on physical information, and predicting to obtain a wind speed prediction value of the wind power plant in the next time period; and optimizing the arrangement of the wind turbine group according to the wind speed distribution in the reconstructed speed field. According to the method, the two-dimensional flow velocity field of the wind power plant can be reconstructed by training one PINN, and accurate wind speed ultra-short-term prediction is realized based on the trained model.

The invention provides a model training method and device, electronic equipment and a storage medium, and relates to the technical field of distributed computing. The method comprises the steps of dividing a context into context tasks based on structural analysis of a to-be-trained model; performing dependency relationship analysis on the context task to determine a task dependency graph; configuring priorities for the context tasks according to the task dependency graph, distributing the context tasks with the configured priorities to corresponding computing nodes, training according to an execution sequence, and obtaining load information and task execution information of the computing nodes; and according to the load information and the task execution information of the computing nodes, scheduling context tasks among the computing nodes, adjusting the execution sequence of the context tasks in the computing nodes, and outputting a training result. According to the real-time load condition of each node, task allocation is adjusted, load balancing of the computational nodes is realized, the utilization efficiency of computational resources is improved, a context task scheduling strategy is optimized, the utilization rate of the computational resources is improved, and the training time is shortened.

The present application discloses a static scanning method and system for unmanned driving C++ code under ROS system, which relates to the field of computer software white box testing, including: when the unmanned driving system C++ code is pushed to the Git Lab code repository, the Jenkins Job is automatically triggered to perform the code scanning task through the integration of Git Lab and Jenkins Job. Jenkins Job pulls the latest code from Git Lab to the scanning server, configures the compilation environment of the unmanned driving system, and compiles the C++ code using CMake or Catkin tools. During the compilation process, the compilation information is intercepted by the build wrapper plug-in of Sonar Qube to generate structured data of the code. Then, the Sonar Scanner tool and its configuration parameters are used to scan the compiled C++ code, and the scanning results are uploaded to the Sonar Qube server. The entire scanning process is automatically scheduled through the Pipeline script of Jenkins Job. In view of the low efficiency of static analysis of C++ code of unmanned driving systems existing in the prior art, the analysis efficiency is improved.

The present invention relates to the field of biological and chemical analysis, and discloses a surface extraction device, a mass spectrometry analysis system and a method. The surface extraction device of the present invention comprises: a sampling unit, an injection unit, a back-suction unit, a six-way valve with a quantitative loop, an electrically controlled multi-way valve and an air supply unit; the sampling unit is coaxially provided with a back-suction channel, an injection channel and a dry gas channel from the inside to the outside; the same side ends of the back-suction channel, the injection channel and the dry gas channel are respectively connected to the back-suction port, the injection port and the dry gas port of the sampling unit, and the other ends jointly form a sampling port; a sampling cavity is formed at the sampling port; the dry gas channel is provided with an air hole at a position higher than the sampling cavity; the injection unit is respectively connected to the six-way valve and the injection port through an electrically controlled multi-way valve; the back-suction unit and the back-suction port are respectively connected to the six-way valve; and the air supply unit is connected to the dry gas port. The scheme of the present invention can maintain the liquid flow stability of continuous analysis, optimize the operation platform, reduce the difficulty of operation, and improve the ionization effect.

The application discloses an online identification method for fractional order parameters of a transformer based on an Adam-SGD fractional order neural network, and comprises the following steps: acquiring a fractional order state equation representing dynamic characteristics of capacitors and inductors of a DC-DC transformer; establishing a fractional order discrete recursive model based on GL fractional order calculus definition; constructing a neural network structure according to a recursive relationship of capacitor voltage and inductor current in the discrete model, taking actually collected voltage and current values as reference values, taking neural network output as estimated values, and constructing a mean square error loss function; performing gradient calculation and weight updating on the loss function by using an Adam-SGD optimization algorithm, and online adjusting neural network weights; and calculating and outputting parameters of the capacitors and the inductors in real time according to the established fractional order discrete model and the obtained weights. When the weights are updated, the application adopts an adaptive switching mechanism of Adam-SGD, and has the rapid convergence ability of Adam and the stable and fine optimization ability of SGD, so that the efficiency and reliability of online training of the fractional order neural network are significantly improved.

This invention discloses a method for diagnosing self-discharge anomalies in energy storage battery packs based on short-term charging data, belonging to the field of battery safety monitoring technology. Addressing the shortcomings of existing methods that rely on complete charge-discharge cycles and are difficult to adapt to multi-stage varying operating conditions, this invention extracts features from multi-stage charging data, combines an autoencoder to achieve adaptive extraction of features under multiple operating conditions, and constructs a data-driven model fused with GCN-BiLSTM (Graph Convolutional Network-Bidirectional Long Short-Term Memory) to accurately estimate the RCC (Remaining Charging Capacity) of individual battery cells. Finally, based on the size, distribution, and changes of the RCC of each individual cell in the battery pack, we can diagnose overall battery pack inconsistencies, individual cell SOC inconsistencies, and self-discharge faults. This provides a guarantee for the safe operation of energy storage systems.

The application provides a machine-generated text detection method, system, device and medium, and belongs to the field of natural language processing. The method comprises the following steps: performing sentence segmentation on an original text; generating a perturbed text through a four-layer structure perception perturbator, wherein the perturbator dynamically selects a perturbation type and intensity through a learnable strategy; reorganizing the text through a pre-trained encoding-decoding model and extracting a feature vector; fusing the multi-source semantic and structural features of the original text, the perturbed text and the reorganized text; and inputting a classifier to output a text detection result in combination with a comprehensive loss function. The problem of low detection accuracy of the text after machine polishing is solved.

The invention relates to the technical field of power system monitoring, and discloses an icing torsion inversion method, system and device based on a physical constraint neural network, and a medium, and the method comprises the steps: collecting the torsion angle data, environmental parameter data and historical icing thickness data of a transmission conductor; establishing a physical constraint model between the torsion angle and the icing thickness based on a wire mechanics principle; constructing a first neural network based on the physical constraint model, and defining a composite loss function including a data driving loss function and a physical constraint loss function; training a first neural network by using historical icing thickness data, and optimizing network parameters through a back propagation algorithm to minimize a composite loss function; and inputting the torsion angle data and the environmental parameter data into a trained first neural network for prediction to obtain an icing thickness prediction value of the current wire. According to the method, the mechanical physical model is embedded into the neural network training process, so that accurate inversion from the wire torsion angle to the icing thickness is realized.

This invention relates to an artificial intelligence-based data inspection quality control method, system, device, and medium. The method includes: cleaning and determining the distribution of raw inspection data to generate a qualified dataset; adaptively learning key model parameters such as the effective data range and optimal smoothing coefficient through iterative calculation and simulation optimization based on the dataset and its distribution type; training a long short-term memory neural network time-series prediction model using the qualified dataset, and setting quality control limits by combining the learned benchmark dispersion to construct a real-time quality monitoring model; inputting real-time inspection samples into the model, generating alarm signals by calculating an exponentially weighted moving average monitoring statistic and comparing it with the control limits; iteratively optimizing the model and parameters based on historical alarm signals and new data to achieve continuous self-updating of quality control, enabling automatic learning of quality control parameters and intelligent optimization of the model, and improving the real-time monitoring capability of minor system deviations during the inspection process.

The present invention discloses an earthquake fault identification method, storage medium, and device, which relate to the field of image processing. The method comprises: dividing a three-dimensional earthquake synthetic data set into a training set and a test set; constructing an earthquake fault identification model, including a Patch Embedding layer, an encoder, a multi-layer feature fusion module, a decoder, and a linear projection layer; the encoder and the decoder both include axial, spatial, and channel attention mechanisms. After passing through the Patch Embedding layer, the three-dimensional seismic data enters the encoder; the multi-layer feature fusion module fuses the feature map output by each encoding unit, and the output feature map of the multi-layer feature fusion module is input to the decoder and then passes through the linear projection layer to obtain the final fault identification result. The real earthquake data to be identified is input into the final earthquake fault identification model after training and testing to obtain the fault identification result. The present invention effectively improves the model's ability to identify small faults.

The application discloses a sputum suction pressure optimization method and system combining multi-parameter cooperative feedback, and relates to the technical field of intelligent control of medical equipment. The method comprises the following steps: predicting and obtaining a physiological state parameter sequence and a sputum viscosity sequence of a wounded person in a preset time zone; based on the body parameters of the wounded person and the transfer scene parameters, a predicted body position feature sequence is constructed through simulation; a body position fluctuation coefficient is obtained according to the body position feature sequence, and an adaptive adjustment time window is dynamically set; taking the maximization of sputum removal efficiency and the minimization of airway mucosa injury risk as double objectives, an adaptive sputum suction negative pressure control scheme is optimized and generated in the time window by combining a generative adversarial network and a genetic algorithm; and finally, the sputum suction operation is performed according to the scheme. Through multi-parameter cooperative prediction and intelligent optimization, the application solves the poor adaptability problem of traditional sputum suction pressure control in a dynamic transfer environment, and significantly improves the accuracy, safety and reliability of sputum suction operation.

The invention relates to the technical field of polyurethane adhesives, in particular to a solvent-free polyurethane adhesive and application thereof. According to the solvent-free polyurethane adhesive, a PET-A component, a PET-B component, an ECR component and a polyurethane curing agent are prepared respectively, the polyurethane curing agent serves as the component A, the mixture of PET-B and ECR serves as the component B, the R (-NCO / OH) value between the component A and the component B is 2.2-3, the component A and the component B are evenly mixed, and the epoxy cinnamon ricinoleic acid polyurethane adhesive is obtained. The solvent-free polyurethane adhesive is prepared from a bio-based material, and has the advantages of high temperature resistance, high humidity resistance, environmental protection and low cost.

The embodiment of the invention provides an offline development environment deployment method and device, equipment and a storage medium, and relates to the field of financial science and technology or other related fields. The method comprises the steps of displaying a workflow corresponding to a to-be-deployed environment on a main interface in response to a received first operation, then determining target configuration information of the to-be-deployed environment in response to a received second operation, and completing offline deployment of the to-be-deployed environment based on the target configuration information and configuration steps. According to the method, the problem that the program development environment is difficult to effectively deploy in the intranet offline environment is effectively solved, the operation and maintenance cost of development environment deployment is reduced by providing an automatic and visual deployment mode, the limitation and network dependence of a traditional scheme are overcome, and the deployment efficiency and the user experience are improved.

The invention discloses a bearing fault identification method based on deep learning under noise interference, and the method comprises the steps: constructing and preprocessing a bearing fault vibration signal data set, training an original CNN model through the preprocessed bearing fault vibration signal data set, and storing the trained original CNN model; constructing a noise-added bearing fault vibration signal data set and carrying out preprocessing; the data set is formed by adding noise on the basis of a bearing fault vibration signal data set; establishing a bearing fault identification model; the model is formed by sequentially connecting an encoder, a decoder and a trained original CNN model; and training a bearing fault recognition model by using the noise-added bearing fault vibration signal data set, and storing the trained bearing fault recognition model for classifying and recognizing bearing fault signals. The bearing fault recognition accuracy is improved, and the problems that a traditional method depends on professional knowledge and the bearing fault recognition accuracy is affected by noise are solved.

The present application relates to a kind of full-band beat vibration identification methods of rotating equipment, the method comprises: using vibration velocity sensor to collect the vibration velocity signal of rotor;Using variable mode decomposition VMD to decompose vibration velocity signal, obtain multiple intrinsic mode function;The FFT transform of all intrinsic mode functions is obtained, and corresponding frequency spectrum is obtained;While the analytic signal of all intrinsic mode functions is obtained after Hilbert transform, the corresponding amplitude envelope is obtained, and the FFT transform of each amplitude envelope is carried out, and corresponding envelope spectrum is obtained;Envelope spectrum is analyzed, and whether intrinsic mode function can be beat vibration signal is judged, and whether it is beat vibration signal is further determined after analyzing frequency spectrum, and excitation component is obtained.Compared with prior art, the present application can effectively avoid the uncertainty caused by subjective judgment, improve the robustness when tracing analysis, and realize full-band beat vibration detection under the condition of no rotation frequency prior.

This application relates to the field of groundwater monitoring, and in particular to a multi-parameter groundwater depth determination device. The device includes an integrated template on which a water level gauge, a thermometer, a conductivity meter, a pH meter, and a GPS positioning device are respectively installed. These components are modularly integrated into the template. A main control unit is mounted on the template, and the main control unit is connected to the water level gauge, thermometer, conductivity meter, pH meter, and GPS positioning device via a corresponding data transmission bus. A cable encoder and a depth-determining float are also installed on the template, and the float is lowered via a guide rod. The purpose of this application is to provide effective support for practical production work such as groundwater pollution prevention and control and ecological restoration.