222results about How to "Avoid error accumulation" patented technology

The invention discloses a child massage teaching auxiliary system based on artificial intelligence, and particularly relates to the technical field of massage auxiliary teaching, a Kalman filtering technology is adopted to effectively filter out large-amplitude interference in hand motion data, time-frequency analysis is carried out in combination with fast Fourier transform or wavelet transform, and frequency domain characteristics of microfibrillation signals are accurately captured; through fine micro-tremor signal analysis, noise interference is removed, the fine change of micro-tremor of the hands of a student is accurately identified, and the evaluation precision of the massage action stability is remarkably improved; and the integrated learning progress management module automatically adjusts teaching contents according to the learning progress and skill mastering conditions of the students, provides real-time feedback and forward excitation, provides excitation when the students have good performance, prompts and optimizes follow-up actions in real time when deviation occurs, helps the students to continuously improve skills, avoids error accumulation, and effectively improves the teaching effect.

The invention discloses a fumigation treatment multi-parameter comprehensive monitoring system which comprises a multi-parameter sensor monitoring module, a multi-modal data processing and analyzing module, a system intelligent monitoring and adjusting module, a fumigation environment control module, a remote maintenance and management module and a system interaction and feedback module. The multi-parameter sensor monitoring module is used for collecting parameter data. The multi-modal data processing and analyzing module is used for fusing and analyzing data. The system intelligent monitoring and adjusting module is used for intelligently monitoring and adjusting fumigation strategies. The fumigation environment control module is used for adjusting environment parameters. The remote maintenance and management module is used for system maintenance and management. And the system interaction and feedback module is used for display interaction of data and a control state. According to the fumigation treatment multi-parameter comprehensive monitoring system, a multi-modal data fusion algorithm based on information channel decomposition is provided for multi-parameter data fusion and analysis; and a multi-parameter intelligent monitoring algorithm based on a multi-layer feedforward neural network is provided to intelligently monitor the fumigation environment.

The invention provides a seamless rapid fusion method and system for multivariate three-dimensional data, a product and a medium, and relates to the technical field of electric digital data processing. By acquiring multi-source heterogeneous data and unifying the multi-source heterogeneous data to a coordinate system of a geographic base map layer, a datum reference for data fusion is established. And the geographic base map layer is used as a benchmark to provide a stable spatial reference for subsequent data fusion. The surface smoothing algorithm is adopted to preprocess the live-action three-dimensional model and the monomer model data, so that the influence of data noise on the fusion precision is reduced. Through cooperative use of feature point extraction and an iterative nearest point algorithm, accurate alignment of different types of three-dimensional data is realized. According to the feature-based iterative optimization mode, the error accumulation problem caused by direct overall fusion is avoided, the fusion precision is improved, and the geometric consistency of the fused scene is ensured.

The invention provides a cascade pump station pre-station water level real-time prediction method based on optimized input and rolling correction, and relates to the field of pump station operation scheduling and water conservancy informatization research. According to the cascade pump station front-station water level real-time prediction method based on optimization input and rolling correction, a method including collected data cleaning, input factor simplification, time-phased prediction and rolling error elimination is provided, firstly, collected related data are cleaned, and then normalization preprocessing is adopted; calculating the correlation of input factors by using a Pearson's correlation coefficient, carrying out sensitivity analysis, calculating the lag time of the water flow propagation process, carrying out time-phased prediction according to the lag time, introducing an error correction (EC) technology, inputting training set data into an LSTM-EC neural network model, and continuously adjusting and optimizing a weight matrix and a deviation item, so as to obtain a prediction result; and completing training optimization of the LSTM-EC neural network model, inputting prediction set data, and finally realizing real-time rolling prediction of the pre-station water level of the cascade pump station.

The invention discloses a method and device for detecting the center position in a laser spot image, and the method comprises the steps: obtaining a laser spot image which is a fuzzy circle / circular ring light spot, and solving the center-of-gravity coordinate of the fuzzy circle / circular ring light spot as an initial center position; calculating radial distances from the initial circle center position along a plurality of different directions of the fuzzy circle / circular ring light spot, and calculating an average value of the radial distances in the different directions as an average radius; based on the average radius, utilizing a least square method to fit a new circle center position; and taking the new circle center position as an initial circle center position, repeating iterative calculation until the difference between the current fitting new circle center position and the previous fitting new circle center position is smaller than a preset threshold value, and determining the current fitting new circle center position as a final circle center position. The method is more accurate in circle center position detection and higher in calculation speed.

The invention relates to the technical field of fluid mechanics, in particular to a method for optimizing a fluid mechanics simulation model based on a deep parallel neural operator, and the deep parallel neural operator comprises an encoder and a decoder; the optimization process of the fluid mechanics simulation model comprises the steps that 1, physical quantities of fluid mechanics are expressed as input functions through an encoder in an original space, and the input functions are preprocessed; 2, mapping the data obtained in the step 1 into a plurality of potential spaces through an encoder to solve a fluid mechanics simulation model, and compensating high-frequency information loss caused by modal truncation; and step 3, recovering the solved data of each potential space into the physical quantity of the fluid mechanics in the step 1 through a decoder, and projecting the physical quantity to the original space. Through the technical scheme, the method has the advantage of solving the problem of error accumulation in the iteration process.

The invention discloses an intelligent guidance system and method for predicting ship navigation disturbance release based on machine learning. The intelligent guidance system comprises a data acquisition module, a machine learning analysis module and a ship navigation prediction feedback module. A channel water pollution prediction model is constructed through a machine learning algorithm, the concentration of pollutants in overlying water of a channel after a ship passes is predicted in real time, and an optimal route is recommended after water environment protection and economic benefits are comprehensively considered, so that the ship is guided to sail. The problem that endogenous pollution release caused by ship navigation disturbance is difficult to predict and control is solved, ecological protection is considered while ship transportation is developed, the method has the advantages of automation, high efficiency, accuracy, flexibility and the like, and technical support is provided for controlling water environment risks in a navigation channel.

The invention discloses a typhoon MPI multi-factor cooperative calculation evaluation method, a medium and a program product, and belongs to the technical field of typhoon intensity evaluation numerical calculation, and the typhoon MPI multi-factor cooperative calculation evaluation method comprises the steps: collecting sea-air power thermal field and typhoon kernel structure data, and generating a multi-source data fusion feature matrix; extracting multi-factor key features related to typhoon enhancement; constructing a typhoon MPI physical calculation model, and calculating MPI and appearance time thereof; introducing an attention mechanism based on a convolutional neural network, constructing a typhoon MPI multi-factor collaborative evaluation machine learning model, and optimizing a loss function through a physical constraint term for training; and finally, combining results of the physical calculation model and the machine learning model, dynamically correcting the results by adopting a weighted fusion strategy, and carrying out uncertainty analysis and error evaluation. According to the method, collaborative application of multi-source data fusion and a physical mechanism and machine learning is realized, and the calculation evaluation accuracy of the typhoon rapid enhancement stage is remarkably improved.

The invention discloses an electro-absorption modulated laser chip and a manufacturing method thereof, and relates to the technical field of semiconductor devices.The chip comprises a substrate, a modulator layer, a passive waveguide layer, a laser layer, an amplifier layer, a wrapping layer and a contact layer, the passive waveguide layer is arranged in a spacer region between the laser layer and the modulator layer in a butt-joint growth mode, the amplifier region is arranged on the other side of the modulator layer, the laser layer is arranged in the laser region on one side of the passive waveguide layer in the spacer region in a butt-joint growth mode, and the amplifier layer is arranged above the passive waveguide layer in the amplifier region. The cladding layer and the contact layer are sequentially arranged above the functional layer, a mask is used for covering a specific area in the butt-joint growth process, and the mask is removed after growth; according to the technical scheme, the active gain material of the amplifier region can be manufactured on the passive waveguide material while the manufacturing steps are simplified, the optical signal loss is reduced, and the chip saturated light output power is improved.

The invention relates to the technical field of spectral analysis, and discloses a glucose solution concentration detection method and system based on a near infrared spectrum. The method comprises the following steps: acquiring transmission light intensity of a sample and a reference liquid by using a narrow-band near-infrared LED light source, calculating absorbance characteristics, screening a dynamic wavelength set based on sensitivity, and constructing a concentration prediction model to realize accurate detection of concentration. Compared with the prior art which depends on single wavelength or lossy detection, especially under the condition of temperature and humidity fluctuation or unstable light source power, the technical problem that high-precision concentration measurement cannot be realized is solved. High-precision glucose concentration detection under complex environmental conditions is realized, so that the problem of error accumulation caused by environmental interference is avoided, the detection precision, stability and environmental adaptability of the system are improved, and the reliability and accuracy of glucose solution concentration detection are improved.

The invention provides a time sequence state data multi-step prediction method, device and equipment. A time sequence state data multi-step prediction model comprises a time sequence state data encoder structure and a multi-step prediction decoder structure; the time sequence state data multi-step prediction model is constructed based on a Transform model architecture; the structure of the time sequence state data encoder is based on a Transform encoder model architecture, and a long short-term memory network model is inserted between a multi-head attention sub-layer and a feed-forward network sub-layer of the Transform encoder for construction. The multi-step prediction decoder structure is constructed on the basis of a Transform decoder model architecture, and comprises a plurality of Transform decoders. According to the invention, a long-short-term memory network model is introduced, and the prediction performance is enhanced by combining global modeling and local time sequence dependence capability. A multi-decoder structure is designed to process a multi-step prediction task, and the problem of error accumulation is effectively solved.

The invention discloses a lithium battery combined health state estimation and service life prediction method, which comprises the following steps of: 1, predicting the health state and the residual service life of a lithium battery by carrying out pulse test on the lithium battery and establishing a Thevenin model at the aging normal operation stage of the lithium battery; 2, after the lithium battery enters a deep aging stage, performing an EIS test on the lithium battery to obtain impedance spectrum data, and correcting the Thevenin model into a variable order RC model; and 3, estimating the health state of the deep aging stage by using the impedance spectrum data and parameters identified by the variable order RC model, and when the prediction error of the two parameters is smaller than an error threshold value, considering that a battery health state and residual life prediction model suitable for the deep aging stage is successfully established. Otherwise, adjusting the order of the variable-order RC model to reduce the error. According to the invention, the reliability of lithium battery life prediction and estimation in a complete life cycle can be ensured, so that the control complexity and cost of a test working condition can be effectively reduced.

The present invention proposes a method and system for integrating Tibetan word segmentation and part-of-speech tagging, and relates to the field of electronic information. The method provided by the present invention obtains Tibetan text information input by a user, calls an integrated model and segments Tibetan syllables and non-Tibetan character blocks, performs CRF prediction, obtains the optimal label prediction, and sorts the writing form of each Tibetan syllable according to the result of the label prediction to obtain the corresponding tagging result. In this application, by proposing to establish a corresponding integrated model, Tibetan word segmentation and part-of-speech tagging are performed based on the corresponding integrated network. Compared with previous solutions, due to the integrity of the integrated model introduced in this application, the problem of cumulative part-of-speech tagging errors caused by word segmentation errors due to the two-stage execution of domain word segmentation and part-of-speech tagging tasks is avoided, and the corresponding Tibetan word segmentation and part-of-speech tagging can be processed more accurately, further increasing the practicality of the solution.

The invention discloses an LLMs combined compression method based on joint optimization, an electronic device and a storage medium, a quantization method and a rarefaction method are combined, errors caused by quantization and rarefaction are minimized at the same time, the method comprises the following steps that 128 fragments are sampled from a WikiText2 training corpus, and each fragment comprises 2048 tags; 2, initializing a diagonal transformation matrix by using a statistical index, carrying out preliminary quantization on a weight by using an equivalent transformation operation, determining a quantization step length and a zero point, and calibrating the quantization step length and the zero point through the data sampled in the step 1; calculating a sparse mask, rearranging weights by using an importance index # imgabs0 #, and preferentially sparsening unimportant weights; carrying out joint optimization; and evaluating the compressed model by using different data sets, and verifying the performance improvement of the model.

The invention relates to the technical field of intelligent report processing, and discloses an intelligent design method for an online form, which aims at improving the flexibility of report production and the intelligence of data processing, and comprises the following steps of: constructing a report structure comprising a plurality of dynamic data grids, extracting keywords in report names, forming a report name keyword set, and constructing a report structure comprising a plurality of dynamic data grids; the method comprises the steps of obtaining keyword information, classifying similar reports to form similar report sets, obtaining corresponding similar report sets according to the keyword information, generating keyword field sets, generating a new report according to the keyword field sets and the keyword information input by a user, and marking the newly generated report as a first report, executing a row title data conflict strategy for the first report, executing a row data arrangement strategy, and optimizing the data grid position information edited by the user.

The invention provides a machining device and method for a main and auxiliary connecting rod of a diesel engine. The machining device comprises a fine boring base, a connecting rod large end positioning assembly and a connecting rod small end positioning assembly. The connecting rod big end positioning assembly comprises a positioning block and two positioning plates which are oppositely arranged, and the positioning block, the two positioning plates and the fine boring base jointly form a first positioning gap used for containing the connecting rod big end; the connecting rod small end positioning assembly comprises a positioning block, a positioning plate and a floating supporting block, a V-shaped groove is formed in the side wall, close to the connecting rod small end, of the positioning block, the V-shaped groove of the positioning block, the positioning plate and the fine boring base jointly form a second positioning gap used for containing the connecting rod small end, and the floating supporting block is aligned to the positioning plate. According to the machining device and method, the connecting rod can be positioned, pressed and fixed for fine boring machining, error accumulation of connecting rod machining is eliminated by adjusting the height of the floating supporting block, the machining device and method are suitable for machining connecting rods of various types and specifications, fixing is reliable, dismounting and mounting are convenient, and the positioning precision and machining precision of the connecting rod are guaranteed.

The invention discloses a same commodity normalization method and system based on a large model, a storage medium and computer equipment. The method comprises the following steps: obtaining commodity titles and self-filling commodity codes corresponding to a plurality of commodities; clustering the commodity titles to determine a commodity class, determining a target commodity code corresponding to the commodity class according to the self-filling commodity code of each commodity in the commodity class, and filtering the commodities, of which the self-filling commodity codes belong to a commodity code cluster, in each commodity of the commodity class according to the commodity code cluster to which the target commodity code belongs, commodities corresponding to the commodity codes in the commodity code cluster are the same; and determining a first to-be-normalized commodity of which the self-filling commodity code does not belong to the commodity code cluster in each commodity of the commodity class, and according to a commodity title corresponding to the first to-be-normalized commodity and a standard title in a standard library, performing same-commodity normalization on the first to-be-normalized commodity through a large model so as to determine a correct commodity code corresponding to the first to-be-normalized commodity. And efficient and accurate same-product normalization of the commodities is realized.

The invention discloses a crank guide rod transmission pneumatic needle suction type small-particle-size bean sprout seed automatic sowing device which comprises a vacuum air chamber, a swing air cylinder, a seed conveying pipe, a guide rod, a fixing frame, a pneumatic vibration seed disc and a seedling disc. The device adapts to seeds of different sizes, thicknesses and shapes, and adsorption failure or seed damage is avoided; according to the shape characteristics of irregular leaf-shaped seeds, the specific suction needle caliber, hole shape and air flow distribution are designed, the adsorption efficiency is improved, and it is ensured that the seeds can be stably adsorbed; an airflow channel is optimally designed by virtue of airflow dynamics, so that uniform acting force on the leaf-shaped seeds in the adsorption process is ensured, falling or dislocation of the seeds caused by non-uniform adsorption positions is avoided, the seeding requirements of seeds of different types, sizes and shapes are met, and the universality and the service life of equipment are improved.

The invention relates to the technical field of air conditioner copper pipes, in particular to a multi-station synchronous correction control system for an air conditioner copper pipe. Comprising a copper pipe straightening unit used for straightening a to-be-treated copper pipe; the detection unit is used for collecting structural parameters of the to-be-treated copper pipe; the central control unit comprises a first processing module and is used for establishing a copper pipe analysis model; the first processing module is further used for setting a primary correction strategy according to the structure parameters of the copper pipe to be processed and the copper pipe analysis model; the first processing module is further used for setting working parameters of the copper pipe correction unit according to the primary correction strategy; the second processing module is used for establishing a linkage monitoring model; the second processing module is further used for judging whether a compensation correction instruction is generated or not according to the linkage monitoring model. By analyzing the structural parameters of the to-be-treated copper pipe, the correction efficiency of the system on the air conditioner copper pipe is improved. And a linkage monitoring model is additionally arranged, local deviation in the correction process of the to-be-treated copper pipe is early warned in time, and error accumulation is prevented.

The present application provides an image processing method, device and electronic device. In the technical solution of the present application, an image to be processed is obtained; the image to be processed is processed by the first branch of a pre-trained neural network model to obtain a first type of parameter, and the first type of parameter is used to perform global color processing on the image; the image to be processed is processed by the second branch of the neural network model to obtain a second type of parameter, and the second type of parameter is used to perform local color processing on the image; the image to be processed is color processed according to the first type of parameter and the second type of parameter to obtain a color-processed image. In the above technical solution, the input of the neural network model is the image to be processed, and different branches output different types of parameters, which can avoid the error accumulation problem in the parameter calculation process and obtain more accurate parameters. Furthermore, the image to be processed is processed according to the obtained parameters, which can improve the image color correction effect.

The invention belongs to the technical field of remote sensing image processing, provides a remote sensing image blocking effect elimination method fusing adaptive weight and an anchoring mechanism, and aims to eliminate the blocking effect from a global level, take radiation consistency as a constraint condition, introduce the adaptive weight to carry out differential adjustment on constraint intensity of different overlapped regions, and improve the blocking effect elimination efficiency. Meanwhile, an anchoring mechanism is adopted, stabilization constraint is carried out on an overall solution space, finally, a global equation set is constructed, an iterative least square optimization strategy is used for solving, image radiation consistency and visual continuity are integrally ensured, and the quality of remote sensing images is improved.

The invention discloses an automatic control device for the press-fitting height of a pin shaft, which relates to the technical field of component control devices and comprises an upper mounting seat, wherein the bottom of the upper mounting seat is rotatably connected with a shaft rod, the bottom of the shaft rod is fixed with a mounting disc, and a positioning seat for placing the part to be press-fitted is arranged below the mounting disc; a punching component is mounted on the mounting disc, and the punching component is used to punch holes in the part to be press-fitted; a pressing component is also mounted on the mounting disc, and the pressing component is used to place and position the pin shaft; a measuring component is also mounted on the mounting disc, and the measuring component is used to measure the length of the pin shaft extending out of the part to be press-fitted; a conversion component is mounted on the bottom of the upper mounting seat, and the conversion component is used to drive the mounting disc to rotate, so as to switch the punching, pressing and measuring processes, and drive the operation of the punching component, the pressing component and the measuring component; the error accumulation caused by the difference in equipment accuracy between different processes is effectively avoided, the excess tolerance rate is reduced, and the error adjustment becomes simple and quick.

This invention discloses a navigation method and system for offshore wind power booster stations based on multimodal information interaction, belonging to the field of artificial intelligence and robot navigation technology. It achieves visual-language modality alignment through the CLIP model, constructing an end-to-end navigation parameter generation model. This model can adaptively generate AGV motion parameters based on natural language commands and scene images, enabling efficient and accurate inspection of offshore wind power booster stations. By establishing a multimodal dataset and associating language commands with AGV motion parameters, this invention solves the error accumulation problem caused by multi-stage processing in traditional methods, improving navigation accuracy and adaptability. Furthermore, this invention employs a multi-task loss function optimization model, including contrastive loss and motion parameter regression loss, enabling the AGV to adaptively adjust motion parameters according to environmental changes, thus improving the efficiency and safety of offshore wind power booster station inspections.

The present application relates to the technical field of precipitation forecast, solves the problem that the prior art is prone to underestimate the peak value of precipitation or produce false alarm under strong convective weather, and particularly relates to a short-impending precipitation forecast method fusing GNSS and water conservancy rain measurement radar, GNSS observation data and water conservancy rain measurement radar base data in a monitoring area are acquired, pretreated, corresponding atmospheric precipitable water distribution field and radar reflectivity factor distribution field are generated, based on a preset time sliding window, time sequence characteristics of the atmospheric precipitable water distribution field and the radar reflectivity factor distribution field are extracted, and the two are superimposed in channel dimension to construct a multi-channel spatio-temporal fusion input tensor. The present application overcomes the defect that the traditional extrapolation method rapidly decays with the prolongation of the forecast time, significantly improves the prediction accuracy of the rainstorm center intensity and the falling area, reduces the missed report and underestimation of the precipitation event, and solves the problem of low prediction accuracy of the traditional linear extrapolation algorithm under strong convective weather.

The invention relates to the technical field of fiber bragg grating inscribing, in particular to a reflectivity-based real-time monitoring inscribing single-wavelength continuous weak grating system, and a to-be-inscribed fiber is used for bearing the inscribed fiber and executing the inscribing task of the grating; the light path inscribing system is used for inscribing a single-wavelength weak grating on the optical fiber to be inscribed; the broadband light source is used for detecting reflected light of the single-wavelength weak grating; the precise 3D transmission platform is used for fixing and moving an optical fiber to be inscribed to a continuous inscribing position; the spectrograph is used for measuring spectral data reflected by the optical fiber; the data collector is used for collecting reflected optical signal data in real time and carrying out signal processing and analysis; the display terminal is used for displaying reflected spectral data and grating inscribing quality information in real time, the accuracy and efficiency in the optical fiber single-wavelength weak grating inscribing process are improved through a dynamic feedback mechanism, continuous grating inscribing is supported, and meanwhile high inscribing quality and high inscribing efficiency are met.

The invention provides a side wall positioning and mounting method and device for an assembly trolley of an overlapped assembly type subway station, and belongs to the technical field of underground engineering construction. The method comprises the following steps: an environment perception and initial positioning step: scanning a construction environment, and determining spatial positions of a side wall and an H-shaped steel groove relative to a preset BIM model through a target identification and point cloud registration technology; a transverse moving and coarse positioning step: planning a path based on a positioning result, and controlling the assembly trolley through real-time pose estimation and visual tracking to enable the side wall to move to a position above the target; an accurate positioning and leveling step: obtaining deviation and real-time attitude through visual detection and multi-sensor fusion technology, and driving an adjusting mechanism to carry out accurate alignment and leveling; a gap detecting and filling step: after the side wall falls in place, measuring the shape of the gap and guiding filling construction; according to the invention, global absolute position calibration is realized, error accumulation is avoided, construction safety is ensured through non-contact measurement, and positioning precision and reliability are ensured by using real-time feedback control.

The invention discloses a strip network railway base station data processing method and system. The method comprises the following steps: obtaining observation data of a plurality of IGS base stations; obtaining observation data of a plurality of standard reference stations in a preset range around the strip network railway reference station, wherein the standard reference stations comprise a national station, a provincial station and a Chinese tectonic environment monitoring network station; and based on the observation data of the IGS reference station, combining the observation data of the standard reference station to perform hierarchical calculation on the plurality of railway reference stations to obtain space coordinate data of each railway reference station. Observation data of an IGS base station and standard base stations such as surrounding national stations and provincial stations are obtained, and a rich data basis is provided for railway base station coordinate measurement and calculation. A hierarchical resolving mode is adopted, the coordinate precision is improved, error accumulation is avoided, and the result reliability is enhanced; the method can adapt to different precision requirements and complex environments, considers local differences, and meets the requirement of a strip network railway for high-precision coordinate data.

The application provides a pressure regulating hydrogen storage device, a pressure regulating hydrogen storage method and a composite hydrogen delivery system, which can deliver hydrogen under pressure for a long time and meet the filling demand of hydrogen for hydrogenation load with large capacity. The pressure regulating hydrogen storage device comprises a hydrogen delivery pipeline, a pressure regulating hydrogen storage unit and a temperature regulating liquid bath unit. The pressure regulating hydrogen storage unit comprises a first pressure regulating hydrogen storage tank group filled with a first metal hydride hydrogen storage material and a second pressure regulating hydrogen storage tank group filled with a second metal hydride hydrogen storage material, which are sequentially connected in series along the hydrogen delivery direction of the hydrogen delivery pipeline. The platform pressure of the first metal hydride hydrogen storage material at room temperature is less than that of the second metal hydride hydrogen storage material at room temperature. The temperature regulating liquid bath unit comprises a first temperature regulating bath chamber containing the first pressure regulating hydrogen storage tank group, a second temperature regulating bath chamber containing the second pressure regulating hydrogen storage tank group and a temperature regulating liquid delivery tank in circulation communication with the first temperature regulating bath chamber and the second temperature regulating bath chamber, respectively.

The embodiment of the invention provides a multi-site collaborative electromagnetic spectrum situation prediction method based on multi-source data fusion and an attention mechanism, and the method comprises the steps: 1, collecting spectrum observation data, spatial data and meteorological data of multiple sites, and carrying out the data preprocessing; step 2, based on prior knowledge and GIS data, calculating a theoretical field intensity value of a known emission source at each monitoring station by using a radio wave propagation model; step 3, constructing a multi-branch fusion neural network MFNN based on 3D-CNN, LSTM, GCN and MLP, extracting joint feature representation of multi-source data in the step 1 and the step 2, and introducing an adaptive attention mechanism to perform weighted fusion on each branch feature; and 4, generating a predicted value of a future frequency spectrum state through a prediction output layer, and performing data inversion to obtain a physical power value to realize channel state inference. Structured information capable of directly driving a decision is output through data inversion and channel state inference.

The present invention relates to a partition welding method, which belongs to the field of welding technology, and includes the following steps: preparing tooling; assembling an outer ring plate, an inner ring plate and blades; a red sleeve positioning sleeve; welding; welding a cover plate; welding a radial steam seal body; heat treatment; removing the cover plate and the positioning sleeve; separating the partition along the center dividing surface; performing surfacing welding on the center dividing surface; post-heat treatment and the like. The present invention welds two semicircular partitions together to obtain a partition with better integrity, avoids error accumulation caused by separate welding, and has a higher matching accuracy after being installed in a steam turbine. The present invention uses a positioning sleeve to position the two semicircular outer ring plate bodies, so that the outer ring plate bodies are tightly spliced ​​together as a whole, with better integrity, and the outer ring plate is not easily deformed during the welding process. When welding the radial steam seal ring, the cover plate is used to completely cover the blade to effectively prevent welding slag from splashing onto the blade surface and causing damage to the blade.