32results about How to "Realize complementary advantages" patented technology

The invention discloses a private domain user management method and system for multiple touch channels, a medium and a program product, and the method comprises the steps: obtaining the historical behavior data of a user in each touch channel; identifying the behavior relevance of the user among the arriving channels; constructing an inter-channel behavior mapping relation of migration paths of the user among the touch channels based on behavior relevance; according to the behavior mapping relation between the channels, the overlapping ratio of user behaviors between the reaching channels, the difference of conversion effects and the overlapping ratio of time sequences are calculated; based on the overlapping rate, the difference and the coincidence degree, determining the reaching channel complementation degree; grouping and clustering the reaching channels according to the reaching channel complementation degree to obtain a plurality of channel collaboration units; and the optimal time difference of the arriving users in each channel cooperation unit is calculated, and the marketing resources are allocated among the arriving channels based on the optimal time difference, so that the accuracy of private domain operation management is improved.

The invention discloses an adaptive timestamp coding enhanced complex equipment incomplete state monitoring data time sequence interpolation method, and relates to the technical field of industrial sensing data processing, and the method comprises the steps: carrying out the standardization of obtained online incomplete state monitoring time sequence data, carrying out the normalization processing of online timestamp sequence data, and carrying out the normalization processing of the online timestamp sequence data; then, a timestamp prior feature coding module is used for outputting online timestamp prior coding features, and effective utilization of timestamp information in the monitoring data in the incomplete state in an interpolation task is achieved; using the trained timestamp-local interpolation model joint framework to obtain an online interpolation result of the monitoring data in the incomplete state; according to the scheme, advantage complementation between a locality time sequence interpolation model and a timestamp probabilistic interpolation model is achieved, the accuracy and robustness of complex equipment state monitoring data time sequence interpolation are effectively improved, and the method can be widely applied to a data enhancement link in various complex equipment operation state analysis tasks.

The invention relates to the technical field of concrete materials, and particularly discloses a composite mineral admixture, a preparation method thereof and concrete. The preparation method of the composite mineral admixture comprises the following steps: mixing and modifying dolomite powder, gypsum and calcium hydroxide to obtain pre-modified dolomite powder; then adding hexadecyl trimethyl ammonium chloride and octylphenol polyoxyethylene ether to obtain modified dolomite powder; and adding nickel slag, metakaolin, desulfurized fly ash and a grinding aid, and grinding to obtain the composite mineral admixture. The concrete prepared from the composite mineral admixture is excellent in compressive strength, shrinkage resistance, permeation resistance and the like, and the strength and durability of the concrete can be remarkably improved.

The utility model provides a light offshore booster station and offshore hydrogen production and hydrogenation fusion platform. A fusion platform is arranged in an edge area of an offshore wind power plant; the fusion platform is provided with five layers, the first layer to the second layer of the fusion platform are hydrogen production function areas, and the third layer to the fifth layer of the fusion platform are power transformation function areas; a cable shaft is jointly arranged on the hydrogen production functional area and the power transformation functional area of the fusion platform, and a power supply cable, a hydrogen pipeline and a cooling water pipe are arranged in the cable shaft; the power supply cable is connected with the hydrogen production functional area and the power transformation functional area, the power transformation functional area is provided with a main transformer radiator and a main transformer body below the fifth layer of the fusion platform in a staggered manner, and the power transformation functional area is provided with a storage battery chamber on the fourth layer of the fusion platform. The fusion platform integrates the functions of offshore hydrogen production and hydrogenation, can effectively relieve the contradiction between the rapid increase of offshore wind power and the slower construction speed of a power grid without additionally using a sea land, solves the problem of offshore wind power consumption, and improves the utilization rate of wind energy.

The invention provides an IMU (Inertial Measurement Unit)-based multi-sensor fusion positioning method and device and a storage medium. The positioning method comprises the following steps: obtaining an IMU pre-integration result based on a constructed IMU pre-integration model; adding the system positioning result at a moment and the IMU pre-integration result to obtain a system positioning predicted value at the current moment; obtaining original data of the plurality of sensors, and performing preprocessing and synchronization to obtain observation values of the plurality of sensors at the current moment; independently filtering the observation values of the plurality of sensors and the system positioning predicted value at the current moment by using the constructed filtering subsystem so as to obtain updated system positioning predicted values of the plurality of sensors at the current moment; according to the system positioning observation values of the plurality of sensors and the confidence coefficients of the plurality of sensors at the previous moment, updating the confidence coefficients of the plurality of sensors at the current moment; and according to the confidence of the plurality of sensors at the current moment, carrying out weighted summation on the updated system positioning predicted values of the plurality of sensors to obtain a system positioning result at the current moment.

This invention belongs to the field of soil conditioner technology, and relates to a method for preparing a soil conditioner, the resulting soil conditioner, and its application method. The preparation method includes the following steps: mixing brewing sludge and biogas residue with sorghum straw particles to obtain a mixture; adjusting the carbon-to-nitrogen ratio to 20:1 to 40:1; adding an organic material composting agent; aerobic fermentation at 55℃ for ≥15 days; turning and cooling the pile at temperatures >65℃; and fermenting for 15-30 days after cooling to below 50℃ to obtain the soil conditioner. By utilizing the nitrogen source of the biogas residue to neutralize the high-carbon characteristics of the straw, and further enhancing the synergistic decomposition of the two by the organic material composting agent, this invention promotes rapid aerobic fermentation and stable high-temperature fermentation for over 15 days. This ensures harmlessness while avoiding the high-temperature burning of beneficial bacteria in the composting agent and the disruption of the synergistic decomposition of the biogas residue and straw, achieving complementary advantages between the biogas residue and straw, and producing a loose, breathable, nutrient-balanced, and safely composted soil conditioner.

The invention relates to a residential energy system low-carbon scheduling method considering a hydrogen-terrestrial heat coordination effect. The residential energy system low-carbon scheduling method comprises the following steps: step (1), constructing a basic physical model for representing energy flow and equipment physical characteristics in a residential energy system; (2) converting the basic physical model into a coupling constraint model based on a dynamic coupling constraint between energy supplies of a set type in the residential energy system; step (3), on the basis of the coupling constraint model, according to a set scheduling period, constructing a low-carbon scheduling objective function to obtain a complete optimization model; and step (4), solving the optimization model to obtain an operation strategy of each device in the residential energy system in the set scheduling period.

The invention relates to a bi-crosslinking cyanate ester-epoxy blended adhesive and a preparation method thereof, and relates to the technical field of high polymer materials, the adhesive comprises the following components in parts by weight: 45-85 parts of cyanate ester resin, 5-35 parts of epoxy resin, 0-20 parts of a reactive diluent, 2-6 parts of a bi-crosslinking curing agent and 0-50 parts of a functional filler, the bi-crosslinking curing agent has the following structure: R is alkyl, aliphatic amine or aryl; r'is a primary amino group and / or a secondary amino group, R ''is-C = N and / or-OCN, m is an integer from 1 to 3, and n is an integer from 1 to 2. By adding the double-crosslinking curing agent containing the amino group and the cyano group, the technical problems of high curing temperature, large stress after curing, poor temperature resistance of epoxy resin and the like of the existing cyanate ester resin system are solved, and the curing temperature, the bonding strength and the temperature resistance of the adhesive are balanced. The adhesive can be cured at the temperature lower than 90 DEG C, and has high glass transition temperature, low thermal weight loss and excellent adhesive property.

The application relates to an independent hybrid frequency modulation energy storage power station capacity configuration method and device, computer equipment, a computer readable storage medium and a computer program product, relates to the field of electric power energy storage, and can improve the resource utilization rate of a hybrid energy storage system. The method comprises the following steps: acquiring total power control instructions of a hybrid energy storage system in a target frequency modulation scene and operation characteristic parameters of each energy storage unit; based on the operation characteristic parameters and the total power control instructions, determining first charge and discharge instructions and second charge and discharge instructions of different energy storage units according to a real-time power distribution strategy; based on the first charge and discharge instructions and a life attenuation model, determining a device replacement cycle of an energy type energy storage unit; based on system frequency modulation response and periodic resource consumption, determining a comprehensive performance model in a full life cycle; under the premise of meeting preset operation constraint conditions, taking the comprehensive performance value of the comprehensive performance model as a target, and determining capacity configuration parameters through a Bayesian optimization algorithm.

The invention belongs to the technical field of satellite communication, and relates to a satellite-ground cooperative beam fault recovery method and system, and the method comprises the steps: carrying out the multi-dimensional monitoring of a downlink signal of a satellite through a ground station, and building a health baseline model for each beam; the satellite-borne equipment executes self-inspection, generates a preliminary alarm data packet when monitoring that the parameter of the key part exceeds a preset threshold value, and reports the preliminary alarm data packet to the ground station through the telemetering link; the ground station performs fusion analysis on the health baseline model, the preliminary alarm data packet, other beam data and space environment data to obtain a fault root cause; the fault root causes comprise temporary attenuation, beam pointing offset and permanent performance recession; the ground station generates a recovery strategy based on the fault root cause; a ground station is taken as a core dominant, and rapid and accurate recovery of on-satellite beam faults is realized through cooperative work with a satellite.

The invention discloses a method for simulating regional runoff based on a coupling model VMD-CNN-LSTM-Attention of irrigation water consumption of main crops. The method comprises the following steps: 1, calculating daily reference crop evapotranspiration ET0 according to meteorological station data of a target region; 2, calculating the irrigation water consumption of different crops in different periods by utilizing the crop growth coefficients of the different crops; and 3, taking the main crop irrigation water consumption, rainfall, temperature, evapotranspiration, soil water content and other monthly elements in the target area calculated in the step 2 as driving factors, and inputting the driving factors into the coupling model for monthly runoff prediction. The method has the following beneficial effects that the irrigation water consumption of main crops with high correlation with the runoff in the research area is calculated, and driving factors with higher correlation with the runoff can be found out according to local conditions as input, so that the prediction precision is improved.

The invention provides a degradation-resistant wastewater treatment system, which comprises: a chemical oxidation system, which comprises an acid adjusting tank, an electro-Fenton reaction tank, an alkali adding neutralization tank and a solid-liquid separation tank which are sequentially connected along a water flow direction; and the biochemical system comprises a denitration selection tank, an anaerobic tank, an anoxic tank, at least one aerobic tank and an eMBR membrane tank which are sequentially connected along the water flow direction. Compared with the prior art, the non-degradable wastewater treatment system has the advantages that the denitration selection tank is additionally arranged at the front end of the anaerobic tank, nitrite nitrogen returned by sludge is removed by using a raw water carbon source, the influence on the growth of phosphorus-accumulating bacteria is reduced, the phosphorus removal effect is improved, an anaerobic ammonia oxidation generation environment is created, and the denitrification effect of the system is enhanced.

The invention discloses a full-crosslinking layered fluorine-containing epoxy super-hydrophobic coating and a preparation method thereof, and belongs to the technical field of super-hydrophobic coatings. The coating comprises the following raw material components: epoxy resin, fluorine-containing epoxy acrylic resin, KH560 modified nano silicon dioxide, a curing agent, an accelerant and the like. The preparation method of the coating comprises the following steps: carrying out surface grafting modification on the nano silicon dioxide particles by using a KH560 siloxane coupling agent; the preparation method comprises the following steps: synthesizing fluorine-containing epoxy acrylate low-surface-energy resin by using tridecafluorooctyl methacrylate, glycidyl methacrylate and isobornyl methacrylate; the coating is prepared by adopting a step-by-step spraying method, the primer is epoxy resin, the surface wear-resistant super-hydrophobic layer is fluorine-containing epoxy acrylate low-surface-energy resin blended modified nano silicon dioxide particle filler, and the three components are integrally crosslinked through epoxy groups. According to the method, the adhesive force and durability of the coating can be improved, and the coating has good super-hydrophobic performance and is suitable for various working conditions.

The invention provides a medical insurance fraud cross identification method based on fusion of multiple machine learning algorithms, and belongs to the field of machine learning technology and insurance risk management cross technology in information technology. The technical problems that in medical insurance fraud recognition, the recognition precision is not high, characteristics are complex, behaviors are hidden, and distinguishing is difficult are solved. According to the technical scheme, the method comprises the following steps of S1, performing data preprocessing on an original data set D0; s2, finding out feature information with high correlation by using the heat map; s3, constructing a medical insurance fraud detection model based on multiple machine learning algorithms; s4, checking the model; and S5, extracting main medical features influencing fraud discrimination through an algorithm. The method has the advantages that the accuracy and efficiency of medical insurance fraud detection are improved, and a more effective risk management solution is provided for insurance companies.

The application discloses a low-temperature liquid carbon dioxide battery system applied to SOFC power generation, and relates to the technical field of carbon dioxide energy storage.The system comprises a solid oxide fuel cell-gas turbine module, which is used for making the solid oxide fuel cell perform an electrochemical reaction based on air, natural gas and water, generating electric energy, and discharging tail gas; a waste heat recovery module, which is connected with the solid oxide fuel cell-gas turbine module, is used for absorbing waste heat in the tail gas discharged by the solid oxide fuel cell-gas turbine module through two circulating loops and recovering electric energy; and a low-temperature liquid carbon dioxide battery module, which is connected with the waste heat recovery module, is used for storing electric energy and heat energy of the waste heat recovery module and supplying power for electric equipment.The application realizes the complementary advantages among the modules, thereby increasing the operation efficiency of the overall system, meeting the power consumption scene demand of medium and small scale peak load shifting, and realizing power supply stability in a region.

The application discloses a water-rich soft rock foundation pit adjacent tunnel stability control method based on artificial intelligence, which comprises the following steps: prior prediction based on a finite element model, monitoring sensor arrangement, construction and training of a neural network model, scheme safety verification, initial decision based on the neural network model, initial grouting scheme execution, real-time monitoring-based cyclic intelligent decision and online incremental learning; training data is generated by establishing a three-dimensional finite element model containing a water-soil coupling mechanism, a neural network model capable of mapping the complex nonlinear relationship between monitoring parameters and optimal grouting schemes is constructed and trained, dynamic decision is made, and the grouting scheme that has passed safety verification is output and the model is continuously optimized; the method realizes the whole-process closed-loop intelligent control from intelligent prediction, real-time decision, safety control to self-optimization, and significantly improves the guarantee capacity and risk control level of the stability of the adjacent tunnel during foundation pit excavation in the water-rich soft rock complex stratum.

The invention relates to the technical field of energy supply, and discloses a plateau village and town wind, light, generation and storage multi-energy complementary micro-grid dynamic optimization system and method, and the system comprises a photovoltaic power generation module which employs solar energy to generate power so as to supplement daytime load demands; the wind power generation module converts wind energy into electric energy as renewable energy source base load; the biomass energy module is used for generating power by utilizing agricultural wastes so as to provide controllable base load; the energy storage module is used for stabilizing power fluctuation, storing surplus electric energy and releasing the surplus electric energy at a load peak; the dynamic complementation module is used for generating biomass energy generating capacity, energy storage charging and discharging capacity and interaction capacity with a large power grid on the basis of wind speed / illumination prediction and a load curve of the next day by taking minimization of total cost as a principle; the technical problems that in plateau villages and towns, wind power and photovoltaic power generation are intermittent and unstable, and stable energy supply cannot be met are solved.

The invention discloses a strong background light license plate recognition system and method, relates to the technical field of optics, and aims to solve the problem that a driving vehicle license plate is difficult to recognize under the conditions of strong light interference and low light in traffic monitoring. The system is composed of a laser illuminator, a polarizing film, a visible light color camera, a visible light color polarization camera, an attenuation sheet, a data acquisition system, a power supply module, a controller, an image processor, a laser beam expanding lens group, a visible light imaging lens group and a visible light polarization imaging lens group. Compared with a traditional traffic monitoring probe, the system can effectively restrain strong light interference generated by an automobile headlamp at night, and the imaging contrast ratio and recognition precision of a license plate in the strong light background are improved. The visible light color camera and the visible light color polarization camera are combined for use, the vehicle contour and the surrounding background image obtained by the visible light color camera are reserved, detail information and background information lost by the polarization image can be effectively supplemented, a basis is provided for judgment of management personnel, and advantage complementation is achieved.

The invention relates to the technical field of nutritious sports food, and discloses cashew nut and sea cucumber peptide exercise recovery powder and a method, the cashew nut and sea cucumber peptide exercise recovery powder comprises the following components by mass: 40%-55% of a composite peptide composition, 25%-35% of a rapid energy supply matrix, 5%-10% of an anti-inflammatory active component, and 10%-15% of a synergistic auxiliary agent, the composite peptide composition is formed by compounding a cashew nut extract and sea cucumber peptide according to a mass ratio of (4-6): (3-5), the molecular weight distribution of the cashew nut extract is 200-800Da, the purity of the cashew nut extract is greater than or equal to 92%, the molecular weight distribution of the sea cucumber peptide is 300-1000Da, the purity of the sea cucumber peptide is greater than or equal to 90%, the rapid energy supply matrix is formed by compounding maltodextrin and fructo-oligosaccharide according to a mass ratio of (3-4): 1, and the anti-inflammatory active component comprises curcumin microcapsules and resveratrol. Through the specific compounding design of the cashew nut extract and the sea cucumber peptide, in combination with precise molecular weight control, advantage complementation of amino acid composition is achieved, and the gastrointestinal tract absorption efficiency of nutritional ingredients is remarkably improved in cooperation with a targeted slow release technology of curcumin microcapsules.

The application discloses a new-born brain function evaluation and neural development prognosis prediction system, relates to the technical field of new-born brain function evaluation, and comprises an EEG electrode module, an fNIRS optical electrode module and a prediction module. The EEG electrode module is used for collecting EEG signals of a new-born in real time within a set time period. The fNIRS optical electrode module is used for collecting rSO2 signals of the new-born in real time within the set time period. The prediction module is used for performing signal preprocessing on EEG signals of the whole brain of the new-born and the rSO2 signals of the new-born, obtaining preprocessed EEG signals and preprocessed rSO2 signals, generating a dynamic brain function connection network sequence capable of representing the time evolution characteristics of brain function connection structures by using the preprocessed EEG signals and the preprocessed rSO2 signals, obtaining a time-varying feature vector by using the dynamic brain function connection network sequence, and obtaining a prediction result of new-born brain function evaluation and neural development prognosis based on the time-varying feature vector and a prediction model established in advance.

The invention discloses a PVT combined bathing wastewater and air source multi-source heat pump water heating device and an operation method. The device comprises a bathing wastewater waste heat collecting loop, a PVT hot water collecting loop and a multi-source heat pump loop, wherein the bathing wastewater waste heat collecting loop, the PVT hot water collecting loop and the multi-source heat pump loop are used for heating municipal water supply; the heated municipal supplied water is stored in the heat supply water tank, and the heat supply water tank is used for providing hot water for users; the multi-source heat pump loop is further used for conducting secondary waste heat recovery on bathing waste water through the bathing waste water waste heat collecting loop and conducting secondary waste heat recovery on low-temperature water through the PVT hot water collecting loop. According to the system, bathing wastewater waste heat, PVT heat production and air energy are comprehensively utilized, and the bathing wastewater waste heat can fully guarantee that domestic hot water is stably produced under the conditions of outdoor low temperature, poor solar radiation or no solar radiation; pVT heat production and an air source can improve the energy efficiency level and stability of hot water production under the conditions of high outdoor temperature, good solar radiation conditions or insufficient bathing wastewater, and complementary advantages are achieved.

The utility model provides a spliced door butt jointer, which comprises a first guide rail installed on the top of a shutter leaf, a second guide rail fixed on the top of a door frame, a movable butt joint seat installed at the end of the first guide rail, a fixed butt joint seat installed at the end of the second guide rail, the movable butt joint seat and the fixed butt joint seat being mutually butted when the shutter leaf is closed, so that the first guide rail and the second guide rail are in communication, a locking block arranged in the movable butt joint seat, the locking block being linked to a linkage seat arranged on the door frame through a linkage mechanism, the linkage seat driving the linkage mechanism to make the locking block disengage from the first guide rail when the shutter leaf is closed, the linkage seat being separated from the linkage mechanism when the shutter leaf is opened, the locking block being inserted into the first guide rail to limit the sliding of a pulley set. The utility model completely solves the problem that the traditional flat door can only be opened by half, limits the entry and exit of articles, and provides a large-size operation space and an unobstructed view for the user.

The invention discloses an intelligent processing and crowd counting method based on photo-thermal image enhancement, and relates to the technical field of computer vision, and the method comprises the steps: 1, collecting a multi-modal image of a target scene, and constructing a training data set; 2, constructing a bimodal calculation architecture based on a deep learning network, and carrying out training optimization by adopting the training data set to obtain a crowd counting model; the crowd counting model comprises an all-weather multi-modal fusion crowd counting network and a weak supervision multi-modal graph convolution counting network; and 3, collecting a to-be-counted multi-modal image, and inputting the to-be-counted multi-modal image into the crowd counting model to obtain a counting result. Visible light image and thermal infrared image information captured by acquisition equipment is deeply fused through a bimodal calculation architecture, and the environment illumination condition is adaptively sensed, so that the crowd counting accuracy and generalization ability in various dense public places are effectively improved.

The invention relates to the technical field of smart homework, in particular to a test paper information input method applied to dot-matrix pen smart homework, and the method comprises the steps: carrying out the question segmentation and question structured labeling based on an electronic format test paper, determining a test paper structured analysis result, and then, storing the test paper structured analysis result into a database; based on the test paper structured analysis result and the test paper in the electronic format, question analysis and information input are carried out, online test paper is obtained, and the content of the online test paper is consistent with that of the test paper in the electronic format. And then, based on the test paper structured analysis result and the code laying format, carrying out code laying processing on the test paper in the electronic format to obtain a code laying test paper on which a dot matrix pattern is laid. According to the method, question segmentation and question structured labeling of the electronic-format test paper are realized through an automatic means, question analysis and information input are automatically performed based on segmentation and question structured labeling results, and the online test paper is automatically generated, so that time consumption and error risks caused by manual input are avoided, and the efficiency is improved. Rapid and accurate input of test paper information is realized.

The application discloses a coalbed methane well gas production NP-LSTM combined prediction method and belongs to the technical field of oil and gas field development prediction. A two-level model fusion prediction framework is constructed by adopting a meta-learning idea. Firstly, in the base model layer, a NeuralProphet model is used to capture the seasonality and event regularity of gas production and other drainage parameters, and an LSTM model is used to learn deep nonlinear time series dependence between multiple parameters. Then, in the meta-model layer, the prediction results of the two base models are spliced with static geological parameters such as coal seam depth, thickness and gas saturation, and input into a random forest regression model for nonlinear fusion, and finally, high-precision gas production prediction values are output. The advantages of the two types of base models are integrated in a soft fusion manner, and the geological static information is effectively combined, so that the accuracy and robustness of the coalbed methane well gas production prediction are significantly improved, and reliable data support is provided for accurate development and production control of the coalbed methane field.

The invention relates to the technical field of organic synthesis and discloses a Sr-composite molecular sieve catalyst as well as a preparation method and application thereof. The prepared ZSM-5 and SAPO-34 are subjected to composite modification, and the binder, the filler and the peptizing agent are introduced, so that the mechanical strength, the thermal stability and the forming performance of the catalyst are improved; by introducing the Sr metal salt, the too strong acidity of the molecular sieve can be reduced, side reactions caused by the too strong acidity are reduced, generation of carbon deposition can be inhibited, the catalyst can keep good stability in the long-term reaction process, and the service life of the catalyst is prolonged. The Sr-composite molecular sieve catalyst prepared by the method has the characteristics of excellent catalytic activity, high low-carbon olefin selectivity, strong stability and long service life, is highly in line with large-scale industrial production requirements, and shows remarkable economic value and wide industrial application prospects.

The application provides a fuel cell hybrid electric vehicle energy management method based on a sand cat group optimization algorithm fusion optimization, and comprises the following steps: in an offline optimization stage, an improved sand cat group optimization algorithm SCSO is used to globally optimize membership function parameters and rule weights of a fuzzy controller, so that a high-performance basic fuzzy controller is obtained; in an online adjustment stage, the basic fuzzy controller obtained in the offline optimization stage is taken as a starting point of a deep deterministic policy gradient algorithm DDPG, an equivalent factor adjustment amount is output by a DDPG agent according to real-time vehicle states, and the output of the basic fuzzy controller is dynamically corrected, so that adaptive optimization of the strategy is realized. The application realizes the consideration of global optimality and real-time adaptability, thereby effectively reducing hydrogen consumption of the whole vehicle, improving fuel cell durability, and improving system dynamic response performance.

The invention provides an intelligent pneumatic conveying system and a fault diagnosis method thereof, and relates to the technical field of industrial automation control, and the method comprises the steps: dividing multi-channel original time sequence data, constructing a first data set and a second data set, according to the sensor abnormal data detection and restoration method based on the improved VAE, abnormal data detection and restoration are carried out on data of the first data set, and noise reduction processing is carried out on data of the second data set based on a self-adaptive noise reduction processing method. According to the method, by constructing the double-path deep learning network, the system macroscopic state and the component microscopic dynamic state are deeply fused, accurate early diagnosis is achieved, in combination with an innovative intelligent data preprocessing mechanism, the data quality and the model robustness are effectively improved, the adaptive feature fusion and confidence evaluation functions are achieved, and the reliability of the system is improved. The system has excellent generalization ability and decision safety, the non-planned shutdown and maintenance cost can be remarkably reduced, and the reliability and economical efficiency of system operation are improved.

This invention relates to the field of non-contact gesture recognition technology, and more specifically, to a lightweight multimodal fusion gesture recognition method based on Rayvision Cloud. The method includes: acquiring video data, micro-Doppler data, and point cloud sequences; calculating the normalized values ​​and effective acquisition time thresholds for the three modalities, and dividing the effective acquisition time into uniform periods; inputting the effective acquisition data of the three modalities into corresponding feature extraction networks to obtain modal features and then performing feature fusion; and inputting the acquired multimodal fusion features into a gesture classification network for training. This invention uses video data, micro-Doppler data, and point cloud sequences as modal data, reducing the impact of single-modal data defects on gesture recognition results, achieving complementary advantages between various modalities, and improving the accuracy of gesture recognition in complex application scenarios.

This invention discloses a method for predicting geological risks in TBM construction based on multimodal historical data fusion, belonging to the field of tunnel and underground engineering technology. This method includes: aligning and unifying the format of the first multimodal geological data; constructing an initial TBM construction geological risk prediction model, including a text semantic feature extraction module, a graphic structure feature extraction module, a multimodal feature fusion module, and a risk classification prediction module; training the parameters of the initial TBM construction geological risk prediction model using a sample set to obtain a final TBM construction geological risk prediction model; inputting the acquired second multimodal geological data into the final TBM construction geological risk prediction model to output the geological risk type and probability of the section to be predicted. This invention achieves advanced prediction of geological risks ahead by deeply exploring the potential value of historical geological exploration reports and design drawings.