284results about How to "Efficient extraction" patented technology

The invention discloses a small target detection method based on a Fourier mixed attention mechanism. An improved small target image detection network model based on RT-DETR is researched and designed, and a Basic Block module in a backbone network is replaced by a self-developed Fourier mixed attention enhancement module (FTABlock). The module is composed of a Fourier transform attention module (FTAModule) and a hybrid dynamic convolution feedforward network (DKMixFFN). The FFTModule calculates the attention weight in the frequency domain through Fourier transform, and strengthens small target feature expression in combination with position coding and multi-head attention; the DKMixFFN adopts a dynamic convolution kernel and a multi-scale dynamic convolution layer to realize adaptive modeling of multi-scale features. According to the method, under the synergistic effect of frequency domain attention and dynamic convolution, the small target feature extraction and perception capability of the RT-DETR model is effectively enhanced, and the small target detection precision is improved.

The invention provides a multi-reflection superposition enhanced microwave displacement measurement method and system, and the method comprises the steps: S1, constructing a microwave multi-reflection propagation path, enabling a microwave radar to transmit electromagnetic waves to the microwave multi-reflection propagation path as a transmission signal, and receiving a corresponding echo signal; s2, extracting length change information of a microwave multi-reflection propagation path according to the echo signal; s3, establishing a linear superposition relation model between the total length of the microwave propagation path and respective displacements of W reflection surfaces in the microwave multi-reflection propagation path; w2; and S4, based on the linear superposition relation model, displacement information of the corresponding reflection surface is obtained and output. According to the invention, a larger phase variation can be obtained under the condition of small displacement or weak vibration, so that the measurement sensitivity is improved and the signal-to-noise ratio is improved.

The invention provides a teaching data conversion method and device oriented to intelligence and electronic equipment, and relates to the technical field of intelligence. The teaching data conversion method for the intelligent teaching system comprises the steps of obtaining to-be-converted teaching data in a first data format; reading metadata information of the to-be-converted teaching data declared in the configuration file; calling a target interface method aiming at the first data format according to the metadata information by utilizing a conversion engine, and obtaining feature data of the body agent during execution of the target task from the to-be-converted teaching data; wherein the conversion engine is a class which is defined by a parent class and is used for converting teaching data in a non-second format into teaching data in a second format; and obtaining the to-be-converted teaching data in the second format according to the feature data by using an aggregation method provided by the conversion engine. According to the method, standardized conversion of the intelligent teaching data with different sources can be realized efficiently at low cost.

The application discloses a data crawling analysis method and device based on an industrial internet, and a medium, and belongs to the technical field of industrial internet data processing, and is used for solving the technical problems that the industrial internet data crawling in the prior art cannot be compatible with multiple types of web pages and dynamic loading technology, lacks customized rule configuration, has weak anti-crawling capability, and has low data structuring degree. The method comprises the following steps: collecting user demand data of a user; performing customized configuration on the initial crawling rule under the relevant rule adjustment to obtain a customized rule; performing crawling processing on a target website under relevant dynamic web pages and static web pages, and based on an anti-crawling mechanism, monitoring and controlling corresponding crawling actions to obtain target webpage data crawled; performing data structure analysis processing on the target webpage data to obtain a data structure format; performing distributed storage on the data structure format to obtain data storage result information; and feeding back and adjusting the customized rule.

The utility model relates to the technical field of automobile door panel processing, and discloses a welding tool for assembling an automobile door panel, which comprises a machine body, a fixed adjusting component and a protective component are arranged above the machine body, the fixed adjusting component comprises a rectangular frame, and a door panel body is placed on the upper surface of the rectangular frame. According to the welding tool for assembling the automobile door panel, through combination of an air cylinder and a rubber pad, stable clamping of a door panel body is achieved, the rubber pad protects the surface of the door panel from being damaged, the adaptability of equipment to door panels of different sizes is enhanced through a limiting rod and a swing piece, the height of a rectangular frame is adjusted through a traction rope, and the door panel is accurately positioned; the welding assembly supports double-face welding, the efficiency is greatly improved, a protective curtain and a collecting frame are matched with a breathable grid plate, welding smoke is effectively extracted and filtered, the air quality is improved, the health of operators is guaranteed, residues are conveniently collected and discharged through the design of a scraper, impurity retention is avoided, the welding process is optimized through the overall structure, and the production efficiency, the environmental protection property and the safety are improved.

The invention belongs to the technical field of biochemical engineering, and relates to an industrial extraction method and application of L-hydroxyproline adapted to the scale of a 30T fermentation tank, and the extraction method comprises the following steps: preparing raw materials; recycling membrane filtration and mother liquor; decoloring is performed; decarburization; evaporating, concentrating and crystallizing; performing centrifugal separation; drying and crushing. The method solves the technical problems that the yield is low (generally less than 70%), the process flow is long (multiple times of chromatography / recrystallization is needed), the product purity is difficult to reach the standard (most 98-99.5%) and the like, the L-hydroxyproline fermentation liquor produced by a 30T fermentation tank is used as a treatment object, efficient extraction is realized through a pretreatment-purification-refining three-stage process system, and the method is suitable for industrial production. According to the method, the primary mother liquor generated by centrifugal separation is innovatively applied to the next batch of membrane filtration process, so that raw material recycling is realized, the purity of the final product stably reaches 99.9%, the overall yield is increased to 75%, and the method has the advantages of simplified process, stable yield and controllable quality, and can be directly applied to large-scale industrial production of L-hydroxyproline.

The application discloses the technical field of power load prediction, and discloses a power load prediction method based on a CNN network and an AdaRNN model, and comprises the following steps: acquiring an externally input power load data set; dividing the power load data set into a plurality of most dissimilar data periods according to a time distribution rule; extracting the space-time correlation between various features in the plurality of most dissimilar data periods by using a CNN network to obtain corresponding features of the power load; and inputting the corresponding features of the power load into a pre-trained Conv-AdaRNN model to perform power load prediction. The application improves the generalization of power load data prediction and effectively extracts feature points related to the power load.

The invention provides a non-contact life signal processing method in a strong clutter environment, and belongs to the field of data processing.The method comprises the steps that extreme points in a signal data sequence of a user and first and last signal data are recorded as to-be-corrected data; obtaining an amplitude adjustment parameter and a slope adjustment parameter of each piece of to-be-corrected data according to a numerical value difference and an acquisition time difference between each piece of to-be-corrected data and the to-be-corrected data on the left side and the right side of the to-be-corrected data; and combining the numerical value of the to-be-corrected data to obtain a corrected value of the to-be-corrected data, thereby obtaining upper and lower envelope lines. The invention aims to solve the problem that the signal data of a user is processed by an EMD (empirical mode decomposition) algorithm due to the fact that an upper envelope line and a lower envelope line which are obtained through extreme points in a signal data sequence are greatly influenced by noise because the noise in an environment where the user is located has a great influence on the signal data of the user. And the accuracy of the obtained heart rate signal and respiration signal is low.

This application provides a method, apparatus, equipment, and medium for predicting mercury concentration at the outlet of an ammonia-based desulfurization system. The method includes collecting historical operating data of the ammonia-based desulfurization system and corresponding outlet mercury concentration data, preprocessing the data to obtain time-series training samples of key feature subsets; constructing a dual-stream parallel prediction model, including a first prediction branch and a second prediction branch; constructing a dynamic confidence gating unit to extract state features reflecting the fluctuation of slurry composition and the severity of load changes in the time series, and calculating the dynamic confidence weights of the prediction output of the first prediction branch and the calculation results of the second prediction branch, then weighted and fused to obtain the final predicted outlet mercury concentration value; during training, training the model based on a constructed hybrid loss function containing prediction error terms and physical consistency constraint terms; collecting real-time operating data of the ammonia-based desulfurization system, inputting it into the trained prediction model, and outputting the predicted outlet mercury concentration value for future times.

The utility model relates to the technical field of medicinal plant raw material extraction, especially relates to a low-temperature press of sea buckthorn seed oil. A low-temperature press of sea buckthorn seed oil has a support, a press cylinder, a first servo motor, a spiral oil press roller, an oil storage tank and an arc-shaped heat dissipation fin group, the press cylinder is installed at the top of the support, the right side of the press cylinder is provided with the first servo motor, the output shaft of the first servo motor is fixedly connected with the spiral oil press roller, the space inside the press cylinder and the spiral oil press roller gradually decreases from right to left, the lower part of the press cylinder is fixedly connected with the oil storage tank, and the upper part of the press cylinder is fixedly connected with the arc-shaped heat dissipation fin group in a symmetrical manner around the periphery. The utility model adopts the design that the spiral oil press roller and the internal space of the press cylinder gradually decrease, and is combined with the arc-shaped heat dissipation fin group, so that the oil extraction rate is improved, the effective heat dissipation system is matched to maintain a low-temperature operation environment, the nutritional ingredients in the sea buckthorn seed oil are maximally reserved, and the quality of the product is improved.

The invention provides a traditional Chinese medicine composition for preventing hair loss, growing hair and removing dandruff as well as a preparation method and application thereof, and belongs to the technical field of traditional Chinese medicines. The invention discloses a traditional Chinese medicine composition for preventing hair loss, growing hair and removing dandruff. The traditional Chinese medicine composition is prepared from the following raw materials in parts by weight: 8-12 parts of lithospermum, 8-12 parts of paris polyphylla, 15-25 parts of duckweed, 8-12 parts of golden larch bark, 8-12 parts of lucid ganoderma, 15-25 parts of cocklebur fruit, 15-20 parts of gallnut, 8-12 parts of radix sophorae flavescentis, 5-10 parts of schizonepeta, 8-12 parts of radix saposhnikoviae, 4-8 parts of notopterygium root, 5-10 parts of schisandra chinensis, 5-10 parts of elecampane, 5-10 parts of eclipta, 5-10 parts of ginseng, 5-10 parts of angelica sinensis, 5-10 parts of saponin, 5-10 parts of soapberry and 5-10 parts of broadleaf holly leaf. , 8-12 parts of fructus ligustri lucidi, and 4-8 parts of herb of wiss mannii The invention provides a traditional Chinese medicine composition for preventing alopecia, growing hair and removing dandruff, aiming at improving the problems of alopecia, dandruff and scalp discomfort.

The present application belongs to the technical field of data cleaning, and particularly relates to a redundant data cleaning method, device, equipment and medium. The method comprises the following steps: obtaining power supply redundant data; performing filtering processing on the power supply redundant data to obtain a filtered data sequence; calculating the information entropy value of each group of filtered data in the filtered data sequence, and calculating the fusion weight corresponding to each group of filtered data according to the information entropy value of each group of filtered data; and superimposing the filtered data according to the fusion weight to obtain fusion data and outputting the fusion data. The present application firstly performs primary filtering processing on the redundant data to filter certain noise data; on this basis, the fusion weight and the information entropy value are calculated, and finally the fusion data is obtained, thereby further improving the accuracy of the fusion data.

This invention relates to the field of relay protection technology for tugboat DC power grids. It provides a microgrid short-circuit fault protection method based on the difference in current change rate, comprising the following steps: continuously collecting current data; setting two sliding windows to record historical current information; fitting the current data of the first sliding window according to a fitting formula; calculating the current change rate of the first sliding window using the fitting formula; obtaining the operating quantity of the longitudinal differential protection based on the current change rate; calculating the sum of the maximum value of the differential current and the average absolute value of the current at both ends of the cable based on the current data of the second sliding window; obtaining the braking quantity of the longitudinal differential protection from the sum of the maximum value of the differential current and the average absolute value of the current at both ends of the cable; and determining that a bipolar short-circuit fault has occurred in the cable and disconnecting the faulty line when the operating quantity and the braking quantity meet predetermined conditions. This method can expand the compatibility range with changes in transition resistance while shortening the fault detection time, thereby improving the reliability and stability of the DC microgrid.

The invention discloses a slice laser based on a topological structure dodging layer. The slice laser comprises a pumping system, a pumping light injection module and a composite dodging waveguide structure, pumping light generated by the pumping system enters the composite dodging waveguide structure through the pumping light injection module; the front end face of the composite dodging waveguide structure is plated with a pumping light wavelength high-reflective film and a laser wavelength antireflection film, and the rear end face of the composite dodging waveguide structure is plated with a pumping source wavelength and laser wavelength high-reflective film; the composite dodging waveguide structure comprises a slice gain medium, a topological structure dodging layer and a middle refractive index layer which are sequentially arranged from the front end face to the rear end face. The topological structure dodging layer is composed of a low-refractive-index material layer and a high-refractive-index material topological graph arranged on the low-refractive-index material layer; and the pump light injection module is fixed above the high refractive index material of the middle refractive index layer or the topological structure dodging layer. Pumping uniformity is improved fundamentally, high-order mode suppression and efficient fundamental mode extraction are facilitated, and output power and light beam quality are improved at the same time.

The invention belongs to the technical field of wind power generation, and particularly relates to a wind turbine generator state evaluation method and a related device, and the method comprises the steps: obtaining an electromagnetic transient response signal of a to-be-evaluated wind turbine generator; performing empirical mode decomposition on the electromagnetic transient response signal of the wind turbine generator to be evaluated to obtain a plurality of image forming model components; performing feature extraction on each image forming model component, and performing normalization processing to obtain a feature vector; classifying the feature vectors by using a pre-constructed state classification model to obtain a state evaluation result of the to-be-evaluated wind turbine generator; wherein the pre-constructed state classification model is a pre-trained support vector machine; according to the method, the fault features in the electromagnetic transient response signals of the wind turbine generator can be effectively extracted, and the accuracy and robustness of state evaluation are improved; meanwhile, non-linear and non-stationary signals can be decomposed in a self-adaptive mode, the time domain features of the signals are reserved, and the information amount and accuracy of state evaluation are improved.

The invention discloses rhubarb polysaccharide as well as a preparation method and application thereof, the total sugar content of the rhubarb polysaccharide is 76.25%, the protein content is 3.4%, and the rhubarb polysaccharide contains active components (such as glucose, arabinose and galacturonic acid) in a specific proportion. The preparation method disclosed by the invention can be used for rapidly and efficiently extracting the rhubarb polysaccharide, and is low in energy consumption, simple and convenient to operate, low in process cost and suitable for large-scale industrial production. Meanwhile, the process ensures the stability and activity of the polysaccharide, and a wide prospect of providing drugs or food additives for preventing pancreatitis and regulating intestinal bacteria for the development and application of the rhubarb polysaccharide is provided. The rhubarb polysaccharide can obviously regulate intestinal flora disorder of mice with pancreatitis induced by arginine so as to up-regulate tight junction protein and promote tissue repair and regeneration, and has a great application prospect in the fields of regulating intestinal bacteria, preventing and treating hypertriglyceridemia type acute pancreatitis, metabolism-related new drugs and special medical food development and the like. Wide development and application prospects are realized.

The invention belongs to the technical field of environmental engineering and chemical separation, and particularly relates to a deep-eutectic solvent composition for efficiently extracting phenolic compounds and a preparation method of the deep-eutectic solvent composition. The deep-eutectic solvent composition for efficiently extracting the phenolic compounds, provided by the invention, consists of a hydrogen bond acceptor and a hydrogen bond donor in a molar ratio of (1-2): (1-9), the hydrogen bond donor comprises n-caprylic acid or n-hexyl alcohol; the hydrogen bond receptor comprises 1-ethyl-3-methylimidazole chloride or tetrabutylammonium bromide. The eutectic solvent composition has the advantages of high partition coefficient, low viscosity, high selectivity, rapid phase separation, environmental friendliness and the like, can efficiently extract phenol, resorcinol, o-cresol, m-cresol and p-cresol, and has single-stage extraction efficiency gt; the regeneration rate is high, and the method is suitable for high-concentration phenolic wastewater, polyhydroxy phenolic wastewater, continuous process and low-toxicity emission requirement scenes.

The application provides an RGB-D image feature collaborative fusion method based on transfer learning and width learning, and comprises the following steps: obtaining an RGB-D data set, performing preliminary training through a neural network, and performing retraining in the data set after modifying the structure; after feature extraction, performing correlation analysis and fusion on RGB image features and depth image features; and using width learning to classify and identify the fused features. The application can reasonably fuse the features of RGB images and depth images, ensure that the feature information of color images and depth images can complement each other, improve the running speed of the system by using width learning, and finally make the classification result have higher accuracy and reliability.

The invention discloses a method for recovering valuable elements in liquid obtained after bauxite wet desulphurization, and relates to the technical field of bauxite resource utilization. Comprising the following steps: carrying out ceramic membrane filtration pretreatment on the bauxite wet desulfurization liquid to remove impurities; the silicon element is recycled in a targeted manner through a compound flocculation-calcination process to obtain the white carbon black; the iron element is recycled through ozone oxidation-ammonium bicarbonate precipitation, and high-purity iron oxide is prepared; separating and purifying an aluminum element through a seed crystal induction-calcination process to obtain metallurgical-grade aluminum oxide; a fluorine-containing precipitant is used for selectively precipitating and recovering magnesium element to obtain a high-quality magnesium salt product; and neutralizing the tail liquid, and discharging after reaching the standard. According to the method, accurate separation and high-valued utilization of four elements of aluminum, iron, silicon and magnesium are realized through a synergistic process of targeted recovery of silicon, gradient separation of iron and aluminum and selective precipitation of magnesium, and the problems of silicon resource waste, low element separation efficiency, insufficient product purity, secondary pollution and the like in a traditional process are thoroughly solved; all-component resource comprehensive utilization of desulfurized liquid resources is achieved.

The invention relates to the technical field of ultra-precision machining, in particular to an ultra-precision cutting machining surface roughness online measuring method and system. First, a machining process signal is read. Secondly, calculating the weight of the processing process signal; the processing process signals are weighted according to the weights, and the weighted processing process signals are fused to serve as first input data; the machining process signal is used as additional input data. Then, obtaining a first input result and an additional input result according to the first input data and the additional input data, and obtaining a first fusion result according to the first input result and the additional input result; thirdly, inputting a high-dimensional feature fusion branch according to the first fusion result and additionally extracted first high-dimensional feature information and additional high-dimensional feature information to obtain a second fusion result; and finally, obtaining a surface roughness measurement result. According to the method, surface roughness prediction of ultra-precision cutting machining is more accurate.

The invention belongs to the field of space target imaging, and discloses a space target three-dimensional reconstruction method for inhibiting dark color interference through combination of brightness and transparency, which is used for solving the problems of poor imaging quality and inaccurate pose estimation of a space target. According to the method, multiple requirements of improving the quality of an input image, adjusting a pose and training a three-dimensional model are jointly completed, and low-quality imaging of an incomplete imaging system is preprocessed based on lucky imaging; and then, dense feature points are extracted by using a pre-training network based on no detector, estimation of a three-dimensional initial point cloud and an initial pose of a camera is carried out, and pose joint adjustment is carried out by combining RANSAC and Bundle AdJustlement. During training, an MLP-based optimizer is used for calculating the deviation of pose adjustment, meanwhile, training of parameters of three-dimensional Gaussian primitives is carried out, and after various physical constraints are added, a one-stop assembly line can be implemented. According to the method, the good basic shape and the sharp edge of the model are kept, the good anti-noise capability is achieved, and meanwhile the real-time performance and the robustness are achieved.

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 application discloses an electrochemical granulation electrode deintercalation lithium extraction membrane stack electrolysis equipment, which comprises an electrolysis tank and an electrode assembly; the electrolysis tank comprises a first side plate, a piece of side plate rubber pad, a plurality of tank bodies, another piece of side plate rubber pad, a second side plate and a first tank body rubber pad, an anion membrane and a second tank body rubber pad arranged in sequence between two adjacent tank bodies; the tank bodies are divided into anode tanks and cathode tanks, the anode tanks and the cathode tanks are arranged alternately in an X direction from the first side plate to the second side plate, and the number of the anode tanks is equal to that of the cathode tanks; and an elastic porous pad is laid on the inner cavity bottom wall of the tank body. The electrolysis equipment can realize efficient extraction of lithium resources in salt lake brine, has the advantages of high lithium recovery rate and current efficiency, and is convenient to install, maintain and observe and monitor in the lithium extraction process.

The utility model relates to a forced air cooling type energy storage battery PACK, which comprises a shell, two battery modules are arranged in the shell, each battery module comprises a plurality of battery cells arranged along a first direction, a ventilation duct is formed between two adjacent battery cells, a sealing partition plate is arranged on one side surface of each battery module along a second direction, and the two ends of each battery module are respectively connected with the shell in a sealing manner. A heat dissipation cavity which is used for accommodating the battery module and is communicated with the ventilation air duct is formed in the shell; a plurality of groups of air holes are formed in the sealing partition plate at intervals in the first direction, and each group of air holes comprises one or more through holes formed at intervals in the height direction; the air holes are correspondingly communicated with the ventilation air duct; an adjusting valve for adjusting the ventilation area is arranged in the through hole; a main air duct is formed between the two battery modules through the two sealing partition plates, and heat dissipation fans are arranged at outlets in the two ends of the main air duct respectively; and the shell is provided with a hollow area for ventilation. According to the utility model, the heat dissipation effect is improved, and the problem of poor temperature consistency of the battery module is solved.

The embodiment of the invention discloses a communication signal multi-parameter intelligent parallel extraction method. The method comprises the following steps: firstly, acquiring a to-be-measured signal and extracting an I / Q component; processing the I / Q components through a multi-scale feature extraction structure to obtain high-dimensional features; extracting shared features from the high-dimensional features through a plurality of residual block groups, inputting a plurality of parallel task branches, and enhancing the shared features in the branches by using a channel attention mechanism to obtain output features; the output features of all task branches are transformed through a learnable linear layer to form a key matrix and a value matrix, the current task features are transformed through a learnable linear layer to form a query matrix, and the input features are segmented into a plurality of heads by means of a multi-head attention mechanism; each head obtains output of a single head based on zoom dot product self-attention operation, and results of all the heads are integrated to obtain enhanced features of the current task; and determining a communication signal parameter prediction label according to the enhanced features, and realizing multi-parameter efficient parallel extraction.

The utility model relates to an aluminum industry waste emulsion sludge extraction device which comprises a lower cylinder body, the top of the lower cylinder body is connected with an attaching plate through bolts, an upper pipe body is arranged above the attaching plate, the outer wall of the upper pipe body is connected with a plurality of evenly-distributed stand columns through bolts, and the stand columns are connected with the lower cylinder body through bolts. The bottom ends of the multiple stand columns are connected to the top of the attaching plate through bolts, a butt joint plate is slidably connected to the position, close to the bottom, in the upper pipe body, an outer pipe body is connected to the top of the upper pipe body through bolts, an inner pipe body is arranged in the outer pipe body in a sleeved mode, and an extraction cylinder is arranged in the inner pipe body in a sleeved mode. A plurality of uniformly distributed liquid leakage holes are formed in the inner wall of the extraction barrel. According to the technical scheme, parts such as the inner pipe body, the arc-shaped groove, the poke rod, the side plate and the penetrating groove are matched with one another, so that the inner pipe body can be driven to move up and down and rotate, intermittent feeding can be performed, sludge can be effectively and uniformly intercepted and filtered, and the sludge extraction efficiency is improved.

This invention relates to the field of resin asphalt concrete mixing technology and discloses a resin asphalt concrete mixing device for steel bridge deck paving, including a base frame, support rollers, a mixing tank, and a dual-shaft motor. The mixing tank is rotatably connected above the base frame and is drivenly connected to the outer surface of the support rollers. A spiral mixing blade and a material lifting plate are fixedly connected to the inner wall of the mixing tank. A tank cover is detachably connected to one end of the mixing tank. A rotating shaft is rotatably connected to the inner wall of the end of the mixing tank away from the tank cover. A multi-tooth support frame is fixedly connected to the outer surface of the rotating shaft. A shearing rod is fixedly connected to the outer surface of the multi-tooth support frame. A vacuum pump is fixedly installed on the inner bottom wall of the base frame. A support rod is fixedly connected to the upper surface of the base frame. A rotating vent is fixedly connected to the outer surface of the support rod. This device uses a dual-shaft motor to drive the tank body to rotate and lift and the center to shear at high speed to achieve three-dimensional uniform mixing. It is equipped with a hollow rotating negative pressure degassing device to remove bubbles, which solves the defects of resin asphalt clumping and material gas content.

This application discloses a wide-bandgap perovskite solar cell based on fullerene derivatives, a tandem solar cell, and a method for fabricating the same, belonging to the field of perovskite solar cell technology. A wide-bandgap perovskite solar cell based on fullerene derivatives includes a conductive substrate, a first hole transport layer, a self-assembled monolayer, a wide-bandgap perovskite light-absorbing layer, a fullerene derivative electron transport layer, a first electron modification layer, and a first metal electrode layer, sequentially stacked. The molecular formula of the wide-bandgap perovskite is FA. 0.5‑1.0 Cs 0.1‑ 0.3 PbI 1.5‑2.0 Br 1.0‑1.5 The fullerene derivative is 5'-Phenyl-5'. H -[5,6]fullereno-C 60 - I h -[1,9:2',3']pyrido[3,2- b [indole. Through the synergistic effect of electron transport, defect passivation, and thin film stability enhancement, the optoelectronic performance and long-term operational stability of wide-bandgap perovskite and all-perovskite stacked devices are significantly improved.]

This invention discloses a dismantling device and method for the substructure of a bridge, belonging to the technical field of bridge dismantling equipment. It includes self-locking casters, a lifting platform, a support platform, a support panel, an insertion device, a rotating device, a drilling device, a top-support device, a release device, a transmission device, and a removal device. The insertion device and the rotating device are mounted on the support platform. The drilling device is connected to the insertion device. Two top-support devices are symmetrically arranged on the support panel. The release device and the removal device are both mounted on the support panel. The transmission device is also mounted on the removal device. This device provides a convenient method for dismantling square timber and back bracing by using screws or similar methods to limit their position on the side of the bridge abutment. By saving time and effort and releasing most of the required pulling force to the ground, the stability and operational efficiency of this device are improved.

This invention discloses a container ship trajectory prediction method based on EMD-PSO-GRU-RBFNN, belonging to the field of intelligent shipping technology. The method takes ship AIS data as input, including time, longitude, latitude, speed, and heading angle information; preprocesses the AIS data to divide it into training and test sets; constructs a trajectory prediction model, calculates the difference between the predicted values ​​of the training set and the actual ship position coordinates to obtain the residual signal; adaptively decomposes the residual signal using empirical mode decomposition to obtain intrinsic mode components and residual components; constructs prediction models for high-frequency and mid-to-low-frequency components, reconstructs the prediction results of each component to obtain the residual prediction value, and superimposes it with the baseline prediction results of the test set to obtain the corrected test set output; the corrected test set output is then evaluated. This invention can model and compensate for trajectory errors at different frequency scales, improving the stability and reliability of trajectory prediction.