1163results about How to "Improve distribution" patented technology

The invention discloses a Transform large model training method based on cloud edge collaboration. The method comprises the following steps: S1, constructing a data set; s2, preprocessing the data set; s3, segmenting the preprocessed data set by using a data parallelism strategy, and performing distributed training by combining data parallelism and model parallelism technologies; s4, monitoring computing resources and network conditions of the edge equipment and the cloud server in real time, dynamically adjusting a data distribution strategy, and optimizing computing load distribution; s5, a multi-agent dominant actor-commentator algorithm is combined with a multi-arm tiger machine strategy, and the layered structure of the Transform model is adjusted in real time; s6, performing partial model reasoning tasks on the edge equipment in combination with the edge reasoning capability; s7, protecting data privacy by adopting a federal learning strategy and utilizing a differential privacy technology; and S8, performing hardware acceleration by utilizing heterogeneous computing resources. According to the method, a cloud edge cooperative computing and intelligent scheduling method is adopted, and large model training and reasoning with efficient privacy protection are realized.

The invention relates to the technical field of powder metallurgy forming, and discloses a compressor valve plate forming device for powder metallurgy and a forming method thereof.The forming device comprises a lower die body, a guide support is fixed to the lower die body, an upper die body is slidably connected to the guide support, and a servo hydraulic cylinder is fixed to the top of the guide support; a servo hydraulic cylinder is arranged on the lower die body, the telescopic end of the servo hydraulic cylinder is fixedly connected with the upper die body, a forming groove is formed in the lower die body, a forming head is arranged on the upper die body, a feeding box is arranged on the lower die body, and a push rod assembly used for adjusting the working position of the feeding box is fixed to one end of the lower die body. The metal powder can be uniformly distributed in the forming groove, so that the thickness of a formed workpiece is uniform, a lubricant can be conveniently and quantitatively sprayed into the metal powder, the flowing property of the metal powder is improved, and the machining quality of the workpiece is improved.

The invention discloses a method and a system for realizing mine communication signal improvement based on edge computing, and relates to the technical field of mine communication signal improvement, and the method comprises the following steps: integrating a real-time monitoring system on each edge node, so as to continuously detect the fluctuation of communication flow and data request quantity, the real-time monitoring system identifies a potential communication peak period by collecting real-time data and generates an early warning signal to notify a network administrator or automatically trigger a load management program. According to the invention, through the real-time monitoring and early warning system on the edge node, active monitoring and peak prediction of mine communication are realized, the overload risk caused by traffic surge is reduced, and the stability of data transmission is ensured. And intelligent scheduling and dynamic resource allocation are combined, so that efficient execution of tasks under high load is ensured, and node overload is prevented. A self-adaptive load balancing and feedback mechanism enables the system to dynamically adjust distribution, quickly respond to a bottleneck and improve self-adaptability and stability.

The present application provides a method for diagnosing and / or evaluating the presence or absence, severity or degree of the improvement of a retinal disease in a subject, which comprises determining and / or evaluating circulatory parameters, retinal function, retina morphology and / or visual relating quality of life (QOL).

The invention discloses an information transmission attack event intelligent prediction model training optimization method, which comprises the steps of a data preparation stage, a feature engineering stage, a model selection and construction stage, a model training stage and a model evaluation and optimization stage, and is characterized in that data are collected from multiple channels, noise is cleaned and removed, missing values and abnormal values are processed, and accurate labeling is carried out; then key features are extracted, redundant irrelevant features are removed through selection, standardized transformation is carried out, a proper model is selected based on data features, an integrated learning combination model is adopted, then a training set, a verification set and a test set are divided, a proper optimizer is selected, parameters are adjusted to be over, training is carried out through the training set, and over-fitting is prevented through the verification set; and finally, using index evaluation of accuracy and recall rate, and adjusting hyper-parameters and continuous optimization of the model structure through cross validation. According to the method, data collection, data imbalance processing and data noise interference processing are adopted, data diversity is improved, and interference of data noise is reduced.

The invention relates to the technical field of temperature regulation and control, in particular to an intelligent temperature regulation and control method and system based on an intelligent building. The method comprises the following steps: performing functional area air flow simulation on intelligent building structure design data to obtain functional area air conditioning flow simulation data; performing air circulation regulation of different functional areas on the functional area air regulation flow simulation data to obtain temperature coupling air circulation regulation data; indoor temperature circulation adjustment compensation is conducted on the temperature coupling air circulation adjustment data according to the human flow real-time monitoring data, and indoor temperature circulation adjustment compensation data are obtained; performing regulation and control logic training optimization on the indoor temperature circulation regulation and compensation data to obtain temperature regulation and control logic optimization data; and performing automatic firmware design based on the temperature regulation and control logic optimization data to obtain temperature regulation and control logic firmware. According to the invention, the temperature regulation and control technology is optimized, so that the temperature regulation and control technology is more perfect.

The invention relates to the technical field of planer type milling machines, and discloses a planer type milling machine with an air cooling mechanism, the planer type milling machine comprises a planer type milling machine, the top of the planer type milling machine is provided with a planer type cross beam, the outer wall of the planer type cross beam is slidably connected with a sliding seat, the surface of the sliding seat is provided with a ram, and the outer wall of the ram is fixedly connected with a mounting frame; the top of the mounting frame is fixedly connected with a driving motor; and a circumferential mechanism is arranged at the bottom of the mounting frame. According to the planer type milling machine with the air cooling mechanism, the circumferential mechanism is arranged, so that sprayed air rotationally blows towards a cutting tool bit, the sprayed air rotationally blows around the tool bit, chippings around the tool bit can be effectively helped to be removed, the chippings are pushed outwards through centrifugal force generated by rotating airflow, and the chippings can be effectively removed; by means of the design, it can be guaranteed that a cutting area is kept clean, the cutting efficiency is improved, and the situation that the cutting quality is reduced due to accumulation of the chippings is avoided.

The invention discloses a CFD-based liquefied natural gas after-condenser structure optimization method and device. The method comprises the steps that an initial model is established based on an assembly drawing, and geometric pretreatment is completed; extracting a fluid channel, segmenting and selecting a computational domain, and carrying out local encryption on a key region; generating a polyhedral grid by using a grid division tool, and completing grid independence verification; importing the polyhedral mesh into a solver, and carrying out model setting; boundary and wall surface conditions are set, and solver parameters are configured; setting an initial flow field through a hybrid initialization method, and establishing a flow monitoring window and a temperature monitoring window to adjust simulation until convergence; analyzing the temperature cloud picture and the flow vector diagram, and adjusting the structure of the baffle plate; the tower diameter and the height of the filler layer are adjusted according to a physical quantity field change curve of the filler layer, and the performance of the recondenser is continuously improved. The problem of how to deeply study the flow heat transfer rule in the after-condenser and obtain the flow field distribution characteristics is solved, so that the working performance of the after-condenser is improved.

The invention discloses a capacitive voltage transformer for a gas insulated substation, which comprises a capacitive voltage divider and an electromagnetic unit, the capacitive voltage divider comprises a high-voltage capacitor C1 and a medium-voltage capacitor, the medium-voltage capacitor comprises a medium-voltage capacitor C21 and an externally connected medium-voltage capacitor C22, and the capacitance of the medium-voltage capacitor C22 is greater than that of the medium-voltage capacitor C21; the electromagnetic unit is connected with the capacitive voltage divider, an intermediate transformer and a compensation reactor are arranged in the electromagnetic unit, a full-shielding high-voltage shielding cover is arranged outside a high-voltage coil of the intermediate transformer, a damper and an overvoltage protector are arranged outside the electromagnetic unit, and the damper and the overvoltage protector are placed in an external junction box. The problem that in the prior art, an extra-high voltage transformer is low in lightning impulse overvoltage tolerance capacity is solved, the impulse overvoltage tolerance capacity is improved, and the breakdown risk is reduced.

The invention belongs to the technical field of power battery cooling, and particularly relates to a flow control method and system for a liquid cooling system of an electric vehicle-mounted power battery, and the method comprises the steps: carrying out the parameterization of a geometric model of the liquid cooling system; establishing an equipment resistance model based on the equipment flow resistance characteristics; initializing flow distribution, constructing a dynamic flow distribution model, calculating the flow of each node and judging whether the flow meets a target requirement or not, and if not, readjusting the flow; calculating the resistance of each section, determining the total resistance of the pipeline, and checking the water pump lift; the flow uniformity is judged by calculating the pipeline resistance between the nodes, and if the flow is not uniform, the flow is redistributed or the pipeline design is optimized until the flow is uniform. Through flow control, the problems of non-uniform flow distribution and poor temperature consistency of multiple battery packs in the prior art are solved, the optimal working state of the power battery is ensured, overheating and overlarge temperature difference are avoided, and the operation efficiency, stability, durability and safety are improved.

The invention relates to the technical field of numerical simulation, in particular to a grid model generation method for the fluid displacement process of well cementation engineering. The method comprises the following steps: acquiring geological data of a well cementation project area, and constructing an initial geological model; well cementation engineering design data are obtained, well cementation engineering topological structure modeling is carried out, and a well cementation engineering topological structure model is obtained; according to the initial geologic model and the well cementation engineering topological structure model, fluid dynamic simulation is carried out, and a fluid characteristic matrix is obtained; performing fluid displacement process simulation according to the fluid characteristic matrix to obtain fluid displacement process data; performing fluid displacement process optimization on the well cementation engineering design data based on the fluid displacement process data to obtain optimized fluid displacement process data, and performing resource allocation redistribution on the well cementation engineering design data according to the optimized fluid displacement process data to obtain optimized well cementation engineering design data. According to the invention, the accuracy of fluid displacement simulation can be improved.

The invention provides a marine environment multi-source data fusion processing method and system, and relates to the field of data processing, and the method comprises the steps: carrying out the fusion of marine environment data and edge node data, so as to obtain a fused data set; dividing the fused data set into fragmented data sets, and preliminarily distributing the fragmented data sets to each heterogeneous node to obtain a distribution result; setting performance evaluation indexes, performing performance testing on each heterogeneous node according to the performance evaluation indexes, and recording the performance indexes of each node when processing different fragment data sets to obtain a test result; analyzing a test result, and identifying an abnormal node and an abnormal fragment data set; and performing optimization according to the abnormal nodes and the abnormal fragment data set to obtain an optimization result. According to the method, data distribution is optimized through data fragmentation and dynamic allocation, and the processing efficiency is improved.

The invention discloses a multi-machine parallel operation control method for a network-forming type energy storage converter, and relates to the field of electrical engineering. Establishing a sparse communication network; obtaining the frequency, the voltage effective value, the active power and the reactive power of the node and the active power and the reactive power of a neighbor node; constructing a network construction type converter control strategy combining active support and cooperative control, and calculating the output frequency and voltage effective value of each network construction type converter; and obtaining a three-phase AC voltage waveform through dq conversion, modulating the three-phase AC voltage waveform into a three-phase PWM switch control signal, and applying the three-phase PWM switch control signal to the network-forming converter to achieve a control effect. The active support of the frequency and the voltage aims at the deviation of the frequency and the voltage to rated values in the VSG control, the contradiction between the power control and the frequency and voltage control is eliminated by the cooperative control strategy, the defects of the VSG control are made up by the proposed optimization control method, the frequency and the voltage of the network construction type converter are controlled to be the rated values, and meanwhile, the efficiency of the network construction type converter is improved. And the active power and the reactive power can be accurately and reasonably distributed.

The present invention relates to the field of grain transport technology, and in particular to an agricultural grain transport device, comprising a base, a shell fixedly installed on the top of the base, a feed hopper and a discharge hopper respectively arranged at both ends of the shell, a cylindrical screen arranged inside the shell, a refrigeration mechanism arranged inside the shell, a mounting plate fixedly installed on the side of the discharge hopper away from the shell, a motor fixedly installed on the top of the mounting plate, a rotating shaft fixedly connected to the output end of the motor, and a spiral hollow disk and a solid disk fixedly connected to the outer wall of the rotating shaft. The advantages are: the inside of the shell maintains a low temperature, the low temperature helps to increase the hardness difference between grain and impurities, improve the efficiency of impurity screening, reduce the stickiness between grains, ensure the smooth flow of grain, the screen effectively screens out impurities in the grain, improves the purity of the grain, pushes the grain in layers, prevents excessive accumulation of grain, and reduces the damage of grain due to friction and extrusion.

The invention provides a bucket for testing the digging resistance of a wheel bucket excavator, and solves the problems that the conventional wheel bucket excavator is difficult to directly and accurately measure the resistance in the tangential direction, the lateral direction and the normal direction of the bucket and the distribution condition of the lateral digging resistance in a lug plate of the bucket cannot be tested. According to the technical scheme, a bucket is connected with a bucket wheel body through three or four sets of hinge parts, each set of hinge part is composed of two bucket lug plates and one bucket wheel body lug plate in a matched mode, and the bucket lug plates and the bucket wheel body lug plates are connected through bidirectional force measuring pin shaft sensors penetrating through lug plate shaft holes; pressure ring sensors are arranged between the two sides of the bucket wheel body lug plate and the two opposite bucket lug plates respectively, pin shaft sensors penetrate through shaft holes of the pressure ring sensors, and the pressure ring sensors are fixed to the opposite faces of the two bucket lug plates through annular baffles. And the pressure ring sensor is embedded in the annular baffle plate. By arranging the bucket for testing, the resistance distribution of the hinged part in the bucket is accurately measured, and the reliability and the testing efficiency are high.

The invention relates to a wafer heater and a chemical vapor deposition system, which are used for solving the technical problem of non-uniform distribution of inert gas on the back surface of a wafer. The wafer heater comprises a heating part used for supporting a wafer and heating the wafer, the upper surface of the heating part is provided with a gas distribution channel, the gas distribution channel comprises a gas inlet and at least one first annular groove surrounding the gas inlet, at least one of the at least one first annular groove is provided with an annular wave line contour, and the gas inlet is communicated with the gas inlet; the plurality of first connecting grooves are radially connected between the air inlet and at least one first annular groove along the radial direction of the first annular groove; and the interface part is arranged at the bottom of the heating part, and the gas inlet is configured to be in fluid communication with an inert gas source through the interface part.

The invention discloses a purification device for producing electronic-grade hydrofluoric acid, which relates to the technical field of hydrofluoric acid production, and comprises a rectifying tower body, tower plates movably connected in the rectifying tower body at equal intervals, overflow weirs fixedly connected on the tower plates, a rotating speed convenient regulation and control mechanism and a state monitoring feedback mechanism, the follow-up rod and the tower plates installed on the outer surface of the follow-up rod can be driven to synchronously rotate in the rectifying tower body, so that the gas-liquid contact area is increased, the gas-liquid contact time is prolonged, the mass transfer process between gas and liquid phases is promoted, in the rectifying process, the separation efficiency and purity can be improved, impurities in raw materials can be removed more easily, and the product quality is improved. And the rotation of the tower plate can enhance the mixing of the liquid on the tower plate, so that the liquid is more uniformly distributed on the tower plate, the residence time and dead zone of the liquid on the tower plate are favorably reduced, and the separation efficiency is further improved.

The invention relates to a combined truncated cone type gas-liquid redistributor. The combined truncated cone type gas-liquid redistributor comprises a conical wall and a connecting bottom plate. Theconnecting bottom plate is fixedly arranged at the bottom of the conical wall, and is provided with a plurality of strip-shaped folding plates; the strip-shaped folding plates are of inverted-V structures; a plurality of air holes are formed at equal intervals in both sides of each strip-shaped folding plate; a stop plate is fixedly arranged on the lateral side of the strip-shaped folding plate above each air hole, and an angle is formed between each stop plate and the lateral side of the corresponding strip-shaped folding plate; the angles between the stop plates and the lateral sides of thestrip-shaped folding plates are proportional to the distances from the stop plates to the center of the connecting bottom plate. The combined truncated cone type gas-liquid redistributor has the advantages that the strip-shaped folding plates and the stop plates are adopted to achieve a good distribution effect of washing liquid, and accordingly, the area and time of contact between gases and liquid can be increased and prolonged; the included angles between the stop plates and the lateral sides of the strip-shaped folding plates are proportional to the distances from the stop plates to the center of the connecting bottom plate, so that the differences among the flowing speeds of gases from each air hole are small, the gas distribution effect is good, poor gas-liquid distribution in a packing tower is reduced, and the gas-liquid mass transfer efficiency is improved.

The invention relates to the technical field of valve control, and discloses a flow-controllable water-saving valve which comprises a control module used for receiving crop water demand data, soil humidity data and environmental parameters, calculating irrigation demands according to the data and generating control signals to adjust irrigation time, frequency and flow; the executing mechanism is connected with the control module and used for adjusting opening and closing of the valve according to the control signal; the sensor assembly is used for collecting real-time data of an irrigation area and sending the real-time data to the control module; the valve assembly comprises a valve element capable of adjusting flow and a flow adjusting device, and the flow adjusting device adjusts the opening and closing degree of the valve element according to instructions of the executing mechanism. By adjusting the irrigation time and the unit time flow, the soil moisture penetration depth can be accurately controlled, too deep or too shallow growth of root systems is prevented, root system distribution is optimized, the water absorption and drought resistance of crops are improved, meanwhile, excessive irrigation or deep layer leakage can be avoided, and water resource waste is remarkably reduced.

The invention discloses a touch-type interactive all-in-one machine, and belongs to the field of interactive all-in-one machines, the touch-type interactive all-in-one machine comprises a machine body, a touch screen is fixedly connected to the outer surface of the upper end of the machine body in an embedded manner, a fixing frame is fixedly connected to the outer surface of the lower end of the machine body, a moisture-proof assembly is arranged on the inner side of the fixing frame, and the moisture-proof assembly comprises a movable part and a sealing part; the movable part comprises an air inlet groove formed in the outer surface of the fixing frame, and a first cavity is formed in the inner side of the fixing frame in an embedded mode. The moisture-proof assembly is arranged, and the air inlet groove is blocked and sealed through the movable plate, so that dust and moisture entering the cavity I can be reduced to a certain extent, and the probability of generating condensation in the cavity I can be reduced; and meanwhile, the probability that metal elements in the interaction all-in-one machine are corroded by moisture can be effectively reduced, so that the service life of the interaction all-in-one machine can be remarkably prolonged, and the adaptability and the working stability of the interaction all-in-one machine are improved.

The invention belongs to the field of high polymer materials, and discloses a surface flame-retardant compatibilization glass fiber reinforced PPE / PS alloy material and a preparation method thereof, the alloy material comprises the following raw materials by weight: 10-84.98% of bulk flame-retardant MPPE resin; 5-35% of polystyrene resin; 10 to 40% of flame-retardant glass fiber; 0.01 to 10% of a flame retardant; 0.01 to 2% of a synergistic flame retardant; and 0-3% of other auxiliary agents. The prepared PPE / PS alloy material has excellent mechanical properties, processability and flame retardance, and can be used in the fields of household appliances, electric appliances, automobiles, optical fiber parts and the like.

The invention discloses a shaft fluid production profile joint inversion method based on DTS and DAS data, and relates to the technical field of oil and gas exploitation monitoring. According to the technical key points, a DTS forward model and a DAS forward model are established respectively, a multi-target optimization model is established based on the minimum joint error of model calculation data and measured data of DTS and DAS, DTS and DAS signal joint inversion is performed through a multi-target optimization algorithm, the optimal parameters of the wellbore liquid production profile are obtained, and the interpretation precision of the underground injection-production profile is improved.

The embodiment of the invention provides an ionization layer electron density inversion method, device, system and equipment and a medium. The method specifically comprises the following steps: determining a target area of electron density inversion; obtaining observation data corresponding to the signal link in the target area; the signal links comprise a first signal link from a GNSS satellite to a ground station, a second signal link from the GNSS satellite to a low earth orbit satellite, and a third signal link from the low earth orbit satellite to the ground station; determining the slant total electron content corresponding to the signal link according to the observation data corresponding to the signal link; solving an ionized layer chromatography function according to the oblique total electron content corresponding to the signal link so as to obtain an electron density value of a grid in the target area; the ionized layer chromatography function is used for representing a mapping relation between the oblique total electron content and the electron density value of the grid. According to the embodiment of the invention, the inversion precision of the ionosphere electron density can be improved.

The invention discloses a lithium iron phosphate positive electrode material rich in oxygen vacancies and a preparation method and application thereof.The preparation method of the lithium iron phosphate positive electrode material rich in oxygen vacancies has the advantages that the number of heating times can be set in a programmable mode, the heating and cooling speed is high, and the heat treatment time is short; the volatilization amount of the lithium salt is small, and the method is more economical and environment-friendly; according to the method, a programmable constant-current power supply is used for electrifying and heating the carbon-based Joule thermal material, and the carbon-based material is used as a heating stage to realize ultrafast synthesis of the lithium iron phosphate material through a physical contact heat transfer effect; and the oxygen vacancy content can be adjusted by regulating and controlling parameters such as temperature and time. The heat treatment is performed in a specific atmosphere (such as vacuum, inert gas, reducing gas and oxidizing atmosphere), so that the reoxidation of the oxygen vacancies can be reduced, and the formation and the stabilization of the oxygen vacancies are facilitated; the adjustable and controllable range of the oxygen vacancy concentration in the lithium iron phosphate material prepared by the preparation method disclosed by the invention is 2.198 E12 to 7.382 E13Spins / mm < 3 >.

The invention belongs to the technical field of geotechnical engineering, and particularly relates to a nanocarbon reinforced superfine cement-based blocking material as well as a preparation method and application thereof. The superfine cement-based blocking material prepared by the invention is prepared by uniformly mixing 45-50 parts of cement, 0.5-2.5 parts of a solid waste sintered active material, 0-0.05 part of a water reducing agent, 0.015-0.1 part of carbon nanotubes, 0.5-2.5 parts of a 1% solution of carbon quantum dots and 45-50 parts of water. The material has relatively low porosity and is not dispersed; the high groutability is achieved, and the material can permeate into tiny holes; and it is guaranteed that the material has high strength, the crushed rock bonding and reinforcing strength is high, and the material has wide application prospects.

The invention provides a community electric vehicle charging station target optimization method and system based on an evolutionary game, and the method comprises the steps: building an information transmission mechanism between an agent and a user, obtaining an agent operation mode based on the information transmission mechanism, and formulating the electricity price of each time period of the next day; establishing an evolutionary game model between the agent and the user demand according to the electricity price in each time period of the next day and the user charging demand feedback; performing target optimization on the evolutionary game model to obtain an optimal community electric vehicle charging electricity price; and dynamically adjusting the number and distribution of the charging piles, the charging speed and / or the power grid load according to the optimal community electric vehicle charging electricity price. According to the invention, the evolutionary game between the agent and the user is established to obtain the global optimal pricing and power purchasing strategy, so that the number and distribution of the charging piles are optimized, the charging speed is improved, the power grid load is balanced and stabilized, and the benefits of the charging station and the satisfaction degree of the user are improved.

The invention relates to a fermentation tank for microbial reaction, in particular to a forced internal circulation type microbubble reactor with a guide cylinder, which comprises a tank body (1), a cylindrical guide cylinder (3) is arranged in the tank body (1), an overflow weir (32) is arranged at the upper part of the guide cylinder, and more than one liquid outlet pipe is arranged at the lower part of the guide cylinder; the upper part of the tank body (1) is provided with a plug flow stirrer (2), the lower end of the plug flow stirrer is provided with a plug flow paddle (24), and the plug flow paddle is positioned at the upper part of the guide cylinder (3); and at least one microbubble generator (4) is arranged in the guide cylinder (3) and is communicated with the air guide pipe. By means of the built-in guide cylinder, the device is large in flow, small in power consumption, mild in flowing, small in shearing force on liquid, uniform in airflow mixing, convenient to distribute and more suitable for a biological fermentation system.

The invention relates to a semiconductor structure and a preparation method thereof. The semiconductor structure comprises a substrate; the base comprises a substrate and an epitaxial layer which are stacked; the main junction is positioned in the epitaxial layer; at least one field limiting ring located in the epitaxial layer; the junction termination extension region is located in the epitaxial layer; the junction termination expansion region and the field limiting ring have an overlapping region; the cut-off ring is located in the epitaxial layer; wherein the surfaces of one sides, far away from the substrate, of the main junction, the field limiting ring, the cut-off ring and the junction termination extension region are flush with the surface, far away from the substrate, of the epitaxial layer; the main junction, the field limiting ring and the cut-off ring are arranged at intervals along a first direction, and the first direction is perpendicular to a direction pointing to the main junction along the substrate; the field oxide layer is at least located on the sides, away from the substrate, of the epitaxial layer, the junction termination extension region and the cut-off ring; the field plates are respectively positioned on one sides, far away from the substrate, of the main junction and the field oxide layer; wherein the doping types of the substrate and the cut-off ring are the same, and the doping types of the main junction, the field limiting ring and the junction termination expansion region are the same. The semiconductor structure has a high withstand voltage level.

The invention discloses a food freezing system based on electrode coupling and a multi-stage electric field and a freezing method thereof. The system comprises an electric field generating device (1), a refrigeration house (2), a space plate type discharge electrode (3), a metal frame discharge electrode (4) and the like. According to the freezing method, a discharge electrode is used for being coupled with an electric field, a special'superposed electric field effect 'is generated, the freezing efficiency is remarkably improved, heat conduction inside and on the surface of food can be accelerated, uniform and rapid crystallization of water is promoted, large ice crystals are effectively prevented from being formed, and therefore damage to cell structures is reduced. Compared with a traditional freezing method, the freezing time can be shortened by 15%-30%, and the juice loss rate is reduced by 3%-6%. Due to the synergistic effect of the double electric fields in different stages, the food can be quickly frozen under lower voltage, and the comprehensive energy consumption can be reduced by 15-30% while the safety is ensured. The method is suitable for large-scale food freezing application, and remarkable technical progress and economic benefits can be brought to the food freezing industry.

The invention belongs to the technical field of computer-aided diagnosis, and particularly relates to a dynamic brain network feature learning method based on Mamba, which comprises the following steps of: performing data processing, including data collection, data preprocessing, dynamic function network sequence construction and state conversion sequence construction; performing sample enhancement by taking the processed data as sample data; constructing a Mama model comprising a linear embedding layer, a Mama layer, a linear mapping layer I, an activation function and a loss function as a classification model; training the Mama model by using the state conversion sequence, and optimizing hyper-parameters of the Mama model; and based on the trained Mama model, predicting the enhanced sample data, and determining a classification result of the sample data through a label output by the Mama model. According to the method, the relation hidden in the dynamic brain network state transition sequence can be efficiently captured, and high-precision classification of each brain network sequence is realized.