35results about How to "Improve output quality" patented technology

The invention relates to the field of agricultural product processing equipment, in particular to a pot body overturning structure and an automatic discharging device based on the pot body overturning structure, the pot body overturning structure comprises a machine body and a pot body assembly, and the pot body assembly is rotationally connected to the machine body; the pot body overturning assembly drives the pot body assembly to overturn around the rotation center of the pot body assembly, and the pot body overturning assembly comprises a servo motor, transmission teeth and an arc-shaped gear ring; the controller is arranged on the machine body; the sensor group is used for detecting the position state of the pot body assembly after overturning and resetting, and the controller is used for controlling the pot body overturning assembly to execute overturning and resetting actions according to a signal fed back by the sensor group. After processing is completed, the pot body overturning assembly automatically enables the pot body assembly to conduct overturning discharging, manual participation in discharging operation is not needed, the production efficiency is improved, safety is higher, the discharging process is controllable and low in impact through a buffer block and a buffer plate, and the complete shape and quality of discharged tea leaves are effectively kept.

This invention provides an output impedance calculation device, which includes: a control unit, an adjustment unit, a LORA transmitting unit, a first impedance calculation unit, a second impedance calculation unit, a matching unit, and a transmitting antenna. The control unit is used to transmit wireless signals; the LORA transmitting unit is used to convert wireless signals into LORA signals; the first impedance calculation unit is used to calculate the input impedance based on the LORA signals; the second impedance calculation unit is used to calculate the output impedance based on the LORA signals; the control unit is used to calculate the reflection coefficient based on the input and output impedances, and to calculate the standing wave ratio based on the transmission coefficient; the adjustment unit is used to adjust the impedance in the circuit based on the standing wave ratio and the reflection coefficient, thereby achieving impedance matching and balance between the input and output terminals of the line, improving the output quality of radio signals, reducing the overall output power consumption, and reducing harmonic interference.

The utility model discloses organic solid waste pretreatment equipment of multistage crushing and magnetic separation integration, including support board, the upper end fixed connection of support board is connected with the pulverization frame and transmission case, the upper end fixed connection of pulverization frame has the pulverization box, the right side of pulverization box is provided with the feeding port, first crushing mechanism, first crushing mechanism includes the first rotation axis rotation connection in the front and rear inner wall of pulverization box, and the outside equal interval fixed connection of first rotation axis has a plurality of crushing pestles, second crushing mechanism, second crushing mechanism includes two shearing type crushing rollers rotation connection in the front and rear inner wall of pulverization frame. When using the equipment, first adopt hammer type crushing and carry out rough crushing, then utilize shearing type crushing and carry out fine crushing, improve actual crushing effect, improve the overall service life of crushing part simultaneously, in addition, can also screen the metal after crushing, avoid metal and enter carbonization system and damage equipment.

The invention discloses a heavy hydrocarbon-containing natural gas dehydration lean glycol purification device which comprises a carrying assembly and a purification discharging mechanism, a feeding and discharging bearing assembly assembled through bolts is arranged above one end of the carrying assembly, and a condensation gas outlet component installed in a sleeving mode is arranged at the output end of the feeding and discharging bearing assembly; and the output end of the condensation air outlet component is provided with a purification discharging mechanism which is mounted in a sleeving manner. According to the device, gas is output to the bottom end of a condensation cabin through a gas outlet pipe and a bridging valve on a condensation gas outlet component, and after the gas is input through the bottom end, the gas is input from bottom to top, so that under the mutual cooperation of an inner ring pipe, an inner fin, an opposite frame, a refrigerating machine, a compression pump and a circulation pipeline, uninterrupted circulation adjustment can be conducted, and the energy consumption is reduced. Due to the fact that the multiple sets of fin plates are distributed in parallel, the fin plates can make full contact, the dehydration effect of products is improved, and the output quality of the products is improved.

The present application relates to a semantic chain feedback and reinforcement learning driven image fusion method, belonging to the technical field of computer vision and image processing. It includes: inputting the preprocessed input image into the frozen dual-path hierarchical encoder, extracting multi-scale visual features and shallow prior features in the prior injection branch; on each intermediate scale, cross-modal collaborative fusion is performed on the infrared features, visible light features, the last layer propagation features and shallow prior features in the multi-scale visual features; chain semantic enhancement is performed on the multi-scale fusion features: obtaining several candidate fusion images; optimizing the unfrozen model parameters; inputting the infrared image and visible light image to be processed into the optimized network, and selecting the noise-free reference candidate as the final fusion image output. This method cooperatively optimizes through chain semantic progressive modulation and difference driven reinforcement learning, improving the visual quality and information expression ability of the fusion image.

This application discloses a rule-constrained decision generation method for a centralized procurement governance hub based on AI. The method includes: receiving a procurement governance task and obtaining a set of rule constraints related to the task; determining a text generation format based on the rule constraint set and calling a decision content generation model to generate initial decision text under the constraints of the rule constraint set and the text generation format; performing multi-dimensional information verification on the initial decision text based on the rule constraint set, and performing rollback and regeneration operations on initial decision texts that fail verification to obtain corrected decision text; associating and storing the verified and corrected decision text with the corresponding rule application path, data source information, and multi-dimensional verification information to form a traceable decision record; and generating a procurement governance decision report based on the decision record. This method ensures the reliability, auditability, and traceability of the procurement governance decision report, reducing procurement governance risks.

The application discloses an interactive form design method and system based on a language large model, and belongs to the technical field of human-computer interaction and form automatic generation. In view of the problems that a traditional form design tool is tedious to operate, highly depends on manual experience, and an existing method based on a large model has high training cost, weak interaction ability and difficulty in real-time collaboration, the following scheme is used for improvement: defining a JSON Schema specification of a form and constructing a visual form designer, integrating a language large model interaction entrance in an interface, dynamically assembling a request containing user instructions, a JSON Schema specification and existing form data according to whether the form is empty, and submitting the request to the language large model; receiving a response in a streaming manner, and real-time analyzing and progressively updating a form designer preview interface, so that a user can directly generate or modify a form through natural language instructions, and multi-round interaction and manual subsequent adjustment are supported. The method is mainly used for reducing a form design threshold, improving design efficiency and intelligent collaboration level.

This invention discloses a method, apparatus, device, and medium for sample type classification based on multi-dimensional data. The method includes acquiring an initial dataset; binning the data in each dimension to generate an N-dimensional grid structure; performing K-1 region search steps to progressively divide the grid cells in the N-dimensional grid structure into K different type regions, where K is a positive integer greater than 1; and setting a corresponding sample type label for each type region to generate a first type classification result. This method possesses automated fitting and optimization capabilities, supports feature inputs of arbitrary dimensions and binning granularity, and completes type classification and model construction without manual intervention, thereby effectively improving recognition accuracy and output quality while significantly reducing labor costs and substantially improving modeling and iteration efficiency.

The utility model relates to a kind of for preventing the adjusting mechanism of mould hunting when winding material, belong to amorphous strip winding material equipment technical field.The utility model is adjusted by setting walking module drive's adjusting stabilizer component near winding disc to realize the regulation of winding material, adjusting stabilizer component includes adjusting stabilizer support, adjusting stabilizer cylinder, adjusting stabilizer tray and the balance plate of lever formation, the gap distance between adjusting stabilizer tray lifting, fine adjustment winding material and winding disc central shaft can be controlled by linear motion module and adjusting stabilizer cylinder, and then solve the unstable problems such as up and down floating caused by the cooperation gap of winding material and central shaft, effectively improve discharging quality.

The invention discloses a wool spinning carding machine needle tooth density self-adaptive control method and system based on fiber fineness monitoring, and relates to the field of mechanical intelligent control, and the method comprises the steps: obtaining fiber fineness monitoring data of a target scene and carding area operation characteristic data in real time; acquiring the needle tooth density of the target working condition through the fiber fineness monitoring data; constructing a working condition density distribution model of the needle tooth density of the target carding machine, and initializing the working condition density distribution model; in combination with the working condition density distribution model, the carding area operation characteristic data and a pre-constructed second relation model, calculating expected working condition pin tooth density, and adjusting the working condition density distribution model; and based on the working condition density distribution model, outputting an optimal apparent needle tooth density set value, generating a control instruction and driving a needle tooth density execution mechanism of the carding machine to act. The problems that in the prior art, manual sampling and off-line detection are dependent, information is seriously lagged, the needle tooth density adaptability is poor, and the dynamic operation working condition of the carding machine is not considered are solved.

The invention relates to the technical field of deep learning, and particularly discloses a museum exhibit deep research system fused with a multi-modal large model. The system comprises a multi-modal data aggregation module, a cross-modal semantic alignment module, a research path generation module, a dynamic argumentation construction module and a research conclusion generation module, through multi-modal data fusion, semantic alignment and intelligent path planning, an argumentation structure is automatically constructed, and a standard research conclusion is generated. The whole-process intelligent support of exhibit research is realized, and the research efficiency and systematicness are improved.

This application discloses a multi-agent test case generation method and apparatus based on a test intent graph, relating to the fields of artificial intelligence and software testing. The method includes: acquiring multiple heterogeneous data sources; performing data standardization, entity alignment, and cross-validation on the heterogeneous data sources to obtain a fused perception result; constructing a test intent graph based on the fused perception result; for each node in the test intent graph, calling N test case agents to generate N candidate test case sets; having a review agent quantitatively score the N candidate test case sets and selecting a reference test case set based on the scoring results; validating each test case in the reference test case set, and automatically repairing test cases that fail verification until the repaired test cases pass verification. This application solves the technical problem of low generation efficiency caused by the lack of a systematic guiding framework in the test case generation process in existing technologies.

The application provides an integrated quantum bit output link and relates to the technical field of input / output device links. The application aims to solve the problem that the space is limited during quantum bit transmission, and efficient heat treatment is difficult to achieve. The integrated quantum bit output link comprises three circulators, a load, a Josephson parametric amplifier and a connector. The load is arranged on the three circulators and is used for absorbing reverse transmission signals. The Josephson parametric amplifier is internally provided with the connector, and the Josephson parametric amplifier is connected with the three circulators through the connector. The integrated quantum bit output link can maintain good performance in a super low temperature environment, achieves the effects of high integration, high heat conduction and magnetic shielding, is easy to assemble and mass-produced, and effectively reduces the volume of the integrated quantum bit output link.

The application discloses a kind of based on raw material information feedback's weighing target value regulation and control system, specifically related to industrial automatic control technical field, raw material characteristic identification module exports characteristic feature group, its internal pre-set correlation screening model is according to feedback screening high correlation feature, constructs change matrix and generates dynamic compensation coefficient, and optimizes characteristic feature group output.Target value determination module determines weighing target value accordingly.After real-time weighing, deviation calculation module calculates the deviation of actual weighing and target value.Regulation and control execution module adjusts feeding equipment according to deviation, and uses dynamic compensation coefficient to correct adjustment amount, so that actual weighing approaches target value;The application constructs change matrix and generates dynamic compensation coefficient by screening high correlation feature, improves compensation pertinence;Regulation and control result is fed back to optimize characteristic identification, realizes self-learning closed loop;Dynamic compensation coefficient is used to correct adjustment amount, so that feeding equipment is self-adapting adjustment, significantly improves weighing precision, dynamic stability and repeatability.

The present disclosure provides a media data processing method and system, device, electronic equipment, storage medium and computer program product. The media data processing method comprises receiving media data, wherein the media data comprises a plurality of first audio frames, a plurality of first video frames, and a display time stamp (PTS) of the plurality of first audio frames and the plurality of first video frames on an original time axis; dividing the media data into a plurality of media segments, wherein each media segment in the plurality of media segments comprises a video segment and an audio segment; for each media segment, transcoding and multiplexing the video segment and the audio segment based on the PTS on the original time axis corresponding to the video segment and the audio segment respectively to obtain a transcoded media segment; and merging the transcoded media segments based on the PTS on the original time axis corresponding to the video segment and the audio segment respectively.

The utility model relates to dry grinder technical field discloses dry grinder with wear -resisting sealing structure, including installation platform, the top of installation platform is installed with organism, the rear side intercommunication of organism has the discharge pipe, the rear side bottom of installation platform is installed with collection box, the rear side middle part of collection box is installed with second motor, the output fixed connection of second motor has driving bevel gear, the inside rear side top fixed connection of installation platform has connecting plate, the bottom rotatory connection of connecting plate has two -way screw rod, the middle part fixed connection of two -way screw rod has driven bevel gear. In the utility model, material is ground after organism, enters collection box through discharge pipe, starts second motor, driving bevel gear drives driven bevel gear to rotate, two -way screw rod rotates, the filter plate is restricted along two -way screw rod linear sliding under the restriction of spacing rod, filters material, so that material can be filtered, improves the product quality of product.

The utility model provides a horizontal formula piece device relates to photovoltaic, semiconductor production technical field is used to solve the problem that the piece of silicon wafer and similar wafer is not continuous and low stability. Horizontal formula piece device, including the piece mechanism and the piece mechanism of going out. The piece mechanism includes: a pair of mounting plate, at least a pair of first transmission roll and at least one second transmission roll, and sucking disc. A pair of first transmission roll is spaced apart between a pair of mounting plate along the first direction, and the second transmission roll is adjustably arranged between a pair of mounting plate up and down, and is located above a pair of first transmission roll, and the outer periphery ring of a pair of first transmission roll and second transmission roll is equipped with first conveyer belt. The downside surface of sucking disc forms the adsorption part. The piece mechanism of going out is close to the piece mechanism setting, and the piece mechanism of going out is located below the piece mechanism. The piece mechanism still includes the blocking mechanism, and the bottom height of blocking mechanism is lower than the downside height of first conveyer belt.

The invention provides a model training method and device for a text open task, on one hand, a reward model is trained through a training text pair marked with a relative quality advantage relation, original complex absolute quality scores of the reward model are converted into simpler and more stable relative advantage judgment, the model training cost and subjective influence are reduced, and the training efficiency is improved. The accuracy and confidence of the reward signal are improved, one first reply in the output replies of the existing model is set as a reference text reply, and it is ensured that the output quality of the trained text model is superior to the output quality of the existing model; on the other hand, in the training process of the to-be-trained model, a second reply output by the to-be-trained model is monitored, and the target reward model is adjusted according to the monitoring result; and / or, the reward value output by the target reward model for the second reply is monitored, and the reference text reply is adjusted according to the monitoring result, so that the optimization direction of the text model is ensured, and the training effect of the text model is improved.

The utility model provides a water treatment corrosion inhibitor preparation dosing device which comprises a water treatment dosing storage tank, a hollow inner column liquid storage bin is arranged between the upper inner wall and the lower inner wall of the water treatment dosing storage tank, and the inner side wall of the water treatment dosing storage tank is fixedly connected with a plurality of partition bin sealing spacer ring plates. A plurality of first electromagnetic liquid outlet valves are fixedly connected to the front side of the hollow inner column liquid storage bin, corrosion inhibitor liquid is accurately injected into the hollow inner column liquid storage bin through a liquid guiding guide pipe, the interior of the water treatment dosing storage tank is divided into equal-volume dosing bins through the multiple partition bin sealing partition ring plates, and the multiple first electromagnetic liquid outlet valves are fixedly connected to the front side of the hollow inner column liquid storage bin; a vertical electric guide rail drives an inner partition displacement ring to slide up and down along the inner side wall of a hollow inner column liquid storage bin, a liquid containing space communicated with a first electromagnetic liquid outlet valve can be flexibly regulated and controlled, then the amount of corrosion inhibitor liquid entering all dosing bins is accurately controlled, and the diversified requirements for storage and taking of corrosion inhibitors under different working conditions are met.

This utility model discloses an air supply system, which includes a main pipe and at least two sets of air supply units. Each air supply unit includes an air compressor and a dryer. A first branch pipe is provided connecting the air compressor and the dryer, and a second branch pipe is connected in parallel with the first branch pipe. The main pipe is sequentially connected to the air supply units. When one air compressor or dryer fails, the equipment in other air supply units can continue to work. If the air compressor of one air supply unit fails, the air compressors of other air supply units can continue to supply air to the dryer of that air supply unit through the main pipe. This avoids the entire system from shutting down due to a single equipment failure, reduces downtime caused by equipment maintenance or failure, improves the overall operating efficiency of the system, avoids the situation of some equipment operating under overload for a long time in the traditional one-to-one single-pipe mode, and improves the output quality of compressed air.

The application discloses a mouth shape retrieval model optimization method, system and device based on time sequence context modeling, which is applied to the technical field of data processing. Multi-frame audio feature coding is completed through a time sequence window, one-dimensional convolution and weighted fusion, so that audio representation with time sequence consistency is generated. A retrieval preference model is constructed in combination with historical retrieval data, time sequence smoothing constraints and state transition rules are formed. A retrieval sorting model is built relying on a multi-objective optimization function, candidate screening is completed by comprehensively matching multi-dimensional matching items and superimposing various types of penalty constraints. Meanwhile, boundary adaptive weight regulation and time sequence stable reasoning strategies are added, and dynamic optimization is performed for pronunciation boundary, high-risk and low-confidence scenarios. The whole scheme runs in a frame-by-frame streaming mode, upgrades traditional single-frame optimal retrieval to time sequence context constraint optimal retrieval, effectively solves the mouth shape jitter problem, and improves the overall time sequence stability and output naturalness.

The invention provides a method and device for suppressing common-mode voltage of a motor and a storage medium, and relates to the technical field of motor control, and the method comprises the steps: obtaining a current operation state parameter of the motor; based on the operation state parameters, generating enhanced modulation waves of each phase for improving the voltage utilization rate of the motor; obtaining the duty ratio of each phase generated by the motor controller based on the enhanced modulation wave; obtaining a target carrier signal corresponding to the phase to which the intermediate value of the duty ratio of each phase belongs, and performing phase shift processing on the target carrier signal, the phase shift processing being used for eliminating a zero voltage vector generated by a motor controller; and generating each phase PWM signal used for driving an inverter in a motor controller based on each phase carrier signal after phase shift processing so as to suppress the common-mode voltage of the motor. According to the method, on the basis of not changing the whole electric driving system of the electric vehicle and topological structures of all parts, all phases of PWM signals acting on an inverter are generated by changing a modulation strategy, and the common-mode voltage of a motor is suppressed.

The utility model provides bean product processing and fermenting equipment. The bean product processing and fermenting equipment comprises a box body, a fermentation cavity is formed in the box body, multiple sets of placement parts are oppositely arranged on the two opposite inner side faces of the fermentation cavity, fermentation discs are arranged on the multiple sets of placement parts in a replaceable mode, and breathable pads are arranged in the multiple sets of fermentation discs; two groups of assembly parts are oppositely arranged on the box body close to one side of the fermentation cavity, and a protective door assembly is rotatably arranged on the two groups of assembly parts; a temperature and humidity control assembly is further arranged on one side of the box body. The temperature sensor and the humidity sensor can continuously collect temperature data in the fermentation cavity, the data can be instantly transmitted to the display panel to be clearly displayed, and when the monitored temperature or humidity deviates from a preset standard value, the heating plate can start the heating function according to the temperature condition so as to recover the proper temperature. And the steam humidifier can perform spraying operation according to the humidity requirement to adjust the humidity, so that the stable proceeding of the bean product fermentation process is guaranteed.

The embodiment of the invention discloses a data processing method and device, equipment, a storage medium and a product. The method comprises the steps of obtaining input data of a large language model, calling the large language model to perform iterative prediction based on the input data and a first decoding mode to obtain a first prediction result of the input data, and calling the large language model to perform iterative prediction based on the input data and a second decoding mode in response to the fact that the first prediction result meets a correction condition to obtain a second prediction result of the input data. And obtaining a second prediction result of the input data. Therefore, when the correction condition (such as receiving negative feedback) is met, the big language model is called to perform iterative prediction based on the input data and the same or different decoding modes, so that the probability that the big language model generates a user satisfactory result can be improved, and the purpose of improving the output quality of the big language model is achieved.

The invention discloses a coaxial double-motor slip-ring-free winding rotor direct connection slip excitation asynchronous power generation system, and belongs to the technical field of wind power generation. The system adopts a coaxial direct connection structure, the main generator (DF) and the auxiliary excitation motor (DJ) operate coaxially, and the capacity of the excitation motor is 25% of that of the main generator. Two rotors are axially and directly connected, and excitation energy is provided by the slip of the excitation motor. The number of turns of the excitation motor rotor winding is set to be larger than that of the generator rotor winding, so that the flow direction of excitation current is reasonable. The double generators run coaxially, the exciter is driven, the generator generates power, frequency, voltage and phase are stabilized automatically by means of electromagnetic negative feedback, and power frequency sine wave electric energy meeting the network access condition is directly output. A traditional IGBT and a high-power converter device are omitted, the system failure rate is reduced, the equipment number complexity and the manufacturing cost are remarkably reduced, the cost per kilowatt-hour is greatly reduced, and the device has the advantages of being free of slip rings, high in abrasion efficiency, long in service life, capable of being connected to and disconnected from a grid and capable of being used in a micro-grid.

A multi-agent context sharing method, system, and medium are disclosed, comprising: parsing the context content output by a first agent to generate corresponding metadata; associating the context content with the corresponding metadata and storing it in a distributed cache; establishing a topic set indexed by a session identifier in the distributed cache and adding the topic identifiers to the topic set; responding to a task to be executed by a second agent, obtaining the topic set corresponding to the current session identifier, reading the topic descriptions corresponding to each topic identifier in the topic set, and generating a list of topic descriptions; inputting the task to be executed and the corresponding list of topic descriptions into a large language model, receiving the topic identifier set returned by the large language model as the target topic identifiers semantically related to the task to be executed; and reading the corresponding context content from the distributed cache according to the target topic identifiers, associating the context content with the task to be executed, and inputting it into the second agent.

An image processing method, applied to an electronic apparatus, includes performing feature extraction processing on a source image to obtain a feature data set corresponding to a source file, determining an image processing parameter corresponding to the feature data set according to the feature data set, and processing the source image according to the image processing parameter.

The utility model relates to the technical field of injection molds, in particular to a double-cone three-roller precision calendering sheet extruder which comprises a machine base, a rolling mechanism, a feeding mechanism, a distribution box, a control box, a water circulation mechanism and a returned glue conveying mechanism, the rolling mechanism, the feeding mechanism and the distribution box are all arranged on the machine base, a roller base is arranged on the rolling mechanism and fixed on the machine base, and the control box is arranged on the machine base. A feeding base is arranged on the feeding mechanism, the feeding base is fixed to the machine base, the control box is fixed to a roller base, the water circulation mechanism is fixed to one side of the machine base, the feeding extrusion equipment and the sizing material tabletting equipment are of an integrated structure, the sizing material returning and feeding functions are added, sizing materials are forcibly fed through double cones to be tableted, and the sizing material tabletting efficiency is improved. And finally, the processing of the rubber sheet is completed through an edge cutting function, so that the working efficiency and the rubber sheet discharging quality are integrally improved, and meanwhile, the production cost is reduced.

The utility model relates to the technical field of cotton thread production, and discloses a humidifying device for pure cotton thread production, which comprises a uniform humidifying mechanism, a sliding humidifying mechanism and a controller, one end in the sliding humidifying mechanism is slidably connected with the uniform humidifying mechanism, one end of the sliding humidifying mechanism is provided with the controller, and the other end of the sliding humidifying mechanism is connected with the controller. The uniform humidifying mechanism comprises a moving frame, hanging sliding blocks are fixedly connected to the two ends of the upper end face of the moving frame, a sliding groove is fixedly connected to the lower end of the end face of one side of the moving frame, a sliding water conveying pipe is slidably connected to one side of the outer wall of the sliding groove, and a plurality of nozzles are formed in the lower end face of the sliding water conveying pipe. According to the device, the controller drives the first servo motor to drive the eccentric wheel to rotate, the eccentric wheel pushes the sliding water conveying pipe and the spray head to reciprocate along the sliding groove through the pull rod, an external water source is supplied to the spray head through the water inlet, cotton threads are evenly humidified in the reciprocating spraying process, and therefore the quality of the cotton threads and the practicability of the device are improved.

This utility model relates to the field of material feeding equipment technology, specifically a material feeding device for product testing. It mainly includes a mounting plate, on the top surface of which a feeding tray and a conveying assembly are mounted. An assembly assembly and a testing assembly are also fixedly mounted on the top surface of the mounting plate. The device further includes a material feeding mechanism positioned above the mounting plate for feeding products, and an opening mechanism positioned above the mounting plate. When a product fails quality testing, a cylinder is activated, moving a movable plate. The movable plate moves the fixture and product a certain distance, and an electric push rod is activated, moving a push plate and a push block. The push block promptly pushes the product off the fixture, causing it to fall into a collection box for collection. This prevents defective products from being moved to the acceptable area due to worker fatigue or inattention, ensuring consistent product quality and improving the overall output quality.