24results about How to "High quantitative accuracy" patented technology

The invention provides a fire extinguisher tank coating process parameter collaborative optimization control method and system. The method comprises the following steps: acquiring infrared radiation data on the surface of a paint film of a fire extinguisher tank in a curing oven, and generating a temperature distribution matrix through thermal gradient analysis; quantifying temperature difference characteristics of different positions of the surface by using the matrix; a multispectral infrared thermal imager is annularly arranged in the curing oven, the optical scanning path of the multispectral infrared thermal imager is dynamically adjusted based on temperature difference characteristics, temperature abnormal points are marked, and a three-dimensional thermal field distortion atlas is constructed; analyzing temperature abnormal point distribution in the atlas, and generating a local cooling compensation vector; and converting the compensation vector into a heating unit power adjustment instruction by applying a sliding mode control mechanism, and adjusting output power to inhibit paint film deformation. Dynamic balance of the curing temperature of the paint film and temperature range compression are realized, and the deformation defect of the paint film is eliminated.

This invention relates to the field of atmospheric particulate matter analysis and discloses a deep learning-based method for calibrating organic aerosol concentrations. This invention addresses the technical problems in existing organic aerosol concentration calibration techniques, such as insufficient utilization of multi-source information, inadequate use of temporal evolution information, and fixed feature fusion methods lacking dynamic weight adjustment. By leveraging multi-source observation vectors, it fully utilizes observation information from different physical properties to establish a stable nonlinear mapping relationship under complex atmospheric conditions and mixed aerosol backgrounds. Through multi-source observation features and historical state features, the model can characterize the dynamic evolution of organic aerosol concentrations. Employing a bidirectional long short-term memory network and SE attention mechanism to process multi-source observation features and historical state features, it can automatically learn and dynamically adjust the importance weights of feature channels, effectively suppressing noise and redundant feature interference. Furthermore, it utilizes forward and backward time-series modeling to fully capture the contextual relationships between consecutive time points.

The utility model relates to the technical field of disinfection dosing, in particular to a disinfection quantitative dosing device. The disinfection quantitative dosing device comprises a base, a dosing shell is fixedly installed on the base through a support, a rotating disc in rotating connection is installed in the dosing shell, and a plurality of quantitative cavities distributed annularly and radially are formed in the outer ring wall of the rotating disc; and an intermittent driving piece for driving the rotating disc to intermittently rotate for feeding is mounted on the rotating disc on the back plate. According to the quantitative dosing device for disinfection, provided by the utility model, the radial quantitative cavities which are annularly distributed are designed on the rotating disc, the volume of each cavity is fixed, and the intermittent driving part is matched to ensure that only one station is moved during rotation each time, so that the dosage height of medicaments falling into the discharge pipe each time is consistent, and the quantitative precision is remarkably improved; and the obvious influence of material level and fluidity fluctuation is avoided.

The application discloses a kind of quantitative methods of migratory birds based on dual-polarized weather radar and bird detection radar.The application is jointly observed by dual-polarized weather radar and bird detection radar, carefully selects the original observation data of the two radars, eliminates the false alarm and clutter track of the bird detection radar, eliminates the meteorological echo and insect echo of the polarization weather radar, then constructs effective quantitative relationship using the data in the airspace monitored jointly, the present method carefully considers the observation airspace of different radars, has higher quantitative accuracy, and is more suitable for China.

The utility model discloses a kind of sodium caseinate is used to quantitative loading device, it belongs to quantitative equipment technical field, including anchoring structure, its upper surface one side fixedly connected pipe valve stable structure, pipe valve stable structure is fixedly connected three-way branch ware in, anchoring structure upper surface other side fixedly connected reciprocating power module cabin, reciprocating power module cabin inner wall one side rotationally connected adjustable phase driving disc, three-way branch ware one side fixedly connected quantitative assembly, reciprocating power module cabin upper surface fixedly connected sterile raw material bin, sterile raw material bin upper surface fixedly connected flow state flow guide dome.The utility model is through setting quantitative assembly, ensure that the amount of dairy product that is inhaled from sterile raw material bin and pushed to soft package each time is constant, to improve the quantitative accuracy of packaging, while device operation process is simple and clear, reduce operation difficulty and labor cost, improve production efficiency, quickly, accurately complete the filling work of a large number of soft package, satisfy production line demand.

An automatic charging machine for combined firework production relates to the technical field of combined firework production and comprises a charging box, a supporting frame is fixed to the side face of the charging box, a PLC is installed on the side face of the supporting frame, a rotating roller is rotatably connected to the inner side of the charging box through a bearing, and quantitative grooves are evenly formed in the rotating roller. A quantitative groove is formed in the upper portion of the charging box, a discharging auxiliary unit is installed on the inner side of the quantitative groove, a feeding hopper communicated with the interior of the quantitative groove is fixed to the top of the charging box, a plurality of charging pipes communicated with the interior of the quantitative groove are fixed to the bottom of the charging box, and a driving mechanism for driving a rotating roller to rotate is installed on the side face of the feeding hopper. The charging quantity of each time is kept consistent, the quantitative precision of combined firework charging is improved, the stability of the product quality is ensured, meanwhile, charging operation can be carried out on a plurality of firework cylinders at a time, and therefore the production efficiency is improved.

The application provides a voltage deviation detection circuit and a voltage regulation system. An offset power supply is coupled to a reference voltage module and a feedback voltage module to output a first voltage or a reference voltage at a first voltage output end and output a second voltage or a feedback voltage at a second voltage output end. The first voltage is obtained by superimposing an offset voltage on the reference voltage, and the second voltage is obtained by superimposing an offset voltage on the feedback voltage. By coupling the offset power supply to the reference voltage module and the feedback voltage module, the first voltage or the reference voltage is output at the first voltage output end, and the second voltage or the feedback voltage is output at the second voltage output end. The similar value of the feedback voltage generated by using a buffer structure is avoided, thereby avoiding the introduction of non-ideal factors, improving the quantization accuracy of the difference between the reference voltage and the feedback voltage, and further improving the accuracy of the output voltage.

The invention discloses a meat freshness rapid detection device, and belongs to the technical field of food nondestructive detection. The device aims at solving the problems that existing detection equipment is poor in portability and difficult to use on site, the detection range is limited due to single excitation wavelength, and the detection process depends on manual operation and is insufficient in intelligence. According to the technical scheme, the device is characterized by comprising a sample placement area, an annular multi-wavelength excitation light source module, a spectrum acquisition module, a single-chip microcomputer module, a Bluetooth module and a power supply and discharge module which are integrated in a device shell; the external intelligent terminal is provided with detection analysis software, and the software controls the excitation of the LED lamp beads with specific wavelengths through the Bluetooth module, automatically identifies the type of the freshness indication label based on the collected spectral characteristics, and calls a corresponding prediction model to calculate the freshness index. The device is mainly used for carrying out rapid and portable on-site detection on meat samples with freshness indication labels.

A kind of multi-threshold phase encoding-based all-optical quantization system belongs to all-optical analog-digital conversion field.It includes high repetition frequency femtosecond pulse light source,1×M optical splitter,phase-shifted optical quantization module, soliton self-frequency shift module and multi-threshold encoding module;The phase-shifted optical quantization module is M-path electro-optic modulator;The soliton self-frequency shift module includes first M-path high nonlinear fiber,M-path single-mode fiber and second M-path high nonlinear fiber;The multi-threshold encoding module includes M wavelength division multiplexer,M×N optical delay line array,M×N photodetector array and encoding module.The present application uses multi-threshold encoding module of phase encoding,realizes simultaneous judgment of multiple threshold points on power transmission curve of M electro-optic modulators, solves code type repetition problem of all-optical quantization system,realizes more effective quantization states with less channel number,improves quantization precision.Meanwhile,quantization bit number of the present application is improved to log2(2MN).

The application discloses a gradient and precision self-adaptive weighted federated learning fairness optimization method and system. In view of the problems of uneven model performance distribution among clients and large communication overhead of traditional federated learning in a heterogeneous environment, an information amount weighting mechanism is constructed based on the training accuracy and participation frequency of the clients on the server side, the aggregation weight of the clients is dynamically determined, the aggregation weight is used for weighted aggregation of the client gradient, the adaptive weighting of the client gradient is realized, and the performance fairness is improved. Meanwhile, the global quantization precision is adaptively adjusted according to the global loss change in the training process, and the corresponding local quantization precision is distributed according to the client aggregation weight, so that the communication overhead is reduced under the premise of ensuring the model performance. The application realizes the joint optimization of federated learning fairness and communication compression, can effectively alleviate the uneven performance problem among the clients, and improves the communication efficiency of the federated learning system.

The utility model discloses a quantitative feeding device for conductive master batch granulation processing, and belongs to the technical field of conductive master batch granulation. The quantitative feeding device for conductive master batch granulation processing comprises a storage tank, a feeding hopper is installed on one side of the top end of the storage tank, a supporting frame is installed on the outer side of the storage tank, three supporting columns are installed at the bottom end of the supporting frame, a discharging opening is formed in one side of the bottom end of the storage tank, a discharging mechanism is arranged in the storage tank, and a collecting hopper is arranged at the bottom end of the discharging opening. A pair of hinges is installed on one side of the bottom end of the collecting hopper, a bottom plate is installed at the bottom ends of the hinges, an electronic scale is installed at the bottom end of the bottom plate, a supporting plate is installed at the bottom end of the electronic scale, a connecting column is installed on one side of the top end of the supporting plate, and the top end of the connecting column is fixedly connected with the bottom end of the storage tank. The quantitative accuracy is improved, and the practical value is relatively high.

This invention provides a drug nebulization device and method, relating to the field of medical device technology. It includes a housing, within which a reciprocating motor, a transmission connector, and a negative pressure generator are housed. The cylinder of the negative pressure generator is connected to a suction tube with a one-way valve and an outlet tube. The suction tube extends into the drug receiving area, and the outlet tube connects to the nebulizer. The device also includes a drug bottle fixing device, a drug inlet / outlet, a detachable mask, a handle with a built-in battery compartment, a nebulization button, a positioning device, and an audible alarm. The nebulization method is as follows: the drug bottle is placed and fixed, the reciprocating motor is started, driving a piston to reciprocate. The change in cylinder volume, combined with the one-way valve, achieves negative pressure quantitative drug intake and positive pressure delivery of the drug solution for nebulization. After operation, the piston returns to its original position. This invention relies on motor drive to replace manual energy storage, making operation simple, providing accurate drug dosage, stable transmission, and also featuring a device location function, resulting in excellent overall performance.

The application relates to a lubricating oil precise quantitative control system, which comprises an oil tank, a hose and a PLC controller. An oil pump is arranged in the oil tank. A pressure sensor is arranged at the output end of the oil pump and connected with two oil conveying pipes arranged at the bottom of the oil tank through a shunt device. Electromagnetic valves are arranged on the oil conveying pipes respectively. The input end of the hose is connected with the oil conveying pipes, and the output end is connected with a roll opening oil part and a bottom oil part respectively. An electric pinch valve and a dropper are sequentially arranged on the hose. The visual oil quantity control pipeline can realize real-time and direct observation of the conveying state of the lubricating oil, and timely adjustment can be carried out. The point-drip infusion type structure can realize precise quantitative control of the lubricating oil, effectively avoid the continuous conveying of the residual oil in the oil pipe after starting and stopping, cause the oil quantity to be excessive, ensure that the product will not cause bad use experience due to the lubricating oil residue when delivered to consumers, improve the overall quality and user satisfaction of the product, and improve the production efficiency and reduce the production cost, so that good economic benefits and social benefits are achieved.

The invention discloses a material elastic constant array ultrasonic in-situ detection method, and belongs to the technical field of ultrasonic nondestructive testing and material characterization. The method comprises the following steps: carrying out primary orthogonal coding data acquisition on a detected material, determining a refraction point under a preset elastic anisotropy constant C of the detected material based on the Fermat theorem, and carrying out delayed superposition imaging on a reflector in a target area of the detected material by adopting a synthetic aperture focusing or full-focusing algorithm, and calculating the ratio F (C) of the maximum amplitude value of the synthetic image to the entropy value of the image as an evaluation objective function, and obtaining an optimal elastic constant corresponding to the global maximum value of the F (C) by adopting optimization tools such as a genetic algorithm, namely the elastic constant of the detected material. The array ultrasonic device is used for one-time orthogonal coding data acquisition, the detected material and the ultrasonic probe do not need to be moved or rotated for many times, the detection efficiency is improved, the method is not affected by the parallelism and thickness uniformity of the surface interface of the detected material, and in-situ, lossless and rapid detection of the elastic constant of the complex component in a service environment can be achieved.

The invention discloses a rock and ore powder anti-splashing sampling device for acid dissolution test, which comprises a temporary storage cylinder for temporarily storing rock and ore powder needing to be added into an acid solvent, a piston at the top end of the temporary storage cylinder is connected with a piston rod, and a taper pipe for sucking and discharging the powder is arranged at the bottom end of the temporary storage cylinder in a penetrating manner; a sampling piece for sucking powder or a sample introduction piece for discharging the powder into an acid solvent in the digestion container is clamped on the outer side of the taper pipe, a protection piece for preventing acid liquor from splashing is arranged on the sample introduction piece, and a supporting piece is arranged on the protection piece; through the sealed sampling and sample introduction structural design of the temporary storage cylinder, the sampling piece and the sample introduction piece, the problem of flying of rock ore powder in the sampling and sample introduction stages is completely eradicated, meanwhile, the immersed feeding and flushing process ensures that no sample residue exists, the sample quantification precision is greatly improved, the detection data deviation caused by sample loss is avoided, and the detection accuracy is improved. And an accurate and reliable basis for the to-be-detected liquid is provided for rock and ore component analysis.

This invention relates to a method for weight quantization of large language models based on statistical distribution, belonging to the field of artificial intelligence technology. This method adaptively selects the quantization threshold by analyzing the skewness, kurtosis, and dynamic range of the statistical weight distribution, achieving high-precision sparse preservation of important weights and low-bit compression of minor weights. Simultaneously, it introduces mean error compensation and variance alignment mechanisms, as well as a block-level parallel quantization strategy, to further reduce quantization errors and improve efficiency. This invention addresses the challenges of existing large language models with their massive parameter scale, requiring high-precision computation and substantial resources for inference, making them difficult to deploy on resource-constrained edge devices; and traditional quantization compression techniques suffer from significant accuracy loss, high computational overhead, and poor adaptability. This method achieves 8-bit and 4-bit quantization without retraining, achieving performance close to FP16 accuracy, significantly reducing model storage space and computational resource consumption, making it suitable for efficient deployment of large language models at edge devices.

The application provides a respiratory flow signal processing method based on dynamic calibration and normalization, individual differences are eliminated through individual treatment course calibration, temperature and humidity interference is eliminated through environment adaptive preprocessing, and zero point calibration precision is improved through adaptive Kalman filtering, so that the quantization precision (error is less than or equal to 5%) and cross-device consistency (error is less than or equal to 3%) of the respiratory flow signal are significantly improved. The method can be widely applied to the respiratory rehabilitation training scenes of children, the elderly and adults, and provides accurate signal input for VR respiratory interaction.

A processor and electronic equipment are applied to the field of data processing. The processor includes a scaling factor processing module including a smoothing pre-processing unit configured to receive the first tensor, the second tensor, and a smoothing control parameter, and a processing unit configured to process the first tensor and the second tensor by a plurality of arithmetic logic units and special function units in the processor. Executing smoothing operation of the first tensor and the second tensor to obtain a third tensor and a fourth tensor; the processing unit is configured to determine scaling factors of the third tensor and the fourth tensor and execute matrix multiplication and accumulation operation using the scaling factors based on the scaling factors of the third tensor and the fourth tensor to obtain a matrix multiplication operation result. The processor provides efficient smoothing preprocessing hardware, the processing does not need to be executed in special hardware devices such as a vector core, hardware-level efficient execution of smoothing preprocessing is achieved, hardware resource scheduling overhead and calculation delay are effectively reduced, and hardware execution efficiency is improved.

The invention discloses an unmanned aerial vehicle-mounted methane sampling and quantifying method and system, and belongs to the technical field of gas pollutant detection. The unmanned aerial vehicle-mounted methane concentric circle sampling and quantification method is realized based on the tunable laser detection technology, a grid package emission area is sampled and measured by adopting a concentric circle closed sampling method, and compared with other sampling methods, the non-dominant air methane flux is not easy to miss calculation, and the sampling accuracy is improved. According to the method, the sampling precision and stability are further improved, the tolerance to wind direction changes and headwind errors can be improved, the method can be suitable for the situation that the wind direction is unstable, the quantification precision of the methane emission rate is further improved, and compared with other similar unmanned aerial vehicle-mounted methane emission detection quantification methods, the method has the advantage that the quantification precision is improved. The relative error of the quantification precision is reduced by about 20%, and the high requirement of the unmanned aerial vehicle-mounted methane emission detection quantification method on the measurement climate condition is reduced.

The invention discloses a mycoplasma pneumoniae rapid quantitative detection kit based on nucleic acid extraction-free and application thereof, and belongs to the technical field of nucleic acid detection. The kit comprises a lysis solution, a detection premixed solution, a hydrogel monomer, a positive reference substance, a negative reference substance and a planar reaction chamber. The lysis system provided by the invention does not influence the activity of polymerase while ensuring effective lysis of pathogens, and realizes direct amplification from a clinical swab sample. By combining efficient LAMP amplification with a space-limited hydrogel microenvironment, absolute quantification is realized by fluorescent dot counting. The total detection time of the kit provided by the invention is shortened to be within 30 minutes, and the practicability of the kit in POCT and resource-limited regions is greatly improved.

The invention relates to the technical field of ecological system service function evaluation. The invention relates to an icing microtopography characteristic value quantification method. The method comprises the following steps: S1, acquiring environment data, icing data and topographic data corresponding to a detection area; s2, setting standard scenes of three types of microtopography types, performing microtopography type judgment on the detection area in combination with the standard scenes, judging and acquiring the microtopography type of the detection area, and extracting difference topography data of topography data and the standard scenes; according to the method, the self-adaptive correction model is constructed by counting the association rule of the difference topographic data and the icing characteristics, so that the icing characteristic parameters can fully reflect the actual topographic difference of the detection area, and meanwhile, the quantitative mapping formula of the icing influence factors is constructed based on the linear weighting method; in combination with a dynamic weight matrix, multi-dimensional influence factors such as seasons, daily hours, terrains and icing characteristics are integrated, dynamic optimization of weight distribution is realized, and a quantification process can better reflect a multi-factor synergistic effect.

This invention discloses a method, apparatus, device, medium, and program for quantizing a visual language model. The method includes: quantizing the visual modules of the visual language model during an offline quantization stage; compensating for quantization errors in the visual modules using a quantization compensator of the visual language model; and performing staged quantization processing on the language modules of the visual language model. The technical solution of this invention can improve the quantization accuracy of large-scale visual language models, thereby improving the inference efficiency of large-scale visual language models and reducing the inference time and computational cost of large-scale visual language models.

The invention discloses an air leakage positioning and quantifying method and system based on dynamic schlieren imaging and a space-time neural network, and the method comprises the steps: constructing a schlieren imaging system, introducing a local controllable heat source into a leakage region, and obtaining a schlieren image sequence of a leakage flow field and leakage source position information; preprocessing the schlieren image sequence to obtain a processed image sequence with unified resolution and normalized pixel values; the processed image sequence is input into a pre-trained CA-ResFpn-ConvLstm neural network model, and a leakage rate prediction value is obtained; and outputting a leakage rate prediction value, and synchronizing visual leakage source position information to complete air leakage detection. Through the integrated design of optical visualization and intelligent quantification, the technical problems that in existing industrial air leakage detection, sensitivity is insufficient, quantification is difficult, and real-time performance is poor are solved, an efficient and reliable technical scheme is provided for industrial safety monitoring, and the industrial application value is wide.

This invention relates to the field of marine ecological monitoring technology, specifically to a method for monitoring planktonic larvae communities of *Sinocyclocheilus spp.* in nearshore waters. The method includes the following steps: S1, using a continuous plankton sampler with layered trawls to obtain samples containing *Sinocyclocheilus spp.* larvae and microscopically acquiring raw image data; S2, preserving the samples at a constant temperature and protected from light, and generating digital atlases containing texture features, edges, and confidence masks under microscopy; S3, inputting a multi-stage feature network to construct embedding vectors and outputting classification labels and instance masks; S4, performing morphological comparison and confidence backtracking correction based on the labels and masks; S5, combining filtered water volume to calculate abundance vectors and stage structure matrices to output ecological monitoring data. This invention achieves accurate identification, quantity statistics, and community abundance expression of *Sinocyclocheilus spp.* planktonic larvae, improving the intelligence and data assetization level of nearshore marine ecological monitoring.