192results about How to "Achieve balance" patented technology

The invention discloses a coal mine heat energy monitoring and cooling regulation and control method based on the Internet of Things, and relates to the technical field of coal mine heat energy monitoring and cooling regulation and control. The collected temperature data are transmitted to a central monitoring system in real time through an Internet of Things gateway; redundant temperature sensors are arranged at key positions, and data of the temperature sensors in the same area are regularly compared and verified. According to the invention, the monitoring robustness is improved through a redundant temperature sensor and a data fusion algorithm, a monitoring blind area caused by a single-point fault is avoided, and the accident risk is reduced; a self-adaptive learning algorithm is adopted to dynamically optimize a detection threshold value and a compensation strategy, so that the anomaly detection accuracy is improved, and false alarms and missing alarms are reduced; the temperature trend is pre-judged through a predictive control algorithm, cooling is conducted in advance, balance between safety and energy efficiency is achieved, overheating accidents are prevented, and energy consumption is reduced.

The embodiment of the invention relates to the field of trajectory tracking, and discloses a target detection tracking method, system and device and a storage medium, and the method comprises the steps: obtaining a plurality of image frames in a video or image sequence; performing detection by using a target detection model to obtain feature detection information of each image frame; inputting the feature detection information into a trajectory tracking model for analysis and processing, and establishing and tracking a motion trajectory of the target object; wherein the method specifically comprises the following steps of: associating the target object by utilizing the frame similarity among the plurality of bounding boxes; and establishing and tracking a motion track of the target object based on an association result. According to the method disclosed by the invention, the technical problems that the tracking object is difficult to accurately detect and the calculation time is relatively long and relatively large computing power resources need to be occupied are solved, the computing power can be reasonably distributed, the balance of high-precision detection and real-time monitoring is realized, the target object is quickly sensed in a complex shielding environment, and the target object detection and tracking effect is improved.

The invention relates to the technical field of nuclear power plant safety, in particular to a nuclear power plant fault diagnosis method based on data balancing, which establishes a data set balancing technology for deep learning through SA-TimeGAN, can effectively solve the problem of data imbalance caused by large normal data volume and small fault data volume in a nuclear power plant, and improves the fault diagnosis accuracy of the nuclear power plant. According to the method, firstly, small sample fault data are preprocessed, then, SA-TimeGAN is used for conducting adversarial training on the multiple types of small sample fault data, generated fault samples are added into an original training data set to achieve balance of sample data, and finally, the balanced data set is input into a CNN-BiLSTM diagnosis model for training, so that fault diagnosis is achieved. According to the method, the recognition capability of the model on minority classes of fault samples is effectively enhanced, and real data distribution is approached through adversarial training, so that high-quality synthetic data is generated.

The invention discloses an intelligent agricultural cloud platform water and fertilizer control method, and relates to the technical field of agricultural planting. Through the real-time monitoring unit, farmland environment information can be monitored in real time, environment temperature, humidity and soil state conditions can be monitored in real time, accurate calculation can be performed, antagonistic reactions among various elements can be avoided, balance among the elements can be realized, the yield and quality can be remarkably improved, and the method is suitable for large-scale popularization and application. The capability of resisting severe weather of crops is enhanced, and meanwhile, the effects of saving water and reasonably applying fertilizer are achieved; under the cooperation of the data analysis unit, the data processing unit and the automatic control unit, the amount of water and fertilizer needed by planted crops can be comprehensively judged, water and fertilizer are supplied to the planted crops in a demanded mode in a real sense, the planted crops grow to the maximum extent in the corresponding stage, and therefore systematic intelligent and accurate management is achieved, and the management efficiency is improved. And waste of water resources and negative effects caused by improper application of fertilizers can be avoided.

The invention provides a container scheduling method and device in a cloud environment, electronic equipment and a storage medium. According to the application, the sub-cluster manager and the resource cluster object are introduced into the sub-cluster, the main cluster manager is introduced into the main cluster, and the sub-cluster manager, the resource cluster object and the main cluster manager cooperate with one another. The real-time resource sensing and resource balancing capabilities of each sub-cluster in the cloud data center in a federated scene are realized, the real-time performance of detecting whether the container is normal or available is improved, the container can be immediately scheduled to other nodes or clusters under the condition that the container cannot obtain actually required system resources, and the resource utilization rate is improved. The unavailable duration of the container is reduced, the influence on the quality of the cloud service provided by the container to the outside is avoided, and the container availability of each sub-cluster in the cloud data center in the federated scene and the comprehensive resource utilization rate of the sub-clusters are improved.

The invention relates to the technical field of air purification and ventilation, and discloses a circulation control method and system for a pressure difference channel of a clean laboratory and a storage medium, the method is based on a clean conveying line and a dirty conveying line, each line comprises a conveying channel, a buffer cabin and a sealing door, and an air feeder and an exhaust fan are arranged in the channels respectively. The rotating speed and the working time of the fan are adjusted according to the volume and the number of the flow devices by obtaining the flow instruction, and the conveying line can be disinfected according to the cleaning instruction. According to the method, pollution can be effectively avoided, circulation requirements can be accurately matched, the conveying line is regularly cleaned, the operation quality of a clean laboratory is comprehensively guaranteed, and it is ensured that the requirement for high cleanliness of an experiment space is met.

The invention discloses an environment-friendly efficient cosmetic hose and a preparation method thereof.The cosmetic hose is prepared from, by weight, 60-80 parts of low-density polyethylene, 9-15 parts of high-density polyethylene, 15-20 parts of metallocene polyethylene, 30-40 parts of a PCR regenerated material, 1-3 parts of an antioxidant, 5-10 parts of a plasticizer, 1-3 parts of a coupling agent and 2-5 parts of a graphene auxiliary. The cosmetic hose is prepared by reasonably proportioning the raw materials, mixing and melting the raw materials at high temperature, extruding the raw materials into a hose shape, and then cooling, shaping, cutting, polishing and packaging the hose. Compared with the prior art, the prepared environment-friendly efficient cosmetic hose has excellent environment-friendly performance and also has excellent barrier property, tensile strength, antibacterial property and ultraviolet aging resistance, long-term use stability of the cosmetic hose is facilitated, and good quality control is guaranteed in the storage process of packaged contents.

The invention discloses an edge-end cooperative task division and resource allocation method based on model segmentation, relates to the field of edge computing video analysis, and integrates CNN channel pruning, bilinear sampling and affine quantization multi-dimensional compression of intermediate transmission feature data; a parameter-shared dynamic neural network is trained based on a sandwich rule and a knowledge distillation technology, and real-time switching of characteristic data compression amplitude is realized; modeling user reasoning performance indexes in edge-end cooperative task division, and constructing an optimization problem of maximizing reasoning accuracy meeting user performance demand constraints and edge resource constraints; a solution algorithm based on dynamic programming is adopted, optimal configuration such as model break points and intermediate feature compression parameters is selected for users, and edge communication resources and computing resources are efficiently allocated, so that the reasoning accuracy under the multi-user condition is improved.

The invention relates to the technical field of inverter control, in particular to a neutral-point voltage full-range control system and method when a T-type three-level inverter breaks down, the T-type three-level inverter is important equipment in an energy router, and when a switch tube of the T-type three-level inverter breaks down, the problems of neutral-point voltage imbalance, current distortion and the like can be caused. According to the invention, the control capability of the neutral-point voltage is greatly improved by decomposing modulation waves when a switch tube fails, neutral-point voltage deviation elimination and oscillation suppression in a full modulation degree and full power factor range are realized, and fault-tolerant operation and neutral-point voltage balance control when the T-type three-level inverter fails are realized on the whole; safe and stable operation of the energy router system is realized, and safe operation of large and important electric power systems such as an electric power system, aerospace, metallurgical mining industry, rail transit and the like is guaranteed.

The invention discloses an intensive culture precise feeding control method based on fuzzy-MPC control, and relates to the technical field of aquaculture intelligent control, and the method comprises the steps: employing a sensing system, collecting multi-source sensing data including fish activity images, healthy growth states and water quality environment data in real time, and forming a comprehensive state data set; and based on the acquired multi-source sensing data, constructing an MTL-LSTM-SAT multi-task learning model. According to the method, the multi-task learning model fusing the LSTM and the soft attention mechanism is constructed, the long-term dependency relationship and dynamic association among the water quality, the feeding and the fish state can be captured, the attention mechanism enables the model to be automatically focused on a key time step, the perception ability of the water quality fluctuation and the feeding influence stage is enhanced, and the accuracy of the water quality fluctuation and feeding influence stage is improved. Meanwhile, double-task collaborative prediction of the growth speed and the disease proportion is achieved through the shared feature layer, a prospective state sequence of the future 7 days is output, and a quantitative and reliable prediction basis is provided for feeding decision making.

The invention discloses a multifunctional indoor storage rack which comprises a base with universal wheels, a vertically telescopic rod, a storage rack main body arranged above the telescopic rod and a dustproof soft cover arranged above the storage rack main body. And a plurality of hanging rods which can slide left and right and are locked through fasteners are arranged on the storage rack main body, so that the hanging or placing distance can be flexibly adjusted according to the width of articles. One end of the dustproof soft cover is wound on the reel in the left frame rod, and the other end of the dustproof soft cover is provided with a magnetic suction rod, can be pulled out and wound on the supporting framework above, and is finally fixed on the right frame rod in a magnetic suction manner, so that a dustproof structure covering the upper part of the storage area is formed. According to the design, dust mainly from the upper portion settles, effective dust prevention is guaranteed, meanwhile, the structure is simplified, and operation is convenient. The device has the advantages of being convenient to move, adjustable in height and width, efficient in dust prevention and the like, reasonable in structure, high in function integration degree, suitable for various indoor storage scenes and high in practicability.

The utility model discloses an air floating platform for optical lens detection, and belongs to the technical field of precision optical detection and air floating supporting, the air floating platform comprises a base, an air floating platform substrate is arranged above the base, a plurality of groups of air floating holes are formed in the air floating platform substrate, and pressure sensing valves are arranged in the plurality of groups of air floating holes; laser interferometer adjusting mechanisms are arranged on the periphery of the air floating platform base plate, the four laser interferometer adjusting mechanisms are slidably connected with the periphery of the air floating platform base plate, a honeycomb matrix air cushion array is arranged in the base, four supporting adjusting mechanisms are arranged at the bottom of the base, and each supporting adjusting mechanism comprises a supporting frame. A threaded rotating shaft is rotationally connected into the supporting frame. When the device is used, the overall balance of the base can be better adjusted when the device is used or after the position of the device is changed, other tools are not needed during adjustment, the overall balance is accurate through the thread adjustment mode, and the practicability of the device is improved.

The invention relates to the technical field of LED display, in particular to an LED fan screen, a heat dissipation control method and a display device, and aims to solve the technical problems of how to accurately obtain the overall working temperature of the LED fan screen and how to automatically adjust the rotating speed of the LED fan screen based on the overall working temperature. In order to achieve the purpose, according to the technical scheme, the method comprises the steps that the working temperature of each position is obtained through temperature sensors arranged in a plurality of heating areas, the overall working temperature of the LED fan screen is obtained based on the temperature influence weight of each heating area, and the rotating speed of the LED fan screen is controlled according to the overall working temperature. Through real-time temperature monitoring and intelligent heat dissipation control of main heating devices of the LED fan screen, the consistency of display colors and the uniformity of brightness of the LED fan screen can be improved, the service life of a core device is effectively prolonged, and meanwhile balance between heat dissipation and energy saving can be achieved.

The application relates to the technical field of yaw slip ring state monitoring, and specifically discloses a multi-dimensional state operation monitoring method and system based on a yaw slip ring, which comprises the following steps: obtaining an original operation data set of the yaw slip ring, wherein the original operation data set contains operation parameters in the electrical, mechanical, attitude and environmental dimensions which are collected via sensors and subjected to redundancy checking; constructing a noise suppression matrix based on the physical mechanism of the yaw slip ring, filtering isolated fluctuations and cross-modal abnormal values in the original operation data set by using the noise suppression matrix, and inputting the noise suppression matrix into a convolutional neural network model to obtain a fusion feature vector. By introducing the noise suppression matrix constructed based on the physical mechanism and the channel interaction mechanism guided by the physical topology, the technical problem that a lightweight model is difficult to extract cross-modal strong coupling features at an edge end is effectively solved, and the sensitivity and accuracy for early composite faults are significantly improved without increasing the computing power burden.

This application relates to the field of organosilicon polymer materials, specifically disclosing a method for preparing and applying a functionalized MQ silicone resin. A method for preparing a functionalized MQ silicone resin includes the following steps: (1) raw material pretreatment; (2) simultaneous addition reaction: mixing and reacting a hydrogen-containing MQ silicone resin, a terminal olefin monomer containing a benzocyclobutene functional group, and a terminal olefin monomer containing a C-F bond; (3) post-treatment to obtain the functionalized MQ silicone resin; wherein the molar ratio of Si-H bonds in the hydrogen-containing MQ silicone resin, C=C bonds in the terminal olefin monomer containing a benzocyclobutene functional group, and C=C bonds in the terminal olefin monomer containing a C-F bond is 1:(0.2-0.6):(0.1-0.4). This application enables the synergistic effect of the rigid benzocyclobutene group and the fluorinated flexible segment, one-step grafting, and structural controllability, ultimately preparing a functionalized MQ silicone resin with excellent dielectric properties, high temperature resistance, low-temperature flexibility, good hydrophobicity, and excellent mechanical strength.

This invention relates to the field of photovoltaic monitoring technology, and discloses a photovoltaic monitoring adaptive sampling method, system, device, and medium based on multi-parameter coupled decision-making. The method includes: constructing a multi-dimensional environmental feature vector; dynamically setting a sampling threshold; determining the current operating mode and the corresponding baseline sampling period; introducing a hysteresis comparison mechanism and a delay confirmation mechanism to optimize the switching control of the operating mode; and introducing a correction coefficient to correct the baseline sampling period to obtain the final sampling period; performing mixed verification on the collected data and outputting data confidence labels; monitoring the operating status in real time; and judging and handling real-time operating faults based on the data verification results. This invention, by constructing a multi-dimensional environmental state vector, performing adaptive sampling based on the environmental state vector, and calculating the sampling period in real time, can output a clean dataset with confidence labels and back-optimize the threshold parameters of the decision layer.

The present application relates to an online batch data processing method and device for training deep neural networks, the method comprising the following steps: sampling a data batch from a given dataset; selecting a subset S from the data batch by using an orthogonalization-based objective function; training a target deep neural network based on the subset S. Compared with the prior art, the present application uses an orthogonalization method to remove the redundancy between subsets, provides an objective function based on orthogonalization de-redundancy correlation to measure the diversity and representativeness of the subsets at the same time, and provides an efficient optimization method based on the objective function and a corresponding data selection algorithm, which can efficiently select diverse and representative subsets.

This invention discloses a Z-axis micromechanical accelerometer and its control method, belonging to the field of acceleration measurement technology. An asymmetric seesaw sensing structure is constructed using mass blocks and a torsional elastic beam. Two symmetrical multifunctional capacitors are formed by designing electrodes below the two mass blocks. The torsion generated by the seesaw structure under Z-axis acceleration can be detected by the change in the differential capacitance of the capacitors, where the capacitors serve as the detection mechanism. Based on the detected change in torsion angle and the PID controller, a PWM voltage for force balance control can be calculated and applied to the symmetrical capacitors via a push-pull method, together with a DC tuned voltage. This generates a constant electrostatic negative stiffness, thereby reducing the equivalent stiffness of the seesaw sensing element, and also generates an electrostatic resultant force that can be modulated by the PWM, maintaining the position of the seesaw sensing structure constant, achieving closed-loop detection. The PWM voltage reflects the magnitude of the external acceleration.

The present application provides a kind of porous carbon and carbon nanotube combined preparation system and method, including activation fluidized bed, carbonization fluidized bed and carbon nanotube preparation fluidized bed;Activation fluidized bed is filled with energy carbon carrier, after oxygen and activating gas are imported, energy carbon carrier and oxygen occur combustion heat release, make activation fluidized bed temperature rise, first tail gas formed is imported into carbonization fluidized bed filled with carbon particle precursor, carbon particle precursor is carbonized under the action of high temperature carried by first tail gas, and generates carbide particle, and the second tail gas containing C1-C 30 ;Carbide particle is transported to activation fluidized bed, by controlling the gas flow rate of first gas inlet, so that carbide particle is always in the upper part of activation fluidized bed, and makes carbide particle activate in activating gas, and forms porous carbon particle;Second tail gas enters carbon nanotube preparation fluidized bed, and under the action of nanometer metal catalyst, C1-C 30 In it, is converted into carbon nanotube to realize low-cost co-production of porous carbon and carbon nanotube.

The invention provides audio automatic detection equipment, and relates to the technical field of audio detection. The audio automatic detection equipment comprises a to-be-tested sound source assembly, and further comprises an air nozzle configured to provide air flow required by a wind noise test for the sound source assembly, and the air nozzle is in fluid communication with a pump set through an air guide pipeline; the artificial ear assembly is correspondingly arranged under the wind nozzle, a main body of the artificial ear assembly is a tubular outer shell, a limiting bayonet is formed in the top end of the outer shell and used for fixing the sound source assembly to be detected, and an audio detection piece is arranged in the bottom end of the outer shell to collect acoustic signals of the sound source assembly in the wind noise environment. According to the invention, a fitting degree quantification structure of deformation of the elastic soft film is matched with an airflow channel switching structure driven by the reversing plug, a test condition close to a real use scene is constructed, the structure can accurately reproduce actual working conditions of earphone wearing loosening and airflow interference, a multi-factor coupling test of ANC intervention and airflow injection under a fitting gap is realized, and the test efficiency is improved. And the influence degree of the airflow on the ANC noise reduction effect and the tone quality is quantified.

The invention discloses a traffic dynamic shortest path distribution method and system based on a triangular fuzzy number weight, and relates to the technical field of intelligent traffic systems.Fuzzy information in a path evaluation attribute is represented through a triangular fuzzy number, a comprehensive weight is analyzed and calculated in combination with deviation maximization and a moisture value, and the decision reliability is effectively improved; a K-Means algorithm is adopted to perform dynamic path search, and through a representative vertex subset pre-calculation and increment updating strategy, the path retrieval efficiency is remarkably improved; a travel time prediction model related to departure time and real-time flow is established, and accurate response to accidental congestion is realized; and finally, outputting an optimal path scheme through a multi-objective decision-making method. Compared with a traditional algorithm, the speed is improved on a real road network data set, the path distribution accuracy is improved, meanwhile, the path safety, comfort and efficiency are effectively balanced, real-time path planning under a large-scale road network is supported, and an efficient solution is provided for intelligent traffic management.

The invention provides a hydrostatic driving method and device for a snow pressing vehicle and a hydraulic system, and relates to the technical field of snow pressing vehicles. The method comprises the steps that the actual rotating speed of an engine is detected and compared with a preset optimal rotating speed range, and a rotating speed deviation signal is generated; based on the rotating speed deviation signal, the displacement adjusting amount of a variable pump and the displacement adjusting amount of a variable motor are calculated through an incremental PID control algorithm; dynamically adjusting the displacement of a variable pump and a variable motor according to the displacement adjusting quantity so as to realize stable control on the rotating speed of the engine; wherein when the load is increased, the variable displacement motor is adjusted to increase the displacement, and if the displacement of the motor reaches the maximum, the variable displacement pump is adjusted to reduce the displacement; when the load is reduced, the variable displacement pump is adjusted to increase the displacement, and if the pump displacement reaches the maximum, the variable displacement motor is adjusted to reduce the displacement. According to the method, the pump and the motor are cooperatively controlled through the incremental PID, the rotating speed of the engine is stabilized, efficient operation of the engine is maintained, and therefore the power utilization rate and the fuel economy are improved.

This invention discloses a method and system for protecting privacy in wearable devices for human activity recognition, aiming to solve the problems of existing methods struggling to balance privacy and utility, and high resource consumption on the edge. The method first acquires multi-channel time-series signals from the wearable device and constructs fixed-length feature vectors. Based on a sensitive attribute inference model and the number of high-risk features, it divides the device into high-risk and low-risk feature sets. Only the high-risk feature sub-vectors are subjected to perturbation noise to generate perturbation latent representations, which are then uploaded to an edge server along with the low-risk feature set. The edge server generates the final feature vectors and assesses the risk of privacy leakage. An offline evaluation is used to construct a privacy-utility trade-off, dynamically adjusting the perturbation dimension and noise amplitude. This invention achieves targeted protection of high-risk features, maximizing the preservation of key recognition information, reducing edge computing power and communication overhead, adapting to the differentiated needs of different scenarios, and balancing privacy security and activity recognition performance.

The invention relates to a flood-proof drainage system driven by a mining explosion-proof motor and a leakproofness detection method thereof. When running in a dry environment, the main drainage pump runs in a dry mode, and the functions of the main drainage pump are the same as those of a mine main drainage system; when the mining anti-explosion dry-type motor is submerged, the waterproof sleeve is subjected to first-layer sealing, the oil chamber is subjected to second-layer sealing, an outer-static and inner-dynamic multi-sealing structure is formed, and reliable operation of a unit in a diving state is guaranteed. Besides, both the interior of the waterproof sleeve and the interior of the oil chamber have a pressure monitoring function and a pressure adjusting function, the internal pressure of the waterproof sleeve and the internal pressure of the oil chamber are monitored in real time, liquid supplementing or air supplementing pressurization is carried out, dynamic air supplementing and air-liquid pressure self-balance of a motor cavity are achieved, the pressure difference between the two sides of a static seal or a dynamic seal is eliminated, the sealing service life is prolonged, and the sealing performance is improved. Reliable operation of the motor in a deep diving state is ensured, and normal starting of the unit during disaster relief, drainage and mine recovery is ensured. The leakproofness detection method of the flood-proof drainage system is simple and safe to operate.

This invention relates to the field of image data processing technology, and more particularly to a motion capture method and system based on multi-view images. The method includes: acquiring all views containing the target marker and the gaze direction vector; constructing a spatial information matrix based on the gaze direction vector and identifying the principal uncertainty axis to determine geometric weights; acquiring the imaging distance and spot circularity of the target marker to determine imaging quality weights; constructing a comprehensive weight matrix based on the geometric weights and imaging quality weights, and combining this with a regularization term constructed from the eigenvalues ​​of the weighted Hessian approximation matrix to construct a system of linear equations; and solving the system of linear equations to obtain the three-dimensional coordinates of the target marker. This invention reduces depth oscillations in narrow baseline or highly occluded scenes by identifying the principal uncertainty axis and combining it with physical imaging quality for weighting, and by utilizing spectral damping technology to apply adaptive constraints in the direction of missing information.

The present invention provides a heat exchanger and a method for optimizing the arrangement of heat exchange tubes thereof, wherein four adjacent heat exchange tubes in a tube bundle are arranged in a square, wherein the outer diameter of the heat exchange tube is d, the center distance between adjacent heat exchange tubes is P, and the angle between the square edge and the water flow direction is the inflow angle α, α≤90°; for the above arrangement: firstly obtain the effect of P / d, α and inlet flow rate U on the shell resistance C D and turbulence C L The response relationship between them; then C D Minimum, C L The maximum is the goal, and multi-objective optimization is performed; finally, according to the optimization results, C is selected D and C L , and the values ​​of P / d and α are inferred based on the response relationship. The tube bundle arrangement optimized by the above method can achieve a balance between shell resistance and turbulence, while reducing shell resistance and increasing turbulence as much as possible to maintain a high heat exchange efficiency.

This invention discloses a health insole containing Atractylodes lancea and its manufacturing process, comprising a breathable surface layer, an Atractylodes lancea functional layer, and an elastic base layer arranged sequentially from top to bottom. The Atractylodes lancea functional layer contains microcapsules containing active ingredients, enabling the slow release of these active ingredients and their efficacy. The Atractylodes lancea functional layer employs a "locally thickened Atractylodes lancea functional layer" design in key areas such as the sole and heel of the insole to enhance the herbal care effect. The breathable surface layer has ventilation channels to achieve "strong antibacterial effect during use and energy-saving and efficacy-preserving effect when idle." This application solves the problems of easy attenuation of the efficacy of traditional Chinese medicine in existing insoles, rapid release of medicinal properties, lack of targeted care for key areas, and poor prevention and treatment effects. It precisely matches the foot health function needs under different states of idle and active use, balancing wearing comfort and health practicality.

This application relates to the field of industrial furnace temperature control technology, and discloses an energy-saving brazing furnace heating process temperature adaptive control method and system. The heating process of the brazing furnace is divided into at least three continuous temperature ranges, including a first temperature range, a second temperature range, and a transition temperature range. A first power control strategy and a first temperature fluctuation tolerance are configured for the first temperature range, and a second power control strategy and a second temperature fluctuation tolerance are configured for the second temperature range. The current temperature range of the brazing furnace is determined based on the current temperature. When it is determined that the brazing furnace is currently in the first temperature range, the first power control strategy is executed to output heating power, ensuring that the characteristic temperature is within the first temperature fluctuation tolerance. When it is determined that the brazing furnace is currently in the second temperature range, the second power control strategy is executed to output heating power, ensuring that the characteristic temperature is within the second temperature fluctuation tolerance. When it is determined that the brazing furnace is currently in the transition temperature range, a smooth switching logic is executed.

The present application relates to the technical field of infrared image processing, and provides an end-side lightweight infrared image denoising method and device and electronic equipment.The method comprises the following steps: acquiring an infrared image with noise to be processed; inputting the infrared image with noise into a target lightweight network trained in advance to output a predicted denoising image; the target lightweight network is obtained by optimization in a training stage based on a joint loss function; the joint loss function at least comprises a pixel-level reconstruction loss, a gradient domain sparse regularization loss and a vertical direction smoothing constraint loss; the gradient domain sparse regularization loss constrains a residual image between a reference image and the predicted denoising image, so that the residual image remains sparse in the gradient domain; the vertical direction smoothing constraint loss is used to calculate a difference value of adjacent pixels in the vertical direction of the predicted denoising image, and the difference value is punished to suppress horizontal stripe noise in the infrared image.The present application solves the contradiction between limited computing power and high-fidelity denoising.

The application discloses a fair and sequential model predictive control multi-objective optimization method for a parallel T-type three-level rectifier, and aims at solving the problems of power distribution, midpoint voltage balance, circulating current suppression, DC voltage stability and grid current tracking in a parallel T-type three-level rectifier system in a high-power rectifier occasion. The application adopts double closed-loop control, the outer ring adopts adaptive droop control with voltage feedforward, realizes the stability of DC voltage and load power distribution, the inner ring adopts fair and sequential model predictive control, through the fair priority order, not only the advantage that the sequential model predictive control does not need to select a weight factor is reserved, but also the problem that the optimal solution cannot be selected due to the fixed priority order of the sequential model predictive control is solved, the balance of the midpoint voltage, the circulating current suppression and the tracking of the grid current are realized, and the overall scheme realizes the multi-objective optimization control of the parallel system.