76results about How to "Inhibition effect" patented technology

The embodiment of the invention relates to a suspension collaborative steering and fault-tolerant method for a wave adaptive unmanned ship, and the method comprises the steps: collecting the environment perception information and attitude data of the unmanned ship in real time, and carrying out the fusion processing, and generating a multi-source data matrix; pre-judging the front sea condition based on the multi-source data matrix, and when the sea condition characteristic parameters meet preset dynamic adjustment conditions, adjusting the expansion and contraction amount of each group of electromagnetic suspensions in a pre-judgment time window so as to form a draught gradient adapting to the sea condition change; when a steering instruction is received, the draft difference of floating bodies on the two sides of the unmanned ship and the thrust difference of propelling systems on the two sides are synchronously adjusted, so that the steering track is controlled, and transverse inclination of the ship body is restrained; in the sailing process, the working states of the electromagnetic suspension and the propelling system are monitored in real time, faults are classified, when it is detected that the execution units are abnormal, a fault-tolerant reconstruction strategy is triggered according to the fault types, and stress distribution or propelling thrust is reconstructed by locking abnormal parts and utilizing the remaining execution units, so that the sailing capacity of the unmanned ship is maintained.

The invention provides a low-flow-resistance underwater sound insulation device which is characterized by comprising a plurality of air bags with the longitudinal sections in the shape capable of being densely laid. A set gap is kept between every two air bags, and the air bags are perpendicular to the water surface The device has the advantages that the broadband sound insulation performance is excellent, the main frequency band of underwater noise measurement is covered, and the influence of various noise sources on the test environment can be effectively inhibited; the hydrodynamic compatibility is good, the through-flow design of air bag array arrangement can maintain low flow resistance while realizing sound isolation, and significant disturbance to a test environment is avoided; and the modular design of the air bag is convenient to install, maintain and expand, and large-scale deployment in various water area environments is facilitated.

This invention discloses a ship dynamic positioning anti-interference control method considering vertical dynamic optimization, comprising: establishing a four-degree-of-freedom nonlinear dynamic model of the ship and obtaining the error vector by combining the desired pose; defining a virtual control law by backstepping and processing the velocity error using a first-order filter; approximating the uncertainties in the model based on an improved Actor-Critic reinforcement learning framework and designing an interference observer to estimate external disturbances; calculating the compensated desired acceleration by combining the approximation output of the uncertainties, the estimated value of the external disturbances, and the velocity error; obtaining the thruster pitch ratio based on the desired acceleration and generating the final control command through a thrust distribution model. This invention can effectively handle model uncertainties and external environmental disturbances in ship dynamic positioning, especially optimizing vertical dynamic performance and improving control accuracy, response speed, and system robustness.

The application belongs to the technical field of positioning and measurement, and particularly relates to a target positioning method and device combining CSI coarse positioning and Y-type CCD fine correction, which comprises the following steps: S1, CSI acquisition equipment is arranged in a target area, and CSI data between a sending end and a receiving end is acquired; S2, a Y-type CCD sensor is geometrically calibrated and a coordinate system is corrected, and coordinate reconstruction is completed; S3, the CSI data is preprocessed, noise is removed, and input is normalized; S4, time domain features of the CSI are extracted by using a time domain branch; S5, frequency domain features of the CSI are extracted by using a frequency domain branch; S6, a gating fusion module is used to fuse the time domain and frequency domain features, and the approximate position coordinates of a target are output; S7, the Y-type CCD sensor is controlled to optically acquire a target area, and point light source coordinate information is acquired; and S8, error correction is performed on the data acquired by the Y-type CCD, and the accurate position coordinates of the target are output.

The utility model discloses a kind of hydrogen energy power generation systems of quick heat dissipation, including outer frame, be set on base and be provided with hydrogen energy power generation module inside;Several open and close doors, hinged on outer frame and be provided with electric louvers on open and close door;Several heat dissipation fans, fixed in outer frame top and into positive pressure air to outer frame inside;The base is equipped with cavity, and the circuit of the hydrogen energy power generation module is laid in base cavity.This hydrogen energy power generation system can realize power generation function by setting power generation module inside outer frame, simultaneously, by setting heat dissipation fan and electric louvers, the ventilation inside can be guaranteed, it is convenient to quick heat dissipation, and by laying circuit in the cavity inside base, the temperature inside outer frame can be reduced, effectively guarantee the normal operation of hydrogen energy power generation system.

A transmission characteristic parameter identification method for a ball screw feeding system comprises the following steps: in the normal operation process of the ball screw feeding system, acquiring voltage, current, rotating speed and other signals through a servo motor and a driving system thereof, and realizing online identification of key electromagnetic parameters and mechanical parameters in stages; under the static working condition of the motor, controlled voltage excitation is applied, stator current is measured, and identification of stator winding resistance and stator inductance is completed; under the steady-state operation condition, the flux linkage amplitude of the permanent magnet is calculated based on a q-axis voltage equation; under a dynamic operation condition, a friction torque model is established through segmented constant-speed operation, a rotor motion equation is combined, a recursive least square algorithm with a forgetting factor is introduced, and online identification of the equivalent rotational inertia of the system is realized; according to the method, complete and accurate transmission characteristic parameters can be obtained in the ball screw feeding system, the identification process is stable, the calculation complexity is low, and the method is suitable for being embedded into a servo control system.

The application provides a band gap reference voltage source circuit, which comprises a starting circuit, a core circuit and a current compensation circuit. The core circuit comprises a first NMOS transistor, first and second transistors, a first operational amplifier, first to third resistors and a compensation resistor. The drain of the first NMOS transistor is connected to a power supply, the source is connected to the first end of the first resistor, the base of the first transistor and the collector of the second transistor, and the gate is connected to the output end of the first operational amplifier. The second end of the first resistor is connected to the collector of the first transistor and the base of the second transistor, and the first end of the first resistor serves as the output end of the core circuit. The emitter of the first transistor is connected to the first end of the second resistor and the same input end of the first operational amplifier. The emitter of the second transistor is connected to the first end of the third resistor and the opposite input end of the first operational amplifier. The second end of the second resistor and the second end of the third resistor are respectively connected to the first end of the compensation resistor and the output end of the current compensation circuit, and the second end of the compensation resistor is grounded.

The invention discloses a load reference measurement system and method based on two-dimensional grating diffraction. The method comprises the following steps: designing a layout geometric scheme of a collimation light path and key parameters of an imaging camera according to the measuring range and precision requirements of a load reference measurement system; structural parameters of a two-dimensional circular hole grating arranged at the front end of the imaging camera are designed according to the LED light source parameters and the key parameters of the imaging camera; at the front end of the imaging camera, enabling the collimation light path to form multi-stage diffraction light spots through a two-dimensional circular hole grating, obtaining the equivalent center-of-mass coordinates of the double light paths by adopting a multi-light-spot center-of-mass averaging method, and calculating the propagation direction of each light beam in an imaging coordinate system based on the equivalent center-of-mass coordinates; and in combination with a collimation light path layout geometric scheme, a load reference under an imaging camera coordinate system is calculated through inversion. According to the invention, the influence of LED light direction fluctuation, imaging noise and single light spot positioning error on the measurement result can be effectively inhibited, and the on-orbit high-precision measurement of the load reference is realized.

The application provides an inland water body extraction method fusing a cavity space pyramid pooling. The method comprises the following steps: constructing a water body extraction network; the water body extraction network adopts a SegFormer network as a framework, and a new decoder is obtained by improving a decoder in the SegFormer network; the new decoder comprises a first 1x1 convolution layer, an ASPP module, a first upsampling module and a semantic fusion module which are sequentially connected from top to bottom; after a remote sensing image is subjected to feature extraction by an encoder, the remote sensing image is subjected to processing by the new decoder; wherein the semantic fusion module takes the feature map extracted by the bottommost layer of the encoder and the feature map output by the first upsampling module as inputs, and takes the output after fusion as a water body extraction result; the water body extraction network is trained to obtain a water body extraction network model; and the trained water body extraction network model is used to extract water bodies from input remote sensing images. The method has good effects on water body edge integrity detection, reduction of small water bodies, branch leakage detection and false detection.

The application provides a PIV-oriented multi-scale pyramid optical flow estimation method, comprising: performing adaptive image preprocessing on an image pair; inputting the processed image into a lightweight feature extraction network to obtain multi-scale feature maps and context feature maps of two frames of images; using a traditional dense optical flow algorithm to calculate an initial optical flow and down-sample the initial optical flow to the deepest layer resolution; using the initial optical flow to locally sample the second frame feature map, and calculating a dot product of the first frame feature map and the second frame feature map to obtain a correlation body; fusing the correlation body and the optical flow information through a multi-scale motion encoder to regress a residual optical flow; performing weighted up-sampling on the residual optical flow based on neighboring pixels; iteratively performing the above steps layer by layer until a full-resolution optical flow estimation is output. Through the fusion of a traditional optical flow prior and a multi-scale pyramid iterative network, the application solves the problem of low optical flow estimation accuracy caused by large changes in PIV environment illumination, similar foreground and background, and transparent objects, reduces the calculation complexity, and meets the real-time requirement.

The present application belongs to the technical field of data processing, and relates to a mental health detection method and device based on large model fine-grained enhancement and a medium. A large language model is used to convert a mental health scale into an extended questionnaire and generate a discrimination standard based on mental health diagnosis rules. For each group of situation, clue, thinking cognition and behavior data in the mental knowledge data set, the situation and clue are combined and mapped into a key vector, and the thinking cognition and behavior are combined and mapped into a value vector. An individualized question set is obtained based on the interview records of the extended questionnaire, the question and answer results of the individualized question set are combined with preset prompt words and input into the large language model to extract mental health representation evidence. The mental health representation evidence is mapped into a feature vector and the similarity with each key vector is calculated. The value vector corresponding to the K key vectors with the highest similarity is combined with the mental health representation evidence, and then the combined result is input into the large language model together with the discrimination standard to output the mental health score of the user to be evaluated and the determination reason.

The application provides a kind of fan blade state monitoring system and method for production in transport process, belong to wind turbine fan blade online monitoring field, the system includes multidimensional sensing data acquisition terminal and blade state early warning terminal;Multidimensional sensing data acquisition terminal includes vibration acceleration acquisition module, multi-node strain monitoring module, first control module and first wireless transmission module;Multi-node strain monitoring module includes a plurality of strain monitoring nodes distributed in blade vulnerable loss site, each strain monitoring node includes two groups of mutually perpendicular strain monitoring components, each group of strain monitoring components includes monitoring strain sensor and compensation strain sensor, monitoring strain sensor is arranged on blade, and compensation strain sensor is arranged on compensation plate of same material with blade.By the monitoring system provided in the application, the strain signal can be compensated, the influence of environmental temperature and other factors on strain monitoring can be effectively inhibited, and the blade state estimation accuracy can be effectively improved.

The present application relates to the technical field of medical equipment, and more particularly to a pancreatic anastomosis intelligent negative pressure drainage system for pancreatic fistula monitoring, which comprises: a baseline construction module for constructing an individual reference baseline data set within a preset time period; an enzyme concentration detection module for detecting whether the pancreatic amylase concentration in the drainage fluid exceeds an abnormal trigger threshold; a tube state identification module for determining a tube state identification result through tube state visual identification and flow rate consistency verification when the enzyme concentration abnormal duration exceeds a threshold; a saturation compensation module for extracting the plateau of the pancreatic amylase concentration curve and judging whether the saturation sequence of each detection point meets the requirements; and a correction module for correcting the initial negative pressure when the saturation sequence requirements are not met. Through the cooperation of multiple modules, the present application realizes real-time and accurate monitoring of the pancreatic anastomosis fistula, improves the reliability and response speed of postoperative complication identification, and provides intelligent drainage state judgment and early warning support for clinical use.

This invention relates to the field of current measurement technology and discloses a method and system for improving the accuracy of NV center current measurement in diamond. The method includes: applying an axial magnetic field outside the diamond sample; comparing the energy level difference between two characteristic states with the target spin; determining whether energy level anti-crossing exists based on the comparison result; determining the fluorescence intensity based on the energy level anti-crossing; determining a normalized fluorescence signal based on the initial fluorescence signal and the target fluorescence signal; determining the coupling relationship between the magnetic field and fluorescence intensity based on the fluorescence intensity, the normalized fluorescence signal, and the placement probability; applying a bias magnetic field opposite to the axial magnetic field based on the open-loop transfer function and the microwave frequency; modulating the measurement microwave frequency with a sine wave; and measuring the NV center current based on the Fourier coefficients, the adjusted signal-to-noise ratio, and the bias magnetic field. This invention improves the accuracy of current measurement.

The invention relates to the field of indoor mapping, discloses a mobile robot indoor mapping method based on multi-sensor fusion, and relates to the technical field of mobile robot navigation. The method comprises the following steps: an upper computer processing module collects data of a wheel type odometer, inertia measurement and a laser radar and carries out time alignment; calculating a state prediction value based on a kinematic model, and generating a posterior state estimation value by using extended Kalman filtering and fusing multi-source observation data; the current active local sub-graph is obtained by taking the current active local sub-graph as an initial value, a continuous probability map model based on bicubic interpolation is constructed, and an optimal matching pose is solved by utilizing a nonlinear optimization algorithm; and updating the probability grid map based on the optimal pose, and generating an indoor map through multi-resolution loopback detection and global pose map optimization introducing a robust loss function. According to the method, slipping and zero offset errors are effectively compensated, the problem that gradient optimization cannot be carried out on a discrete map is solved, and the robustness and precision of indoor mapping are improved.

The present application provides an optical concentration measuring device which suppresses the influence of distortion of an optical image caused by a processing error. An optical concentration measuring device (1) includes a substrate (2), an emitting section (3) provided on a main surface (20) of the substrate and emitting light, a light receiving section (4) provided on the main surface and receiving at least a portion of the light emitted by the emitting section, and a light guiding section (5) guiding at least a portion of the light emitted by the emitting section to the light receiving section, the light guiding section having a first reflecting section (51) and a second reflecting section (52), the first reflecting section having a main reflecting surface (53) and two sub reflecting surfaces (54), each of at least a portion of which is a quadric surface, a plurality of light spots being generated on the main reflecting surface when viewed from the front of the second reflecting section, the first reflecting section having a correction surface correcting at least one of the plurality of light spots, the correction surface being provided between the main reflecting surface and at least one of the two sub reflecting surfaces and extending in a direction opposite to the second reflecting section.

This invention relates to a storage I / O multi-path routing method and system based on Cgroup tracing, belonging to the field of data storage technology. It solves the problem of performance degradation caused by the inability of existing multi-path routing to perceive business priorities and isolate abnormal traffic. The method includes: obtaining the Cgroup identity information to which the I / O request belongs, and then determining the path channel type corresponding to the I / O request; path channel types include: dedicated channels, isolated channels, and general channels; determining the physical path set corresponding to the path channel type from multiple pre-divided physical path sets, where different path channel types correspond to different physical path sets; selecting a physical path from the physical path set as the target path, and sending the I / O request to the storage device through the target path. This achieves business-aware, accurate multi-path routing, improving business performance.

The application discloses a kind of inhibiting ICI radar communication integrated system signal processing method, compensates ICI in echo signal using SIDFnT algorithm, removes the influence of random communication information loaded in echo signal using Hadamard division, the phase matrix of Hadamard quotient is sought, remove the influence of excessive amplitude produced by division, finally using the characteristics of two-dimensional FFT energy accumulation, target energy is concentrated, so as to estimate target distance and speed by the position of maximum amplitude in matrix, and the estimated result is used to complete the compensation of interference caused by speed distance. The application uses the energy focusing characteristics of SIDFnT, suppresses echo ICI, can effectively improve radar target estimation accuracy, and directly uses radar estimation result to complete communication end compensation, greatly reduces the complexity of communication end processing method.

This application primarily protects a method for recognizing human movement intentions, comprising the following steps: data acquisition: acquiring and processing electromyographic signals of human movement using an sEMG signal acquisition system to obtain raw electromyographic signal data; data processing: using filters to process interference in the raw data, extracting features from the raw data to obtain a data feature set; establishing the desired objective function and evaluation function; probabilistic estimation of kernel function parameters and regularization parameters; and distribution estimation of membership kernel function parameters. To address the problems in human movement intention perception based on electromyographic signals, improve the model's recognition efficiency and robustness, and enable it to handle nonlinear data, this invention combines the excellent predictive ability of a low-order fuzzy model with a robust probabilistic model, employing expectation operations to predict movement intentions to eliminate random interference in the model and improve the model's ability to handle time-varying dynamic features and external disturbances.

The invention provides a trailer hinge angle real-time estimation method and system based on feature point tracking. The method comprises the steps of obtaining a first image frame of a target trailer at a current moment and a second image frame of the target trailer at a previous moment; determining a first feature point list of the first image frame according to a second feature point list in the second image frame; according to the first feature point list and a preset feature point template, calculating to obtain a visual estimation angle and confidence; a first Kalman filter at the current moment is updated according to the visual estimation angle and the confidence coefficient, an updated second Kalman filter is obtained, and a hinge angle estimation value at the current moment is output based on the second Kalman filter; and within a preset time period after the current time, according to the motion data of the target trailer, outputting the hinged angle estimated value of each time within the preset time period through the second Kalman filter, thereby improving the accuracy and real-time performance of real-time estimation of the hinged angle of the trailer.

The present invention addresses the problem of suppressing the influence of a sulfur component derived from a salt of an electrolyte solution when recovering a positive electrode active material from a positive electrode mixture. The present invention provides a method for recycling a lithium ion battery using LiFSI as a salt of an electrolyte solution, the method comprising: a treatment agent mixing step for mixing lithium carbonate with a positive electrode mixture recovered from a waste lithium ion battery; and a heating step for heating a mixture obtained by mixing lithium carbonate with the positive electrode mixture to a heating temperature that is equal to or higher than the melting point of LiFSI and equal to or lower than the melting point of lithium carbonate.

The application discloses a kind of based on artificial intelligence's consumer goods harmful substance intelligent screening method, it is related to artificial intelligence technical field, it is difficult to effectively identify and analyze the technical problem that a large number of natural language, commercial name or commonly known as etc. non-standardized description exists in ingredient list, it is also difficult to accurately convert these fuzzy text information into the standardized terminology that can be accurately matched by regulation database;Including the following steps: the fuzzy ingredient list containing natural language or commercial name description is preprocessed by multi-task text regularization model;Then, utilize embedding semantic similarity model and the multi-evidence fusion presumption unit of improved attention mechanism, fuzzy component is intelligently analyzed and matched, and standard substance identification and confidence are output;Finally, through multi-agent risk decision framework combines real-time regulation database, dynamic compliance screening and risk assessment are carried out;The application improves the intelligent level of consumer goods safety compliance management.

The invention relates to the technical field of vehicles, in particular to a vehicle safety certification and control system, a vehicle and a control method.The vehicle safety certification and control system comprises a multi-mode biological characteristic quantization processing module which is used for obtaining at least one biological characteristic of a user and conducting coding fusion on the biological characteristics in a hardware operation environment to generate a quantum-state identity label; the dynamic key management and vehicle control module is used for completing identity authentication according to the quantum state identity credential, generating an encryption key of a vehicle control instruction according to the traffic situation information if the identity authentication is successful, and outputting the vehicle control instruction after safely packaging the vehicle control instruction by using the encryption key in response to user operation or a preset event, and the vehicle is controlled to execute corresponding operation. Therefore, the problems that in the related technology, biological characteristic transmission of a vehicle-mounted safety system is prone to being stolen, secret key management is static and rigid and cannot respond to road conditions in real time, and quantum hardware is difficult to stably work in the vehicle-mounted vibration and high-temperature environment are solved.

The invention provides a power distribution network station area user-transformer relation identification method based on an admittance regression algorithm, and belongs to the technical field of user-transformer relation identification. The problems of high cost, low precision, poor robustness and the like in the identification of the user-transformer relationship in the low-voltage distribution area are solved; the method comprises the following steps of: acquiring three-phase voltage and three-phase current of a distribution transformer in a low-voltage distribution transformer area of which the user-transformer relationship is to be identified and voltage and current of each branch user; respectively calculating the total admittance of the transformer and the admittance of each user, and carrying out standardization processing on the calculated admittances; constructing a regression model between the total admittance of the transformer and the admittance of all users, and taking a regression coefficient in the regression model or a decision coefficient based on the regression coefficient as a basis for identifying the transformer area user-transformer relationship; estimating a regression coefficient by adopting a least square method, and setting a threshold value to judge a user-transformer relationship; the application is applied to a low-voltage distribution area.

The application discloses a novel high-temperature-resistant coreless linear motor with a water-cooling heat dissipation structure and belongs to the technical field of linear motors. The motor comprises upper and lower stator assemblies arranged in parallel and a mover component arranged in the passages of the two stator assemblies. The stator assembly is formed by fastening a stator yoke plate, a permanent magnet array and a water-cooling support plate to form a sealed liquid-cooling chamber. The permanent magnet array is completely immersed in the cooling liquid. An S-shaped flow guide rib is arranged in the chamber. The end of the water-cooling support plate is provided with a liquid inlet and a liquid outlet and is externally connected with a water-cooling machine. The water-cooling support plate is made of a non-magnetic high-thermal-conductivity material. The mover component comprises high-temperature-resistant winding coils and is wound by inorganic insulation foil wires and is integrally solidified by a high-temperature-resistant material. The application can ensure that the permanent magnet always works in a stable magnetic performance interval at a temperature of 300 DEG C or above, ensures large thrust output, significantly suppresses thrust fluctuation and improves the operation reliability of the system.

The utility model provides a driving device for mouse disease prevention, which relates to the technical field of mouse disease prevention and comprises a shell, a sound wave mouse repelling device is fixedly connected in the shell, an electricity storage module is fixedly connected in the shell, and the electricity storage module is electrically connected with the sound wave mouse repelling device through a wire. The sound wave rat repelling device is arranged in the shell, sound waves generated by the sound wave rat repelling device can be used for repelling rats, so that the function of preventing rats is achieved, a power source can be provided for the sound wave rat repelling device through cooperation of the electricity storage module, the sound wave rat repelling device can operate continuously, the prevention effect is guaranteed, and rat repelling and killing bait can be placed in the food box, so that the sound wave rat repelling device is convenient to use. The device can further play a role in preventing rats, the rat repelling effect is improved, the device can be conveniently fixed by means of matched use of the arranged sliding rail, the movable block and the fixing device, the device can be flexibly installed at different positions by moving the movable block in the sliding rail, and the device is convenient to use.

The invention discloses a channel deposition thickness online measurement method and system based on multi-source fusion self-checking. The system can synchronously collect the water level, water temperature and ultrasonic sounding data of a target section in a water operation state, inverts the bed surface elevation after sound velocity correction, and then obtains the deposition thickness through the difference between the bed surface elevation and a reference bed surface. According to complex working conditions such as muddy water and turbulent fluctuation, a self-checking system containing echo quality, repeatability, neighborhood consistency, re-checking point errors and drift indexes is constructed, a four-level re-testing mechanism of points, testing lines, fracture surfaces and a system is set, and corresponding calibration is automatically triggered according to index overrun. And finally, the deposition statistics, the thickness uncertainty and the credibility are output, and original data, processing parameters and retest records are associated and archived, so that the traceability of the whole process is realized. The method can realize stable and reliable on-line monitoring of the deposition thickness under the condition of not cutting off water, and is suitable for channel operation management under complex working conditions of sand, muddy water and the like.

The present application relates to the technical field of robots, and relates to a self-adaptive path decision optimization method and system of an intelligent logistics robot, comprising: acquiring a plurality of initial images of the intelligent logistics robot, wherein the initial images comprise images of different angles at different times; determining movement information of the intelligent logistics robot by using the plurality of initial images; determining a non-instructional movement range of an abnormal movement structure of the intelligent logistics robot by using the movement information, and determining a target path; adjusting an operation process of the intelligent logistics robot by using the target path, and executing the operation of the intelligent logistics robot by using the adjusted path. According to the present application, a three-dimensional model is established by multi-angle image analysis, and the movement characteristics of a mechanical arm connecting rod are identified, so that the instructional movement and the non-instructional shaking can be effectively distinguished, the compound movement state of the robot base movement and the mechanical arm movement can be accurately distinguished, and the influence of the non-instructional mechanical arm shaking on path planning can be effectively inhibited.

The invention discloses an underground ore grade spectrum fusion detection method based on deep learning, and relates to the technical field of ore spectrum detection, and the method comprises the steps: taking a convolutional neural network as an infrastructure, taking a spectral line conflict relation set as a constraint condition for model training, and obtaining a spectral line conflict recognition model; carrying out conflict judgment and self-adaptive digestion on the multi-modal spectrum sample set by utilizing a spectral line conflict recognition model to obtain a purified multi-source spectrum feature vector set; and performing quality evaluation and uncertainty estimation on the purified multi-source spectral feature vector set to obtain a feature reliability weight vector, and performing weighted fusion based on the feature reliability weight vector to generate a spectral fusion feature vector. According to the invention, regression prediction is carried out on the spectrum fusion feature vector by using the multi-layer feedforward neural network, deep learning driving detection and prediction credibility quantification of the ore grade are realized, and the stability and practical value of an underground ore grade detection result are improved.

This application discloses a silicon carbide CMOS integrated circuit structure, applicable in the field of semiconductor technology, comprising: a substrate, a first type of transistor, a second type of transistor, and an isolation structure; the first type of transistor includes a first well region, a first gate structure, a control structure, and a first source / drain region; the first gate structure and the control structure are spaced apart along a first direction, with the control structure disposed on both sides of the first gate structure; the second type of transistor includes a second well region, a second gate structure, a control structure, and a third source / drain region; the second gate structure and the control structure are spaced apart along the first direction, with the control structure disposed on both sides of the second gate structure. The substrate, the first well region, and the third source / drain region are of a second conductivity type; the second well region and the first source / drain region are of a first conductivity type. This application, by setting the control structure, suppresses the threshold voltage drift of the device at high temperatures, thereby improving the high-temperature operating limit of the silicon carbide CMOS integrated circuit.