27results about How to "Avoid chance" patented technology

The invention discloses a water quality detection device and method for a drainage pipe network, and belongs to the field of water quality detection. The water quality detection device for the drainage pipe network comprises a shell and a detection probe, and further comprises a flow dividing pipe which is fixedly installed at the top of the shell, a flow guiding plate is rotatably installed in the shell through a first rotating shaft, two symmetrically-arranged first baffles are rotatably installed in the shell through a second rotating shaft, and the flow guiding plate is fixedly installed at the top of the shell; a first torsional spring is fixedly mounted between the second rotating shaft and the inner wall of the shell; the fence is fixedly connected to the interior of the shell, a filter plate is installed in the fence, and the detection probe is arranged between the filter plate and the fence; water flow is guided through rotation of the flow guide plate, then water falls on the first baffle to rotate the first baffle and tighten the first torsional spring, the flow speed is reduced through the elastic force of the first torsional spring and guiding of the flow guide plate, liquid impurities are filtered through the filter plate, and therefore the detection precision is improved.

The invention discloses a water conservancy project water quality detection sampling method and device, and relates to the technical field of water conservancy project water quality detection.The water conservancy project water quality detection sampling device comprises a ship body and a mounting table, a sampling depth adjusting mechanism is fixedly arranged on the top end face of the mounting table, and a sampling mechanism is arranged on the bottom end face of the sampling depth adjusting mechanism; when the sampling mechanism descends to the sampling depth and needs sampling, the servo motor rotates to drive the cam to adjust the angle, at the moment, the protruding position of the cam makes contact with the end face of the sliding rod, the sliding rod drives the connecting block and the sliding block to move, and when the sliding block moves to expose the upper water inlet to the outside, the water level sensing mechanism senses the water level. When the lower water inlet is formed in the sampling box, sampling operation can be started, the sampling accuracy is improved, the automation degree during sampling is improved, the problem of sampling difficulty can be solved, the sampling number is two, and the contingency of samples can be effectively prevented.

The invention discloses a method and a system for relieving illusion of a large language model. The method comprises the following steps of: obtaining a plurality of candidate words of a tth time step in a large language model; for each candidate word, obtaining a prediction probability of a high-layer neural network in each layer of the large language model; for each candidate word, performing normalization processing on the prediction probability of a middle-layer and high-layer neural network in each layer of the large language model to form cross-layer prediction distribution, and calculating the entropy of the cross-layer prediction distribution as the cross-layer entropy of each candidate word; based on the cross-layer entropy of each candidate word, correcting the initial prediction probability of each candidate word under the t-th time step to obtain the correction probability of each candidate word under the t-th time step; and outputting the candidate word with the highest value of the correction probability as an output result corresponding to the t-th time step. By utilizing the method, the output illusion of the large language model can be effectively eliminated.

The invention belongs to the field of sea surface ship driving, and particularly relates to a ship lane changing intention recognition method and device, electronic equipment and a storage medium. According to the ship lane changing intention identification method, the lane changing intention of the continuous multi-frame driving images is identified and comprehensively analyzed to obtain the final lane changing intention, so that the contingency of identifying the intention by adopting the multi-frame images can be avoided by adopting the continuous multi-frame driving images for analysis, and the accuracy of the ship lane changing intention identification is improved. And accurate identification of the lane changing intention is realized. According to the method provided by the invention, for the ship with the lane changing intention, surrounding ships are determined by analyzing the trajectory of the ship. Whether other ships belong to surrounding ships or not is determined by calculating the distance and the speed between the other ships and the current ship, and a threshold value is set to filter ships far away. And a reliable basis is provided for analyzing the lane changing intention of the ship by combining a plurality of images subsequently.

The invention relates to a raindrop distribution trueness evaluation method and device oriented to site simulated rainfall. The method comprises the following steps: collecting microscopic data of real rainfall at each level; on the basis of microscopic data of real rainfall, calculating an average raindrop distribution matrix of different grades to serve as an evaluation benchmark; microcosmic data of simulated rainfall of a site are collected, and a raindrop distribution matrix of the simulated rainfall is calculated; and calculating the similarity between the raindrop distribution matrix of the simulated rainfall and the average raindrop distribution matrix of each grade, and determining a truth degree evaluation result of the simulated rainfall of the field according to the similarity. Compared with the prior art, the method provided by the invention starts from the level of the microscopic raindrop distribution matrix of rainfall, constructs the evaluation criterion based on the raindrop distribution matrix of the real rainfall sample, and can realize the quantitative evaluation of the raindrop distribution trueness of the simulated rainfall by quantifying the similarity between the simulated rainfall sample and the evaluation criterion.

The invention provides an environmental state novelty evaluation method and device and a storage medium. The environmental state novelty evaluation method comprises the following steps: based on multiple groups of random noises and an environmental state at a current moment, acquiring a low-dimensional environmental characteristic predicted value at a next moment corresponding to each group of random noises by using an environmental state transfer function; the environment state transfer function is used for predicting low-dimensional environment characteristics at the next moment according to the environment state at the current moment; and calculating an average error according to the real value of the low-dimensional environment feature at the next moment and the predicted value of the low-dimensional environment feature at the next moment corresponding to each group of random noise, and taking the average error as the novelty of the environment state. According to the environment state novelty evaluation method and device and the storage medium provided by the invention, the average error of the environment state characteristics is obtained through multiple groups of random noises to serve as the environment state novelty, so that the contingency of a single model and the high energy consumption of multiple parallel models are avoided, and the measurement accuracy of the environment state novelty is improved.

The present invention discloses a composite metal wire coating thickness measurement method, evaluation method, terminal and medium, wherein the measurement method comprises: obtaining a cross-sectional image of the composite metal wire to be measured from which the coating contour and outer contour have been extracted; marking a first number of pixel points along the coating contour, and marking a second number of pixel points along the outer contour, wherein the second number is greater than the first number; applying a KD tree algorithm, finding the pixel points marked on the outer contour closest to the pixel points marked on the coating contour, and obtaining a first number of pixel point pairs; taking the distance between two pixel points in the pixel point pair as the coating thickness, calculating and saving the distance corresponding to the first number of pixel point pairs. The coating thickness is calculated by finding the distance between the coating contour and the nearest pixel point pair on the outer contour, thereby avoiding the coating thickness measurement error caused by the center position error and improving the coating thickness measurement accuracy; the closest pixel point pair is found based on the KD tree algorithm, which greatly improves the operating efficiency.

The invention discloses an image processing device and method based on fluorescence microscope collection, and the method comprises the steps: obtaining a to-be-processed image collected by a fluorescence microscope, and generating a collection environment distribution feature vector of the to-be-processed image; inputting the to-be-processed image and the collection environment distribution feature vector thereof into a pre-trained multi-class image processing model, and outputting a target image corresponding to the to-be-processed image; wherein the pre-trained multi-class image processing model comprises an image attention distribution module, an image distributed processing module and an image feature fusion output module, and the image distributed processing module comprises a plurality of classes of image processing models. Therefore, differential image processing is carried out on the to-be-processed images under different acquisition conditions, and the adaptability and the processing effect of the image processing model are improved.

The present invention relates to the field of wind turbine tower transportation technology, and in particular to a large-scale wind turbine tower clamp-type dump transport vehicle. An actual amplitude average value is calculated based on the amplitude recorded by a vibration detector, and whether a tower transportation process meets standards is determined by comparing the actual amplitude average value with a preset amplitude average value. For a situation where the tower does not meet standards, the reason why the tower transportation process does not meet standards is determined by comparing the average amplitude in the vertical direction or the horizontal direction, and the preset amplitude average value is corrected to a corresponding value using a corresponding coefficient according to the reason. Based on the comparison result of the corrected preset amplitude average value with the actual amplitude average value, it is determined again whether the tower transportation process meets standards. If the tower still does not meet standards, the vehicle is stopped urgently and the tightening torque of a first hydraulic adjustable distance flange and a second hydraulic adjustable distance flange is adjusted.

The embodiment of the invention discloses a human body posture recognition method and device, electronic equipment and a readable storage medium. The human body posture recognition method comprises the steps of obtaining an initial posture and an original signal of a detection target, and generating a plurality of frames of point cloud data corresponding to the detection target according to the original signal; acquiring a tracking algorithm and a posture detection algorithm; inputting each frame of point cloud data into a tracking algorithm for centroid detection to obtain centroid information of the detection target in each frame of point cloud data; inputting the point cloud data and the centroid information corresponding to the point cloud data frame by frame into an attitude detection algorithm for attitude detection, and obtaining multiple frames of point cloud data in the sliding time window; and when the multi-frame point cloud data in the sliding time window satisfies a preset condition, the detection target is transferred from an initial attitude to a target attitude.

The invention discloses a human body use state dynamic evaluation method based on artificial intelligence. The method comprises the steps that state information used for reflecting the overall human body use state is acquired; analyzing the state information based on an artificial intelligence model to judge whether the human body use state is suitable for continuing or not; outputting a judgment conclusion according to an analysis result, wherein the judgment conclusion is used for indicating whether the human body use state is allowed to continue; in the continuous use process of the human body, the steps are periodically repeated to keep the consistency and stability of judgment. The invention further discloses a human body use state dynamic evaluation system based on artificial intelligence. By using the method or the system, the overall state in the continuous use process of the human body can be dynamically tracked and evaluated, the long-term consistency and stability of judgment are ensured, and running state judgment parameters of the human body use state lacked in an existing diagnosis system are complemented; the judgment result has a blocking type control characteristic; and the judgment layer and the execution scheme are structurally decoupled.

The application provides a multi-module power supply parallel current sharing control method and system, and particularly relates to the technical field of power electronics. The method comprises the following steps: collecting output current time sequence data of a plurality of parallel power supply modules, calculating current sharing residual errors of each parallel power supply module at different sampling moments; identifying a plurality of historical load events and constructing a plurality of residual error trajectories of each parallel power supply module, constructing a plurality of residual error slot occupation trajectory of each parallel power supply module according to the plurality of residual error trajectories, constructing an average residual error trajectory of each sampling moment in each type of historical load event, extracting slot occupation distribution data of each parallel power supply module at each sampling moment, constructing an identity slot trajectory of each parallel power supply module and generating a current sharing compensation trajectory of each parallel power supply module, and using the current sharing compensation trajectory for current sharing control of the plurality of parallel power supply modules. The application can improve the current sharing control pertinence and stability of the plurality of parallel power supply modules in the load change process.

The invention relates to the technical field of textile machinery component performance testing, in particular to a textile machinery component performance testing device and method.The textile machinery component performance testing device comprises an L-shaped mounting plate, a first motor is mounted on the front side face of the vertical section of the L-shaped mounting plate through a motor box, and a mounting unit is arranged at the output end of the first motor; a pushing unit, a movable testing unit and a fixed testing unit are sequentially arranged on the front side face of the vertical section of the L-shaped mounting plate from right to left and located at the circumferential position with the output end of the first motor as the circle center. According to the invention, the sharpness of the movable blade assembly is tested in a manner that the cloth for sharpness test moves in the direction perpendicular to the movable blade assembly and is cut off by the movable blade assembly, and the cloth for sharpness test is in a wrinkled state through the wrinkling element; therefore, the cutting effect of the blade of the movable blade assembly on the cloth which is in a wrinkled state and is used for sharpness testing is tested, and the test result has more practical significance.

This invention discloses an automatic detection method, device, equipment, and medium for vehicle parking status at intersections. The method includes: when a vehicle is detected entering the intersection, acquiring trajectory point data consisting of multiple continuous trajectory points, including the current trajectory point; collaboratively determining whether the vehicle is parked based on the velocity, acceleration, and jerk of each trajectory point in the event window trajectory point data; and determining the parking status based on the parking duration when the vehicle is determined to be parked and its travel direction is consistent with the main traffic direction corresponding to the intersection. This invention accurately distinguishes between effective parking and low-speed creep scenarios through multi-point collaborative analysis, and eliminates interference by combining directional consistency verification. Without relying on additional hardware and high-precision maps, it achieves low-cost, high-precision, low-interference, and wide-coverage automatic detection and hierarchical early warning of parking behavior at intersections.

The invention provides a valve control method and device, an engine and a vehicle, and belongs to the technical field of vehicles, when it is monitored that an air inlet bypass valve is stuck and lagged and cannot give an alarm immediately, the air inlet bypass valve is controlled to conduct self-cleaning, oil stains and foreign matter attached to the air inlet bypass valve are cleaned through self-cleaning, accidental foreign matter is avoided, and the service life of the air inlet bypass valve is prolonged. And fault misjudgment caused by oil gas coking and the like after pressure is avoided. If self-cleaning fails, a fault code is reported, an alarm prompt is not directly given according to the fault code, fault judgment contingency is avoided by determining the continuous period number of clamping stagnation of the air inlet bypass valve, the fault judgment accuracy is improved, and only when the continuous period number is larger than or equal to the preset safety frequency, the fault code is reported. According to the method, the fault of the air inlet bypass valve can be determined, alarm prompt is given, contingency of single judgment is avoided through continuous fault judgment, alarm prompt when the air inlet bypass valve is intact is avoided, the use cost is saved, and the use experience is improved.

The present invention relates to a method for making a PVC highlight film, and to the field of making PVC highlight films. In the process of making the PVC highlight film, it also specifically relates to a PVC highlight film making device, including an inverted shaped frame, and a detection table evenly arranged on the left and right is arranged between the left and right vertical sections of the inverted shaped frame. The detection table is composed of a left-right symmetrical fixing frame and a support plate symmetrically installed between the left and right fixing frames. The upper ends of the detection tables are all provided with fixing units, and the horizontal sections of the inverted shaped frame are provided with detection units. The present invention can not only perform impact resistance test on the PVC highlight film, but also perform tensile strength and bending strength test on it. One machine has multiple uses, which can not only save the time and energy consumed by replacing different detection equipment in traditional detection, improve work efficiency, but also reduce the production cost required for detection.

The application discloses a three-component geophone azimuth correction method and device based on multi-event fusion and probability weighting, equipment, medium and product, and relates to the technical field of geophysical exploration. The method is as follows: after at least two seismic signals recorded by a target three-component geophone and generated by at least two active source events are obtained, firstly, a preliminary azimuth calibration angle corresponding to each active source event is obtained, then a corresponding weight is assigned to the calibration angle according to a signal quality index calculated based on the corresponding seismic signal, then all calibration angles are fused according to all weights to obtain a final azimuth calibration angle, finally, the original horizontal component data is corrected by coordinate rotation using the final azimuth calibration angle to obtain corrected horizontal component data. Thus, by introducing a weighting mechanism based on signal quality, more accurate and reliable azimuth correction can be realized, the quality of seismic data and the reliability of subsequent interpretation are improved, and the application and promotion are facilitated.

The invention relates to the technical field of measuring devices, in particular to a sliding bearing flange assembly height detecting device which comprises a base and a bearing seat, a support is welded to the outer wall of the base, a bearing bush is fixed in the bearing seat, a flange end is arranged at the end of the bearing bush, and the sliding bearing flange assembly height detecting device further comprises a circumference measuring mechanism. According to the invention, through the arrangement of the circumference measuring mechanism, during measurement, a driving lever firstly lifts a measuring rod upwards, then a turntable drives the measuring rod of a digital dial indicator to rotate around the axis of a bearing seat, so that the switching of measuring point positions is realized, after the driving lever is separated from a starting lever, the starting lever drives the measuring rod to move downwards under the action of a spring I, and then the measuring rod is driven to move downwards; the measuring rod abuts against the flange end, the measuring result is read through the digital dial indicator and input into the control system for recording, and the operation is repeated so that the rotating disc rotates by one circle, the effect of multi-point measurement of the flange end is achieved, and the contingency of the measuring result is avoided.

The invention relates to a group of urine endogenous metabolic markers and application thereof, and the urine endogenous metabolic markers comprise two or more than two metabolic markers of 4-pyridoxine, phosphate PA (27: 6), 5, 6-dihydrothymine and glycodeoxycholic acid. The relative expression level of the marker provided by the invention is closely related to the capecitabine-related hand-foot syndrome, the marker is screened through machine learning, an intersection is taken, a refined and stable Logistic regression prediction model is constructed, the occurrence of the hand-foot syndrome is accurately predicted in advance, and the method has the advantages that the effect of accurately predicting the occurrence of the hand-foot syndrome is achieved while the effectiveness of chemotherapy is ensured. And the method has important significance in improving the chemotherapy safety.

The invention is suitable for the technical field of photovoltaic monitoring, and provides a photovoltaic module resonance monitoring and adjusting method and system, and the method comprises the following steps: collecting vibration data and wind power information of a photovoltaic module based on a sensor, the wind power information comprising wind speed, wind direction and time, and the vibration data comprising a timestamp; analyzing the vibration data through Fourier transform, determining a resonance time period, calling a wind speed and a wind direction corresponding to the resonance time period, analyzing the called wind speed and wind direction, determining wind speed and wind direction information of resonance generated by the photovoltaic module, and optimizing the wind speed and wind direction information based on the vibration data and the wind power information; and generating an adjustment suggestion scheme based on the optimized wind speed and wind direction information and the proportion duration of the resonance time period every day. The wind speed and wind direction information can be optimized, accidental factors in the wind speed and wind direction information are avoided, a targeted adjustment suggestion scheme is more accurately generated, and scientific basis and decision support are provided for operation and maintenance of a photovoltaic power station.

The utility model discloses a water conservancy reservoir flood pressure detection device, which comprises a detection carrier, a sensor group, a propelling assembly and a snorkeling propeller fixedly arranged at the front end of the detection carrier, two pipe bases arranged in parallel are fixedly arranged at the tail end of the detection carrier, and the sensor group used for water conservancy reservoir flood pressure detection is carried in the detection carrier. And side wheel frames are symmetrically arranged on the two sides of the detection carrier, the propelling assembly comprises propelling motors, advancing paddles and rotary drum wheels, and the two propelling motors are fixedly installed on the inner sides of the two pipe containing bases correspondingly. According to the utility model, the mobile carrier is used for carrying and moving the detection sensor, the propelling assembly is used for moving and driving, the snorkeling propeller is matched for controlling the snorkeling depth in water, the mobile carrier can move to a proper detection position to monitor the water level and pressure of flood in real time, and multi-point information acquisition and detection are realized; the sample data contingency caused by single sample data is avoided, and the flood evolution condition is mastered in time.

The present invention discloses a method and system for alleviating the hallucination of a large language model. The method comprises the following steps: obtaining multiple candidate words at the t-th time step in the large language model; obtaining the prediction probability of the high-level neural network in each layer of the large language model for each candidate word; normalizing the prediction probability of the high-level neural network in each layer of the large language model for each candidate word to form a cross-layer prediction distribution, and calculating the entropy of the cross-layer prediction distribution as the cross-layer entropy of each candidate word; based on the cross-layer entropy of each candidate word, correcting the initial prediction probability of each candidate word at the t-th time step to obtain the corrected probability of each candidate word at the t-th time step; outputting the candidate word with the highest value of the corrected probability as the output result corresponding to the t-th time step. By using the present invention, the output hallucination of the large language model can be effectively eliminated.

The invention discloses a connecting band and a production process thereof, the connecting band comprises a main body and a protective layer, one side of the main body is provided with a connecting structure, a plurality of fixing rods are arranged above the connecting structure, connecting buckles are arranged above the fixing rods, the other side of the main body is provided with the protective layer, the protective layer comprises adhesive glue and a protective band, and the protective band is arranged on the main body. One side of the adhesive is attached to one side of the protective belt, the other side of the adhesive is attached to the main body, the connecting belt extrudes the ball on the connecting structure of the other connecting belt into the gap of the ball of the connecting belt of the fixing column, the ball and the connecting belt are clamped together through the ball, and then the protective layer on the later supporting structure is removed. And the adhesion layer is exposed to adhere to the surface of another object, so that two objects which cannot be adhered originally can be connected.

The present invention relates to the field of data processing technology, and in particular to a learning effect evaluation system and method based on a Logistic regression model. The present invention determines whether a user has suspected behavior of catering to evaluation through the time difference between each question time of the user and a preset question time period of the current course, draws a histogram through the single viewing time of the user, determines the user's learning habit category according to the number of data corresponding to each data group in the histogram, determines the corresponding habit feature parameters according to different learning habit categories, and calculates the dominant feature representation value, and finally adds the dominant feature representation value and the preset general feature parameters to the Logistic regression model to calculate the probability of the user's qualified learning performance, thereby realizing the analysis of the user's suspicious behavior according to the user's learning process information, and extracting data with strong representation for suspicious users to add to the model calculation analysis to avoid misjudgment of the user's suspicious behavior, thereby improving the accuracy of model evaluation.

The utility model discloses a distribution box explosion-proof performance detection device, which comprises an installation box, the front surface of the installation box is hinged with a box door through a hinge, a support plate is fixedly connected between the two sides of the inner wall of the installation box, and a limiting mechanism for limiting the distribution box is arranged in the installation box. The limiting mechanism is located between the bottom of the inner wall of the installation box and the bottom of the supporting plate, and an adjusting mechanism used for adjusting the position of the limiting mechanism is arranged between the limiting mechanism and the installation box. According to the distribution box explosion-proof performance detection device, through cooperative arrangement of the adjusting mechanism and the driving mechanism, the adjusting mechanism drives the distribution box to move transversely, and the driving mechanism drives the impact hammer to move back and forth, so that full-coverage test of the impact hammer on the top surface of the distribution box is realized, and multi-point impact test on the distribution box is facilitated; the contingency of the impact test result is effectively avoided, and the test precision is further improved.

The invention discloses an intelligent lighting control system, which belongs to the technical field of lighting control, and solves the technical problem in the prior art that the reference data of the lighting effect designed by the lighting engineer is relatively one-sided, the lighting effect does not meet the expectations of the audience, and the viewing effect of the audience is affected; the invention obtains historical data, numbers the stage lights according to the atmosphere type and obtains the lighting labels, so as to facilitate the management of the stage lights; analyzes the vote rate of each stage light through simulation experiments, and includes the stage lights with a vote rate greater than a set vote threshold into the alternative experimental schemes; calculates the viewing rates of different stage lights through historical data analysis, and includes the stage lights with a viewing rate greater than the set viewing threshold into the alternative actual schemes, calculates the evaluation coefficient of each stage light by combining the two alternative schemes, and selects the maximum value of the evaluation coefficient as the operation scheme; avoids the one-sidedness and contingency of the reference data, and makes the analyzed data more accurate.

The wet filter material detection platform is designed for solving the possible problems of a filter bag of a wet filter dust collector in actual production and application, and experimental parameters required in actual production are better integrated by the platform. The system comprises a visual semi-closed experiment unit, a data processing and analyzing unit and an environment control unit which are all arranged on a movable experiment platform, the lower end of the experiment platform is connected with a structure protection box body, and wheels are arranged on the protection box body to facilitate movement of the experiment platform. The platform simulates various relatively complex actual working environments through mutual cooperation of the modules in the experiment unit, and performs comprehensive testing on filter material performance parameters such as dust removal efficiency, cleaning period and service life of the filter material.