57results about How to "Balanced weight" patented technology

The invention relates to a detection vehicle of an electric car charging facility. The detection vehicle comprises a cab and an operation room, and is characterized in that the operation room comprises a charging facility detection platform formed by a power supply, a load and a management device, wherein the management device comprises a collection module, an auxiliary module and an analysis module. Compared with the prior art, the maneuverability of the detection vehicle of charging equipment is strong; the detection vehicle is applicable to driving on various roads and can be conveniently moved to a field to detect; and the configuration of the equipment on the vehicle is reasonable, the balance weight is uniform and the performance of the vehicle is guaranteed to be good.

The invention discloses a thyroid nodule Tirads grading automatic identification model construction method and device. The method comprises the following steps: S1, carrying out data preprocessing on a thyroid ultrasound image; s2, performing ellipse fitting conversion on the preprocessed thyroid ultrasound image to generate a segmentation mask as a training label; and S3, building a multi-task model based on a convolutional neural network, inputting the training label into the multi-task model, and performing training by adopting a preset loss function to obtain a trained thyroid nodule Tirads grading automatic identification model. According to the invention, in thyroid nodule TI-RADS grading automatic identification, the latest convolutional neural network framework is used for carrying out automatic identification tasks on thyroid ultrasound images in an end-to-end mode, TI-RADS grading is completed, detection work can be completed in the early stage of thyroid cancer, and therefore expensive work such as puncture does not need to be used; and a doctor is assisted to complete thyroid cancer screening.

The invention relates to the field of sleeping aid supplies, provides a quilt and a making method thereof and aims to solve the problems that a thinner quilt is too light and a quilt with larger massexceeds the warm-keeping demand. The quilt comprises a bottom layer and a surface layer which are provided with sealed edges, wherein a first middle layer and a second middle layer are arranged between the bottom layer and the surface layer, a filling layer is arranged between the first middle layer and the second middle layer and comprises granular filler, and density of the filler is larger thanthat of the bottom layer, the surface layer, the first middle layer and the second middle layer. The making method of the quilt comprises steps as follows: the bottom layer, the first middle layer, the filling layer, the second middle layer and the surface layer are laid in sequence, uniform filling units are formed in filling areas of the quilt by quilting, and the filler is fixed in the corresponding areas by the filling units. The quilt is thin and large in mass and can realize pacifying and sleeping aid effects under the condition of lower warm-keeping demand, overall counter weight and body stress are uniform, the body feels comfortable, and the making method is simple and convenient.

The invention relates to a network intrusion detection method based on data mining. In the prior art, problems that the classification accuracy is reduced due to the defect of sample weight updating,the classification speed is low due to a redundant weak classifier, and the calculation cost is high exist. According to the method, in the weak classifier training stage, the Adaboost algorithm of the improved weight updating method is adopted for weak classifier training, the sample weights are updated according to the weighted average accuracy of all samples in previous t times of training, infinite expansion of the noise sample weights is restrained, and weight updating of all the samples is more balanced. In a weak classifier combination stage, a new mode for measuring the similarity between weak classifiers is provided; selective integration is performed based on the similarity measurement mode and the hierarchical clustering algorithm, weak classifiers with the similarity exceedinga threshold value are classified into one class, the weak classifier with the highest classification accuracy in each class is selected to be combined into a strong classifier, and therefore redundantweak classifiers are removed, the classification speed is increased, and calculation expenditure is reduced.

The invention relates to a solid rocket ground test free boundary construction device. The solid rocket ground test free boundary construction device comprises a rocket product and a state monitoringinstrument, and is characterized in that the solid rocket ground test free boundary construction device further comprises a product bracket trolley, a gas source, a bag type gas pillow, an additionalgas chamber and a product dimension tool; the rocket product is carried by the product bracket trolley; the gas source provides air pressure for the bag type gas pillow; the product dimension tool isdesigned according to the external size of a rocket structure, the upper portion is used for being attached to the bearing portion of the outer surface for surface contact, and the bottom is connectedwith the bag type gas pillow by a flange; the upper portion of the additional gas chamber is connected with the bag type gas pillow by a flange, and the bottom is fixedly connected with the ground; and the state monitoring instrument is installed on the product dimension tool. The solid rocket ground test free boundary construction device is not limited by a test field. A bag-type gas pillow suspension system can be built quickly to simulate the freedom-freedom boundary of a product when a rocket body horizontal bracket vehicle is in a parking state.

The invention relates to the technical field of artificial intelligence, and discloses an information recognition method and device based on semantic enhancement, a medium and equipment, and the method comprises the steps: obtaining to-be-recognized text information and keywords; obtaining similar words of the keyword in a pre-trained word vector library; adopting a preset context coding model to code the similar words to obtain hidden vectors of the similar words, calculating word vector weights of the similar words according to the hidden vectors, and obtaining semantic enhancement information of the keywords through calculation; calculating the information weights of semantic enhancement information and hidden vectors relative to keyword semantic understanding through a preset gating model, calculating semantic feature parameters of keywords according to the information weights, calculating semantic prediction sequence tags corresponding to the semantic feature parameters, and performing semantic recognition on text information to be recognized. Therefore, semantic enhancement of the keywords is realized, the weights of the context environment and the semantic enhancement information are balanced, and the text information to be identified is accurately identified.

The invention relates to the technical field of image recognition and provides a data sampling method and a data sampling device for solving the problem of long tail of a training data set in the prior art, and aims to solve the problem of incomplete network feature learning caused by the fact that training samples are the same in each network training process from the beginning to the end of network training, and the method comprises the following steps: acquiring the current iteration round of a network; updating the sample weight of each type of samples based on the current iteration roundof the network and the sample number of each type of samples, with the increase of the current iteration round of the network, the weight of the tail type of samples being gradually increased; and according to the updated sample weight of each type of samples, sampling samples meeting a preset condition as target samples. Through the increase of network iteration turns, the sample weights in the training data set are increased, but the increase amplitude of the weights of the tail class samples is large, the probability that the tail class samples are selected as the training samples is increased, the problem of long tails can be effectively relieved, and the network feature learning effect is good.

The invention relates to a vineyard mechanization planting management device and belongs to the technical field of grape planting device structures. The device comprises a left post, a right post and a plurality of vineyard posts arranged between the left post and the right post; parts of the left post, the right post and the vineyard posts are embedded into the soil according to standards, and the left post and the right post are provided with an irrigation device, a grape trunk height positioning device and a grape vine fixing device used for fixing grape vines. The device is convenient to mount and easy to operate, the service life is long, labor, material and cost consumption can be reduced, and grape growth is benefited.

The invention provides a high-rise suspension detecting structure micro-vibration control device. The device comprises a steel rigid T-shaped table (1) which is a horizontal platform and is used for placing a radar inspected device and supporting a frame weight to maintain the balance thereof, a stopper system which comprises a class I stopper (2) and a class II stopper (4) and is used to preventimpact vibration and a sudden load from causing the steel rigid T-shaped table (1) to be separated from a supporting device to improve the stability. The class I stopper (2) is used to limit the relative position of the steel rigid T-shaped table (1) and an air spring in an air floating system (3). The class II stopper (4) is used to limit the relative position of a supporting rigid frame column (5) and an air spring in the air floating system (3). The device further comprises the air floating device (3) which is provided with 6 air springs to ensure stable horizontal face of the steel rigid T-shaped table (1), and a building structure basic floor (10) which is used to withstand the full load of a building and transfer the load to the foundation to realize anti-vibration.

The invention provides a single-output-shaft hub-type DC brushless motor. The single-output-shaft hub-type DC brushless motor includes a rotor, a stator, a connection shaft, bearings, a Hall element, a Hall element mounting frame, an oil seal, a dustproof ring and an O-shaped seal ring. The rotor includes a first shell, a second shell and a steel magnet. The steel magnet is connected with a rim fixedly. The first shell includes the rim, spokes and a bearing sleeve, wherein the rim and the spokes are formed integrally. The Hall element mounting frame is fixed arranged on the stator. The Hall element is fixedly mounted on the Hall mounting frame. The connection shaft is a hollow shaft. According to the invention, the rotor is convenient in assembly, high in precision and good in adaptability. The single-output-shaft hub-type DC brushless motor is good in use safety and reliability and quick and convenient in maintenance and repair.

The invention relates to a crane capable of adjusting a counterweight position. The crane capable of adjusting the counterweight position comprises a crane main body and a counterweight; guide rails are arranged on the crane main body; hanging lugs are arranged on the counterweight; the counterweight is guided and supported on the guide rails through the hanging lugs; rolling ball sets used for supporting the hanging lugs in the vertical direction are arranged between the guide rails and the hanging lugs; and either each of the guide rails or each of the hanging lugs is provided with rolling ball retaining holes allowing corresponding rolling balls in the rolling ball sets to be partially embedded, and the other one of each of the guide rails and each of the hanging lugs is provided with arolling fit face used for being in rolling fit with the rolling balls. The counterweight is supported on the guide rails in a guiding and moving mode through the hanging lugs, and the rolling balls are positioned through the rolling ball maintaining holes, so that the stable operation of the rolling balls is guaranteed; and the guide rails achieve guiding and supporting of the hanging lugs through the rolling ball set, so that the rolling ball abrasion is small, guiding is reliable, the stable operation of the counterweight is facilitated, and the operation safety of the crane is guaranteed.

The invention discloses counterweight combined water tanks for the pressure test of foundation tubular piles and a pressure test method, and belongs to the technical field of pile foundation detection of a building. The combined water tanks include the four bottom-layer water tanks which are arranged on the foundation tubular piles in a Chinese character 'qi' shape and the four upper-layer water tanks which are overlapped on the four bottom-layer water tanks in a Chinese character 'chuan' shape. The pressure test method includes the steps that (1) a pressure test and load bearing combined device is built; (2) a supporting strip-shaped stack for stable arrangement of the water tanks is formed; (3) the four bottom-layer water tanks arranged in the Chinese character 'qi' shape are arranged; (4) water is injected into the four bottom-layer water tanks to increase the weight; (5) the four upper-layer water tanks arranged in the Chinese character 'chuan' shape are arranged on the four bottom-layer water tanks; and (6) a pressure test instrument is started to detect the descending height of the foundation tubular piles under test. The counterweight combined water tanks for the pressure test of the foundation tubular piles and the pressure test method have the advantages that the conveying quantity of counterweight objects is reduced by eight times, the conveying efficiency is improved, the conveying expenses are reduced, the bottom-layer water tanks are prevented from inclining or toppling over in the direction away from the foundation tubular piles, and the purpose that the water tanks can be filled with water to serve as a counterweight and provide pressure for the foundation tubular piles is achieved.

The invention relates to an integration ground resistance measuring device. The integration ground resistance measuring device is structurally characterized in that a voltage line roll and a current line roll are respectively arranged at one end of a box body, and a voltage lead and a current lead are arranged on two independent rolls in a wound mode respectively; a ground resistance megger is arranged at the other end of the box body and firmly connected with one ends of the voltage lead and the current lead, and the other ends of the voltage lead and the current lead are led out of the box body; an interval is formed between the inner wall of the box body and the ground resistance megger to form an accessory cabin; crank inserting holes which correspond to the voltage line roll, the current line roll and the ground resistance megger are formed in the lateral wall of the box body. By means of the integration ground resistance measuring device, time and labor are saved, test data are accurate, attention is paid to actual effects, carrying is easily carried out, connection is conveniently carried out, winding and unwinding are free and flexible, and manpower and material resources are saved; meanwhile, the integration ground resistance measuring device has the advantages of shake prevention, dust prevention, water prevention, heat preservation and the like. The integration ground resistance measuring device is small in occupied area in site and easy to monitor, and guarantees are brought to safety production.

The invention discloses an SVM-based credit default prediction method under differential privacy in the technical field of credit default. The SVM-based credit default prediction method comprises thefollowing steps of S1, preprocessing data; S2, carrying out variable selection; and S3, designing a weighted SVM optimization model under differential privacy according to a differential privacy serial combinatorial property. An effective solution is provided for a differential privacy SVM learning problem under data imbalance, and the solution can specifically solve the data imbalance problem faced when differential privacy SVM learning is used to predict customer default. The method is suitable for application scenes with data imbalance including credit card default prediction, such as disaster prediction, medical diagnosis and other fields, and is also suitable for the technical scheme of the invention.

The invention relates to household articles, and discloses an acarus killing instrument. The acarus killing instrument comprises a shell, a handle arranged on the shell, a dust collecting and filtering device with a fan and a power supply module, the dust collecting and filtering device and the power supply module are arranged in the shell, and the acarus killing instrument is characterized in that the shell comprises a first side part and a second side part opposite to the first side part, the axis of the handle is located between the first side part and the second side part, the first side part and the second side part are oppositely arranged, a draught fan is arranged in the shell in the direction of the first side portion, and a power module is arranged in the shell in the direction ofthe second side portion. Compared with the prior art, the balance weights on the two sides of the acarus killing instrument are reasonably distributed, the acarus killing instrument is convenient totake in use, the use experience cannot be affected by the too heavy side, and the whole acarus killing instrument is more attached to a to-be-cleaned surface in use.

A method for influencing an acoustic noise of an air-conditioning system of a vehicle, wherein a psychoacoustic variable, in particular a sharpness of the sound, of the air-conditioning system is reduced by influencing a low-frequency and / or medium-frequency spectral component of a frequency spectrum of the sound of the air-conditioning system. A regulating or control device for carrying out the method.

The invention discloses an ultrasonic detecting blind-guide stick, a control system and a method, and solves the problems that the detection distance of an electronic blind-guide stick in the prior art is relatively close, the scanning is incomplete and the detection precision is not high. The ultrasonic detecting blind-guide stick can detect obstacles without gaps and accurately detect the distance between the obstacles and a human body, and is convenient in operation. The ultrasonic detecting blind-guide stick comprises a stick rod, a handle, a driving device, a connecting shaft and a detecting device, the connecting shaft is connected with the stick rod, the driving device provides power for one end of the connecting shaft, and the detecting device is arranged at the other end of the connecting shaft; the detecting device comprises an ultrasonic transmitter and an ultrasonic receiver, and a conical beam emitted by the ultrasonic transmitter achieves gapless scanning under the driving of the driving device.

The invention discloses an ecological risk early warning method based on a field environment monitoring station. The method comprises the following steps: monitoring ecological data acquired by a monitoring terminal of a field ecological observation station in a selected area in real time, calculating a total ecological system service index and an ecological risk index value, analyzing the ecological risk grade of the selected area according to the ecological risk index value, and carrying out early warning. According to the method, the ecosystem service is taken as an index, and an ecologicalrisk analysis model is imported to provide quantitative description for the uncertainty risk faced by the ecosystem; the weights of different ecosystem service types are better balanced, and the change of ecosystem output type services is accurately given; the problem that risk estimation is difficult to perform due to the fact that ecological negative effects are often ignored in a traditional ecosystem service calculation model is avoided.

The invention discloses a training method and device of a machine learning model and a readable storage medium. The method comprises the steps of obtaining a plurality of first training samples; performing first random enhancement processing on the first training sample to obtain a second training sample; after it is judged that the second training sample meets a preset enhancement condition based on a preset total enhancement number, second random enhancement processing is carried out on the second training sample to obtain a third training sample, and the preset total enhancement number is a divisor of the batch processing number; and training the machine learning model by using the training samples to obtain a trained machine learning model, the training samples including a second training sample or a third training sample, and the batch processing number being the number of the training samples input to the machine learning model each time. In this way, the convergence speed of the model can be increased.

The invention provides a balance weight balanced stirrer driving system and a stirrer system. The stirrer driving system comprises a magnetic driver, a control box, an operation module and a moving trolley, the moving trolley comprises a trolley frame, an operation part and a first roller set, the trolley frame is provided with a front end and a rear end in the insertion direction, and the controlbox and the magnetic driver are both arranged in the middle of the trolley frame; the operation part and the operation module are both arranged at the rear end of the frame; the first roller set is arranged in the middle of the frame and located below the control box, the control box is located between the magnetic driver and the operation module, and in the insertion direction, the distance fromthe magnetic driver to the control box is smaller than the distance from the operation module to the control box; the stirrer system includes the stirrer driving system. Counterweight balance of thestirrer driving system can be achieved, a worker applies acting force to the operation part, the rear end of the frame rotates downwards or upwards around the first roller set, and the labor-saving effect is achieved.

The invention discloses an all-terrain vehicle and a frame thereof. The frame comprises a main frame body and a third auxiliary frame body, a driving cabin, a power cabin and an oil tank cabin are formed on the main frame body, the power cabin is arranged behind the driving cabin, the oil tank cabin is formed in the position, on the inner side of the front portion of the driving cabin, of the main frame body, and the third auxiliary frame body and the main frame body are of a split structure. The third auxiliary frame body is installed on the upper portion of the oil tank cabin. Therefore, the oil tank is arranged in the cab of the all-terrain vehicle, and the oil tank cabin is arranged at the front part of the cab, so that the space in the cab can be reasonably utilized, and the oil tank and the power device can be separated, so that heat generated by working of the power device can be prevented from being transferred to the oil tank, and the use safety of the oil tank can be ensured. Besides, the oil tank cabin is arranged in the driving cabin, so that the structural complexity of the frame can be reduced, the arrangement difficulty of the oil tank can be reduced, and the balance weight of the whole all-terrain vehicle can be further uniform.