34results about How to "Accurate and efficient completion" patented technology

The invention discloses a neural network-based acoustic glass defect detection method which comprises the following steps: acquiring a knocking signal of a glass sample through a pickup in an actual production environment; preprocessing the knocking signal; performing feature extraction on the pure knocking signal; setting initial parameters of a BP neutral network: taking an extracted feature as the input of the neutral network, setting the number of nodes on an input layer of the BP neutral network to be 7, setting the number of nodes on a hidden layer of the BP neutral network to be 15, setting the number of nodes on an output layer of the BP neutral network to be 2, and setting an output result to be (0, 1) which indicates that the glass sample has a defect, and to be (1, 0) which indicates that the glass sample is defect-free; training the BP neutral network, setting the learning rate of the BP neutral network to be 0.1, setting a target square error value to be 0.1, training the BP neutral network through a LeVenberg-Marquardt algorithm, and stopping training if an error of the neutral network is less than the set target square error value. A signal feature extracted through the method is high in distinction degree, and can complete a glass defect detection task more accurately and efficiently.

The invention discloses a method for controlling defrosting of an air conditioner of a heat pump type electric automobile on the basis of a humidity sensor technology. The method comprises the following steps: acquiring ambient humidity Hamb, ambient temperature Tamb and the surface temperature TEVAP of fins of an outdoor heat exchanger in a running process of the automobile through a temperature and humidity sensor, and storing acquired data once at set time intervals Tset; judging a current automobile running environment is in a common frost region or severe frost region in a theoretical frost region of an air heat source pump according to the ambient humidity Hamb and the ambient temperature Tamb, if the current automobile running environment is in the common frost region or severe frost region in the theoretical frost region of the air heat source pump, carrying out the next step, and if the current automobile running environment is not in the common frost region or severe frost region in the theoretical frost region of the air heat source pump, returning to the previous step; carrying out defrosting action for the time delta T; and judging whether the current surface temperature TEVAP of the fins of the current outdoor heat exchanger is greater than 0 DEG C or not, if the current surface temperature TEVAP of the fins of the current outdoor heat exchanger is greater than 0 DEG C, finishing defrosting control, and if the current surface temperature TEVAP of the fins of the current outdoor heat exchanger is not greater than 0 DEG C, returning to the previous step. By the method, defrosting action of a heat pump air conditioner system can be finished accurately and effectively, and energy consumption can be greatly reduced.

The invention discloses a dye mixing device capable of adding additives along with temperature. The device comprises a box body, a dye adding pipe with a downward opening is arranged on the top wall of the box body, a dye adding pipe plug is arranged at the top end of the dye adding pipe, a stirring mechanism positioned on the right side of the box body is arranged in the box body, a stirring cavity is formed at the bottom wall of the dye adding pipe, a heater is arranged at the bottom wall of the stirring cavity, temperature control mechanisms are symmetrically arranged on the left and rightsides of the stirring cavity, first chutes with openings facing the stirring cavity are symmetrically formed in the left and right end walls of the stirring cavity, an additive inlet pipe is arrangedon the top wall of each first chute, an additive inlet groove with an opening back to the stirring cavity is formed in the top wall of each additive inlet pipe, a second chute with an opening facing the stirring cavity is formed in the end wall, away from the stirring cavity, of each first chute, and additive mechanisms are arranged at the first chutes and the second chutes, the additives needingto be added at different temperatures can be added along with rising of the temperature of a heated dye, and stirring is conducted for a period of time after adding.

The invention provides an improved process for machining a crankshaft oblique oil hole. The improved process includes: firstly a stepwise centering chamfering drill is used for machining a tapered blind hole at the corresponding position on a cylindrical surface of a crankshaft, and then an ordinary gun drill performs machining of the crankshaft oblique oil hole through a lead hole provided by the tapered blind hole. The improved process enables a machining method which drills first and then chamfers in a traditional process to be changed, service life of the ordinary gun drill is one time of that of a special gun drill, using effects of the ordinary gun drill and the special gun drill before coping and after coping have no difference, and therefore the improved process for machining the crankshaft oblique oil hole can reduce cost consumption of cutters greatly, is safe and reliable, and has great popularization value.

The invention provides a CT image and MRI three-dimensional image registration method and device, and the method comprises the steps: obtaining an originally collected CT image and an originally collected MRI three-dimensional image, and carrying out the preprocessing operation, and obtaining a CT processing image and an MRI processing image; performing modal conversion on the CT processing image and the MRI processing image by using a modal conversion generator obtained by pre-training to obtain a CT modal conversion image and an MRI modal conversion image; predicting the CT processing image and the MRI processing image by using a single-mode registration network obtained by pre-training to obtain a single-mode registration confidence coefficient; and performing registration on the CT processing image, the MRI processing image, the CT mode conversion image, the MRI mode conversion image and the single-mode registration confidence coefficient by using a multi-mode registration network obtained by pre-training so as to realize automatic registration of the CT image and the MRI three-dimensional image.

The invention discloses an emergency handling method and application thereof, wherein the emergency handling method is advantageous in that a monitoring center and a remote terminal are arranged; theremote terminal is arranged in field; the monitoring center stores and displays emergency plans of dangerous accidents; the monitoring center has an emergency plan task decomposition area; when an accident occurs, the monitoring center performs matching according to an accident condition and selects the emergency plan, and generates an emergency plan task decomposition area; the emergency plan task decomposition area comprises decomposed tasks of the emergency plan; the monitoring center performs communication with the remote terminal for transmitting the decomposed tasks to the field, therebycommanding field personnel for performing emergency handling. The emergency handling method settles the following problems in prior art: long-time studying of emergency personnel for mastering totalcontents of a manual, time and labor wasting, and incapability of quickly, accurately and efficiently finishing emergency handling.

The invention belongs to the technical field of power systems. The storage device is applied to data of a high-voltage direct-current power transmission system. The system comprises a data collectionunit, a data processing unit and a data storage unit, wherein the data processing unit is in bidirectional connection with the data collection unit and the data storage unit, the data collection unitis used for acquiring operation data in a converter station of the high-voltage direct-current power transmission system and the direct-current power transmission line in real time, and the data processing unit is used for processing and judging the data acquired by the data collection unit. According to the invention, the operation state monitoring task of the high-voltage direct-current transmission converter station can be completed reliably, accurately and efficiently; meanwhile, the operation data can be calculated and processed, the operation state of the high-voltage direct-current power transmission system can be given, and the monitored operation data can be backed up and stored, so that monitoring personnel can know the real-time and historical operation states of the converter station conveniently, and the method has important significance for ensuring safe and stable operation of the high-voltage direct-current power transmission system.

The present application provides a question and answer model training method and device, wherein the question and answer model training method includes: constructing an initial text meaning group corresponding to a sample corpus, and generating a scene-oriented vocabulary corresponding to the sample corpus based on the initial text meaning group space; obtain training samples, and determine the sample phrases corresponding to the training samples; query the scene-oriented vocabulary space based on the sample phrases, determine the target text meaning group corresponding to the training samples according to the query results; use the target The text meaning group and the training samples train the initial question answering model until a target question answering model meeting the training stop condition is obtained.

The invention discloses a real-time road matching method based on driving records, including S1: collecting driving records and obtaining road data; S2: determining candidate positioning points: calculating the positioning area of ​​the collected driving records, and performing positioning according to the acquired road data , determine the candidate positioning points of the driving record; S3: Calculate the evaluation index of the candidate positioning point and the candidate matching road: calculate the evaluation index for the determined candidate positioning point, and then calculate the evaluation index of the candidate matching road; S4: Road matching: according to the candidate The evaluation index of the matching road obtains the best candidate matching road, and determines the best candidate matching road of the last valid driving record as the road matching result. A system includes an acquisition module for collecting driving records and road data, a control module for executing a control program, and a storage module loaded with programs, wherein the program executes a real-time road matching method based on driving records.

The invention provides an intelligent building block system and a guidance method thereof. The system comprises a light guide member which is connected with a bottom interface of the building block and is inserted into the building block; when the building blocks are correctly assembled and stacked on a base according to a preset structure, each light guide member forms an optical path; a light source device is provided on the base to provide a light source to the light path; a collecting device is arranged at the position where the building block emits the light source of the light guide member so as to collect the light source data; a processing device is connected with the collecting device to obtain the identification data after the light source data is processed; a prompting device isconnected with the processing device to perform the identification data to release the guidance information. Therefore, the structure conformity monitoring and guidance for the splicing structure inthe process of building block splicing can be realized.

The invention discloses a robot positioning and calibration method, device and electronic equipment. Wherein, the method includes: determining the reference point of the robot when calibrating the target joint; based on the reference point, tracking the calibration points at multiple angles when the robot rotates the target joint to obtain the calibration parameters; comparing the calibration parameters with the DH parameters of the robot, The comparison results are obtained, in which the DH parameters of the robot refer to the parameters between the end of the manipulator of the robot and the base coordinate system; based on the comparison results, the link parameters of the target joints are compensated to realize the positioning calibration of the robot. The invention solves the technical problem in the related art that it is necessary to search for a large number of points and waste time when the robot is positioned and calibrated.

The invention relates to an abnormal sound detection device for a notebook computer. The abnormal sound detection device comprises a case, a workbench horizontally arranged in the case, a carrying table assembly arranged on the workbench and used for placing the notebook computer, and a detection assembly used for detecting whether the notebook computer makes abnormal sound or not. The carrying table assembly comprises a bearing disc horizontally arranged above the workbench, supporting plates vertically arranged on the two sides of the bearing disc correspondingly and used for connecting the bearing disc and the workbench, a hollow-out part arranged in the middle of the bearing disc and smaller than the bottom area of the notebook computer, two positioning blocks arranged on the bearing disc and located on the two sides of the hollow-out part correspondingly, and a positioning strip arranged on the bearing disc and positioned behind the hollow-out part; the detection assembly comprises a mounting base and a laser vibration sensor which is vertically arranged on the mounting base, and the detection end of the laser vibration sensor penetrates through the hollow-out part and points to the bottom of the notebook computer; the abnormal sound of the bad notebook computer can be accurately distinguished through the laser vibration sensor, manpower is greatly saved, and the testing efficiency is improved.

The present disclosure provides a registration method and device for a CT image and an MRI three-dimensional image, wherein the method includes: acquiring the originally collected CT image and the MRI three-dimensional image and performing a preprocessing operation to obtain a CT processed image and an MRI processed image; The mode conversion generator obtained through training performs mode conversion on the CT processed image and the MRI processed image to obtain a CT mode converted image and an MRI mode converted image; CT processed images and MRI processed images are predicted to obtain single-modal registration confidence; using the pre-trained multi-modal registration network, the CT processed images, MRI processed images, CT modality conversion images, MRI modality Registration is performed on the state conversion image and the single-modal registration confidence level, so as to realize the automatic registration of the CT image and the MRI three-dimensional image.

The invention specifically discloses a surface abnormality detection method based on mixed supervision learning. The method includes the following steps: firstly acquiring normal samples with normal surfaces and abnormal samples with abnormal surfaces and preprocessing them; and constructing a neural network model including an anomaly localization network, a self-attention network and an anomaly discriminating network, and then preprocessing the data. Input the neural network model for training to obtain the abnormality detection neural network model; finally, input the image data to be tested into the abnormality detection neural network model, and then automatically determine whether the image to be tested is abnormal and locate the abnormal area. In the present invention, only a small number of abnormal samples need to be roughly labeled, and there is no need to provide a large number of finely labeled defect samples as training samples, which reduces the need for high-precision labeling of fully supervised learning, greatly reduces labeling costs, improves detection efficiency, and can be accurate and efficient. Complete industrial surface inspection tasks.

The invention relates to the field of data processing, in particular to a computer-based data picture labeling method, system and device. It mainly includes: obtaining multiple pending pictures that need to be marked with data; sequentially encoding multiple pending pictures to form a slicing task queue, performing slicing tasks on the pending pictures according to the coding order, and obtaining several small pictures to be classified ; The small pictures to be classified are sequentially encoded to form a classification task queue, and the classification tasks are performed on the small pictures to be classified according to the coding order, and the classification results are output; the classification results output by the classification tasks are reviewed; output Images for data annotation and corresponding annotation information. Through the data picture labeling method, system and device, the task of labeling a large number of data pictures can be completed accurately and efficiently.

The invention discloses a stereoscopic vision servo-based operating method and mechanism for a pivoting mechanism, and belongs to the technical field of industrial control equipment. The operating mechanism comprises a deep information acquisition unit, an optical acquisition unit, a calibration module and a mechanical arm. The operating method comprises the following steps that the mapping relation is constructed, transformation parameters are obtained, coarse positioning is performed on a target, accurate positioning is performed on the target, three-dimensional coordinates of the rotation center are calculated, and the pivoting mechanism is operated. By the aid of the operating method and mechanism, spatial positioning of the pivoting mechanism is realized, the automation degree of a system is increased, the operation process is simplified, and operation can be completed accurately and efficiently.