52results about How to "Resolve poor accuracy" patented technology

The invention discloses a path determination method, a path determination device, and a navigation instrument. The path determination method comprises the following steps: obtaining a starting point and an ending point of a to-be-determined path; obtaining data of obstacles in corresponding regions of the starting point and the ending point, wherein each obstacle is expressed by a polygon; the data of the obstacles are corresponding data of the polygons; determining general views of the corresponding regions according to the starting point, the ending point and the obstacle data; calculating the adjacent cost matrixes of the general views; and determining the shortest path from the starting point to the ending point according to the adjacent cost matrixes. Through the path determination method and the path determination device, the path is determined on the basis of the general views; the path determination process without a road network region is transformed into the path determination process with the road network region; and the path search accuracy is high in accuracy.

The invention discloses a method and device for acquiring streaming media play time. The method comprises the following steps: controlling a player to play streaming media in response to an operating command of a user, and acquiring the starting time information of streaming media play; monitoring the play state of the player while the player plays the streaming media; if the play state abnormity of the player is monitored, acquiring the time information of the play state abnormity of the player in order to obtain abnormity time information; at the end of the streaming media play, acquiring the end time information of the streaming media play; and computing the actual play time of the streaming media according to the end time information, the starting time information and the abnormity time information. Through the method and the device, the actual time computation of the streaming media play can be more accurate.

The invention provides a method and a device for determining a resident city, and an electronic device. The method comprises: according to characteristic information of a user, characteristic information of each candidate city, and behavior information of a user in each candidate city, through a city probability model, predicting to obtain fitting probability of the user and each candidate city; according to the fitting probability of the user and each candidate city, determining a first probability threshold value; according to the first probability threshold value, determining a resident city of the user from each candidate city. The method and the device solve problems in the prior art that time consumption is relatively long counting according to city classification, and that accuracyto determine a resident city is poor for a user who is active among a plurality of cities. The method and the device can determine the resident city of the user through the characteristic informationof the user and the cities, and through the behavior information of the user in the cities, and a calculation process is simplified, and accuracy is improved.

The invention relates to an action model based on deep learning and a training method thereof. The action model comprises a data completion module, a data encoding and decoding module and a state reasoning module. And after the action model is trained, a planning problem can be solved. According to the action model, the missing part of the original data can be complemented, and the problem of pooraccuracy caused by partial missing of the original data when the action model is trained is effectively solved. The action model learns the expression of the proposition form of the state in the implicit space through data training, and learns the reasoning ability in the implicit space, so that the problem can obtain a solution sequence through a search algorithm. The training process is unsupervised learning, the advantage of summarizing rules in a large amount of data of deep learning is efficiently utilized, and the cost of manual modeling is not consumed.

The invention discloses a method and a device for generating database query sentences. The method for generating the database query sentences includes acquiring a plurality of preset query objects; acquiring target leaf query objects from a plurality of leaf query objects; determining query columns; adding the query columns and the target leaf query objects into root query objects to obtain target query objects; generating the target database query sentences according to the target query objects. The preset query objects include the root query objects and the leaf query objects. The query columns are used for representing columns corresponding to to-be-queried contents. The method and the device have the advantages that the problem of poor accuracy when database query sentences are generated can be solved by the aid of the method and the device; the query columns and the target leaf query objects are added into the root query objects to obtain the target query objects, the target database query sentences can be generated according to the target query objects, and accordingly an effect of improving the accuracy of the generated database query sentences can be realized.

The invention discloses a lane line detection method and device, equipment and a storage medium. The method includes acquiring an initial perspective of a lane; performing perspective transformation on the initial perspective view to obtain a first aerial view; fitting the adaptive ROI matrix of the first aerial view on the basis of the target lane line equation, so that the limitation of manually setting the ROI matrix is solved, the adaptive ROI matrix better fits the lane line, the detection accuracy of the lane line is improved, and the computing power burden is reduced; based on the target inverse perspective transformation matrix, restoring a target ROI region corresponding to the adaptive ROI matrix to the initial perspective image to obtain a target perspective image; and based on the target ROI region of the target perspective image, extracting a first lane line of a lane, reducing noise outside the lane line, increasing adaptability of a large-curvature curve, and further improving lane line detection accuracy.

The invention provides a photosensitive device and a preparation method thereof, a display substrate and a method for detecting light intensity, belongs to the technical field of display, and can at least partially solve the problem of poor accuracy of light intensity detected by the existing photosensitive device. A photosensitive device provided by the invention comprises a substrate; a dual-gate transistor located on the substrate and comprising a first gate and a second gate; a photosensitive structure for sensitization comprising a photosensitive structure layer capable of receiving external light and generating corresponding electric charges according to different light intensities; and a first pole and a second pole respectively connected to the photosensitive structure layer, wherein the first pole is connected to the first gate.

The invention discloses a text key information extraction method and device, electronic equipment and a storage medium, and relates to the field of artificial intelligence. According to the specific implementation scheme, an original text is segmented according to preset segmentation units, and a unit sequence corresponding to the original text is generated; extracting recognition information of at least one target fragment based on the original text by adopting a fragment copying principle according to the unit sequence and a pre-trained information extraction model; and generating key information based on the recognition information of the at least one target segment. According to the technical scheme, the segment comprising the continuous words can serve as the target segment to be copied, in the process of extracting the key information, the frequency of copying can be effectively reduced, accumulated errors are reduced, and then the speed and accuracy of extracting the key information can be effectively improved.

The invention belongs to the technical field of diesel engines, and provides an engine cooling system. The system comprises a water pump, a water tank, an engine water jacket and an expansion water tank which are communicated through a pipeline, and further comprises a pressure sensor, a temperature sensor and an ECU which are assembled in the expansion water tank. The pressure sensor is used for collecting pressure signals in the expansion water tank, the temperature sensor is used for collecting temperature signals of cooling water of the engine cooling system, and the ECU is electrically connected with the pressure sensor and the temperature sensor. According to the engine cooling system, the pressure sensor is installed on the expansion water tank, whether the cooling liquid is boiled or not is comprehensively judged through the pressure, monitored by the pressure sensor, in the expansion water tank in combination with the cooling liquid temperature and the environment pressure, and then whether alarming or torque limiting is needed or not is judged, so that the problem that when the altitude is high or the concentration of the cooling liquid changes, whether the cooling liquid is boiled or not is judged only through the temperature, and judgment is inaccurate is solved.

The invention provides a measuring device and method for minimum technical output force of an extraction condensing-type steamer generator set and belongs to the field of steamer machine heat supply.The problem that as for an existingmeasuring method for the minimum technical output force of the extraction condensing-type steamer generator set, the accuracy is poor or the consumed time is long issolved. An automated mode is adopted to conduct online collection on data needed for generation of a thermoelectric coupling model of the extraction condensing-type steamer generator set, and the collected data are subjected to error correction. According to the data subjected to error correction, a plurality of relation curves of the extraction condensing-type steamer generator set are fit, andthe thermoelectric coupling model of the extraction condensing-type steamer generator set is generated according to the multiple relation curves of the extraction condensing-type steamer generator set. According to the thermoelectric coupling model of the extraction condensing-type steamer generator set and the known heat load of the extraction condensing-type steamer generator set, the minimum technical output force of the extraction condensing-type steamer generator set is measured. The measuring device and method are particularly suitable for measuring the minimum technical output force ofthe extraction condensing-type steamer generator set.

The invention provides a channel estimation smoothing method and device, and the method comprises the steps: carrying out the fast Fourier transform of a to-be-smoothed channel estimation result, and determining a time domain impulse response sequence; based on the time domain impulse response sequence, determining a plurality of candidate accumulated powers corresponding to selection windows with different selection lengths in the first range and noise powers corresponding to the second range; determining a channel length based on the plurality of candidate accumulated powers and the noise power; determining a signal-to-noise ratio by using the channel length and a target accumulated power selected from the plurality of candidate accumulated powers; and selecting a target smoothing filter coefficient corresponding to the signal-to-noise ratio and the channel length from the plurality of candidate smoothing filter coefficients, and filtering the to-be-smoothed channel estimation result by using the target smoothing filter coefficient. By adopting the channel estimation smoothing method and the channel estimation smoothing device, the problem of poor smoothing filtering accuracy when smoothing filtering is carried out on the channel estimation result is solved.

The invention discloses an intelligent control method for low-temperature operation of an air conditioner, and particularly relates to a method for intelligently controlling low-temperature operationof an air conditioner and the air conditioner. The method for intelligently controlling low-temperature operation of the air conditioner comprises the following steps that A, after a master control unit of the air conditioner receives a refrigerating command, an indoor temperature detection device is controlled to detect the indoor temperature, and an outdoor temperature detection device is controlled to detect the outdoor environment temperature; B, the master control unit of the air conditioner corrects the detected indoor temperature according to the indoor temperature, the outdoor environment temperature and the set temperature to obtain the indoor corrected temperature; and C, the master control unit of the air conditioner controls the operation frequency of a compressor according tothe indoor corrected temperature. Meanwhile, the air conditioner is provided. Through correction to the indoor temperature, the operation frequency of the compressor is controlled according to the corrected indoor temperature, under the condition that the air conditioner is in low-temperature refrigerating operation and it is guaranteed that the pipe temperature meets refrigerating, the frequencyof the compressor is in low-frequency operation, and the air conditioner can operate stably.

The invention discloses a parameter stereo matching method and system based on a binocular camera. The method comprises the following steps: acquiring a left image and a right image in a target detection area; based on the weight parameters, performing feature extraction on the left and right images through a multi-level network structure to obtain left and right feature maps with multiple feature dimensions; and performing matching cost aggregation on the left and right feature maps based on a supervised learning strategy to obtain an aggregated cost value. The technical problems that an existing parameter matching algorithm for automatic driving is poor in accuracy and low in calculation speed are solved.

The invention provides an osteotomy precision verification method. The method comprises the following steps of: step S10, selecting a plurality of measurement point positions on an original three-dimensional skeleton model, and measuring the distance between each measurement point position and a designed osteotomy plane corresponding to the measurement point position to obtain the designed osteotomy amount of each measurement point position; step S20, obtaining the actual osteotomy amount of the plurality of measurement point positions; and step S30, calculating the difference value d1 betweenthe designed osteotomy amount and the actual osteotomy amount corresponding to each measurement point position, if the absolute value of the difference value d1 corresponding to one of the measurement point positions is greater than 3mm, determining that the osteotomy precision is unqualified, and if the absolute values of all the difference values d1 corresponding to all the measurement point positions are smaller than or equal to 3mm, determining that the osteotomy precision is qualified. By application of the technical scheme provided by the invention, the problem of poor accuracy of the osteotomy precision verification method in related technologies can be effectively solved.

The embodiment of the invention provides an alternating current voltage sampling circuit. The circuit solves the problem of bad accuracy and reliability of a sampling circuit in the prior art, effectively inhibits temperature drift, realizes accurate detection and has a wide detection range. The circuit comprises a first voltage conversion module, a second voltage conversion module and a sampling module; two input ends of the first voltage conversion module are respectively connected with the zero wire and the fire wire of an alternating current power supply, two output ends of the first voltage conversion module are respectively connected with two input ends of the second voltage conversion module, and the output end of the second voltage conversion module is connected with the input end of the sampling module; the first voltage conversion module is used for converting the first alternating current voltage outputted by the alternating current power supply to a second alternating current voltage, the second voltage conversion module is used for converting the second alternating current voltage to a third alternating current voltage, and the sampling module is used for sampling the third alternating current voltage and obtaining the peak voltage and the peak frequency of the alternating current power supply according the sampling result through calculation. The alternating current voltage sampling circuit and method are applicable to the electronic technology field.

The invention relates to a seismic source nonlinear scanning signal design method based on forced vibration, and belongs to the technical field of geophysical exploration. According to the method, the actual earth surface condition is considered, a target frequency-energy value is determined according to the vibration characteristic information of a near-surface forced vibration system during scanning excitation of a vibroseis, so that the frequency change rate, the instantaneous frequency, the instantaneous phase and the instantaneous output data of nonlinear scanning signals are determined, and design of the nonlinear scanning signals of the vibroseis is achieved. The method can effectively adapt to different earth surfaces, enhance the construction adaptability of the vibroseis in a complex environment, meanwhile, can effectively improve the signal-to-noise ratio of vibroseis construction data; and the method is closer to practical production application.

The invention discloses a solar water heater intelligent analysis and management system based on big data. The system comprises an illumination detection module, a timing accumulation module, an illumination preprocessing screening module, a mapping heat analysis module, a solar energy storage data module, a heat collection pipe heat conversion processing module, a water temperature detection module, a heat absorption obstruction analysis module, an early warning control processing module and an irradiation shielding device. According to the invention, the illumination pretreatment screening module and the mapping heat analysis module are combined with the heat collection tube heat conversion processing module, and the theoretical water temperature variation in the solar water tank and theactual water temperature variation in the water tank, which are converted from the absorbed heat, can be analyzed; and the content of the incrustation in the heat collecting pipe is analyzed in combination with the incrustation content accounting model, the to-be-cleaned coefficient corresponding to the incrustation is judged according to the incrustation content so as to remind a user to clean the water tank and the incrustation in the heat collecting pipe in time, the irradiation shielding device is adopted for shielding the heat collecting pipe, and the service life of the solar water heater is prolonged.

The application belongs to the technical field of the farmland irrigation, and specifically relates to an irrigation water productivity scale conversion method and device. The method comprises the following steps: dividing mother scale region into more than two sub-scale regions, coupling two factors of spatial variability and return water reutilization to obtain an irrigation water productivity scale conversion relation of the sub-scale region and the mother-scale region, thereby accurately and conveniently acquire the mother scale region irrigation water productivity through the collected basic data. Through the method disclosed by the application, the spatial variability and the return water reutilization are coupled at the same time, and the problem that the final result is bad in accuracy when only one factor is used for performing irrigation water productivity scale conversion in the prior art can be solved; the basic data required by the invention is easy to obtain, the method is simple, and the problems that the used time is long and the computing difficulty is large when the complex hydrological model acquires the irrigation water productivity of the mother scale productivity can be solved in the premise of guaranteeing the result accuracy.

The invention discloses a building maintenance work order analysis method, device, storage medium and client. The method includes: acquiring a building maintenance work order set and a BIM model of a building involved in the building maintenance work order set; determining the The structural information of each building maintenance work order in the building maintenance work order set, and the original corpus of the building maintenance work order set; and determine the keyword library according to the BIM model; according to the structured information, the original The corpus and the keyword library perform semantic word segmentation processing and analysis processing to realize the automatic classification of the building maintenance work order collection. The solution of the present invention can solve the problem of poor accuracy of statistical analysis using structured information such as the location, time, and repairer of the maintenance work order, and achieve the effect of improving accuracy.

The invention discloses an inductance parameter determining method, an inductance parameter determining device, a storage medium and a synchronous reluctance motor. The method includes the steps: acquiring a group of torque set values of the synchronous reluctance motor; solving d-axis inductance and q-axis inductance linear equations of the synchronous reluctance motor based on the group of torque set values to obtain a group of d-axis and q-axis inductance changing along with the torque set values. The method can solve the problem that a changing curve of the d-axis and q-axis inductance ofthe synchronous reluctance motor changing along with torque is acquired by finite element simulation fitting, so that accuracy is improved.

The invention discloses a vehicle detection method and device. The method comprises the following steps: acquiring driving data of a vehicle; features of multiple dimensions in the driving data are extracted, and the features are used for representing the thermal runaway state of the power battery of the vehicle; the features of the multiple dimensions are processed through a state evaluation model, a detection result of the vehicle is obtained, the detection result is used for representing whether the vehicle has a thermal runaway phenomenon or not, and the state evaluation model is obtained through machine learning. The technical problem that battery thermal runaway accidents occur frequently due to the fact that early warning of the battery thermal runaway phenomenon is poor in accuracy, not timely enough and poor in real-time performance in the prior art is solved.