34results about How to "Reduce evaluation error" patented technology

The invention discloses a lithium battery health state estimation method based on a multi-factor evaluation model. The method comprises the following steps: firstly, establishing a first-order RC equivalent circuit model of a lithium ion battery; obtaining a specific function relationship between the open-circuit voltage and the state of charge, and constructing an SOC-OCV relationship table between the open-circuit voltage OCV and SOC values of different states of charge; then respectively establishing state observation equations taking ohm internal resistance, polarization internal resistance and polarization time constant as state variables; obtaining ohmic internal resistance, polarization internal resistance and polarization capacitance of the lithium battery in real time through iterative computation; and evaluating the health state of the brand-new battery of the same type by using the weight coefficients respectively corresponding to the ohmic internal resistance, the polarization internal resistance and the polarization capacitance obtained by training. According to the method, the problem that in the prior art, the SOH accuracy is estimated only through the ohmic internalresistance, and large defects exist is solved; by means of the Kalman filtering algorithm, the ohmic internal resistance, the polarization internal resistance and the polarization capacitance of thebattery are estimated online, and the estimation precision of the health state is improved.

The invention discloses a terminal side time-domain video quality evaluation method and apparatus. The method comprises: calculating the remarkable motion area proportion of each video frame, each remarkable motion area proportion referring to the proportion which a remarkable changing area between two neighboring video frames takes up in one video frame area; according to the remarkable motion area proportions, dividing all the video frames into absolute common frames and suspected distortion frames; performing freeze frame detection, scene conversion frame detection, jitter frame detection and ghosting artifact frame detection on the suspected distortion frames; and according to scene conversion frame detection results, performing scene segmentation on a video, calculating the scene information weight of each scene, calculating distortion coefficients according to freeze frame detection results, jitter frame detection results and ghosting artifact frame detection results, and determining terminal side time-domain video quality according to the remarkable motion area proportions, the scene information weights and the distortion coefficients.

The invention discloses a method and a system for evaluating the ageing life of a crosslinked polyethylene insulating cable for smart energy. The method comprises the following steps: accelerating a thermal ageing life estimation test, and introducing an electrical ageing factor in the thermal ageing life estimation test. According to the method and the system, electricity is introduced on the basis of a single factor thermal ageing estimation test, that is, the thermal ageing test is accelerated on a cable test specimen, and meanwhile a rated working voltage is applied to an insulating layer of the cable test specimen, so that the simulated ageing test conditions are in better accordance with an ageing process in a practical operating process of cable insulation; therefore, the evaluated life value which is more close to the actual life of the cable can be obtained; the method and the system have more guidance meaning on the transformation and replacement of a cable project engineering.

The invention provides an SOH-SOC combined online estimation method for an energy storage element. The method comprises the steps of: quantifying transient internal resistance, incremental voltage, standard deviation, sample entropy, peak point number and fundamental wave amplitude characteristic parameters, and establishing an SOH evaluation index system; establishing an evaluation set, establishing a relationship between the quantized indexes and the evaluation set through membership functions, and fuzzifying discrete data of the indexes by using the corresponding membership functions in a fuzzy logic mode; defuzzifying the fuzzy set by adopting a weighted average method to obtain a determined SOH evaluation value; and based on the SOH evaluation value, performing calculating to obtain apower battery SOH and SOC joint estimation result according to the determined current maximum available capacity of the power battery. According to the method, the problem of evaluation dispersion caused by inconsistent battery characteristics is effectively solved, so that SOH evaluation errors are reduced, the problems of mutual influence between parameters and change along with different ratesare solved, and the SOC estimation precision is improved.

The invention discloses a bone age evaluation method for realizing image segmentation and classification based on deep learning. The method comprises the following steps of: 1, processing a data set by using a digital image processing method to obtain sample data with higher quality; 2, manually marking a part of hand bone images, training an image segmentation network U-Net by using the part of images, then segmenting the data set by using the trained U-Net to obtain a background-removed data set, and making a training set, a verification set and a test set according to a certain proportion;3, training an improved image classification network VGG16 by using the processed data set; and 4, testing the trained model by using the test set and evaluating a result. Compared with an original bone age evaluation method, the improvement method provided by the invention effectively improves the accuracy of evaluating the hand bone image by the model, and has higher efficiency at the same time.

The invention belongs to the technical fields of medical treatment and cloud service, and relates to a cloud radiation therapy plan evaluation method, equipment and a storage medium. The method comprises the steps of: inputting a radiotherapy plan data to-be-evaluated into a patient database according to a standard format; using a machine learning method to generate a standard radiotherapy plan corresponding to the radiotherapy plan to-be-evaluated; automatically comparing the radiotherapy plan to-be-evaluated with the standard radiotherapy plan for scoring; and displaying the scoring resultsobtained by the machine learning method and outputting a result report. The evaluation method provided by the invention is beneficial to collecting expert knowledge and forming an expert knowledge base, which reduces artificial subjective factors, improves objectivity and reduces errors.

The invention discloses a multi-scale incremental entropy algorithm for evaluation of time sequence complexity. For an original electrophysiological time sequence, the original sequence is firstly coarse-grained to obtain coarse-grained sequences in corresponding scale; then the incremental entropy of each coarse-grained sequence is calculated to obtain the incremental entropy value of the sequence at the scale, so the fluctuation of the signal complexity with the scale is calculated, and the characteristic of the signal is extracted. With the increase of the time scale, the evaluation error of the entropy value is small, the consistency of the algorithm is good, and the calculation time is short. The shortcomings of increasing of the original multi-scale entropy with time scale, decreasing of the data length and the undefined entropy value are overcome. The method has better recognition performance on two signals than multi-scale sample entropy.

The invention discloses an elevator door running state monitoring method and system based on TOF visual detection. The method comprises the steps that S1, a TOF visual module is used for monitoring door opening and closing actions of an elevator door and continuous displacement data of the elevator door in the door opening and closing processes in real time; S2, a running curve of the elevator door in the door opening and closing process is obtained according to the monitored continuous displacement data; and S3, the running state of the elevator door is evaluated according to the running curve in the door opening and closing process. The elevator door running state monitoring method can achieve monitoring of the elevator door running state without connecting a vertical elevator car door motor system, and has the advantages of being simple in implementation method, low in cost and complexity, high in precision, safe, reliable and the like.

The invention relates to the field of communication, in particular to an integrated comprehensive drive test instrument. The instrument is used for solving the problem of simultaneous competing test of multiple networks. A method comprises the following steps that an antenna matching circuit module, and the receiving module is used for receiving a total downlink signal sent by the same radio frequency antenna; the antenna matching circuit module is used for distinguishing frequency bands of downlink signals, sending the downlink signals subjected to frequency band distinguishing to the terminal multi-module module, and the terminal multi-module module is connected with the antenna matching circuit module, comprises a first type module set, and is used for receiving one path of downlink signals on the corresponding frequency band and testing and comparing service related parameters. Thus, downlink signals are ensured to come from the same radio frequency antenna; the evaluation error caused by different radio frequency antennas is reduced; compared with the prior art, the accuracy of testing and comparing the related parameters is improved, the service related parameters of different frequency bands belonging to various networks can be tested and compared at the same time, the test comparison of various networks is realized, the test overhead is reduced, the operation process issimplified, and the high efficiency of the test comparison is improved.

The invention discloses an evaluation method for concentration effect of hydrogen iodide in iodine-containing hydriodic acid by using an electrolysis-electrodialysis device, which belongs to the technical field of electrochemistry. According to the evaluation method, an electrolysis-electrodialysis cell treated by the electrolysis-electrodialysis device is repeatedly washed by using washing liquid containing iodine ions and deionized water, then the difference of the masses of the washing liquid containing iodine ions and the deionized water before and after washing is calculated, and the mass of residual liquid in the electrolysis-electrodialysis cell is calculated through balance according to the law of conservation of mass; therefore, evaluation errors caused by the residual liquid in the electrolysis-electrodialysis cell are reduced, the electrolysis-electrodialysis cell does not need to be dismounted, the whole process is simple and quick, and the method has strong operability and a high practical value.

The invention provides a device and system for detecting human body striding and relates to the technical fields of human motion measurement, human body function and medical treatment. The device comprises a first longitudinal meter line adjusting rod, a second longitudinal meter line adjusting rod and an infrared detector, wherein the second longitudinal meter line adjusting rod is symmetrical with respect to the first longitudinal meter line adjusting rod; the infrared detector comprises an infrared emitting end and an infrared receiving end; the infrared emitting end is used for emitting infrared rays; the infrared receiving end is used for receiving the infrared rays emitted by the infrared emitting end; the infrared emitting end and the infrared receiving end are respectively arranged on the first longitudinal meter line adjusting rod or the second longitudinal meter line adjusting rod; the position of the infrared emitting end on the first longitudinal meter line adjusting rod or the second longitudinal meter line adjusting rod is as same as the position of the infrared receiving end on the first longitudinal meter line adjusting rod or the second longitudinal meter line adjusting rod. According to the invention, the technical problem in the prior art that the human body striding result cannot be accurately judged through visual inspection for the detection result is solved, the accuracy of the human body striding detection is increased and the evaluation error of human body striding is reduced.

The invention discloses a method for evaluating human health risks based on soil heavy metal forms. The method comprises the following steps: collecting and processing a sample, and determining analyzed heavy metal elements; extracting a heavy metal occurrence form, and determining the content of a heavy metal bio-available state; determining the exposure concentration of the heavy metal accordingto the bio-available state content of the heavy metal; and carrying out health risk evaluation analysis according to the exposure concentration of the heavy metal. According to the method for evaluating the human health risks based on the soil heavy metal forms, the bio-availability of the heavy metal is combined with the health risk evaluation model of the heavy metal so that the uncertainty ofthe health risk model can be reduced from the perspective of reasonably estimating the exposure dose of the heavy metal. The risk evaluation method is convenient to operate, more scientific in assessment, smaller in assessment error and convenient for people to assess and prevent health risks of heavy metals in soil to human bodies.

The invention relates to a street environment quality evaluation method based on VR panoramic visual perception. The invention relates to the technical field of urban design planning, and the method comprises the steps: obtaining main elements and level settings which affect a street environment through pre-investigation; through VR panoramic modeling rendering, a city street environment virtual simulation scheme scene is constructed by means of VR panoramic equipment and a related technology platform; designing a questionnaire, processing questionnaire data by adopting statistical software Stata SEV15.1, and establishing a discrete selection model; constructing a utility function through model fitting, studying and judging the weight relationship of each influence factor, and selecting a walking environment scene with the maximum utility for the pedestrian according to the random utility theory of a discrete selection model; sequentially arranging absolute values of the weight relations from large to small to obtain preference analysis of different types of crowds on street environment element selection behaviors; and ArcGIS software is used for analysis to obtain an overall evaluation map of the actual street environment in the city.

The invention discloses an acoustic index-based dysarthria grading evaluation method. The method comprises the following steps: acquiring voice data of a patient with dysarthria after cerebral apoplexy; extracting objective acoustic feature intelligibility of the voice data, and collecting, extracting and analyzing acoustic indexes of the subject, the objective acoustic feature intelligibility being a pathological voice analysis index which is high in recognition degree and good in sensitivity at home and abroad; determining critical values of slight abnormality, moderate abnormality and severe abnormality of the voice by using the obtained acoustic feature intelligibility based on an automatic clustering algorithm K-means of artificial intelligence; using the intelligibility critical value for clinically evaluating the severity level of the dysarthria of the subject. According to the method, evaluation errors caused by subjective factors such as the technical level of technicians, subjective judgment and regional difference are reduced; meanwhile, operation is easy, and consumed time is short.

The invention discloses an ultrasonic method and system for evaluating the temperature change based on nonlinear thermal expansion, belongs to the field of ultrasonic detection, and aims to solve theproblem that the temperature rise of a target area cannot be effectively monitored in various methods for treating diseases based on in-vivo tissue heating at present. The invention discloses a methodfor evaluating the temperature change of biological tissues by utilizing B-ultrasonic original echo signals, and the method comprises the following steps of locally heating the biological tissues byusing the methods, such as focused ultrasound, radio frequency, microwaves, etc., imaging the target area by using B-type ultrasound, collecting the original echo signal of the target area, selectinga target frame by using a dynamic frame selection algorithm to suppress the living body noise based on a B-type ultrasound time sequence image, and calculating a time delay image when the ultrasound passes through the tissues; and obtaining a temperature change image according to a nonlinear model of the thermal expansion and the sound velocity. According to the method, the error does not exceed 3DEG C within the temperature rise range of 30 DEG C, the application of the B-ultrasonic temperature rise monitoring technology in thermal therapy is promoted, and the safety and effectiveness of thermal therapy can be remarkably improved.

The invention discloses an ultrasonic method and system for evaluating temperature changes based on nonlinear thermal expansion, which belongs to the field of ultrasonic detection, and aims at the problem that the temperature rise of the target area cannot be effectively monitored in the current methods of treating diseases based on heating tissues in the body , established a method for evaluating the temperature change of biological tissue by using the original echo signal of B-ultrasound. This method uses focused ultrasound, radio frequency, microwave and other methods to locally heat biological tissues, uses B-mode ultrasound to image the target area and collects its original echo signals, and uses a dynamic frame selection algorithm to select target frame pairs based on B-mode ultrasound time-series images. Live noise is suppressed, and the time-delay image is calculated when the ultrasound passes through the tissue; then the temperature change image is obtained according to the nonlinear model of thermal expansion and sound velocity; this method has an error of no more than 3°C within a temperature rise of 30°C, which will promote The application of B-ultrasound temperature rise monitoring technology in hyperthermia can significantly improve the safety and effectiveness of hyperthermia.

The invention belongs to the field of steady-state analysis and performance evaluation of AC / DC converters, and provides a modular multilevel converter over-modulation risk evaluation method and system. The method comprises the following steps: defining a dynamic modulation ratio; the method comprises the following steps: constructing a modulation signal of the modular multilevel converter according to a topological structure of the modular multilevel converter, obtaining an amplitude of a corresponding harmonic component according to the modulation signal, and obtaining a dynamic modulation ratio value of the modular multilevel converter based on the amplitude of the corresponding harmonic component; and evaluating whether the modular multilevel converter has an over-modulation risk based on the dynamic modulation ratio value.

The invention discloses an evaluation method and system for an intermediate scheduling scheme of a job shop, and belongs to the technical field related to shop scheduling. According to the invention, only the values of 'head length + tail length + process processing time ' of two specific processes in the neighborhood solution are calculated, so that the evaluation efficiency is improved; The method comprises the following steps: taking a larger value in parameter summation results of specific procedures as a preliminary evaluation result, and dividing the preliminary evaluation result into two classes: comparing the preliminary evaluation result with completion time in an original solution, taking the preliminary evaluation result greater than or equal to the completion time of the original solution as a first class, and taking the preliminary evaluation result as a final evaluation result of the neighborhood solution; and taking the initial evaluation result smaller than the original solution completion time as a second class, and reducing the influence of other procedures without changing the processing sequence on the critical path on the evaluation result by introducing an error correction parameter, so that the initial evaluation result is further corrected. Through the operation, the effects of improving the evaluation accuracy and reducing the evaluation error are finally achieved.

The invention discloses an ultrasonic method for measuring temperature change of biological tissue based on thermal expansion and gating algorithm. Aiming at the problem that temperature increment ofa target area cannot be effectively monitored in methods for treating diseases based on various current body tissue heating modes, a method for evaluating the temperature change of the biological tissue by using B-mode ultrasound RF signals is established. According to the method, methods such as focused ultrasound, radio frequency and microwave are used for locally heating the biological tissue,B-mode ultrasound is used for imaging the target area and collecting the RF signals of the target area, a target frame is selected based on B-mode ultrasound timing sequence images, time delay imagesthat ultrasound passes through the tissue are calculated, and accordingly temperature change images are obtained; according to images outside a heating area, the coefficient of a self-adaption filteris calculated, and the obtained temperature change images are subjected to noise suppression. According to the ultrasonic method, the error in the temperature increment range of 18 DEG C is smaller than or equal to 2 DEG C, the temperature increment monitoring technology of B-mode ultrasound is promoted to be applied to thermal therapy, and the safety and effectiveness of the thermal therapy can be remarkably improved.

The invention relates to a method for diagnosing post-stroke dysarthria tone errors based on a neural network and a diagnosis device therefor, and belongs to a novel dysarthria objective evaluation technology based on acoustic indexes. According to the method, monosyllable word tone data of post-stroke dysarthria patients taking mandarin as a mother language and normal adults of the same age groupare collected and divided into a training group and a test group; the training group is trained by an artificial neural network to obtain a diagnosis model; the correct rate of automatic tone identification of the test group is tested; and parameters are adjusted until the diagnosis accuracy is greater than 90%. According to the method, the tone is evaluated through artificial intelligence, the objective and efficient purposes can be achieved, the evaluation difference by different mechanisms and different therapists is reduced, and the method has certain guiding significance for implementation of rehabilitation training.

The invention discloses a health risk evaluation method for applying biochar-based fertilizer to soil, and relates to the field of soil pollution risk assessment, and the method comprises the following steps: collecting and processing a sample, and determining the types of analyzed heavy metal elements and polycyclic aromatic hydrocarbons; extracting a heavy metal form, and determining the content of the heavy metal in a bioavailable state; measuring the concentration of the polycyclic aromatic hydrocarbon, and determining the equivalent toxicity concentration of the polycyclic aromatic hydrocarbon relative to the benzo [a] pyrene; and constructing a risk evaluation model. According to the method, the risk assessment of the combined pollution is more accurate through the probabilistic method, and the safe application amount and high-influence parameters of the charcoal-based fertilizer can be determined from the risk assessment. The risk assessment method is convenient to operate, more scientific in assessment and smaller in assessment error, people can assess and prevent the health risk of applying the biochar-based fertilizer to pollutants in soil conveniently, a proper fertilization strategy is determined, and the risk influence is reduced.

The invention relates to a dancing data processing method and device, dancing equipment, computer equipment and a computer storage medium. The method comprises the following steps: playing music content of a dance game to a user through a display device, and acquiring beat points of the music content; acquiring footstep action data generated by the dance game of the user through action capture equipment at each beat point position of the music content; calculating overlapping area data of footsteps according to the footstep action data, and calculating a score value corresponding to each footstep action data according to the overlapping area data; and determining evaluation data of the user in the dance game according to the score value of each footstep action data. According to the above technical scheme, the dancing level of a player can be correctly judged, an evaluation error is reduced, and wide propagation and sharing of dance games are promoted.

The embodiment of the invention provides a signal-to-noise ratio evaluation method and device based on an unmanned aerial vehicle remote control signal. The method comprises the steps: obtaining a to-be-evaluated unmanned aerial vehicle remote control signal, and taking the to-be-evaluated unmanned aerial vehicle remote control signal as a to-be-evaluated signal; inputting the to-be-evaluated signal into a pre-established neural network model to obtain a predicted signal-to-noise ratio of the to-be-evaluated signal; and evaluating the signal-to-noise ratio by utilizing double convolutional neural networks and a double-layer long-short-term memory network which are connected in sequence. The accuracy of predicting the signal-to-noise ratio is improved by continuously adjusting the parameters of the neural network model, so that the accuracy of evaluating the signal-to-noise ratio by the neural network model is improved, and the evaluation error is reduced.

The invention belongs to the field of aviation turbofan engine main combustion chamber design, particularly relates to a full-ring combustion chamber test piece flow resistance characteristic data processing method, and provides a combustion chamber flow resistance characteristic data processing method, namely a head estimation correction method on the basis of a direct processing method and a turbine bleed air correction method. That is to say, in combustion chamber flow resistance characteristic data processing, middle ring cavity airflow pressure loss is considered. When the total pressure loss of the diffuser is calculated, the loss part of the middle ring cavity is added, the combustion chamber flow resistance characteristic closer to the real situation can be obtained, and the evaluation error is reduced.