132 results about "Interval estimation" patented technology

The invention discloses a device and a method for fusing sensor information. The sensor-information fusion method comprises the steps of adopting a self-adaptive weighted-data fusion algorithm to allocate corresponding weight numbers for induction information from a plurality of sensors, adopting a Kalman filtering algorithm to perform optimal estimation to the induction information of the corresponding weight numbers so as to acquire the locally fused induction information, adopting a D-S theory estimation algorithm to perform interval estimation to the uncertain induction information in the locally fused induction information, adopting a multi-Bayesian estimation algorithm to perform individual association probability distribution for the induction information after interval estimation so as to synthesize a joint posteriori probability distribution function, and utilizing the joint distribution function to output a final fusion value of the induction information. The invention has the advantages of improving the reliability, confidence and efficiency of sensor information fusion.

The invention discloses an energy storage capacity optimization method of a wind power farm based on min (minimum) component fluctuation of wind power. The method comprises the following steps that an energy storage capacity optimization configuration model of the wind power farm is established; the maximum charge-discharge power of an energy storage system is determined; effect evaluation of energy storage smooth wind power farm power is obtained according to an occupation ratio of a min component of output power of the wind power farm; an approximation degree between a reference output curve and a wind power farm output curve after the action of the energy storage system is analyzed according to correlation coefficients between a reference output of the wind power farm and an output after the action of the energy storage system; effect evaluation of smooth output power of the wind power farm under different time stages; and a smooth output of the wind power is achieved finally. On the basis of grasping the characteristic of the min component fluctuation of the wind power, the model determines the capacity configuration and the maximum charge-discharge power of the energy storage system with an interval estimation theory of probability statistics. The smooth output of the wind power is improved by energy storage equipment with smaller capacity, and adverse effects of random fluctuation of the wind power on an electric system are reduced.

The invention discloses a multi-kind and small-quantity part production process capability index determining method based on features. The multi-kind and small-quantity part production process capability index determining method based on the features is characterized in that firstly, processing features are taken as objects, processing feature quality characteristic values which are different in size and identical in processing technology and tolerance grade are defined as the same processing feature sample, wherein the ratio of the standard deviation and the tolerance of the processing feature quality characteristic values is a constant; secondly, sample individuals are normalized, the mean values of the sample individuals are the same, namely the sample individuals conform to the same type of distribution; finally, a process capability index is calculated based on processing feature samples. According to the multi-kind and small-quantity part production process capability index determining method based on the features, the problem of sample size of multi-kind and small-quantity production is fundamentally solved, and conversion from interval estimation to point estimation of the process capability index oriented to multi-kind and small-quantity production is achieved. The obtained process capability index can be used for calculating the disqualification rate of the processing feature quality characteristic values, the processibility analysis is performed on feature design structures according to the disqualification rate, the proper design structures are selected, 100 percent detection is avoided, and the detection efficiency is improved.

A method and apparatus for scaling video images, which applies more particularly to techniques for converting video sequence formats and deinterlacing. The invention comprises a parameterized interpolation (101) which computes interpolations at each new point of the output. An insensitive interval estimation (102) computes an interval where the probability distribution of the output image value is estimated to have its maximum, and an insensitive loss calculation (103) derives for each interpolation value a negative log likelihood measurement from an interval-insensitive probability distribution. After a spatial regularization (104) the invention includes a minimum loss selection (105) which selects the minimum loss interpolation at each new point and provides it as an output image value. The invention also includes a total variation increment module and a directional matching error module that can penalize the insensitive loss, for the minimum loss selection.

The invention discloses a variable weighting combination landscape water quality early warning method which comprises on-line water quality monitoring, water quality change predicting and water quality early warning, wherein the on-line water quality monitoring comprises the following steps: gathering water samples of all monitoring points of landscape water, and transmitting the monitoring data to the server of a data center for storage and analysis; the water quality change predicting comprises the following steps: performing normalization processing on water quality samples of the data center, establishing a neural network prediction model, a support vector machine prediction model and a variable weighting combination prediction model respectively, and performing interval estimation on the variable weighting combination prediction model; water quality early warning comprises the following steps: performing water quality early warning by utilizing the interval estimation value of the variable weighting combination prediction model, and performing warning report simultaneously. According to the invention, the function of landscape water quality on-line early warning is realized; the variable weighting combination mode of a neural network and a support vector machine is introduced to a landscape water quality early warning system for the first time, which is an innovation in the field of landscape water quality. The variable weighting combination landscape water quality early warning method has an excellent practical value for the water quality maintenance and operation management of a landscape water body.

The invention provides a method for detecting a moving target of a video image. The method comprises the following steps of: A. for previous N frames of images, constructing an initial background image by a codebook model; B. counting the mean value and the variance of the initial background image for an initialized hybrid gaussian background model; C. carrying out interval estimation on a background pixel and setting a confidence interval of the background pixel; and D. updating parameters of the hybrid gaussian background model so as to generate the background image when the pixel meets the confidence interval to obtain the moving target by subtraction of the current frame of image and the background image. According to the method, pixel values are matched so as to get rid of mutational pixel values in initialization of parameters of the model, a foreground image and the background image can be accurately distinguished due to the arrangement of the confidence interval, and relatively mutiple accurate moving target images can be obtained.

The invention discloses a fault detection method for Kalman filtering sensor information fusion. The method includes the steps of: 1. according to a Kalman filtering theory, establishing a state equation and an observation equation of a linear dynamic system; 2. according to the observation equation obtained in step 1, making use of a least square process to acquire state estimation, a corresponding mean square error matrix and an innovation sequence; 3. utilizing the known innovation sequence to acquire different channel-normalized new innovation sequences, and composing an innovation matrix of an m channel parallel sensor; 4. acquiring the spectral norms and spectral norm mean value of the innovation matrix according to the innovation matrix obtained in step 3; and 5. carrying out fault detection on the Kalman filtering sensor information fusion. The method provided in the invention adopts mathematical statistics and interval estimation, simplifies complex calculation, and substantially improves the fault detection speed.

The invention relates to an incomplete data fuzzy clustering method for information feedback RBF network estimations, which comprises the following steps: 1) presenting an information feedback RBF network model; 2) presenting an incomplete data fuzzy clustering method (IFRBF-FCM) of information feedback RBF value estimations; 3) selecting the corresponding training sample set for the incomplete data sample by using the nearest neighbor rule, and training the IFRBF network for each missing attribute by using the nearest neighbor training sample set, thereby realizing the estimation prediction of the missing attribute in the incomplete data sample and obtaining the complete data set after the estimation recovery of the IFRBF network; 4) determining the estimation interval of the attribute ofthe incomplete data to propose an incomplete data fuzzy clustering method (IFRBF-IFCM) of IFRBF interval estimations to obtain fuzzy clustering results. The invention adopts the IFRBF network to estimate the incomplete data set and recovers the intact data set. Compared with the comparison method, the clustering result of the intact data set is more accurate than that of numerical type estimations, and the robustness is better.

The invention discloses a high-proportion censored data reliability interval estimation method based on proportion grouping re-sampling. The method includes the steps that firstly, life test data are grouped into complete data and censored data; secondly, repeated sampling is conducted according to a specific proportion; thirdly, distribution parameters and reliability point estimations are obtained; fourthly, repeated re-sampling is conducted by the adoption of a Bootstrap method, and corresponding distribution parameters and reliability point estimations are calculated; fifthly, the obtained reliability point estimations are sequenced, and then the reliability interval estimation is obtained. The method avoids appearance of zero-failure data in the process of repeated re-sampling, and normal operation of the method is guaranteed; the method needs less theoretical knowledge, is easy and convenient to calculate, is easy to implement, facilitates use of engineering technicians and has good application value.

A biosignal measuring device has a first information acquisition module, a second information acquisition module, a first calculation module, a second calculation module, a third calculation module, a heartbeat interval estimation module, a determination module, and an output module. The first information acquisition module acquires the pulse wave signal. The second information acquisition module acquires an electrocardiogram signal. The first calculation module calculates a pulse wave velocity based. The second calculation module determines a heartbeat interval. The third calculation module calculates a relationship between pulse wave velocity and the heartbeat interval. The heartbeat interval estimation module estimates an estimated heartbeat interval based on pulse wave velocity and calculated relationship. The determination module determines whether the position of the first and second information acquisition modules has changed based on the heartbeat interval and the estimated heartbeat interval. The output module outputs information.

The invention discloses a fault detection method based on robust interval estimation, and the method comprises the following steps: designing a model uncertainty system considering interference factors, and designing an interval observer for the model uncertainty system; establishing an augmented error system in combination with the model uncertainty system and an interval observer; according to the parameters of the model uncertainty system, optimizing the parameters of the interval observer in combination with the l-infinity criterion, and proposing sufficient conditions for the existence ofthe interval observer; converting a bilinear problem in an interval observer design process into two linear problems by applying an iterative algorithm. The application of the interval observer in the field of fault detection is considered, and the output of the interval observer and the real output of the model uncertainty system are utilized to construct a residual error to detect the fault. The method not only solves the problem of estimation of the system state with uncertain parameters in the system, but also has practical significance and practical value when applied to fault detection.

The invention discloses a video rapid defogging method and relates to the technical field of video processing. The method comprises a step of extracting a background image for solving an atmospheric optical value and a medium transmissibility; a step of solving an atmospheric optical value A from the background image by using a dark channel prior principle and an interval estimation way; a step of acquiring the initial estimate t1 (x) and the rough estimate t2 (x) of the medium transmissibility form the extracted background image by using the dark channel prior principle; a step of performing pixel-level fusion, rapid united bilateral filtering, and value domain regulation on the initial estimate t1 (x) and the rough estimate t2 (x) of the medium transmissibility to obtain the final estimate t(x) of the medium transmissibility, and a step of defogging a video. The method has fast processing speed and takes account of defogging effect and speed while guaranteeing a good video processing effect.

The present invention discloses a distribution network planning method based on the industry load prediction technology. The method comprises a step of preparing data which comprises load information, substation alternate point position information, the construction cost of a unit line and a substation, unit load loss cost, unit line damage cost, the practical coefficient needed by the industry load prediction technology, and the database of a phase coefficient, a step of using the industry load prediction technology to predict a load point load based on a load point installation capacity, a step of completing a substation address selection working through a genetic algorithm population operation, and obtaining a selection point set, a step of completing the distribution of a load point based on the Voronoi diagram theory, and using the Monte-Carlo-based interval estimation to calculate line current and distribution substation capacitance, and a step of taking the total cost in a planning period as a minimum target function, calculating the target function, and outputting an optimal result. According to the method, the distribution network planning method is more close to an actual condition, the cost of actual distribution network construction operation is reflected, and a decision support is provided for the investment of the distribution network construction in a new situation.

The invention discloses a hospitalization expense estimation and judgment method for single disease. The method comprises: cleaning abnormal data, screening out qualified samples, and establishing a sample database; neatening the samples and extracting influence factors related with hospitalization expense estimation and judgment of the single disease; performing expense interval estimation according to the number of samples, wherein the number of samples has to be greater than 10; if the number of samples is greater than 50, at first, trying to delete a maximum value or a minimum value through algorithm iteration, wherein the rest samples are reserved verification samples; then verifying a sample distribution pattern, limiting the residual rate of the samples, and normalizing the samples to estimate an expense interval; and if the number of samples is smaller than or equal to 50 and greater than 10, performing t distribution estimation and directly calculating expense estimation of a confidence interval.

The invention relates to a method for establishing a power grid line loss rate prediction model. Firstly, a p-element linear regression model is established, and then the model is sequentially subjected to the significance test of the linear regression and the significance test of the regression coefficient. Finally, estimate the prediction confidence interval. The invention has the advantages of simple structure, fast prediction speed and good extrapolation characteristics.

The invention discloses a photovoltaic interval prediction method and system based on self-coding and an extreme learning machine. A dual-output extreme learning machine model is used for constructinga photovoltaic interval prediction model, dual outputs are the upper limit and the lower limit of a photovoltaic interval respectively, and due to lack of actual values of the photovoltaic interval,a traditional extreme learning machine training algorithm fails accordingly. According to the method, a heuristic algorithm is adopted to optimize a training extreme learning machine, a correspondinginitialization algorithm is provided for improving the calculation result and efficiency of the heuristic algorithm, and the method comprises: carrying out prediction interval initialization based onlinear regression interval estimation; carrying out inputting weight matrix initialization of an extreme learning machine based on self-coding; and carrying out optimization training of a particle swarm-based double-output extreme learning machine. The photovoltaic prediction interval obtained through the model replaces a traditional photovoltaic point prediction value, and more sufficient information can be provided for day-ahead robust scheduling and intra-day economic scheduling of a power system.

The invention provides a method for building an in-orbit X-ray pulsar timing model. According to the method for building the in-orbit X-ray pulsar timing model, the problem of data introduced by weak X-ray signal intensity and strong observation platform dynamic condition is effectively solved by organically combining various calculation methods of X-ray pulsar photon arrival time preprocessing, high-precision pulsar autorotation frequency calculation, timing model initial value interval estimation, observation pulse profile reconstruction, high-precision pulse arrival time estimation and robust timing model estimation, ephemeris information such as high-precision pulsar position and time required by X-ray pulsar navigation can be provided, and accurate navigation is realized. The method is suitable for autonomous operation and control of orbiting satellites or constellation of the earth or other planets.

The invention provides an improved travel time reliability analysis method. On the basis of the traditional travel time reliability index estimation, a Bonferroni-modified K-S detection method is usedto check whether seasonal and monthly differences exist in the travel time; lognormal distribution is used to fit the distribution of the travel time, and the maximum expectation algorithm is used for parameter calibration; a Bootstrap method is adopted to perform sampling with replacement on travel time estimation samples; M rounds of repeated sampling are carried out; the travel time reliability index TTRm for each data subset is calculated; and a TTR meeting normal distribution is subjected to point estimation and interval estimation. The method outputs a reliability index estimation interval, in comparison with a reliability index point estimation value, the description of the travel time reliability by the confidence interval is more comprehensive, and an interval upper limit value has more practical significance in application of guidance release, traffic planning and so on.

The invention discloses a high speed railway bridge vehicle-bridge vibration performance safety early warning method. The method includes acquiring the maximum value of the acceleration response amplitude for each passing train and the corresponding vehicle speed data, pairing the corresponding elements of two sets of sequences, and making an arrangement according to the vehicle sped; decomposing the acceleration response amplitude sequences of a main beam by means of wavelet packet decomposition, extracting the acceleration-vehicle speed median line, and determining the optimal wavelet packet decomposition parameters according to the matching degree of the acceleration amplitude mean of the acceleration sequence within the fixed sequence length; acquiring the fluctuation interval, surrounding the acceleration-vehicle speed median line, of the acceleration amplitude in each subsequence based on an interval estimation theory; and superimposing the acceleration fluctuation interval and the acceleration-vehicle speed median line to obtain the safety early warning interval of the high speed railway bridge vehicle-bridge vibration performance. Compared with the prior art, the method provided has the advantages of comprehensive consideration of factors, accurate and reasonable way, clear physical meaning, easy understanding and implementation, and wide application.

The invention discloses a bridge progressive drift data cleaning method based on a data difference value. The bridge progressive drift data cleaning method comprises the steps of obtaining the sensor original data; calculating a difference value of the adjacent data between the original data; extracting the trend data from the original difference value data through a Super Smart algorithm to obtain a difference value between the original difference value data and the trend data; obtaining a change range of a normal value of a difference value of the original difference value data and the trend data by adopting an interval estimation theory according to a mean value and a standard deviation of the difference value of the original difference value data and the trend data; searching a data position where the difference value of the original difference value data and the trend data is outside the confidence interval, determining the position of the drift abnormal value in the original data, and cleaning the drift abnormal value. According to the present invention, the progressive drift phenomenon in the data with slower signal fluctuation can be accurately identified, the processing efficiency of the original data is improved, the real monitoring data can be restored to the maximum extent, and the effective data is provided for the subsequent structure state analysis.

The invention relates to an evaluation method of uncertainty in the entropy measurement for multisensor neural network of poor information, and belongs to measuring and testing field. The method is characterized in following steps:(1)using RBF(Radial Basis Function) to establish a data fusion model of multisensor data conditions under the condition of poor information, acquiring fusion sequence reflecting the measuring process;(2)using the maximum entropy principle to process obtained fused sequence, to obtain the probability density function of the fusion sequence;(3)using obtained probability density function to process true value and regional evaluation, to realize uncertain measurement with poor information. The method allows small sample of measure data, the unknown probability distribution and can make full use of the existing information of measured data to realize effective evaluation of uncertain measurement.

Medical devices and related patient management systems and methods are provided. A method of presenting information pertaining to operation of a medical device involves obtaining historical glucose measurement data for a patient from a database, identifying, based on the historical glucose measurement data, a first plurality of event patterns within respective ones of a plurality of monitoring periods, determining a respective value for a confidence metric for each respective event pattern of the first plurality of event patterns based at least in part on a detection criterion associated with the respective event pattern, a respective subset of the historical glucose measurement data corresponding to the respective monitoring period of the plurality of monitoring periods associated with the respective event pattern, and an interval estimation metric associated with the historical glucose measurement data, and providing one or more graphical indicia influenced by the confidence metric.

The invention relates to a no-failure data ultra-small sample-based product life distribution assessment method. The method comprises the following steps of: estimating a shape parameter (as shown inthe specification) of product life which obeys Weibull distribution, wherein the shape parameter is selected from 1.5 to 2.5; estimating a position parameter, obtaining a median rank of a parameter estimation success probability according to a life sample value, and obtaining a relationship between a sample amount under a stipulated confidence level and the position parameter; estimating a scale parameter, and expressing a relationship (as shown in the specification) among reliability, the confidence level, the sample amount and a failure amount by using binomial distribution, wherein R is thereliability, C is the confidence level, j is a failure sample number, an equation (as shown in the specification) exists to obtain a scale parameter estimation formula (as shown in the specification)under the given confidence level. Through the method, life distribution, a confidence coefficient of which is 95% is estimated by using a unilateral interval estimation method according to an empirical value of the product life Weibull distribution shape parameter and the estimated position parameter, so that effectiveness of the method is verified.

The invention provides an aircraft route segment fuel consumption range estimation method based on QAR data. The aircraft route segment fuel consumption range estimation method based on the QAR data comprises the steps that the QAR data are classified according to aircraft models and route segments to obtain QAR data samples of specific route segments of specific aircraft models; fuel consumption data are extracted from the QAR data samples and calculated, so that fuel consumption samples of the specific route segments of the specific aircraft models are composed; the mean value and standard deviation of the fuel consumption samples are calculated; a normality test is conducted on the fuel consumption samples; the minimum sample size is calculated with a given significance level and given estimated accuracy; the minimum sample size is checked; fuel consumption range estimation of the specific route segments of the specific aircraft models is conducted. According to the aircraft route segment fuel consumption range estimation method, the QAR data are classified according to the aircraft models and the route segments to conduct fuel consumption range estimation of the specific route segments of the specific aircraft models, and therefore the purpose of determining and checking the minimum sample size in the process of conducting fuel consumption range estimation of the specific route segments of the specific aircraft models through the QAR data is achieved under the circumstance that the sample size of the QAR data is limited.

The invention provides a dynamic system gain estimation method based on a historical data steady-state value, and the method comprises the steps: firstly, dividing an input time sequence and an outputtime sequence into short data segments through employing a linear segmentation representation method, and finding out data segments with input and output being in a steady state at the same time; calculating input and output steady-state values from the data segment under the steady-state condition; elements, in the steady-state values, associated with the same steady-state gain in statistical significance are found, the elements are divided into one group, the steady-state gain of each group is estimated, and interval estimation of estimation parameters is given. According to the method, theinput and output steady-state values can be automatically found in the historical data sample, a plurality of steady-state gains under different working conditions can be accurately and effectively estimated, verification is carried out through a visualization method, and the problems that the time consumption for searching the steady-state values is long, the steady-state values are easily influenced by nonlinearity and the steady-state gain change is difficult to detect are solved.