CN102495947A - Method for expressing and analyzing uncertainty based on virtual reality technology - Google Patents

Abstract

The invention discloses a method for expressing and analyzing uncertainty based on a virtual reality technology, which is used for overcoming the technical defects of high complexity, low accuracy, low efficiency, adversity to quantity value transmission among departments and the like existing in the convention uncertainty analyzing method. In the method, the virtual reality technology is used for assisting in expressing uncertainty, so that metering personnel can judge a measured distribution type understand the variable rule of the uncertainty along with the performance of a measuring instrument. The uncertainty is calculated through a fitting parameter by using a curve fitting method, so that the evaluation accuracy of the uncertainty can be increased. Meanwhile, the virtual reality technology is applied in a quantity value transmission process, so that relevant personnel can rapidly and efficiently understand the uncertainty given by other departments, and the quantity value transmission efficiency is increased.

Description

A kind of based on the expression of virtual reality technology and the method for analysis uncertainty

Technical field

The present invention relates to the metrological testing technology field, relate in particular to a kind of based on the expression of virtual reality technology and the method for analysis uncertainty.

Background technology

The purpose of measuring is in order to obtain measurement result, but it is inadequate only providing measurement result.All there is defective in any measurement, all measurement results all more or less depart from measured true value, therefore when providing measurement result, also must be pointed out the degree of reliability of the measurement result of giving.The degree of reliability of measurement result is then reflected by Measurement Uncertainty.Uncertainty is the key concept in metrology and measurement field.International Bureau of Wieghts and Measurements (BIPM) united with 7 international organizations such as International Organization for Standardization in 1993 and has issued " uncertainty of measurement is represented guide " (Guide to the Expression of Uncertainty in Measurement is called for short GUM).China has also issued JJF1059-1999 " evaluation of uncertainty in measurement and expression " in 1999, so that synchronous with the world, can carry out unified assessment to measuring process, guarantee the reliability and the unitarity of metering field transmission of quantity value.

At present, LPU method (Law of Propagation of Uncertainty) is that China metrological services at different levels analyze the method that uncertainty is the most often used.The LPU method has the advantage that method is easy, the scope of application is wider as the method that GUM recommends; But need infer that in the LPU method criterion of deduction is very complicated but also can only be applicable to the part situation not only to measured probability density distribution type according to a series of criterions.Analyze uncertainty to survey crew and bring very big difficulty.In addition; For the engineering technical personnel of some non-metering specialties, they are proficient in the knowledge of specialty separately, but the principle of uncertainty of measurement are known few; To some complicated measurement models, express uncertainty through mode word merely and possibly beyond one's depth them.

Summary of the invention

In view of this; Fundamental purpose of the present invention is to provide a kind of uncertainty based on virtual reality technology to express and analytical approach, be used to solve existing analysis on Uncertainty method too very complicated, poor accuracy, efficient is low and be unfavorable for technological deficiency such as interdepartmental transmission of quantity value.

For achieving the above object, technical scheme of the present invention is achieved in that

Technical scheme 1, a kind of uncertainty of measurement based on virtual reality technology are expressed and analytical approach, and this method comprises:

A. introduce the measurement model y=f (x that measures input quantity and output quantity ₁, x ₂... .x _n), y is that output quantity is promptly measured in the formula, x ₁, x ₂... .x _nBe input quantity;

B. introduce input quantity x ₁, x ₂... .x _nProbability distribution, represent that wherein the parameter of input quantity probability distribution is designated as a _Ij, i=1...n, n are the number of input quantity, j=1...m, and m is for influencing input quantity x _iThe number of parameters of probability distribution;

C. according to measurement model y=f (x ₁, x ₂... .x _n) and input quantity x ₁, x ₂... .x _nProbability distribution calculate probability density function g (y, a of output quantity y ₁₁..., a _Nm);

D. according to the real needs of measuring, select a key parameter a _Kr(1≤k≤n, 1≤r≤m), confirm a _KrSpan, obtain the value of other input quantity probability distribution parameters according to the measurement instrument performance, this moment, the probability density function of y was designated as g (y, a _Kr), with output quantity y, key parameter a _Kr, probability density value g (y, a _Kr) distinguish X, Y, three coordinate axis of Z under the corresponding three-dimensional virtual reality system, generate diagram of block;

E. under virtual reality system, set up the graphical user interface gui interface and realize control, to realize probability density curve with a to cutting face far away in the said diagram of block and nearly cutting face _KrThe control that changes, survey crew can pass through this graphic user interface thus, observes probability density curve with a with the mode of interaction _KrThe tracing pattern that changes;

F. confirm a according to measuring actual conditions _KrValue, and to y sampling, thereby obtain one group of discrete point set (y that representes the y probability density _t, g _t), t=1.....T, T are the sampled point number, g _tBe y=y _tThe time probability density value;

G. be objective function with each exemplary distribution in the uncertainty evaluation, be discrete point set (y _t, g _t) curve fitting; Exemplary distribution in the uncertainty evaluation comprises: normal distribution, rectangular distribution, triangle distribution, U type distribute; Can obtain the pairing error of fitting statistical value of each exemplary distribution after the curve fitting; The pairing distribution pattern of minimum value is the distribution pattern of measured y in each fitting of distribution error statistics value, this distribution pattern is designated as the L type distributes;

H. be distributed as objective function with the L type, to discrete point set (y _t, g _t) can obtain curve fitting parameter in the process that carries out curve fitting, by the curve fitting parameter uncertainty that can be expanded, thereby provide the uncertainty evaluation report.

Technical scheme 2 is based on technical scheme 1, and the step that generates diagram of block in the steps d is specially:

D1, definite output quantity y and key parameter a of measuring _KrSpan and value at interval, and to y and a _KrSampling, thus y obtained _(p), a _{Kr (q)}, p=1 wherein, 2...P, q=1,2...Q, wherein P, Q are the sampled point number, and calculate g _{(p, q)}(y _(p), a _{Kr (q)}), in the three dimensions of virtual reality, set up point set (y _(p), a _{Kr (q)}, g _{(p, q)}(y _(p), a _{Kr (q)})), note by abridging and be g _{(p, q)}, the corresponding y of the three-dimensional X axle of virtual reality _(p), the corresponding a of Y axle _{Kr (q)}, the corresponding g of Z axle _{(p, q)}(y _(p), a _{Kr (q)});

D2, according to g _{(p, q)}(y _(p), a _{Kr (q)}) value each colouration of concentrating for point, the color map table adopts the Jet mapping table;

D3, above-mentioned point set is decomposed into point set by following principle divides into groups; Each point set constructed in groups becomes a triangular facet; Group forming criterion is: three adjacent at first in twos points constitute 1 point set and divide into groups; The projection of triangular facet on the XY plane of each point set constructed in groups one-tenth can not have intersection in addition, and be following by the point set block form of above principle construction: [g ₍₁₁₎, g _(2,1), g ₍₁₂₎], [g _(1,2), g _(2,1), g _(2,2)] ... .... [g _{(p, q)}, g _{(p+1, q)}, g _{(p, q+1)}], [g _{(p, q+1)}, g _{(p+1, q)}, g _{(p+1, q+1)}] ... ... [g _{(P-1, Q-1)}, g _{(P, Q-1)}, g _{(P-1, Q)}], [g _{(P-1, Q)}, g _{(P, Q-1)}, g _{(P, Q)}] p=1,2...P, q=1, three points in the 2...Q, square brackets promptly are combined as a point set and divide into groups;

D4, concentrate the divide into groups triangular facet that constitutes of color and the point set of each point, according to coloring mode, be each triangular facet colouration, thereby obtain colored diagram of block based on the summit according to point.

Technical scheme 3 is based on technical scheme 1, and step g is specially:

G1, according to the determined a of technical scheme 1 step f _KrValue is set a through graphic user interface in the figure that technical scheme 1 step e is generated _KrValue, thereby obtain the probability density curve of measured y;

G2, because the pairing probability density curve shape difference of different exemplary distribution great disparity, the probability density curve shape of the measured y that obtains according to technical scheme 3 step g 1 can the exclusive segment exemplary distribution;

G3, be objective function, be discrete point set (y with each exemplary distribution of not being excluded _t, g _t) curve fitting, after the curve fitting, the pairing error of fitting statistical value of each exemplary distribution that can be excluded, the pairing distribution pattern of minimum value is the distribution pattern of measured y in the error of fitting statistical value, this distribution pattern is designated as the L type and distributes.

The present invention expresses uncertainty with virtual reality technology for assisting, and helps metrological personnel to judge measured distribution pattern, helps metrological personnel to understand uncertainty with surveying instrument changes of properties rule.And calculate the accuracy that uncertainty can improve uncertainty evaluation through fitting parameter with curve fitting method.Modern surveying need be set up complete transmission of quantity value system simultaneously, and uncertainty of measurement is the transmission of quantity value core work, and transmission of quantity value need be in inter-sectional this information of transmission uncertainty of difference.Expressing uncertainty with interactive patterned mode helps in the transmission of quantity value process; The related personnel understands the uncertainty that other departments provide fast and efficiently, and the present invention can be widely used in the expression and analysis of all kinds of mechanism for testing to uncertainty of measurement.

Description of drawings

Fig. 1 carries out the method flow diagram that uncertainty of measurement is expressed and analyzed for what the embodiment of the invention provided based on virtual reality technology;

Fig. 2 is the method flow diagram that generates diagram of block in the inventive method;

Fig. 3 is the synoptic diagram of the EVM parameter-definition that relates in the embodiment of the invention;

The display effect figure of the diagram of block of the measured probability distribution of expression that Fig. 4 provides for the embodiment of the invention;

The μ that Fig. 5 provides for the embodiment of the invention is 0 o'clock measured probability density curve figure;

The μ that Fig. 6 provides for the embodiment of the invention is 0.162 o'clock measured probability density curve figure;

The μ that Fig. 7 provides for the embodiment of the invention is 0.5 o'clock measured probability density curve figure.

Embodiment

For making the object of the invention, technical scheme and advantage clearer, below lift embodiment, through to EVM parameter Uncertainty Analysis in the digital communication, the present invention is explained further details.

EVM (Error Vector Magnitude) is the most important parameter of digital wireless communication field gauge signal quality.Usually use VSA to measure the EVM parameter of digital modulation signals.To this Analysis of Measurement Uncertainty is the key that guarantees modulation quality parameter magnitude tracing.The method of expression and analysis EVM uncertainty is as shown in Figure 1, and concrete steps are following.

Step 101 is set up measurement model: the definition of EVM is the ratio of Error Vector Magnitude and reference signal amplitude, and referring to Fig. 3, it defines suc as formula shown in (1),

$\mathrm{EVM}=\sqrt{{\left({I-I}_0\right)}^2+{\left({Q-Q}_0\right)}^2}/\sqrt{{{\mathrm{<figure-callout\; id="0"\; label="I"\; filenames="1111151420255.png"\; state="\{\{state\}\}">I</figure-callout>}}_0}^2+{{\mathrm{<figure-callout\; id="0"\; label="Q"\; filenames="1111151420255.png"\; state="\{\{state\}\}">Q</figure-callout>}}_0}^2}---\left(1\right)$

I, Q are the coordinate figure of actual signal constellation point in the formula, I ₀, Q ₀For the coordinate figure of reference signal constellation point is two constants.Formula (1) is the measurement model among the technical scheme 1 step a, and wherein EVM is the measurement model output quantity, abbreviates output quantity as, is also referred to as measured; I, Q are the measurement model input quantity, abbreviate input quantity as.

Step 102 is confirmed measurement model input quantity probability distribution: I, Q are the input quantities of measurement model, and it distributes by the VSA decision of measuring-signal.In the present embodiment, the PSA signal analyzer that uses Agilent company can be known according to the performance index of PSA signal analyzer as surveying instrument: I, Q are separate, and all accord with normal distribution its depart from I ₀, Q ₀Mathematical expectation be μ _I, μ _Q, its variance is σ _IQ ², i.e. I～N (μ _I+ I ₀, σ _IQ ²), Q～N (μ _Q+ Q ₀, σ _IQ ²).Accomplished the step b of technical scheme 1 thus.

Step 103 is found the solution the probability density function of measurement model output quantity: the measurement model that provides according to formula (1) and the distribution of input quantity, and the probability density function that can obtain output quantity EVM through mathematical derivation is:

$g(\mathrm{EVM},\mathrm{\&mu;},\mathrm{\&sigma;})=\frac{\mathrm{EVM}}{{\mathrm{\&sigma;}}^2}\exp \left(\frac{-{\mathrm{EVM}}^2-{\mathrm{\&mu;}}^2}{{2\mathrm{\&sigma;}}^2}\right)B_0\left(\frac{\mathrm{EVM\&mu;}}{{\mathrm{\&sigma;}}^2}\right)---\left(2\right)$

In the formula ${\mathrm{\&mu;}}^2=\frac{1}{{{\mathrm{<figure-callout\; id="0"\; label="I"\; filenames="1111151420255.png"\; state="\{\{state\}\}">I</figure-callout>}}_0}^2+{{\mathrm{<figure-callout\; id="0"\; label="Q"\; filenames="1111151420255.png"\; state="\{\{state\}\}">Q</figure-callout>}}_0}^2}{{\mathrm{\&mu;}}_{\mathrm{<figure-callout\; id="2"\; label="I"\; filenames="1111151420255.png"\; state="\{\{state\}\}">I</figure-callout>}}}^2+\frac{1}{{{\mathrm{<figure-callout\; id="0"\; label="I"\; filenames="1111151420255.png"\; state="\{\{state\}\}">I</figure-callout>}}_0}^2+{{\mathrm{<figure-callout\; id="0"\; label="Q"\; filenames="1111151420255.png"\; state="\{\{state\}\}">Q</figure-callout>}}_0}^2}{{\mathrm{\&mu;}}_{\mathrm{<figure-callout\; id="2"\; label="Q"\; filenames="1111151420255.png"\; state="\{\{state\}\}">Q</figure-callout>}}}^2,$ $\mathrm{\&sigma;}=\frac{{\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="0"\; label="IQ\; I"\; filenames="1111151420255.png"\; state="\{\{state\}\}">IQ</figure-callout>}}}{\sqrt{{I_{}}^{}}},$ B ₀The imaginary part of 0 rank distortion Bessel's function is got in expression.Accomplished the step c of technical scheme 1 thus.

Step 104 selects to influence the key parameter of uncertainty: for present embodiment, according to the performance of the surveying instrument Agilent PSA of company signal analyzer can obtain measuring input quantity probability density parameter μ, the σ value is respectively 0.5 and 0.1.Because the μ value of different measuring instrument and σ value can be different, be the rule that the uncertainty that reflects EVM changes with μ, selecting μ is the key parameter that influences EVM parameter measurement uncertainty, σ is a constant 0.1.Also can be as the case may be, selecting σ is key parameter, and μ is made as fixed value, its analytic process and method are introduced with μ to be elected as key parameter identical here no longer in detail.When electing μ as key parameter, (EVM, diagram of block μ) can express the situation that probability density curve changes with μ clearly, the rule that the uncertainty that helps survey crew to understand EVM changes with μ through in virtual reality system, setting up function g.

Step 105 generates diagram of block: the implementation procedure of step 105 is comparatively complicated, and the concrete grammar of performing step 105 has been explained in the step 201 of accompanying drawing 2～204.

Step 201 generates point set: according to the definition of EVM, the span of EVM is 0 to 1, sets value and is spaced apart 0.01, obtains EVM _(p)=0,0.01,0.02.....1 is p=1 wherein, 2......101.According to the performance of normal signal analyser, the span of μ is 0 to 0.5, sets value and is spaced apart 0.005, obtains μ _(q)=0,0.005,0.01.....0.5 is q=1 wherein, 2......101.Calculate g according to formula (2) _{(p, q)}(EVM _(p), μ _(q)), and in the three dimensions of virtual reality, set up point set (EVM _(p), μ _(q), g _{(p, q)}(EVM _(p), μ _(q))), note by abridging and be g _{(p, q)}, accomplished the steps d 1 of technical scheme 2 thus.

Step 202 summit colouration: according to g _{(p, q)}(EVM _(p), μ _(q)) value each colouration of concentrating for point, the colouration process is: at first set up the Jet type color map table on one 64 rank, calculate g then _{(p, q)}(EVM _(p), μ _(q)) pairing call number in the color map table, can adopt following formula computation index number,

Cindex＝fix[(g-g _min)/(g _max-g _min)×64]+1(3)

G represents g in the formula _{(p, q)}(EVM _(p), μ _(q)) value, g _MaxAnd g _MinBe respectively the minimum and maximum value of g, fix representes according to the principle that rounds up decimal to be turned to integer.Cindex is g _{(p, q)}(EVM _(p), μ _(q)) pairing call number in the color map table.In Jet color map table, get the rgb value that Cindex element can obtain this color.Accomplish the steps d 2 of technical scheme 2 thus.

Step 203 point set divides into groups: the group technology according to the steps d in the technical scheme 23 is introduced obtains [g with aforementioned point set grouping _(1,1), g _(2,1), g _(1,2)], [g _(1,2), g _(2,1), g _(2,2)] ... ... [g _{(p-1, q-1)}, g _{(p, q-1)}, g _{(p-1, q)}], [g _{(p-1, q)}, g _{(p, q-1)}, g _{(p, q)}] ... ... [g _(100,100), g _{(101,, 100)}, g _(100,101)], [g _(100,101), g _(101,100), g _(101,101)].Three points in the square brackets represent that a point set divides into groups, and each point set constructed in groups becomes a triangular facet.The point set that has obtained thus in technical scheme 2 steps d 3 divides into groups.

Step 204 curved surface colouration: concentrate the color of each point and the triangular facet that the point set grouping constitutes according to point,, be each triangular facet colouration, can obtain diagram of block after each triangular facet colouration according to coloring mode based on the summit.So-called coloring mode based on the summit is meant the color according to each summit of three-dimension curved surface, gives the plane colouration between the summit according to the principle of color gradient, accomplishes each step of technical scheme 2 thus, and the steps d of technical scheme 1.The effect of Fig. 4 for being reached after this step of completion, the value of the corresponding measured EVM of coordinate axis of mark EVM among the figure, the value of the corresponding key parameter μ of coordinate axis of mark mu, the corresponding probability density value of the coordinate axis of mark PDF.Different surveying instrument μ values can be different, and Fig. 4 has reflected the situation that the probability density of EVM changes with surveying instrument μ value.Because directly by measured probability density decision, therefore this figure has also just reflected the uncertainty of EVM and the relation of μ to uncertainty.In virtual reality system, therefore the observer can also be more convenient for observing rule and the trend of the probability density of EVM with parameter μ variation through operations such as rotation, translation, convergent-divergent each position from arbitrarily angled observation figure.

Step 106 is controlled the interface through the graphical user and is realized mutual: can in virtual reality system, add the scroll bar control as graphic user interface (GUI), and the slide block currency of scroll bar control and the far and near cutting face of user's finding view are associated.Accomplish the step e of technical scheme 1 thus.Through the deviate of far and near cutting face is set, the probability density curve the when observer at a time can only to be similar to see μ be a certain particular value, and through the text display node μ value is shown.Fig. 5～Fig. 7 is that intercepting μ is 0,0.162,0.5 o'clock probability density curve figure, the value of the corresponding measured EVM of horizontal ordinate among the figure, and the corresponding probability density value of ordinate, scroll bar is positioned at the figure bottom.In fact when the observer controls the GUI control of virtual reality; What the observer saw not merely is many static pictures; Along with the observer spurs scroll bar; During the value of just approximately continuous change μ, figure can change along with observer's action continuously, and this interactive operation meeting brings stronger experience directly perceived to the observer.So that the observer observes the rule that the probability density curve of measured EVM changes with μ.

Step 107 is confirmed input quantity probability density parameter value: performance index input quantity probability density parameter μ, σ value according to the surveying instrument Agilent PSA of the company signal analyzer in the present embodiment are respectively 0.5 and 0.1.

Step 108 pair output quantity probability density function sampling: because input quantity probability density parameter value is definite, the measurement output quantity probability density function that integrating step 103 obtains, can obtain with output quantity EVM is univariate probability density function:

g＝EVMexp[-50(EVM ²+0.25)]B ₀(50EVM)(4)

Span sampling according to EVM obtains one group of data (EVM _t, g _t).EVM wherein _t=0,0.01,0.02.....1, t=1,2 ... 101.g _tAccording to EVM _tValue calculate by formula (4).Accomplish the step f of technical scheme 1 thus.

Step 109 is judged the output quantity distribution pattern: according to the determined input quantity probability density of step 107 parameter value; In the mentioned graphical user's interactive interface of step 106, setting the μ value is 0.5; Obtain the probability density curve of measured y thus, shown in accompanying drawing 7.Probability density curve and normal distribution or the triangle distribution that from Fig. 7, can obviously see measured y are comparatively approaching, and be bigger with rectangular distribution, U type difference in distribution.Based on above-mentioned judgement; Sampled data is carried out the curve fitting of normal distribution and triangle distribution according to methods of numerical; The error of fitting statistic SSE that obtains when wherein doing the curve fitting of normal distribution equals 0.0004; Be the minimum value of each exemplary distribution error of fitting statistic, the distribution pattern of judging measured EVM thus is normal distribution.Accomplish each step of technical scheme 3 thus, realized the function of technical scheme 1 step g.

Step 110 provides the uncertainty report: in the process of step 109 curve fitting, remove and obtain the fitting parameter that the error of fitting statistic also can obtain characterizing curve.The distribution pattern of measured EVM is normal distribution in the present embodiment; When sampled data is carried out the curve fitting of normal distribution; The match objective function is a fitting parameter for a, b in

formula, and x is the fitting function independent variable.Can obtain match value a=0.099 and b=0.51 after the curve fitting.Be the standard uncertainty of EVM according to the match value of the nature parameters a of normal distribution.Thus according to the evaluation uncertainty conventional method (being the LPU method) in the time of can obtaining k=2 the expanded uncertainty of measured EVM be 0.198.Wherein k=2 representes that fiducial probability is 95%.Accomplish the step h of technical scheme 1 thus, obtained the uncertainty of measurement report of EVM, realized measuring the assessment and the expression of quality.

Can see that through present embodiment the present invention is that the evaluation uncertainty judges that particularly measured distribution pattern provides a kind of new, effective method.Wherein expressing uncertainty with the mode of interactive graphicsization not only helps survey crew to judge that measured distribution pattern helps to test the related personnel simultaneously and understands the meaning that uncertainty is contained; The core information that helps this metering field of uncertainty; Between metrological services at different levels, transmit accurately and efficiently; Thereby guarantee carrying out smoothly of transmission of quantity value work, technical support is provided for setting up complete magnitude tracing system.

Because the flow process of evaluation of uncertainty in measurement and measurement parameter are irrelevant.Therefore obviously not only the evaluation of uncertainty in measurement to EVM is effective for this method, and during this method can be applicable to all kinds of Evaluation of uncertainty of measurement such as length, temperature, pressure and expresses, this method had very strong versatility.

The above is merely preferred embodiment of the present invention, is not to be used to limit protection scope of the present invention.

Claims (4)

1. the uncertainty of measurement based on virtual reality technology is expressed and analytical approach, it is characterized in that this method comprises:

Measurement model y=f (the x of input quantity and output quantity is measured in a, introducing ₁, x ₂... .x _n), y is that output quantity is promptly measured in the formula, x ₁, x ₂... .x _nBe input quantity;

B, introducing input quantity x ₁, x ₂... .x _nProbability distribution, represent that wherein the parameter of input quantity probability distribution is designated as a _Ij, i=1...n, n are the number of input quantity, j=1...m, and m is for influencing input quantity x _iThe number of parameters of probability distribution;

C, according to measurement model y=f (x ₁, x ₂... .x _n) and input quantity x ₁, x ₂... .x _nProbability distribution calculate probability density function g (y, a of output quantity y ₁₁..., a _Nm);

D, according to the real needs of measuring, select a key parameter a _Kr(1≤k≤n, 1≤r≤m), confirm a _KrSpan, obtain the value of other input quantity probability distribution parameters according to the measurement instrument performance, this moment, the probability density function of y was designated as g (y, a _Kr), with output quantity y, key parameter a _Kr, probability density value g (y, a _Kr) distinguish X, Y, three coordinate axis of Z under the corresponding three-dimensional virtual reality system, generate diagram of block;

E, under virtual reality system, set up the graphical user interface gui interface and realize control, to realize probability density curve with a to cutting face far away and nearly cutting face in the said diagram of block _KrThe control that changes;

F, confirm a according to measuring actual conditions _KrValue, and to y sampling, thereby obtain one group of discrete point set (y that representes the y probability density _t, g _t), t=1.....T, T are the sampled point number, g _tBe y=y _tThe time probability density value;

G, be objective function, be discrete point set (y with each exemplary distribution in the uncertainty evaluation _t, g _t) curve fitting, obtain the pairing error of fitting statistical value of each exemplary distribution after the curve fitting, the pairing distribution pattern of minimum value is the distribution pattern of measured y in each fitting of distribution error statistics value, this distribution pattern is designated as the L type distributes;

H, be distributed as objective function with the L type, to discrete point set (y _t, g _t) obtain curve fitting parameter in the process that carries out curve fitting, by the curve fitting parameter uncertainty that is expanded, thereby provide the uncertainty evaluation report.

2. method according to claim 1 is characterized in that, the step that generates diagram of block in the steps d is specially:

D1, definite output quantity y and key parameter a of measuring _KrSpan and value at interval, and to y and a _KrSampling, thus y obtained _(p), a _{Kr (q)}, p=1 wherein, 2...P, q=1,2...Q, wherein P, Q are the sampled point number, and calculate g _{(p, q)}(y _(p), a _{Kr (q)}), in the three dimensions of virtual reality, set up point set (y _(p), a _{Kr (q)}, g _{(p, q)}(y _(p), a _{Kr (q)})), note by abridging and be g _{(p, q)}, the corresponding y of the three-dimensional X axle of virtual reality _(p), the corresponding a of Y axle _{Kr (q)}, the corresponding g of Z axle _{(p, q)}(y _(p), a _{Kr (q)});

D2, according to g _{(p, q)}(y _(p), a _{Kr (q)}) value each colouration of concentrating for point;

D3, said point set is decomposed into point set by following principle divides into groups; Each point set constructed in groups becomes a triangular facet; Group forming criterion is: three adjacent at first in twos points constitute 1 point set and divide into groups; The projection of triangular facet on the XY plane of each point set constructed in groups one-tenth can not have intersection in addition, and be following by the point set block form of above principle construction: [g _(1,1), g _(2,1), g ₍₁₂₎], [g ₍₁₂₎, g _(2,1), g _(2,2)] ... .... [g _{(p, q)}, g _{(p+1, q)}, g _{(p, q+1)}], [g _{(p, q+1)}, g _{(p+1, q)}, g _{(p+1, q+1)}] ... ... [g _{(P-1, Q-1)}, g _{(P, Q-1)}, g _{(P-1, Q)}], [g _{(P-1, Q)}, g _{(P, Q-1)}, g _{(P, Q)}] p=1,2...P, q=1, three points in the 2...Q, square brackets promptly are combined as a point set and divide into groups;

D4, concentrate the divide into groups triangular facet that constitutes of color and the point set of each point, according to coloring mode, be each triangular facet colouration, thereby obtain colored diagram of block based on the summit according to said point.

3. method according to claim 2 is characterized in that, in the steps d 2, according to g _{(p, q)}(y _(p), a _{Kr (q)}) value concentrate for point each colouration the time, the color map table adopts the Jet mapping table.

4. method according to claim 1 is characterized in that step g is specially:

G1, according to the determined a of step f _KrValue is set a through graphic user interface in the figure that step e is generated _KrValue, thereby obtain the probability density curve of measured y;

The probability density curve shape exclusive segment exemplary distribution of g2, the measured y that obtains according to step g 1;

G3, be objective function, be discrete point set (y with each exemplary distribution of not being excluded _t, g _t) curve fitting, after the curve fitting, the pairing error of fitting statistical value of each exemplary distribution that is not excluded, the pairing distribution pattern of minimum value is the distribution pattern of measured y in the error of fitting statistical value, this distribution pattern is designated as the L type distributes.

Images