CN105588599A - Adaptive correction method of vibration displacement errors of mobile mapping system - Google Patents

Abstract

The invention relates to an adaptive correction method of vibration displacement errors of a mobile mapping system, and belongs to the technical field of spatial data collection and treatment. The adaptive correction method includes the following steps of: a) enhancing original vibration displacement data as a continuous time domain vibration displacement signal dependent on multiple variable mobile mapping environmental parameters; b) classifying and analyzing characteristic relations of each measurement environmental parameter variation and signal frequency domain distribution, the relations including the mapping relation of a measurement environmental parameter variation threshold and a signal frequency domain distribution scope, and the relation of a measurement environmental parameter variation trend and the frequency domain distribution width; c) using the parameter variation threshold mapping relation to dynamically and adaptively divide the continuous time domain vibration displacement signal into signal time domain intervals with specific frequency domain distribution complexity; and d) using variation trend consistent relation to restrain a value range of internal detail level parameters, and executing vibration displacement error correction treatment in every time domain interval. The adaptive correction method of the vibration displacement errors of the mobile mapping system can solve the problem that the present signal treatment method cannot suit accurate error correction of the vibration displacement of the mobile mapping system under the complex mapping environment, and contributes to supporting the high precision mobile mapping data analysis.

Description

The adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error

Technical field

The invention belongs to space data collection and processing technology field, particularly a kind of vehicle-mounted mobile measuring system vibrationThe adaptive corrective method of displacement error.

Background technology

Vehicle-mounted mobile measuring system (MobileMappingSystem, MMS) is based on the flexible integrated whole world of carrier vehicleNavigation system (GlobalPositioningSystem, GPS), laser scanner, image sensor and inertial navigation unitAll kinds of location such as (InertialMeasurementUnit, IMU), determine appearance sensor and measuring transducer ground moving surveyAmount system. MMS is not only less is subject to the restriction of weather and measured zone, also has the outstanding advantage of field process efficiency, can be effectiveQuick Acquisition and the regular update of supporting target information, become urban planning, road traffic, map revision, monitoring management graduallyDeng the technical way that obtains three-dimensional spatial information in sector application.

The various kinds of sensors entirety of the carrier vehicle of MMS and lift-launch thereof has formed one and has had quality, spring and dampingVibrational system, the periodicity displacement of being introduced by system vibration thereby be extensively present in vehicle-mounted mobile and measure the every sky obtainingBetween in data. In running, the value difference of MMS periodic vibration displacement can approach a decimeter magnitude, and forming can not in measured dataThe component of ignoring, accurately extracting vibration displacement becomes thus in the post processing of vehicle-mounted mobile measurement data and ensures the quality of data and precisionPrecondition.

Having the acceleration transducer of analog input and high measurement accuracy characteristic as the Primary Component of IMU, is to form MMSImportant component part. Utilize acceleration transducer to measure MMS vertical direction acceleration and resolve vibration position by quadratic integralThe method of moving, not only has the ease for operation of measurement device, the intuitive of extraction algorithm, also has especially and can effectively eliminate the effects of the actThe significant advantages such as the random noise of instantaneous acceleration certainty of measurement, thereby become and existingly carry out engineering survey institute extensively based on MMSThe vibration displacement extracting method adopting. But, due to the high-sensitivity characteristic of acceleration transducer, even if select high-precision sensingDevice, system is also difficult to avoid the acceleration signal of actual output to be subject to: the angular deviation of 1. installing, 2. instrument output is intrinsic zero inclined to one sidePoor, the 3. non-zero-deviation of integration initial value, and the many factors such as the instrument scale deviation that produces that 4. raise with serviceability temperatureImpact, makes to resolve the dither displacement that obtains and sneaks into multiple low frequency aberration composition, and especially, above-mentioned error is towards continuous surveyIn the integrated acceleration resolving of amount process, will produce cumulative effect, become the vibration displacement precision that serious reduction is extracted, because ofThis, the vibration displacement extracting mode based on integrated acceleration generally adopting in the face of the existing MMS of each professional domain, how accurately andMultifactor accumulated error in correcting signal efficiently, becomes the pass that accurate extraction vibration displacement and then performance measurement data are worthKey problem.

The existing vibration displacement antidote towards integrated acceleration cumulative errors mainly concentrates on computer based numberWord signal processing (DigitalSignalProcessing, DSP) field. Correlation technique mainly utilize actual vibration signal withThe error signal feature that otherness distributes on frequency domain, adopts signal to process operator, by towards " the specific frequency domain of overall signal is thinGanglionic layer time " decomposition and reconstruction, reject the wherein corresponding multiple low-frequency component of control information. On the whole, prior artThrough correcting the multifactor error in global displacement signal with stable frequency domain character, this art method is at following documentIn all have a discussion: the auspicious duckweed of knob, Cai Baigen. a kind of research [J] of accelerometer error modification method. sensor technology, 2002,21(4): 4-6; Zhang Pengfei, Long Xingwu. the research [J] of Compensating Error Model of Quartzose Flexible Accelerometer. sensing technology journal,2006,19 (4): 1100-1102; Xu Chao, Shen Xiaorong, Li JianJun, etc. the Wavelet Denoising Method technology of vehicle-mounted micro-acceleration gauge signal is groundStudy carefully [J]. sensing technology journal, 2008 (11): 2442-2444; Liu Xinyu, Xiao Chuanyu, Wu Yong. inertial reference measurement of higher degree methodApplication [J] in surface evenness detects. transport information and safety, 2009 (5): 166-169; Qin Fangjun, is permitted Jiangning, LeePeace, etc. the accelerometer noise-reduction method [J] based on small echo Kalman filtering. Wuhan University of Technology's journal: traffic science and engineeringVersion, 2009,33 (1): 49-52; Zhao Baoxin, Zhang Baocheng. the research of polynomial trend item and the MATLAB of signal realize [J]. and colouredEquipment, 2009 (2): 16-19; Xu Shiwei, Shen Haibin. the acceleration transducer accumulated error removing method [J] based on SVM. passSensor and micro-system, 2012 (6): 42-44; Ma Yue, Li Song, Li Ying, etc. the modeling and simulation [J] of accelerometer signal processing.Computer Simulation, 2012,29 (3): 351-354; The old snow winter, Chen Shuohong, xuwei. the small echo shape filtering of accelerometer signal withSample Entropy is analyzed [J]. instrument and meter for automation, 2013,35 (5): 22-25; A kind of absolute Navigation of estimated acceleration meter driftMethod (publication number: 103542853A); A kind of external field environment three-dimensional measurement accelerometer error is (open without unusual method of estimationNumber: 103995152A); MayagoitiaRE, NeneAV, VeltinkPH.Accelerometerandrategyroscopemeasurementofkinematics:aninexpensivealternativetoopticalmotionanalysissystems[J].Journalofbiomechanics,2002,35(4):537-542；HesamiR,McManusKJ.Signalprocessingapproachtoroadroughnessanalysisandmeasurement[C]//TENCON2009-2009IEEERegion10Conference.IEEE,2009:1-6；WuP,GeY,ChenS,etal.De-noisingalgorithmbasedoncompressionofwaveletcoefficientforMEMSaccelerometersignal[C]//InformationandAutomation(ICIA),2010IEEEInternationalConferenceon.IEEE,2010:402-407；KownackiC.OptimizationapproachtoadaptKalmanfiltersforthereal-timeapplicationofaccelerometerandgyroscopesignals'filtering[J].DigitalSignalProcessing,2011,21(1):131-140.

In the conventional method, signal decomposition towards level of detail heavy to closing for accurate extraction effective information compositionWant. But, while utilizing existing method to process the vibration displacement data of MMS collection and parsing, the precision of correction result and stabilityLevel of detail parameter to signal decomposition and reconstruct is very responsive. Its main cause is the complexity of ground moving measurement environmentCause system vibration to change the sequential non-stationary characteristic that produces amplitude and frequency with measurement environment, wherein, affect Vibration stabilityMeasurement environment complexity not only comprise the outside geographical environmental condition that MMS traverse measurement process relates to, also comprise internal systemThe driving conditions of carrier vehicle. Thus, at the original vibration displacement letter based on the collection of complicated traverse measurement environment Integration SolvingIn number, the non-stationary characteristic of vibration signal increases sharply the complexity of frequency domain distribution between effective information and multifactor error, and existingTowards the error correction method of specific detail level overall situation Unified Solution, thereby be difficult to adapt to that traverse measurement obtains shakes continuouslyIn moving displacement signal, the detail of effective information frequency domain character changes, thereby causes the decomposition of owing of different period primary signalsWith cross resolution problem:

1. owe decomposition and will be difficult to fully extract effective vibration displacement from primary signal, and

2. crossing to decompose will cause the vibration displacement restructuring procedure based on decomposition data again to introduce error percentage.

To sum up analyze, art methods is difficult to realize vehicle-mounted traverse measurement system vibration displacement essence under complicated measurement environmentTrue and stable error correction.

Summary of the invention

The object of the invention is to for the deficiencies in the prior art, a kind of vehicle-mounted mobile measuring system (MMS) vibration is providedThe adaptive corrective method of displacement error.

The technical solution adopted for the present invention to solve the technical problems is: vehicle-mounted mobile measuring system vibration displacement errorAdaptive corrective method, comprises the following steps:

Step 1, measurement environment parameter strengthens: the output signal of acceleration transducer is carried out to quadratic integral and obtain vertical sideTo original vibration displacement, and the synchronous acquisition measurement environment information corresponding with original vibration displacement, record measurement environment information and beThe continued time domain vibration displacement signal data object that comprises measurement environment parameter;

Step 2, characteristic change parameter is resolved: resolve each measurement environment parameter and change the feature pass distributing with signal frequency domainBe to comprise mapping relations, the measurement environment parameter variation tendency of measurement environment parameter change threshold and signal frequency domain distributionRelation with frequency domain distribution range;

Step 3, feature time domain interval division: based on reflecting of measurement environment parameter change threshold and signal frequency domain distributionPenetrate relation, dynamically dividing continued time domain vibration displacement signal is the adjacent time domain Interval Set of one group of change threshold constraint;

Step 4, Oscillation error correction: the relation based on measurement environment parameter variation tendency and frequency domain distribution range is approximatelyThe span of bundle adjacent interval level of detail, carries out by time domain interval shaking that optimization is chosen based on level of detail parameter adaptiveMoving displacement error correction process.

Described measurement environment information comprises two classes:

Inner parameter: comprise running status advance and time-out, travel condition under rectilinear motion and curvilinear motion, straight lineThe velocity and acceleration of motion, the angular speed of curvilinear motion;

External parameter: surface roughness and waviness.

Described vibration displacement signal data object:

Wherein, O_SRepresent vibration displacement signal data object, be recorded as one group of time T taking the setting sampling interval as labelThe time domain multiple parameters group of item; S represents the vibration displacement that integrated acceleration obtains, be recorded as one group with obtain moment t, obtain skyBetween position x, the immediate movement s sequence that y is corresponding; P represents the polynary measurement environment parameter item strengthening, and comprises and represents to obtain respectively SInner parameter group { PI} and external parameter group { PO}, each inner parameter PI of traverse measurement environment_mBe recorded as and obtain moment t coupleThe instantaneous parameters value p answering, each external parameter PO_nBe recorded as and obtain locus x, the parameter value p that y is corresponding; S and PI_mWhile passing throughBetween parametric t carry out synchronization association; S and PO_nCarry out synchronization association by location parameter t.

The mapping relations of described measurement environment parameter change threshold and signal frequency domain distribution comprise the following steps:

Step 2.1, statistical parameter value distribution characteristics: extract the value of polynary measurement environment parameter item in data, successively systemThe characteristic value value and the frequency fr that count each parameter item, obtain every parameter p according to the array A of the ascending sequence of value_p：

Step 2.2, initializes threshold interval: to every parameter p, according to A_pSpan initializes threshold interval setSEC_p：

SEC_p＝{[value_{_min},value_{_max}]}；

Step 2.3, assessment decomposition threshold interval: to the threshold interval S set EC of every parameter p_pCarry out following sub-step:

Step 2.3.1, extracts set element number Nr=|SEC_p|；

Step 2.3.2, successively to every section of interval SEC_p(n), n={1 ..., Nr}, carries out following sub-step:

VI) according to A_pThe value value of middle interval characteristic value value_{_min},value_{_max}With mode value value_{_frmax}, drawBy stages is subintervalAnd subinterval

VII) extract respectively O_SParameter p monodrome delta data section { the L1}, { L2} of middle corresponding subsec1 and subsec2;

VIII) respectively to data segment { L1} and { the original vibration displacement signal of the corresponding time domain of L2} S data interval S_L1WithS_L2, carry out effective displacement matching, obtain effective displacement S of corresponding different level of detail_{L1'level_n}And S_{L2'level_m}，

IX) obtain the effective displacement S of each level_{L1'level_n}And S_{L2'level_m}Variances sigma² _{L1_level_n}And σ² _{L2_level_m}；

X) extracting respectively the corresponding level of detail N1 of variance minimum of a value and N2 is that the feature of subsec1 and subsec2 is thinGanglionic layer time; Relatively N1 and N2:

D) as | N1-N2| < 1, finish this segment processing;

E), as | N1-N2|=1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former districtBetween section [value_{_min},value_{_max}], upgrade SEC simultaneously_pWith after Nr, finish this segment processing;

F), as | N1-N2| > 1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former districtBetween section [value_{_min},value_{_max}], and respectively to [value_{_min},value_{_frmax}] and [value_{_frmax},value_{_max}] holdRow step 2.3.2:I)～V);

Through step 2.3, obtain corresponding to feature level of detail N_p＝{N₁,N₂,...,N_Nr},Nr＝|SEC_p| parameter pInterval set SEC_p：

SEC_p＝{[value_{_min},value_{_frmax(a)}],[value_{_frmax(a)},value_{_frmax(b)}],

...,

[value_{_frmax(x)},value_{_frmax(y)}],[value_{_frmax(y)},value_{_max}]}；

Step 2.4, resolves mapping relations based on threshold interval resampling: the SEC that step 2.3 is obtained_pCarry out based on districtBetween boundary value linear interpolation, obtain parameter p and { corresponding the increasing progressively of N} of frequency domain distribution scope Digital Signal Processing level of detail parameterChange threshold interval:

SEC_p＝{[-∞,(value_{_min}+value_{_frmax(a)})/2],

[(value_{_min}+value_{_frmax(a)})/2,(value_{_frmax(a)}+value_{_frmax(b)})/2],

...,

[(value_{_frmax(x)}+value_{_frmax(y)})/2,(value_{_frmax(y)}+value_{_max})/2],

[(value_{_frmax(y)}+value_{_max})/2,+∞]}；

Thus, parameter p change threshold and frequency domain distribution range mappings are closed and are:

$\left\{\begin{array}{c}(valu{e}_{\_min}+valu{e}_{\_frmax\left(a\right)})/2_{\&RightArrow;}{N}_1;\\ (valu{e}_{\_frmax\left(a\right)}+valu{e}_{\_frmax\left(b\right)})/2_{\&RightArrow;}{N}_2;\\ ...;\\ (valu{e}_{\_frmax\left(y\right)}+valu{e}_{\_\max })/2_{\&RightArrow;}{N}_{Nr-1};\\ (valu{e}_{\_fr\max \left(y\right)}+valu{e}_{\_\max })/2_{\&RightArrow;}{N}_{Nr};\end{array}\right..$

The relation of described measurement environment parameter variation tendency and frequency domain distribution range comprises following sub-step:

Step 2.5, resolves { the unidirectional constant interval SEC of N}_N：

Step 2.6, judges SEC_NWhether each interval change direction Sgn (sec_n) is consistent with parameter p incremental variations interval;If consistent, flag parameters p and this level of detail interval are variation relation in the same way, are designated as R=1; Otherwise, flag parameters p and thisLevel of detail interval is inverse change relation, is designated as R=-1.

The described mapping relations based on measurement environment parameter change threshold and signal frequency domain distribution, dynamically divide continuouslyTime domain vibration displacement signal is that the adjacent time domain Interval Set of one group of change threshold constraint comprises following sub-step:

Step 3.1, extracts O according to the increasing order of time tag T_SIn the polynary environment measurement parameter group corresponding with S{P}；

Step 3.2, the interval SEC of each parameter threshold_p, according to sequential increasing order from moment t farthest_minStart when nearestCarve t_maxScanning S, travel through the multiple parameters value that each moment instantaneous vibration displacement s is corresponding { P}, when the variation of arbitrary parameter p surpasses simultaneouslyCross interval threshold value, dividing S is corresponding data section, obtains successively the vibration displacement time domain Interval Set S of corresponding parameter changing value_sec：

Wherein, ψ represents parameter item codomain; C is marked off hop count; M value is object O_SIn polynary environment measurement ginsengSeveral numbers, m=|O_S.{P}|；N_pxBe that the value of x parameter p is at interval Ψ_pxThe feature level of detail shining upon when variation; p₁Change for recording the parameter item of crossing over interval threshold value.

The described relation constraint adjacent interval level of detail based on measurement environment parameter variation tendency and frequency domain distribution rangeSpan comprise following sub-step:

Successively to S_secIn every section of interval (S_X,ψ{P}_X),_X=1 ..., C} carries out following sub-step:

Step 4.1, resolves ψ { P}_XIn the codomain scope ψ { N of the feature level of detail that shines upon of each parameter p_p1,...,N_px,...,N_pm}；

Step 4.2, according to section ordinal number X, carry out following operation:

Step 4.2a, works as X=1, at ψ { N_p1,...,N_px,...,N_pmIn scope:

IV) to S_XCarry out by the matching of level of detail vibration displacement; By to S_XFrom the base of 1 to level_x level frequency domain decompositionOn plinth, the successively cumulative high-frequency data item that represents effective vibration displacement; Wherein, level_x ∈ ψ { N_p1,...,N_px,...,N_pm}；

V) calculate the variances sigma of each level of detail level_x² _{level_x}；

VI) more each level of detail variance, the corresponding level of detail level_x of mark variance minimum of a value (σ²Min) beSegment (SX, ψ { P}_X) optimum level of detail L_X; Preserve towards level of detail L_XFitting result be segment (S_X,ψ{P}_X)Correction result;

Step 4.2b, works as X > 1:

VI) obtain a segment (S_X-1,ψ{P}_X-1) optimum level of detail L_X-1＝level_x(σ²min)；

VII) estimate the optimum level of detail increment Delta level that corresponding each parameter changes:

$\mathrm{\&Delta;}level=\left(\underset{i=1}{\overset{m}{\&Sigma;}}{R}_{m*}\right|{\left(N_{pi}\right)}_X-{\left(N_{pi}\right)}_{X-1}\left|\right)/m;$

Wherein, m=|O_S.{P}|, be object O_SIn polynary environment measurement number of parameters; R_mRepresent that i parameter p i is in districtBetween ψ { (N_pi)_X,(N_pi)_X-1Variation tendency concord value, i.e. SEC_NValue;

VIII) estimation interval section (S_X,ψ{P}_X) optimal Decomposition level be:

L_X＝L_X-1+Δlevel；

IX) difference details of use level L_X-1,L_X,L_X+1, matching S_X; Described approximating method is with step 4.2aI);

X) calculate each level of detail variances sigma²; Judge fitting result:

D) work as σ² _LX-1≥σ² _LX≤σ² _LX+1, label L_XFor segment (S_X,ψ{P}_X) optimum level of detail; Preserve towards carefullyGanglionic layer time L_XFitting result be segment (S_X,ψ{P}_X) correction result;

E) work as σ² _LX-1≥σ² _LX≥σ² _LX+1, upgrade L_X＝L_X+1; Return to execution step 4.2b:IV)～V);

F) work as σ² _LX-1≤σ² _LX≤σ² _LX＋1, upgrade L_X＝L_X-1; Return to execution step 4.2b:IV)～V);

Step 4.3, integrates piecemeal correction result and saves as the vibration displacement value after rectification.

The present invention has following beneficial effect and advantage:

1., in the complicated measurement environment of generally facing for the existing MMS of each professional domain, utilize shaking of integrated acceleration parsingMoving displacement data, provides a kind of adaptive error correction solution of taking time-domain signal self frequency domain complex characteristic into account. Solve existingThere is the error correction method of carrying out specific detail level overall situation Unified Solution based on Digital Signal Processing, for want of to " multipleIn the Non-stationary vibration signal extracting in assorted traverse measurement environment, effectively between vibration and multifactor error, unbalanced frequency domain dividesCloth " solution, and then effective information frequency domain in the vibration displacement signal that is difficult to adapt to obtain in complicated traverse measurement environmentThe detail of feature changes, thereby causes different period primary signals to owe to decompose and cross decomposing restraining error to correct precisionProblem.

2. the inventive method goes for the complicated ground measurement environment changeable towards condition, and accurate and flexible ground is correctedThe vibration displacement error of MMS, can effectively strengthen the adaptability that MMS changes measurement environment, contributes to support further Exact SolutionsAnalyse the three-dimensional spatial information gathering based on MMS.

3., in the complicated measurement environment of generally facing for the existing MMS of each professional domain, utilize shaking of integrated acceleration parsingMoving displacement data, provides a kind of adaptive error correction solution of taking time-domain signal self frequency domain complex characteristic into account. Solve existingThere is the error correction method of carrying out specific detail level overall situation Unified Solution based on Digital Signal Processing, for want of to " multipleIn the Non-stationary vibration signal extracting in assorted traverse measurement environment, effectively between vibration and multifactor error, unbalanced frequency domain dividesCloth " solution, and then effective information frequency domain in the vibration displacement signal that is difficult to adapt to obtain in complicated traverse measurement environmentThe detail of feature changes, thereby causes different period primary signals to owe to decompose and cross decomposing restraining error to correct precisionProblem.

4. the inventive method goes for the complicated ground measurement environment changeable towards condition, and accurate and flexible ground is correctedThe vibration displacement error of MMS, can effectively strengthen the adaptability that MMS changes measurement environment, contributes to support further Exact SolutionsAnalyse the three-dimensional spatial information gathering based on MMS.

Brief description of the drawings

Fig. 1 principle of the invention schematic diagram;

The overview flow chart of Fig. 2 the inventive method.

Detailed description of the invention

Below in conjunction with embodiment, the present invention is described in further detail.

A kind of adaptive corrective method that the present invention relates to vehicle-mounted mobile measuring system vibration displacement error, belongs to SpatialAccording to gathering and processing technology field, technical scheme comprises the following steps: a) strengthening original vibration displacement data is polynary variable carCarry the continued time domain vibration displacement signal of measurement environment parameter-dependent, when characteristic change parameter parsing and continuous signal feature are providedThe basic data of territory interval division; B) each measurement environment parameter change threshold, variation tendency and signal frequency domain distribution are resolved in classificationCharacteristic relation, the constraints of dividing the interval and corrective interval vibration displacement of feature time domain is provided; C) utilize parameter to change thresholdIt is while having the signal of specific frequency domain complex distribution degree that value mapping relations dynamic self-adapting is divided continued time domain vibration displacement signalInterval, territory, the computing unit of correcting as vibration error; D) utilize level of detail parameter between variation tendency concord confining regionSpan, carry out by time domain is interval the vibration displacement error correction processing of choosing based on level of detail adaptive optimization. ThisThe core of invention spirit is: will affect " MMS Vibration stability " and then show associated " vibration displacement signal frequency domain complexity "" vehicle-mounted mobile measurement environment variable element " introduces signal errors correcting process, thereby overall continuous shaking displacement signal is decomposedWith the selected level of detail parameter of reconstruct, dynamically adaptive according to " variation characteristic of measurement environment parameter " between local intervalAnswer and optimize and revise, the present invention can solve existing signal processing method and be difficult to adapt to vehicle-mounted traverse measurement under complicated measurement environmentSystem vibration displacement is error correction problem accurately, contributes to support high accuracy vehicle-mounted mobile measurement data to resolve.

The continuous shaking displacement number that specifically obtains under complicated measurement environment for MMS and utilize integrated acceleration to resolveAccording to, provide a kind of and take time domain vibration displacement signal into account in the MMS of frequency domain complex characteristic vibration displacement error adaptive corrective method.Described " self adaptation " specifically refers to: will affect " MMS Vibration stability " and then show associated " vibration displacement signal frequency domain complexityDegree " " vehicle-mounted mobile measurement environment variable element " introduce signal errors correcting process, thereby make overall continuous shaking displacement letterNumber selected level of detail parameter of decomposition and reconstruction, moving according to " variation characteristic of measurement environment parameter " between local intervalState is optimized and revised adaptively.

Technical scheme of the present invention comprises the following steps:

Step 1, measurement environment parameter strengthens: MMS acceleration sensor outputs signals is carried out to quadratic integral parsing and obtainWhen the original vibration displacement of vertical direction, the measurement environment information that synchronous acquisition is corresponding with instantaneous vibration displacement; Adopt towardsThe method of object, in calculator memory, is recorded as the metrical information of parsing " connecting of " comprising polynary measurement environment parameter item "Continuous time domain vibration displacement signal data object ". Described measurement environment information specifically refers to that MMS is at complicated measurement environment runningIn, affect the variable measurement environmental information of vibration displacement signal frequency domain complexity, close with parameter item form and vibration displacement dataConnection. Preserve the basis that continued time domain vibration displacement signal object that measurement environment parameter strengthens utilizes as step 2 and step 3Data.

Step 2, characteristic change parameter is resolved: Facing Digital signal processing method, classification is resolved each measurement environment parameter and becomeChange the characteristic relation distributing with signal frequency domain. The variation characteristic relation of resolving successively comprises that quantitative " change threshold and frequency domain divideThe mapping relations of cloth scope " and " concord of variation tendency and frequency domain distribution range " qualitatively. Preserve respectively each parameter itemVariation characteristic relation provides constraint information as empirical model for step 3 and step 4.

Step 3, feature time domain interval division: based on the mapping relations of parameter change threshold and frequency domain distribution scope, dynamicallyDividing adaptively continued time domain vibration displacement signal is the adjacent time domain Interval Set of one group of change threshold constraint. Described each time domainThe corresponding specific frequency domain scope of effective vibration displacement signal component in interval, meets specific frequency domain complex distribution degree, therefore canEmploying has the digital signal processing method of specific detail level parameter and carries out decomposition. Preservation has specific frequency domain complex distribution degreeVibration displacement signal time domain interval as in step 4, carry out vibration error correct calculate processing unit.

Step 4, Oscillation error correction: utilize optional network specific digit signal processing method, carry out based on carefully by time domain is intervalThe vibration displacement error correction processing that ganglionic layer subparameter adaptive optimization is chosen. Described self adaptation specifically refer to utilize variation tendency withThe span of the concord constraint adjacent interval level of detail of frequency domain distribution range; Described preferably by by level of detailCorrection result carries out the realization of statistical nature qualitative assessment. The vibration displacement of preserving after correcting is used for supporting high-quality and high accuracy carCarry the post processing of traverse measurement spatial data.

And in described step 1, the variable measurement ambient parameter of associated vibration displacement signal frequency domain complexity comprises followingTwo classes:

Step 1a, MMS inner parameter: the polytropy of registration of vehicle running status, the parameter item that mainly can consider comprises fortuneRow state advance and time-out, travel condition under rectilinear motion and curvilinear motion, straight-line velocity and acceleration, curveThe angular speed of motion etc.;

Step 1b, MMS external parameter: record the diversity of road surface service condition, the parameter item that mainly can consider comprises roadSurface roughness and waviness etc.

The continued time domain vibration displacement signal data object parametric form table thus that comprises polynary measurement environment parameter itemBe shown:

Wherein, O_SRepresent the vibration displacement signal data object that ambient parameter strengthens, be recorded as one group with between particular sampleEvery the time T time domain multiple parameters group that is tag entry; S represents the vibration displacement that integrated acceleration obtains, be recorded as one group withObtain moment t, obtain locus x, the immediate movement s sequence that y is corresponding; P represents the polynary measurement environment parameter item strengthening, bagDraw together the inner parameter group { PI} and external parameter group { PO}, the each inner parameter PI that represent to obtain respectively S traverse measurement environment_mRecordFor the instantaneous parameters value p corresponding with obtaining moment t, each external parameter PO_nBe recorded as and obtain locus x, the parameter that y is correspondingValue p; S and PI_mCarry out synchronization association by time parameter t; S and PO_nCarry out synchronization association by moment parametric t.

Meanwhile, be to ensure the feasibility of analytical analysis, the each parameter item strengthening need be supported by the 1. outside historical letter of systemThe mode of breath Input matching, or the measuring transducer 2. configuring by the MMS mode of detection in real time, obtain corresponding with vibration displacementParameter value.

And in described step 2, each parameter change threshold and frequency domain distribution range mappings are related to the realization side of quantitative resolutionFormula comprises following sub-step:

Step 2.1, statistical parameter value distribution characteristics: extract the value of polynary measurement environment parameter item in data, successively systemThe characteristic value value and the frequency fr that count each parameter item, obtain every parameter p according to the array A of the ascending sequence of value_p：

Step 2.2, initializes threshold interval: to every parameter p, according to A_pSpan initializes threshold interval setSEC_p：

SEC_p＝{[value_{_min},value_{_max}]}；

Step 2.3, the threshold that level of detail (LevelsofDetail) the assessment displacement signal based on vibration displacement decomposesValue is interval: to the threshold interval S set EC of every parameter p_pCarry out following sub-step:

Step 2.3.1, extracts set element number Nr=|SEC_p|；

Step 2.3.2, successively to every section of interval SEC_p(n), n={1 ..., Nr}, carries out following sub-step:

I) according to A_pThe value value of middle interval characteristic value value_{_min},value_{_max}With mode value value_{_frmax}, drawBy stages is subintervalAnd subinterval

II) extract respectively O_SParameter p monodrome delta data section { the L1}, { L2} of middle corresponding subsec1 and subsec2;

III) adopt the Digital Signal Processing current techique towards frequency domain decomposition, choose particular procedure algorithm, comprise applicableFitting of a polynomial, least square fitting etc. that stationary signal decomposes, and the Fourier transform of applicable non-stationary signal, small echo becomeOne or more combinational algorithms in changing etc., respectively to data segment { L1} and { the original vibration displacement signal of the corresponding time domain of L2}S data interval S_L1And S_L2, by the effective displacement matching of level of detail, (the signal decomposition result of each level of detail is divided in executionFor " effectively " and " error " two parts, " effectively displacement matching " extracts the live part in decomposition result), obtain corresponding differentEffective displacement S of level of detail_{L1'level_n}And S_{L2'level_m}，Described fit procedure is logicalCross on the basis of decomposing at each level signal frequency domain, successively add up and represent that the multi-level high-frequency data item of effective vibration displacement is realExisting;

IV) the effective displacement S of each level of parsing gained_{L1'level_n}And S_{L2'level_m}Variances sigma² _{L1_level_n}Withσ² _{L2_level_m}；

V) extracting " variance minimum of a value " corresponding level of detail N1 and N2 is respectively the feature of subsec1 and subsec2Level of detail; Relatively N1 and N2:

G) as | N1-N2| < 1, directly finish this segment processing;

H), as | N1-N2|=1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former districtBetween section [value_{_min},value_{_max}], upgrade SEC simultaneously_pWith after Nr, finish this segment processing;

I), as | N1-N2| > 1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former districtBetween section [value_{_min},value_{_max}], and respectively to [value_{_min},value_{_frmax}] and [value_{_frmax},value_{_max}] holdRow step 2.3.2:I)～V);

Through step 2.3, obtain corresponding each feature level of detail N_p＝{N₁,N₂,...,N_Nr},Nr＝|SEC_p| parameter pInterval set SEC_p：

SEC_p＝{[value_{_min},value_{_frmax(a)}],[value_{_frmax(a)},value_{_frmax(b)}],

...,

[value_{_frmax(x)},value_{_frmax(y)}],[value_{_frmax(y)},value_{_max}]}；

Wherein, value_{_frmax(a)},value_{_frmax(b)}And value_{_frmax(x)},value_{_frmax(y)}Be [value_{_min},value_{_frmax}] local extremum in scope.

Step 2.4, resolves mapping relations based on threshold interval resampling: the SEC that step 2.3 is obtained_pCarry out based on districtBetween the linear interpolation of boundary value, obtain parameter p and { corresponding the passing of N} of frequency domain distribution scope Digital Signal Processing level of detail parameterIncrease change threshold interval:

SEC_p＝{[-∞,(value_{_min}＋value_{_frmax(a)})/2],

[(value_{_min}+value_{_frmax(a)})/2,(value_{_frmax(a)}+value_{_frmax(b)})/2],

...,

[(value_{_frmax(x)}+value_{_frmax(y)})/2,(value_{_frmax(y)}+value_{_max})/2],

[(value_{_frmax(y)}+value_{_max})/2,+∞]}；

Thus, parameter p change threshold and frequency domain distribution range mappings are closed and are:

$\left\{\begin{array}{c}(valu{e}_{\_min}+valu{e}_{\_frmax\left(a\right)})/2_{\&RightArrow;}{N}_1;\\ (valu{e}_{\_frmax\left(a\right)}+valu{e}_{\_frmax\left(b\right)})/2_{\&RightArrow;}{N}_2;\\ ...;\\ (valu{e}_{\_frmax\left(y\right)}+valu{e}_{\_\max })/2_{\&RightArrow;}{N}_{Nr-1};\\ (valu{e}_{\_fr\max \left(y\right)}+valu{e}_{\_\max })/2_{\&RightArrow;}{N}_{Nr};\end{array}\right.$

And in described step 2, each variation tendency and frequency domain distribution range concord R comprise: 1. correlation between the accumulation rate and speed, 2.Inverse change; The implementation of the qualitative parsing of described concord comprises following sub-step:

Step 2.5, resolves { the unidirectional constant interval SEC of N}_N：

Step 2.6, judges SEC_NThe uniformity in each interval change direction Sgn (sec_n) and parameter p incremental variations interval;If unanimously flag parameters p and this level of detail are interval is variation relation in the same way, be designated as R=1; Otherwise, flag parameters p and this is thinJoint level section is inverse change relation, is designated as R=-1;

And, in described step 3, utilizing change threshold constraint, dynamic self-adapting is divided the concrete enforcement of time domain Interval SetMode comprises following sub-step:

Step 3.1, extracts O according to the increasing order of time tag T_SIn the polynary environment measurement parameter group corresponding with S{P}；

Step 3.2, resolves the interval SEC of the each parameter threshold obtaining according to step 2_p, according to sequential increasing order from farthestMoment t_minStart to nearest moment t_maxWhen scanning S, travel through multiple parameters value corresponding to each moment instantaneous vibration displacement s{ P}, when the variation of arbitrary parameter p exceedes interval threshold value, dividing S is corresponding data section, obtains successively corresponding parameter changing valueVibration displacement time domain Interval Set S_sec：

Wherein, ψ represents parameter item codomain; C is marked off hop count; M value is object O_SIn polynary environment measurement ginsengSeveral numbers, m=|O_S.{P}|；N_pxBe that the value of x parameter p is at interval ψ p_xThe feature level of detail shining upon when variation; SpecialNot, p₁Change for recording the parameter item of crossing over interval threshold value.

And, in described step 4, based on correcting by time domain Oscillation displacement error of coherence of changing trend constraintDetailed description of the invention comprises following sub-step:

Successively to S_secIn every section of interval (S_X,ψ{P}_X),_X=1 ..., C} carries out following sub-step:

Step 4.1, resolves ψ { P}_XIn the codomain scope ψ { N of the feature level of detail that shines upon of each parameter p_p1,...,N_px,...,N_pm}；

Step 4.2, according to section ordinal number X, carry out following operation:

Step 4.2a, works as X=1, at ψ { N_p1,...,N_px,...,N_pmIn scope:

I) to S_XCarry out by the matching of level of detail vibration displacement; Described specific detail level level_x ∈ ψ { N_p1,...,N_px,...,N_pmThe concrete grammar of matching is: towards the Digital Signal Processing current techique of frequency domain decomposition, selecting step2.3.2III) the concrete Processing Algorithm adopting, comprises and is applicable to fitting of a polynomial, the least square fitting etc. that stationary signal decomposes,And one or more combinational algorithms in Fourier transform, the wavelet transformation etc. of applicable non-stationary signal, by S_XFrom 1 toOn the basis of level_x level frequency domain decomposition, successively the cumulative high-frequency data item that represents effective vibration displacement is realized;

II) calculate the variances sigma of each level of detail level_x² _{level_x}；

III) more each level of detail variance, the corresponding level of detail level_x (σ of mark " variance minimum of a value "²min)For segment (S_X,ψ{P}_X) optimum level of detail L_X; Preserve towards level of detail L_XFitting result be segment (S_X,ψ{P}_X) correction result.

Step 4.2b, works as X > 1:

I) obtain a segment (S_X-1,ψ{P}_X-1) optimum level of detail L_X-1＝level_x(σ²min)；

II) estimate the optimum level of detail increment Delta level that corresponding each parameter changes:

$\mathrm{\&Delta;}level=\left(\underset{i=1}{\overset{m}{\&Sigma;}}{R}_{m*}\right|{\left(N_{pi}\right)}_X-{\left(N_{pi}\right)}_{X-1}\left|\right)/m;$

Wherein, m=|O_S.{P}|, be object O_SIn polynary environment measurement number of parameters; R_mRepresent that i parameter p i is in districtBetween ψ { (N_pi)_X,(N_pi)_X-1Variation tendency concord value, by contrast step 2.5 and 2.6 SEC that resolve_NAssignment value 1Or-1;

III) estimation interval section (S_X,ψ{P}_X) optimal Decomposition level be:

L_X＝L_X－1+Δlevel；

IV) difference details of use level L_X-1,L_X,L_X+1, matching S_X; Described approximating method is with step 4.2aI);

V) calculate each level of detail variances sigma²; Judge fitting result:

A) work as σ² _LX-1≥σ² _LX≤σ² _LX+1, label L_XFor segment (S_X,ψ{P}_X) optimum level of detail; Preserve towards carefullyGanglionic layer time L_XFitting result be segment (S_X,ψ{P}_X) correction result;

B) work as σ² _LX-1≥σ² _LX≥σ² _LX+1, upgrade L_X＝L_X+1; Return to execution step 4.2bIV)～V);

C) work as σ² _LX-1≤σ² _LX≤σ² _LX+1, upgrade L_X＝L_X-1; Return to execution step 4.2bIV)～V);

Step 4.3, integrates piecemeal correction result and saves as the MMS vibration displacement value after the inventive method adaptive corrective.

The present invention is directed to vehicle-mounted mobile measuring system (MMS) entirety under complicated measurement environment obtains and utilizes acceleration long-pendingDivide the continuous shaking displacement data of resolving, a kind of vibration displacement mistake of taking time domain vibration displacement signal frequency domain complex characteristic into account is providedPoor adaptive corrective method.

Set forth the present invention program's principle below in conjunction with accompanying drawing.

As shown in Figure 1, the principle of technical solution of the present invention is: for the existing vibration position based on Digital Signal ProcessingShift error antidote adopts towards specific detail level overall situation Unified Solution, is difficult to adapt to the company that complicated measurement environment obtainsThe detail variation characteristic of effective information frequency domain distribution in continuous vibration displacement signal, cause local signal interval owe to decompose andCross and decompose and be difficult to realize vehicle-mounted traverse measurement system vibration displacement under complicated measurement environment and accurately ask with stable error correctionTopic. The inventive method is by impact " MMS Vibration stability " in complicated measurement environment and then show associated " vibration displacement signal frequency domainComplexity " " vehicle-mounted mobile measurement environment variable element " introduce signal errors correcting process:

1. by resolving and utilize " mapping relations of parameter change threshold and frequency domain distribution scope ", realize the continuous position of time domainThe interval self adaptation of multiple characteristic values of shifting signal is divided; Meanwhile,

2. by resolving " comformity relation of parameter variation tendency and frequency domain distribution range " and being introduced into interval divisionAlgorithm, the constraints of estimating as adjacent interval level of detail in signal;

Thereby make the selected level of detail parameter of overall continuous shaking displacement signal decomposition and reconstruction, in local intervalBetween optimize and revise according to " variation characteristic of measurement environment parameter " dynamic self-adapting.

Describe technical solution of the present invention in detail below in conjunction with embodiment accompanying drawing.

As shown in Figure 2, the general steps flow process of the inventive method is: first, strengthen original vibration displacement data in complexityThe measurement environment variable element item that shows associated vibration displacement signal frequency domain complexity in measurement environment, obtains polynary measurement environmentThe continued time domain vibration displacement signal of parameter-dependent, for characteristic change parameter resolve and continuous signal feature time domain interval division carryFor basic data; Then,, on the basis of induction parameter variation characteristic, each measurement environment parameter change threshold, change are resolved in classificationThe characteristic relation that change trend and signal frequency domain distribute, as the constraint bar of dividing feature time domain interval and corrective interval vibration displacementPart; Afterwards, utilize parameter change threshold mapping relations dynamic self-adapting to divide continued time domain vibration displacement signal, obtain thering is spyDetermine the signal time domain interval of frequency domain distribution complexity, the computing unit of correcting as vibration error; Finally, according to statistical nature byInterval qualitative assessment optimal Decomposition level also utilizes parameter variation tendency concord constrained optimum decomposition level computer capacity,The upper interval error correction of carrying out primary signal decomposition and useful signal reconstruct in this basis.

The implementation procedure of the embodiment of the present invention adopts computer realization automation processing, comprises following concrete steps:

Step 1, measurement environment parameter strengthens.

MMS acceleration sensor outputs signals is carried out to quadratic integral parsing and obtain the same of the original vibration displacement of vertical directionTime, " the measurement environment information " that synchronous acquisition is corresponding with instantaneous vibration displacement; Described " measurement environment information " specifically refers to that MMS is multipleIn assorted measurement environment running, affect the variable measurement environmental information of vibration displacement signal frequency domain complexity; By synchronous acquisitionMeasurement environment information associated with vibration displacement data with parameter item form. In reality measurement environment, affect vibration displacement signal frequentlyThe variable measurement environmental information of territory complexity is various, and the inventive method, according to the difference of information source, will affect vibration displacementThe variable measurement ambient parameter item of signal frequency domain complexity is summarized as two classes:

Step 1a, MMS inner parameter: the polytropy of registration of vehicle running status, the parameter item that mainly can consider comprises fortuneRow state advance and time-out, travel condition under rectilinear motion and curvilinear motion, straight-line velocity and acceleration, curveThe angular speed of motion etc.;

Step 1b, MMS external parameter: record the diversity of road surface service condition, the parameter item that mainly can consider comprises roadSurface roughness and waviness etc.

Thus, adopt OO method to record in calculator memory and resolve metrical information for comprising polynary measurement ringThe continued time domain vibration displacement signal data object of border parameter item, and data object parametric form is expressed as:

Wherein, O_SRepresent the vibration displacement signal data object that ambient parameter strengthens, be recorded as one group with between particular sampleEvery the time T time domain multiple parameters group that is tag entry; S represents the vibration displacement that integrated acceleration obtains, be recorded as one group withObtain moment t, obtain locus x, the immediate movement s sequence that y is corresponding; P represents the polynary measurement environment parameter item strengthening, bagDraw together the inner parameter group { PI} and external parameter group { PO}, the each inner parameter PI that represent to obtain respectively S traverse measurement environment_mRecordFor the instantaneous parameters value p corresponding with obtaining moment t, each external parameter PO_nBe recorded as and obtain locus x, the parameter that y is correspondingValue p; S and PI_mCarry out synchronization association by time parameter t; S and PO_nCarry out synchronization association by location parameter t. For ensureing to resolveThe feasibility of analyzing, the each parameter item strengthening need support by the 1. mode of the outside historical information Input matching of system, or 2. logicalCross the measuring transducer mode of detection in real time of MMS configuration, obtain the parameter value corresponding with each instantaneous vibration displacement.

Affecting in the variable measurement environmental information of vibration displacement signal frequency domain complexity, " speed " is as MMS inside oneItem is easy to adopt existing sensor to survey and show the travel condition of vehicle information of associated MMS Vibration stability, and velocity amplitude is multipleIn assorted measurement environment, particularly urban road environment, because other operational vehicles of road surface, barrier, restricted driving condition etc. are uncertainFactor impact very easily changes in measuring process; In addition gaseous-pressure when, carrier vehicle engine and transmission system are workedThe energy imbalance that is directly subject to the impact of " road surface fluctuating " with the inertia force cyclically-varying meeting of movement parts and change, and thenAffect vibration displacement signal frequency domain and distribute, and there is between section between even each road surface point road surface waviness in actual pavement conditionsDifference. Therefore, the present invention is with MMS interior vehicle running state parameter " speed " and service condition parameter " road surface, the outside road surface of MMSRising and falling " two traverse measurement processes the most easily change and then the measurement environment parameter that the most directly affects Vibration stability is example,Summary of the invention is described, wherein, following steps are carried out in the concrete enforcement that measurement environment parameter strengthens:

To a of instantaneous signal one by one of acceleration transducer output in traverse measurement process_i(t) adopt without loss of generality logicalSolve original vibration displacement s of each corresponding moment with quadratic integral formula_i：

$s_i=s_0+\underset{k=1}{\overset{i}{\&Sigma;}}(v_0+\underset{j=1}{\overset{k}{\&Sigma;}}{a}_j\mathrm{\&Delta;}t)\mathrm{\&Delta;}t;$

In formula, s₀Represent vertical direction initial displacement observation; v₀Represent vertical direction initial velocity observation; Δ t is for addingVelocity sensor sampling time interval.

Solve original vibration displacement s_iMeanwhile, 1. for example, according to distance-measuring device (the conventional wheel encoder of MMS configurationDevice) obtain the distance increment Δ d of every section of time interval Δ t_i, and then calculate the average speed v in every segment distance increment＝Δd_i/ Δ t, and by moment t and vibration displacement s_iAssociated; 2. will be according to existing laser measurement method, Analytic Calculation Method or(IRI is by one group for the historical evenness of road surface degrees of data IRI that one or more combined method processing in leveling measuring method etc. obtainCorresponding specific location information x, the flatness value iri of y represents), the positional information x obtaining according to GPS alignment sensor, y withVibration displacement s_iCoupling is associated. Create thus velocity dependent V and represent that flatness IRI and parametric form that road surface rises and falls representVibration displacement data object O_S：

$\left\{\begin{array}{c}Os=\{T,S,\{P\left\}\right\};\\ P=\{V,IRI\};\\ S=\left\{{(s,t,x,y)}_i\right\};\\ V=\left\{{(v,t)}_i\right\};\\ IRI=\left\{{(iri,x,y)}_i\right\};\end{array}\right.$

Preserve O_SThe basic data of inputting as step 2 and step 3.

Step 2, characteristic change parameter is resolved.

Towards optional network specific digit signal processing method, O is resolved in classification_SIn each measurement environment parameter change with signal frequency domain and distributeCharacteristic relation. The variation characteristic relation of resolving successively comprises quantitative " mapping relations of change threshold and frequency domain distribution scope "" concord of variation tendency and frequency domain distribution range " qualitatively.

In the selection of described optional network specific digit signal processing method, the conventional general skill of Digital Signal Processing towards frequency domain decompositionArt comprises the fitting of a polynomial, the least square fitting etc. that are applicable to stationary signal decomposition, and the Fourier of applicable non-stationary signalConversion, wavelet transformation etc. Described characteristic relation resolving, can select a kind of or many in existing digital signal processing methodPlanting combination implements; Selected distinct methods is embodied as respectively difference corresponding to the level of detail parameter of signal frequency domain complex distribution degreeParameter item, mainly comprise:

1. the parameter item based on small wave converting method is embodied as " the signal decomposition number of plies ";

2. the parameter item based on polynomial fitting method is embodied as " intercepting multinomial exponent number ";

3. the parameter item based on high-pass filtering method is embodied as " wave filter is by frequency ".

Wherein, preferably, take vibration displacement data into account and there is typical non-stationary characteristic in the time-domain signal overall situation, be more suitable for choosingBy the wavelet transformation class processing method with excellent time domain, frequency domain multiresolution disposal ability;

Further preferably, due to the Local Symmetries that has of periodic feature of vibration, in concrete wavelet transformation classWhen processing method is carried out, be more suitable for selecting the method with symmetrical wavelet function, as Coiflet small echo, Symlet small echo andOne in Biotrhogonal small echo.

Therefore, the embodiment of the present invention specifically adopts wavelet-decomposing method, specifically selects Coiflet wavelet function to carry out parameterVariation characteristic is resolved.

Wherein, each parameter change threshold and frequency domain distribution range mappings are related to that the implementation of quantitative resolution comprises following sonStep:

Step 2.1, statistical parameter value distribution characteristics. Extract the value of polynary measurement environment parameter item in data, successively systemThe characteristic value value and the frequency fr that count each parameter item, obtain every parameter p according to the array A of the ascending sequence of value_p：

The embodiment of the present invention is corresponding specifically carries out above-mentioned feature primary system to the flatness IRI of speed V and the fluctuating of expression road surfaceMeter, obtains respectively array A_VAnd A_IRI。

Step 2.2, initializes threshold interval. To every parameter p, according to A_pSpan initializes threshold interval setSEC_p：

SEC_p＝{[value_{_min},value_{_max}]}；

The embodiment of the present invention is corresponding respectively according to the A of example test data section_VAnd A_IRISpan initializes threshold intervalSet, obtains taking (thousand ms/h) as unit and SEC that logarithm value rounds_VWith taking (rice/km) as numerical value unit and logarithm valueThe SEC rounding_IRI：

SEC_V＝{[15,50]}；

SEC_IRI＝{[2,6]}；

Step 2.3, based on vibration displacement level of detail assessment decomposition threshold interval. To the threshold interval collection of every parameter pClose SEC_p(in the present embodiment respectively to SEC_VAnd SEC_IRI), carry out following sub-step:

Step 2.3.1, extracts set element number Nr=|SEC_p|, initial value is 1;

Step 2.3.2, successively to every section of interval SEC_p(n), n={1 ..., Nr}, carries out following sub-step:

I) according to A_pThe value value of middle interval characteristic value value_{_min},value_{_max}With mode value value_{_frmax}, drawBy stages is subintervalAnd subinterval

II) extract respectively O_SParameter p monodrome delta data section { the L1}, { L2} of middle corresponding subsec1 and subsec2;

III) adopt the Digital Signal Processing current techique towards frequency domain decomposition, choose particular procedure algorithm, respectively logarithmAccording to section { L1} and { the original vibration displacement signal of the corresponding time domain of L2} S data interval S_L1And S_L2, carry out effective by level of detail(" displacement signal " that primary signal obtains comprises " effectively " composition and " error displacement " composition, refers to non-error displacement here in displacementComposition, i.e. effectively displacement) matching, obtain effective displacement S of corresponding different level of detail_{L1'level_n}And S_{L2'level_m}，Described fit procedure is passed through first to decompose at each level signal frequency domain, thereby by yardstickLevel is peeled off the vibration displacement of the effective vibration displacement of high frequency and low frequency combined error, then successively adds up and represents effective vibration displacementMulti-level high-frequency data item realize.

The general Mallat discrete wavelet decomposition and reconstruction algorithm of the concrete employing of the embodiment of the present invention, uses following small echo generalDecomposition and reconstruction formula is realized the wavelet transformation of discrete level of detail:

$\begin{array}{c}{c}_{j+1}\left(n\right)=h\left(n\right)*c_j\left(n\right)=\underset{k}{\&Sigma;}h\left(n\right)c_j(2n-k)\\ d_{j+1}\left(n\right)=g\left(n\right)*c_j\left(n\right)=\underset{k}{\&Sigma;}g\left(n\right)c_j(2n-k)\end{array}$

$c_j\left(n\right)=\underset{k}{\&Sigma;}h(n-2k)c_{j+1}\left(k\right)+\underset{k}{\&Sigma;}g(n-2k)d_{j+1}\left(k\right)$

Wherein, d_jBe the effective vibration displacement of high frequency of j level of detail, c_jMixed for what decomposite at j level of detail low frequencyClose the low-frequency vibration displacement of error percentage, h (n) and g (n) bank of filters for being obtained by orthogonal wavelet transformation is corresponding concreteWavelet function and scaling function thereof, n, k is discrete signal ordinal number.

IV) the effective displacement S of each level of parsing gained_{L1'level_n}And S_{L2'level_m}Variances sigma² _{L1_level_n}Withσ² _{L2_level_m}；

V) extracting " variance minimum of a value " corresponding level of detail N1 and N2 is respectively the feature of subsec1 and subsec2Level of detail; Relatively N1 and N2:

J) as | N1-N2| < 1, directly finish this segment processing;

K), as | N1-N2|=1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former districtBetween section [value_{_min},value_{_max}], upgrade SEC simultaneously_pWith after Nr, finish this segment processing;

L), as | N1-N2| > 1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former districtBetween section [value_{_min},value_{_max}], and respectively to [value_{_min},value_{_frmax}] and [value_{_frmax},value_{_max}] holdRow step 2.3.2:I)～V);

Through step 2.3, obtain corresponding to feature level of detail N_p＝{N₁,N₂,...,N_Nr},Nr＝|SEC_p| parameter pInterval set SEC_p：

SEC_p＝{[value_{_min},value_{_frmax(a)}],[value_{_frmax(a)},value_{_frmax(b)}],

...,

[value_{_frmax(x)},value_{_frmax(y)}],[value_{_frmax(y)},value_{_max}]}；

Embodiment of the present invention test data section is corresponding obtains respectively dividing the speed parameter interval that between back zone, boundary value roundsSEC_VAnd SEC between surface relief parameter region_IRI：

SEC_V＝{[15,20],[20,30],[30,50]}；N_v＝{(5,6),(6,7),(7,8)}；

SEC_IRI＝{[2,4],[4,6]}；N_IRI＝{(5,6,7),(6,7,8)}；

Step 2.4, resolves mapping relations based on threshold interval resampling: the SEC that step 2.3 is obtained_pCarry out based on districtBetween boundary value linear interpolation, obtain parameter p and { corresponding the increasing progressively of N} of frequency domain distribution scope Digital Signal Processing level of detail parameterChange threshold interval:

SEC_p＝{[-∞,(value_{_min}＋value_{_frmax(a)})/2],

[(value_{_min}＋value_{_frmax(a)})/2,(value_{_frmax(a)}＋value_{_frmax(b)})/2],

...,

[(value_{_frmax(x)}+value_{_frmax(y)})/2,(value_{_frmax(y)}+value_{_max})/2],

[(value_{_frmax(y)}+value_{_max})/2,+∞]}；

Meanwhile, parameter p change threshold and frequency domain distribution range mappings are closed and are:

$\left\{\begin{array}{c}(valu{e}_{\_min}+valu{e}_{\_frmax\left(a\right)})/2_{\&RightArrow;}{N}_1;\\ (valu{e}_{-frmax\left(a\right)}+valu{e}_{\_frmax\left(b\right)})/2_{\&RightArrow;}{N}_2;\\ ...;\\ (valu{e}_{\_frmax\left(y\right)}+valu{e}_{\_\max })/2_{\&RightArrow;}{N}_{Nr-1};\\ (valu{e}_{\_fr\max \left(y\right)}+valu{e}_{\_\max })/2_{\&RightArrow;}{N}_{Nr};\end{array}\right.$

Thus, embodiment of the present invention test data section is corresponding obtains respectively speed parameter V and surface relief parameter I RI and changeChange threshold value and frequency domain distribution range mappings are closed:

$V\&RightArrow;N=\left\{\begin{array}{c}17.5\&RightArrow;(5,6);\\ 25\&RightArrow;(6,7);\\ 40\&RightArrow;(7,8);\end{array}\right.$

$IRI\&RightArrow;N=\left\{\begin{array}{c}3\&RightArrow;(5,6,7);\\ 5\&RightArrow;(6,7,8);\end{array}\right.$

Meanwhile, each variation tendency and frequency domain distribution range concord R comprise: 1. correlation between the accumulation rate and speed, 2. inverse change; DescribedThe implementation of the qualitative parsing of concord comprises following sub-step:

Step 2.5, resolves { the unidirectional constant interval SEC of N}_N：

Step 2.6, judges SEC_NThe uniformity in each interval change direction Sgn (sec_n) and parameter p incremental variations interval;If unanimously flag parameters p and this level of detail are interval is variation relation in the same way, be designated as R=1; Otherwise, flag parameters p and this is thinJoint level section is inverse change relation, is designated as R=-1;

The embodiment of the present invention is resolved thus and is obtained speed parameter V and surface relief parameter I RI and frequency domain distribution range and beVariation relation in the same way, is all designated as R=1, and its implication is: along with speed V increases, MMS vibrates non-stationary enhancing, effectively vibrationDisplacement composition occupies in original vibration displacement signal has wider frequency domain scope, in the time that signal is processed, needs more detailsInformation is reduced useful signal; In like manner, along with surface relief parameter I RI increases, also need more details information to reduce effectivelySignal.

Preserve respectively parameter item " speed V " and represent that the variation characteristic relation of " flatness IRI " that road surface rises and falls is as warpTest model and provide constraint information for step 3 and step 4.

Step 3, feature time domain interval division.

Based on the mapping relations of parameter change threshold and frequency domain distribution scope, continued time domain vibration is divided on dynamic self-adapting groundDisplacement signal is the adjacent time domain Interval Set of one group of change threshold constraint, specifically comprises following sub-step:

Step 3.1, extracts O according to the increasing order of time tag T_SIn the polynary environment measurement parameter group corresponding with SP}, resolve in the embodiment of the present invention:

O_S.V＝{(v,t)_i}；

O_S.IRI＝{(iri,x,y)_i}；

Step 3.2, resolves the interval SEC of the each parameter threshold obtaining according to step 2_p, according to sequential increasing order from farthestMoment t_minStart to nearest moment t_maxWhen scanning S, travel through multiple parameters value corresponding to each moment instantaneous vibration displacement s{ P}, when the variation of arbitrary parameter p exceedes interval threshold value, dividing S is corresponding data section, obtains successively corresponding parameter changing valueVibration displacement time domain Interval Set S_sec：

Wherein, ψ represents parameter item codomain; C is marked off hop count; M value is object O_SIn polynary environment measurement ginsengSeveral numbers, m=|O_S.{P}|；N_pxBe that the value of x parameter p is at interval Ψ p_xThe feature level of detail shining upon when variation;Especially, p₁Change for recording the parameter item of crossing over interval threshold value.

Embodiment of the present invention test data section is basis: SEC_V={ [15,20], [20,30], [30,50] } (km/littleTime) and SEC_IRI={ [2,4], [4,6] } (rice/km) for threshold value dividing data section be | S_sec|=15 intervals. Described when eachThe corresponding specific frequency domain scope of effective vibration displacement signal component in interval, territory, meets specific frequency domain complex distribution degree, thereforeCan adopt the digital signal processing method with specific detail level parameter to carry out decomposition.

Preservation has the vibration displacement signal time domain interval of specific frequency domain complex distribution degree as carrying out vibration mistake in step 4The poor processing unit calculating of correcting.

Step 4, Oscillation error correction.

Utilize optional network specific digit signal processing method, carry out the optimization based on level of detail parameter adaptive by time domain interval and chooseThe processing of vibration displacement error correction. Described self adaptation specifically refers to utilize " speed V " and represents " the flatness IRI " that road surface rises and fallsVariation tendency and the span of the concord constraint adjacent interval level of detail of frequency domain distribution range, according to step 2Resolve, when after the one piece of data parameter value that intersects at data the last period increase, in requisition for increase wavelet decomposition levelExtract effective vibration displacement composition; Described preferably by realizing calculating variance minimum of a value by level of detail correction result. SpecificallyEmbodiment comprises following sub-step:

Successively to S_secIn every section of interval (S_X,ψ{P}_X),_X=1 ..., C} carries out following sub-step:

Step 4.1, resolves ψ { P}_XIn the codomain scope ψ { N of the feature level of detail that shines upon of each parameter p_p1,...,N_px,...,N_pm}；

Step 4.2, according to section ordinal number X (X={1 in embodiment, 2 ..., 15}), carry out following operation:

Step 4.2a, works as X=1, at ψ { N_p1,...,N_px,...,N_pmIn scope:

I) to S_XCarry out by the matching of level of detail vibration displacement; Described specific detail level level_x ∈ ψ { N_p1,...,N_px,...,N_pmThe concrete grammar of matching is: the concrete wavelet-decomposing method of choosing in step 2, the tool of adopting in the embodiment of the present inventionBody selects Coiflet wavelet function to process, by S_XOn the basis of 1 to level_x level frequency domain decomposition, successively tiredAdd the high-frequency data item realization that represents effective vibration displacement. In embodiment test data section, level_x ∈ { 5,6,7,8};

II) calculate the variances sigma of each level of detail level_x² _{level_x}; In embodiment test data section, σ² _{level_5}＝349；σ² _{level_6}＝242；σ² _{level_7}＝96；σ² _{level_8}＝115；

III) more each level of detail variance, the corresponding level of detail level_x (σ of mark " variance minimum of a value "²min)For segment (S_X,ψ{P}_X) optimum level of detail L_X, in embodiment test data section, be L₇; Preserve towards level of detail L₇'sFitting result is segment (S_X,ψ{P}_X) correction result.

Step 4.2b, works as X > 1:

I) obtain a segment (S_X-1,ψ{P}_X-1) optimum level of detail L_X-1＝level_x(σ²min)；

II) estimate the optimum level of detail increment Delta level that corresponding each parameter changes:

$\mathrm{\&Delta;}level=\left(\underset{i=1}{\overset{m}{\&Sigma;}}{R}_{m*}\right|{\left(N_{pi}\right)}_X-{\left(N_{pi}\right)}_{X-1}\left|\right)/m;$

Wherein, m=|O_S.{P}|, be object O_SIn polynary environment measurement number of parameters; R_mRepresent that i parameter p i is in districtBetween ψ { (N_pi)_X,(N_pi)_X-1Variation tendency concord value, by contrast step 2.5 and 2.6 SEC that resolve_NAssignment value 1Or-1;

III) estimation interval section (S_X,ψ{P}_X) optimal Decomposition level be:

L_X＝L_X-1+Δlevel；

IV) difference details of use level L_X-1,L_X,L_X+1, matching S_X; Described approximating method is with step 4.2aI);

V) calculate each level of detail variances sigma²; Judge fitting result:

D) work as σ² _LX-1≥σ² _LX≤σ² _LX+1, label L_XFor segment (S_X,ψ{P}_X) optimum level of detail; Preserve towards carefullyGanglionic layer time L_XFitting result be segment (S_X,ψ{P}_X) correction result;

E) work as σ² _LX-1≥σ² _LX≥σ² _LX+1, upgrade L_X＝L_X+1; Return to execution step 4.2bIV)～V);

F) work as σ² _LX-1≤σ² _LX≤σ² _LX+1, upgrade L_X＝L_X-1; Return to execution step 4.2bIV)～V);

Method accordingly, 15 interval corresponding level of detail that embodiment of the present invention test data is divided are respectively: 7,7,8,8,8,7,7,7,6,6,5,5,6,5,6}；

Step 4.3, integrates piecemeal correction result and saves as the MMS vibration displacement value after the inventive method adaptive corrective.Wherein, when each segment data section interval division there is situation about not overlapping in place, by average value processing, ensures seamless between data segmentBe connected. The vibration displacement of preserving after correcting is used for supporting high-quality and high accuracy vehicle-mounted mobile measurement space Data Post.

MMS vibration displacement data after rectification can be used as Given information, further accurately resolves and adopts based on MMS for supportingThe three-dimensional spatial information of collection. Due to the level of detail that vibration displacement data is decomposed, basis " parameter variation " between local signalOptimize and revise impact dynamic self-adapting, thus in significantly reducing between each parameter region for effective vibration component cross decompose withOwe resolution problem, in improving the precision of correcting rear data, also strengthened the adaptability that MMS changes measurement environment, can prop upHold high-quality and high accuracy vehicle-mounted mobile measurement space Data Post; In addition, due to by characteristic change parameter and signal complexityThe consistency constraint of degree is introduced in the multiple characteristic values interval decomposed algorithm of continuous signal, sets up and takes full advantage of parameter and change modelEnclose and the required corresponding relation that carries out decomposition level, thereby the computer capacity of having dwindled actual decomposition level, reduces calculatingRedundancy and complexity, in the time correcting a large amount of measurement data, can effectively promote overall treatment efficiency especially.

Above-mentioned specific embodiment is only to the explanation for example of the present invention's spirit, not the present invention is made to any pro forma limitSystem. When concrete enforcement, carry out software programming by those skilled in the art according to above-mentioned flow process and realize;In implementation process, can not depart from technical solution of the present invention or surmount the defined scope of appended claims equivalent variations,Replace and modify, all belong in the scope of technical solution of the present invention.

Claims (7)

1. the adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error, is characterized in that, comprises the following steps:

Step 1, measurement environment parameter strengthens: the output signal of acceleration transducer is carried out to quadratic integral, and to obtain vertical direction formerBeginning vibration displacement, and the synchronous acquisition measurement environment information corresponding with original vibration displacement, record measurement environment information for comprisingThe continued time domain vibration displacement signal data object of measurement environment parameter;

Step 2, characteristic change parameter is resolved: resolve each measurement environment parameter and change the characteristic relation distributing with signal frequency domain, bagDraw together mapping relations, measurement environment parameter variation tendency and the frequency domain of measurement environment parameter change threshold and signal frequency domain distributionThe relation of distribution range;

Step 3, feature time domain interval division: the mapping based on measurement environment parameter change threshold and signal frequency domain distribution is closedSystem, dynamically dividing continued time domain vibration displacement signal is the adjacent time domain Interval Set of one group of change threshold constraint;

Step 4, Oscillation error correction: the relation constraint phase based on measurement environment parameter variation tendency with frequency domain distribution rangeThe span of neighboring interval level of detail, by the interval execution of the time domain vibration position that optimization is chosen based on level of detail parameter adaptiveShift error correction process.

2. the adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error according to claim 1, its featureBe that described measurement environment information comprises two classes:

Inner parameter: comprise running status advance and time-out, travel condition under rectilinear motion and curvilinear motion, rectilinear motionVelocity and acceleration, the angular speed of curvilinear motion;

External parameter: surface roughness and waviness.

3. the adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error according to claim 1, its featureBe described vibration displacement signal data object:

Wherein, O_SRepresent vibration displacement signal data object, be recorded as one group taking the time T of setting the sampling interval as tag entryTime domain multiple parameters group; S represents the vibration displacement that integrated acceleration obtains, be recorded as one group with obtain moment t, obtain space bitPut x, the immediate movement s sequence that y is corresponding; P represents the polynary measurement environment parameter item strengthening, and comprises and represents that respectively obtaining S movesInner parameter group { PI} and external parameter group { PO}, each inner parameter PI of measurement environment_mBe recorded as with to obtain moment t correspondingInstantaneous parameters value p, each external parameter PO_nBe recorded as and obtain locus x, the parameter value p that y is corresponding; S and PI_mJoin by the timeNumber t carries out synchronization association; S and PO_nCarry out synchronization association by location parameter t.

4. the adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error according to claim 1, its featureThe mapping relations that are described measurement environment parameter change threshold and signal frequency domain distribution comprise the following steps:

Step 2.1, statistical parameter value distribution characteristics: extract the value of polynary measurement environment parameter item in data, statistics is each successivelyThe characteristic value value of parameter item and frequency fr, obtain every parameter p according to the array A of the ascending sequence of value_p：

Step 2.2, initializes threshold interval: to every parameter p, according to A_pSpan initializes threshold interval S set EC_p：

SEC_p＝{[value_{_min},value_{_max}]}；

Step 2.3, assessment decomposition threshold interval: to the threshold interval S set EC of every parameter p_pCarry out following sub-step:

Step 2.3.1, extracts set element number Nr=|SEC_p|；

Step 2.3.2, successively to every section of interval SEC_p(n), n={1 ..., Nr}, carries out following sub-step:

I) according to A_pThe value value of middle interval characteristic value value_{_min},value_{_max}With mode value value_{_frmax}, demarcation intervalFor subinterval subsec1[value_{_min},value_{_frmax}] and subinterval subsec2[value_{_frmax},value_{_max}]；

II) extract respectively O_SParameter p monodrome delta data section { the L1}, { L2} of middle corresponding subsec1 and subsec2;

III) respectively to data segment { L1} and { the original vibration displacement signal of the corresponding time domain of L2} S data interval S_L1And S_L2, holdThe effective displacement matching of row, obtains effective displacement S of corresponding different level of detail_{L1'level_n}With

IV) obtain the effective displacement S of each level_{L1'level_n}And S_{L2'level_m}Variances sigma² _{L1_level_n}And σ² _{L2_level_m}；

V) extracting respectively the corresponding level of detail N1 of variance minimum of a value and N2 is the feature levels of detail of subsec1 and subsec2Inferior; Relatively N1 and N2:

A) as | N1-N2| < 1, finish this segment processing;

B), as | N1-N2|=1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former segment[value_{_min},value_{_max}], upgrade SEC simultaneously_pWith after Nr, finish this segment processing;

C), as | N1-N2| > 1, adopt [value_{_min},value_{_frmax}],[value_{_frmax},value_{_max}] replace former segment[value_{_min},value_{_max}], and respectively to [value_{_min},value_{_frmax}] and [value_{_frmax},value_{_max}] execution stepRapid 2.3.2:I)～V);

Through step 2.3, obtain corresponding to feature level of detail N_p＝{N₁,N₂,...,N_Nr},Nr＝|SEC_p| parameter p intervalS set EC_p：

$\begin{array}{c}{\mathrm{SEC}}_p=\{\&lsqb;{\mathrm{value}}_{\_\min },+{\mathrm{value}}_{\_fr\max \left(a\right)}\&rsqb;,\&lsqb;{\mathrm{value}}_{\_fr\max \left(a\right)},+{\mathrm{value}}_{\_fr\max \left(b\right)}\&rsqb;,\\ \mathrm{...},\\ \&lsqb;{\mathrm{value}}_{\_fr\max \left(x\right)},+{\mathrm{value}}_{\_fr\max \left(y\right)}\&rsqb;,\&lsqb;{\mathrm{value}}_{\_fr\max \left(y\right)},+{\mathrm{value}}_{\_\max }\&rsqb;;\end{array}$

Step 2.4, resolves mapping relations based on threshold interval resampling: the SEC that step 2.3 is obtained_pCarry out based on interval borderValue linear interpolation, obtains parameter p and frequency domain distribution scope Digital Signal Processing level of detail parameter { the incremental variations threshold that N} is correspondingValue is interval:

$\begin{array}{c}{\mathrm{SEC}}_p=\{\&lsqb;-\mathrm{\&infin;},\left({\mathrm{value}}_{\_\min }+{\mathrm{value}}_{\_fr\max \left(a\right)}\right)/2\&rsqb;,\\ \&lsqb;\left({\mathrm{value}}_{\_\min }+{\mathrm{value}}_{\_fr\max \left(a\right)}\right)/2,\left({\mathrm{value}}_{\_fr\max \left(a\right)}+{\mathrm{value}}_{\_fr\max \left(b\right)}\right)/2\&rsqb;,\\ \mathrm{...},\\ \&lsqb;\left({\mathrm{value}}_{\_fr\max \left(x\right)}+{\mathrm{value}}_{\_fr\max \left(y\right)}\right)/2,\left({\mathrm{value}}_{\_fr\max \left(y\right)}+{\mathrm{value}}_{\_\max }\right)/2\&rsqb;,\\ \&lsqb;\left({\mathrm{value}}_{\_fr\max \left(y\right)}+{\mathrm{value}}_{\_\max }\right)/2,+\mathrm{\&infin;}\&rsqb;\};\end{array}$

Thus, parameter p change threshold and frequency domain distribution range mappings are closed and are:

$\left\{\begin{array}{c}(valu{e}_{\_min}+valu{e}_{\_frmax\left(a\right)})/2\&RightArrow;N_1;\\ (valu{e}_{\_frmax\left(a\right)}+valu{e}_{\_frmax\left(b\right)})/2\&RightArrow;N_2;\\ ...;\\ (valu{e}_{\_frmax\left(y\right)}+valu{e}_{\_\max })/2\&RightArrow;N_{Nr-1};\\ (valu{e}_{\_fr\max \left(y\right)}+valu{e}_{\_\max })/2\&RightArrow;N_{Nr};\end{array}\right..$

5. the adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error according to claim 1, its featureThe relation that is described measurement environment parameter variation tendency and frequency domain distribution range comprises following sub-step:

Step 2.5, resolves { the unidirectional constant interval SEC of N}_N：

Step 2.6, judges SEC_NWhether each interval change direction Sgn (sec_n) is consistent with parameter p incremental variations interval; If oneCause, flag parameters p and this level of detail are interval is variation relation in the same way, is designated as R=1; Otherwise, flag parameters p and this detailsLevel section is inverse change relation, is designated as R=-1.

6. the adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error according to claim 1, its featureBe the described mapping relations based on measurement environment parameter change threshold and signal frequency domain distribution, dynamically divide continued time domainVibration displacement signal is that the adjacent time domain Interval Set of one group of change threshold constraint comprises following sub-step:

Step 3.1, extracts O according to the increasing order of time tag T_SIn the polynary environment measurement parameter group { P} corresponding with S;

Step 3.2, the interval SEC of each parameter threshold_p, according to sequential increasing order from moment t farthest_minStart to nearest moment t_maxScanning S, travel through the multiple parameters value that each moment instantaneous vibration displacement s is corresponding { P}, when the variation of arbitrary parameter p exceedes district simultaneouslyBetween threshold value, dividing S is corresponding data section, obtains successively the vibration displacement time domain Interval Set S of corresponding parameter changing value_sec：

Wherein, ψ represents parameter item codomain; C is marked off hop count; M value is object O_SIn polynary environment measurement parameterNumber, m=|O_S.{P}|；N_pxBe that the value of x parameter p is at interval ψ_pxThe feature level of detail shining upon when variation; p₁Be used forRecord changes the parameter item of crossing over interval threshold value.

7. the adaptive corrective method of vehicle-mounted mobile measuring system vibration displacement error according to claim 1, its featureBe getting of the described relation constraint adjacent interval level of detail based on measurement environment parameter variation tendency and frequency domain distribution rangeValue scope comprises following sub-step:

Successively to S_secIn every section of interval (S_X,ψ{P}_X), X={1 ..., C} carries out following sub-step:

Step 4.1, resolves ψ { P}_XIn the codomain scope ψ { N of the feature level of detail that shines upon of each parameter p_p1,...,N_px,...,N_pm}；

Step 4.2, according to section ordinal number X, carry out following operation:

Step 4.2a, works as X=1, at ψ { N_p1,...,N_px,...,N_pmIn scope:

I) to S_XCarry out by the matching of level of detail vibration displacement; By to S_XOn the basis of 1 to level_x level frequency domain decomposition,The successively cumulative high-frequency data item that represents effective vibration displacement; Wherein, level_x ∈ ψ { N_p1,...,N_px,...,N_pm}；

II) calculate the variances sigma of each level of detail level_x² _{level_x}；

III) more each level of detail variance, the corresponding level of detail level_x of mark variance minimum of a value (σ²Min) be intervalSection (S_X,ψ{P}_X) optimum level of detail L_X; Preserve towards level of detail L_XFitting result be segment (S_X,ψ{P}_X) rectifyPositive result;

Step 4.2b, works as X > 1:

I) obtain a segment (S_X-1,ψ{P}_X-1) optimum level of detail L_X-1＝level_x(σ²min)；

II) estimate the optimum level of detail increment Delta level that corresponding each parameter changes:

$\mathrm{\&Delta;}level=\left(\underset{i=1}{\overset{m}{\&Sigma;}}{R}_{m*}\right|{\left(N_{pi}\right)}_X-{\left(N_{pi}\right)}_{X-1}\left|\right)/m;$

Wherein, m=|O_S.{P}|, be object O_SIn polynary environment measurement number of parameters; R_mRepresent that i parameter p i is at interval ψ{(N_pi)_X,(N_pi)_X-1Variation tendency concord value, i.e. SEC_NValue;

III) estimation interval section (S_X,ψ{P}_X) optimal Decomposition level be:

L_X＝L_X-1+Δlevel；

IV) difference details of use level L_X-1,L_X,L_X+ 1, matching S_X; Described approximating method is with step 4.2aI);

V) calculate each level of detail variances sigma²; Judge fitting result:

A) work as σ² _LX-1≥σ² _LX≤σ² _LX+1, label L_XFor segment (S_X,ψ{P}_X) optimum level of detail; Preserve towards levels of detailInferior L_XFitting result be segment (S_X,ψ{P}_X) correction result;

B) work as σ² _LX-1≥σ² _LX≥σ² _LX+1, upgrade L_X＝L_X+ 1; Return to execution step 4.2b:IV)～V);

C) work as σ² _LX-1≤σ² _LX≤σ² _LX+1, upgrade L_X＝L_X-1; Return to execution step 4.2b:IV)～V);

Step 4.3, integrates piecemeal correction result and saves as the vibration displacement value after rectification.